Article pubs.acs.org/est
Model Analysis of Secondary Organic Aerosol Formation by Glyoxal in Laboratory Studies: The Case for Photoenhanced Chemistry Andrew J. Sumner, Joseph L. Woo, and V. Faye McNeill* Department of Chemical Engineering, Columbia University, New York, New York 10027, United States ABSTRACT: The reactive uptake of glyoxal by atmospheric aerosols is believed to be a significant source of secondary organic aerosol (SOA). Several recent laboratory studies have been performed with the goal of characterizing this process, but questions remain regarding the effects of photochemistry on SOA growth. We applied GAMMA (McNeill et al. Environ. Sci. Technol. 2012, 46, 8075−8081), a photochemical box model with coupled gas-phase and detailed aqueous aerosol-phase chemistry, to simulate aerosol chamber studies of SOA formation by the uptake of glyoxal by wet aerosol under dark and irradiated conditions (Kroll et al. J. Geophys. Res. 2005, 110 (D23), 1−10; Volkamer et al. Atmos. Chem. Phys. 2009, 9, 1907−1928; Galloway et al. Atmos. Chem. Phys. 2009, 9, 3331− 306 3345 and Geophys. Res. Lett. 2011, 38, L17811). We find close agreement between simulated SOA growth and the results of experiments conducted under dark conditions using values of the effective Henry’s Law constant of 1.3−5.5 × 107 M atm−1. While irradiated conditions led to the production of some organic acids, organosulfates, and other oxidation products via well-established photochemical mechanisms, these additional product species contribute negligible aerosol mass compared to the dark uptake of glyoxal. Simulated results for irradiated experiments therefore fell short of the reported SOA mass yield by up to 92%. This suggests a significant light-dependent SOA formation mechanism that is not currently accounted for by known bulk photochemistry, consistent with recent laboratory observations of SOA production via photosensitizer chemistry.
■
INTRODUCTION Secondary organic aerosol (SOA) represents a significant fraction of tropospheric particulate matter, though its formation is not yet fully understood.1 Glyoxal, a gas-phase oxidation product of biogenic and anthropogenic volatile organic compounds (VOCs), has been suggested as a direct contributor to SOA formation.2 Despite its high vapor pressure, glyoxal readily partitions to aerosol water. Glyoxal is highly watersoluble, and once in the aqueous phase, it spontaneously undergoes two successive hydrations and may self-oligomerize or form dimers.3 Due to these properties, the uptake of glyoxal to aerosols can lead to SOA formation, even in the absence of the in-particle photochemistry which drives the formation of SOA by glyoxal in cloudwater.4 Kroll and co-workers observed very high uptake of glyoxal onto ammonium sulfate seed under dark conditions in the Caltech dual 28 m3 chamber.5 They showed for the first time that glyoxal uptake by deliquesced ammonium sulfate aerosols at constant relative humidity and inorganic aerosol composition depends linearly on gas phase glyoxal concentration. In other words, in that study, dark uptake of glyoxal to aqueous sulfate aerosols could be described accurately using Henry’s Law. Galloway and others later revisited glyoxal uptake onto ammonium sulfate seed conditions in the Caltech chamber.6 Along with demonstrating reversible SOA formation under dark conditions, they reported the irreversible formation of a small amount (∼0.5 wt %) of nitrogen-containing organic species, identified as imidazole carboxaldehyde and related compounds. The effective Henry’s Law constant, H*, inferred from laboratory studies such as those described above, is often © 2014 American Chemical Society
applied in simulations of glyoxal uptake to aerosols to describe the reversible component of uptake. H* describes enhanced uptake from the gas phase (beyond what would be predicted using the physical Henry’s Law constant) due to reversible aqueous-phase reactions7 and depends on a number of factors including aerosol composition, relative humidity, and pH. Reported experimental values for H* for glyoxal range from ∼105 M atm−1 for bulk solutions8−10 to 108 M atm−1 or higher in aerosol chamber studies.5,6,11,12 Ip et al. showed via a laboratory study of H* of glyoxal in bulk aqueous solutions that sulfate and, to a lesser degree, chloride, have a salting-in effect.10 Yu and co-workers showed that the sulfate ion induces a shift in the hydration equilibrium of glyoxal toward the hydrated gem-diol form, consistent with the observations of Ip et al.13 Kampf et al. systematically investigated glyoxal uptake to aerosols of varying ionic strength and proposed a modified Setschenow relationship between H* and the concentration of ammonium sulfate in the aerosol.14 Laboratory chamber studies have produced conflicting evidence regarding the effect of photochemistry on SOA growth via glyoxal uptake to seed aerosols. Volkamer and coworkers studied the uptake of glyoxal generated via acetylene oxidation to aerosols of varying compositions, including some containing fulvic acid and amino acids.11 Their experiments were performed in a 7 m3 indoor smog chamber at UC Riverside. They found that aerosol growth increased greatly Received: Revised: Accepted: Published: 11919
April 24, 2014 September 10, 2014 September 16, 2014 September 16, 2014 dx.doi.org/10.1021/es502020j | Environ. Sci. Technol. 2014, 48, 11919−11925
Environmental Science & Technology
Article
aerosol-phase chemistry.20 Allowing for differences in H* due to varying seed aerosol conditions, we find it necessary to invoke newly proposed mechanisms for SOA formation involving photosensitizer chemistry16,17,21 in order to achieve closure with the irradiated experiments, particularly when fulvic acid is present in the seed aerosol.
under irradiated conditions, especially when fulvic acid was present in the seed aerosol. They attributed this enhancement to organic photochemistry driven by the OH radical. Quantitatively, the rate of SOA formation under irradiated conditions was found to be a factor of ∼500 times higher compared to dark conditions, a fast enough rate to compete with gas-phase glyoxal photolysis as a sink for glyoxal. That group recently performed a followup study of photochemically enhanced SOA formation by glyoxal in ammonium sulfate/ fulvic acid aerosol, in a different reaction chamber. They again saw rapid growth during periods of irradiation. These quick increases in SOA growth coincided directly with UV exposure and appeared significantly stronger in experiments with ammonium sulfate/fulvic acid seed than experiments with only ammonium acid seed.14 In contrast to the observations of Volkamer and co-workers, in the 2009 study of Galloway et al., overall organic growth was reduced under irradiated conditions as compared to dark conditions, although an OH source was not present.6 Glyoxal oxidation products as well as glycolic acid sulfate were observed, indicative of active radical chemistry; however, increased temperature and gradual particle dehydration during the irradiated experiments may have caused glyoxal to repartition to the gas phase. Glyoxal photolysis in the gas phase is also proposed as a reason for less growth, but reemission of glyoxal species adsorbed on the chamber walls likely compensated at least partially for this loss. Following up on these experiments, they later studied glyoxal SOA formation in the same chamber under photochemical conditions in the presence of OH as well as dark conditions.12 They found that the chemical composition, as measured by an Aerodyne Aerosol Mass Spectrometer (AMS), of glyoxal SOA formed on ammonium sulfate seed aerosol during photochemical experiments with OH present was identical to that formed under dark conditions. They suggested that the identity of the seed aerosol played an important, although poorly understood, role in determining photochemical SOA formation from glyoxal. Recently, experimental evidence has emerged that may shed some light on the apparent discrepancies among these chamber observations. Several studies indicate that light-absorbing organic species, when photochemically excited, may contribute to the formation of SOA.15−18 The photoexcited molecule may either directly oxidize volatile organic material via electron transfer, or it may generate oxidizing radicals in situ.19 Aregahegn and co-workers found evidence that imidazole carboxaldehyde (IC), a known product of dark reactions between glyoxal and ammonia in the aerosol phase,6 can oxidize VOCs such as limonene, isoprene, and toluene under UV irradiation.17 The ability of a VOC to be oxidized by a particular photosensitizer depends on the relative energy levels of the molecular orbitals of the VOC and the photosensitizer. The control experiments of Aregahegn et al.17 suggest that IC may not be capable of oxidizing glyoxalthat is, in a controlled chamber experiment where glyoxal is the only VOC present, the IC-glyoxal reaction would not lead to autocatalytic glyoxal SOA formation. Other conjugated molecules such as fulvic acid, present in the seed aerosols of some of the experiments performed by Volkamer et al.11 and Kampf et al.,14 could also act as photosensitizers. In this study, we analyze the experimental results of Kroll et al.,5 Volkamer et al.,11 and Galloway et al.6,12 using GAMMA (Gas Aerosol Model for Mechanism Analysis), a photochemical box model with coupled gas-phase and detailed aqueous
■
METHODS Details of the model GAMMA and its application to modeling aqueous SOA formation chemistry in ammonium sulfate aerosols can be found in the work of McNeill et al.20 In brief, the aqueous phase processing of glyoxal includes glyoxal self-oligomerization and hydration, glyoxal oxidation reactions with OH to eventually form organic acids (succinic, glycolic, glyoxylic, formic, oxalic) and photochemical formation of organosulfates.4,20,22,23 Henry’s Law equilibration is allowed for oxalic, formic, glycolic, glyoxylic, pyruvic, and acetic acids. Succinic and malonic acids have exceedingly high Henry’s Law constants (>109 M atm−1)24 and therefore are assumed to remain in the particle phase. Ionized acids (e.g., oxalate) are assumed to remain in the particle phase. The sum of all organic acids and their ionized forms are reported as “organic acids” in this study. Relevant initial concentrations and experimental parameters (T, initial glyoxal concentrations, seed concentration and size, RH, hv, OH concentration) were extracted from the literature5,6,11,12 and organized for input to GAMMA. Aerosol liquid water content and the activity coefficient of H3O+ were calculated using E-AIM.25 All of the experiments analyzed were performed with ammonium sulfate (AS) seed except for a subset of the Volkamer et al.11 experiments (11, 17, 20, and 26), which were performed using seed particles composed of 50% AS and 50% fulvic acid. For the purposes of those simulations, fulvic acid was specified in E-AIM, following the work of Kampf et al.14, assuming a molecular weight of 150 g mol−1 and a molar volume of 4.39 cm3 mol−1. In order to account for the variability in H* across experiments with different experimental conditions (T, RH, initial glyoxal concentration, seed aerosol), we aimed to use the results of dark experiments at a given set of conditions to determine H*, then applied that value to model irradiated experiments under similar conditions. Experiment 3 from Galloway et al.6 and experiment 4 from Galloway et al.12 were coupled in this way (the average H* value was used for the remainder of the Galloway et al.12 experiments). Similarly, dark experiments from Kroll et al.5 and Volkamer et al.11 were grouped to calibrate effective Henry’s Law constants and then coupled with irradiated experiments in the work of Volkamer et al.11 The H* values obtained in this manner were consistent with the recommendations of Kampf et al.14 More details are provided in the Results and Discussion section. Information regarding the absolute actinic flux of the aerosol chamber light sources was not available for the experiments analyzed here. In order to investigate the effects of gas-phase glyoxal photolysis further, chamber simulations were run with varying photolysis strength factors. Photolysis strength is scaled to the Caltech dual chamber running with 50% of the UV lights. This condition was calibrated by Paulot et al. to correspond to a gas-phase photolysis rate constant for HOOH of 2.15 × 10−6 s−1.26 In other words, the Caltech chamber running with 50% lights corresponds to Photolysis Factor = 1. Experiments 4 through 7 of Galloway et al.12 were performed with 10% lights in that chamber, corresponding to a Photolysis 11920
dx.doi.org/10.1021/es502020j | Environ. Sci. Technol. 2014, 48, 11919−11925
Environmental Science & Technology
Article
Table 1. Comparison of Experimental Results and GAMMA Output for Dark Experimentsa average H* (2.7 × 107 M atm−1)
varying H*
experimental experiment
[Glyx] (ppb)
RH (%)
T (K)
ΔV (μm3 cm−3)
H* (107 M atm−1)
ΔV (μm3 cm−3)
% error
ΔV (μm3/cm3)
% error
Kroll2005_1 Kroll2005_2 Kroll2005_3 Kroll2005_4 Kroll2005_5 Volkamer2009_15 Volkamer2009_16 Volkamer2009_32 Volkamer2009_34 Galloway2009_2 Galloway2009_3
55 153 126 110 158 189 167 100 173 67 182
47.5 54.5 53.5 44.5 54.5 60 49 50 52 56 56
293 293 293 293 293 298 298 298 298 293 293
3.43 11.12 8.58 4.5 12.6 20.9 42 11.00 11.6 10.7 40.4
2.0 1.6 1.6 1.3 1.7 3.2 3.0 4.4 2.3 5.0 5.5
3.38 11.29 8.65 4.5 12.3 20.4 42 11.04 11.3 10.7 39.7
−1.55 1.45 0.82 0.91 −2.27 2.61 −0.18 0.40 −2.73 0.404 −1.76
4.49 17.4 14.1 9.14 19.0 17.4 38.4 6.98 13.1 6.12 21.1
30.9 56.4 64.7 103 50.4 −16.8 −8.6 −36.5 13.3 −42.8 −47.7
a
Experiment numbers correspond to experiment numbers reported in the work of Kroll et al.,5 Volkamer et al.,11 and Galloway et al.6 Results are reported for GAMMA simulations where H* was varied to match experimental data and when an average value of H* was used.
Factor of 0.2. Note that, since we do not directly simulate OH formation via photolysis of CH3ONO in the gas phase but rather use the authors’ reported gas-phase OH concentrations in our simulations, the photolysis factor primarily impacts the glyoxal photolysis rate.
■
RESULTS AND DISCUSSION Table 1 displays our simulation results for dark experiments and compares them to the experimental data. The effective Henry’s law constant for glyoxal was varied in these simulations in order to achieve agreement with the experimental data to within 3%. The inferred experimental values of H* vary from 1.3 to 5.5 × 107 M atm−1. These values are consistent with the compilation of Kampf et al.,14 which included much of the same experimental data. We note that the ionic content of the aerosols in these studies places them in what those authors referred to as the “kinetically limited” region of the Setschenow plot, suggesting that there should not be significant variation of H* with changes in ionic strength under these conditions (and making it inappropriate to apply their parametrization of H* for this study). However, the simulation results are sensitive to these variations in H*: when an average value of 2.7 × 107 M atm−1 is applied to simulate all of the experiments, errors of up to 103% are introduced (Table 1). This is demonstrated in Figure 1, which shows predicted vs measured volume growth when H* is allowed to vary (blue boxes) and when the average value is used (red dots). The errors are nearly random in sign, with a small bias toward overestimation of particle growth: the average error is 15%, but average absolute error is 36%. Predicted SOA growth fell significantly short of the experimental values for most of the irradiated experiments regardless of the presence of OH precursor, and for all of the experiments for which fulvic acid was present in the seed aerosol (see Table 2 and Figure 2). Using GAMMA as described in the work of McNeill et al.,20 with no photosensitizer chemistry, we predict that SOA growth is dominated by Henry’s Law uptake of glyoxal from the gas phase to the aqueous aerosol phase even under irradiated conditions; that is, significant photochemical SOA formation such as what was observed experimentally, was not predicted. To illustrate this, we can discuss as an example a chamber experiment (Galloway et al.12, experiment 4) performed under irradiated conditions with no source of hydroxyl radicals. Using the value of H* inferred from dark reactions under the nearest corresponding
Figure 1. Volume growth predicted by GAMMA vs measured volume growth for the dark glyoxal uptake experiments listed in Table 1. Results obtained by varying H* to find the best agreement with each experimental data set are indicated by the box symbols, and predictions using the average value of H* obtained from that exercise (2.7 × 107 M atm−1) are shown as red dots.
conditions (5.5 × 107 M atm−1, Galloway et al.6, experiment 3), GAMMA underestimates the experimentally observed SOA formation by 68%. The predicted concentrations of organic species in the aqueous aerosol phase for experiment 4 of Galloway et al.12 are shown in Figure 3. Organic acids (such as succinic and oxalic acids) formaldehyde and trace organosulfates are present in the aqueous phase under irradiated conditions, but their concentrations are several orders of magnitude less than glyoxal and its hydrates and oligomers (shown together on the figure as “Glyoxal (aq)”). In contrast to cloudwater conditions such as those studied by Lim et al.,4 in the absence of additional in-particle sources of OH, the high organic concentrations in aerosol lead to an OH-depleted chemistry regime, decreasing the relative importance of OH photoxidation for SOA production in aerosol water. This is consistent with the results of McNeill et al.20 and the recent simulations of Ervens et al., who also predicted OH-limited conditions during glyoxal SOA formation in aqueous aerosol.27 We predict similar results for the chamber studies performed by Volkamer et al.11 For four of the experiments from that study, volumetric aerosol growth was predicted as a function of the photolysis strength (Figure 4). Photolysis intensity was varied 11921
dx.doi.org/10.1021/es502020j | Environ. Sci. Technol. 2014, 48, 11919−11925
Environmental Science & Technology
Article
Table 2. Comparison of Experimental and Modeled Results for Irradiated Experiments, Using Gamma without the Addition of FA Photocatalysisa EXPERIMENTAL
CALCULATED
Experiment
[Glyx] ppb
RH (%)
T (K)
ΔV (μm3/cm3)
H* (107 M atm‑1)
ΔV (μm3/cm3)
% Error
Galloway2011_4 Galloway2011_5 Galloway2011_6 Galloway2011_7 Volkamer2009_7 Volkamer2009_9 Volkamer2009_13 Volkamer2009_20* Volkamer2009_11* Volkamer2009_17* Volkamer2009_26*
131 38 59 253 143 117 160 70 100 162 121
60 76 76 67 88 49 46 86 54 49 67
293 292 291 292 298 298 298 298 298 298 298
33 5 5 96 10.6 1.5 3.1 11.4 5.5 10 22.8
5.5 2.7 2.7 2.7 4.0 4.4 2.1 4.0 4.4 2.1 3.5
10.6 1.32 1.95 7.33 10.7 1.53 1.73 3.2 2.18 1.99 7.13
−68 −74 −61 −92 1.19 2.0 −44 −72.0 −60 −80 −68.7
a
Experiment numbers correspond to experiment numbers reported in the work of Volkamer et al.11 and Galloway et al.12 Experiments in which the seed aerosol contained fulvic acid are marked with an asterisk (*). Note that H* was assigned for each simulation based on dark experiments at the nearest corresponding conditions (see text for details).
Figure 4. Aerosol growth vs photolysis scaling factor for selected experiments from the work of Volkamer et al.11 See the text for more details.
Figure 2. Volume growth predicted by GAMMA vs measured volume growth for the irradiated experiments listed in Table 2. The black line indicating 1:1 agreement is shown as a guide to the eye.
consistent with the observations of Galloway et al.6 Photolysis destroys gas-phase glyoxal, generating CO, HO2, and HCHO, and leading to the off-gassing of glyoxal from the aqueous aerosol phase following Henry’s Law. This effect increases with increasing light intensity. Besides reducing SOA formation due to Henry’s Law partitioning, glyoxal photolysis also results in the generation of a small amount of OH, but the products of the resulting aqueous-phase glyoxal oxidation do not significantly contribute to SOA mass. Note that, since calibrated photolysis intensity was not available for the Volkamer et al.11 experiments, the values reported in Table 2 are the upper bound predicted volume growth (obtained with a photolysis factor of 0.1−0.25). Galloway et al.12 suggested that their chamber walls served as a glyoxal reservoir and therefore gas-phase depletion of glyoxal due to uptake or photolysis was not likely. Assuming a fixed gas-phase glyoxal concentration in GAMMA, but allowing for OH production via photochemistry in the chamber, we still arrive at a 64% underprediction in ΔV for their experiment 4 (as compared to 68% when gas-phase glyoxal was not held fixed). Based on our analysis as described above, we can eliminate variations in H*, differences in photolysis intensity, and a glyoxal reservoir on chamber walls as potential explanations for discrepancies in growth phenomena observed across laboratory
Figure 3. Aqueous aerosol composition predicted by GAMMA experiment 4 of the work of Galloway et al.12
in order to evaluate its potential impact on the experimental observations. We find that, for each experiment simulated, predicted SOA formation decreased with increasing photolysis intensity in the model due to gas-phase glyoxal photolysis, 11922
dx.doi.org/10.1021/es502020j | Environ. Sci. Technol. 2014, 48, 11919−11925
Environmental Science & Technology
Article
significant formation of the organic acids, malonic acid and succinic acid, by reacting with glyoxal, which in turn led to further Henry’s Law uptake of glyoxal. We found that, assuming that reaction 1 takes place in the aerosol bulk, a photochemical rate constant of ∼10−4 s−1 was sufficient to match the growth observed by Volkamer et al. This corresponds to an OH production rate of ∼10−4 M s−1, similar to the OH production rate calculated by Ervens et al. as being required to overcome OH limitations in box model simulations of SOA formation by glyoxal in aqueous aerosol particles.27 The rate constant that we infer is somewhat higher than the photolysis rate for aqueous hydrogen peroxide in our model (3.77 × 10−6 s−1),28 suggesting that this pathway likely does not account for all the missing photochemical SOA formation. Reaction 2 was also able to account for the observations made by Volkamer et al. with a rate constant ranging from 5 × 10−5 to 10−4 M−1 s−1. In the absence of mechanistic information and measured rate constants, it is difficult to distinguish via modeling whether the proposed photoenhanced chemistry is a surface or bulk process. However, since FA and IC are both surface-active, VOCs are more readily available at the surface, and UV intensity is higher at the surface than in the aerosol interior, it is likely that surface processes are important. This is consistent with assumptions made in modeling studies by Volkamer and co-workers.29,30 Besides the application to this analysis, fulvicand humic-like “brown carbon” substances are common components of atmospheric aerosols31 and the potential impact of their photoenhanced chemistry on ambient SOA formation deserves further study. We note that GAMMA also underestimates particle volume growth in the absence of fulvic acid (e.g., experiment 13 of Volkamer et al.11). Fulvic acid was not used in the experiments of Galloway et al.,12 and they did not observe the dramatic photoenhancement of SOA formation seen by Volkamer et al.;11 however, our detailed analysis reveals a consistent unknown photochemical source of SOA in their experiments (Table 2). While some error is likely attributable to variations in H*, we would expect that error to be randomly distributed or biased positive (see Figure 1) rather than showing the consistent underprediction demonstrated in Table 2. As Aregahegn et al.17 proposed, light-absorbing glyoxal SOA products such as IC could be participating in photoenhanced chemistry. No evidence exists from the experiments of Aregahegn et al. for IC’s ability to oxidize glyoxal either directly or via the oxidation of radicals in situ.19 However, a variety of other light-absorbing, fulvic-like compounds are known to be formed by glyoxal in ammonium sulfate solutions which could potentially act as photosensitizers.32,33 This chemistry, which involves the participation of ammonium, would be enhanced at lower relative humidities (i.e., higher solute concentrations), perhaps explaining the absence of this effect in experiment 7 of the work of Volkamer et al.11 Finally, additional unknown pathways of SOA formation under irradiated conditions may exist besides the photoenhanced chemistry analyzed here−for example, unaccountedfor in-particle sources of OH, such as photolysis of organic peroxides34−38 which would open up the possibility of photochemically driven oligomerization.22,39
studies. The systematic underprediction of SOA growth we see under irradiated conditions reveals a light-dependent formation mechanism that is not accounted for in GAMMA. We therefore explored the potential for photoenhanced chemistry in the aerosol phase. Candidate photosensitizers include fulvic acid, present in the seed aerosols of Volkamer et al.,11 and imidazole carboxaldehyde (IC), one of the products of glyoxal reacting with ammonium sulfate solutions in the dark.6 While Aregahegn and co-workers17 reported photoenhanced SOA production by IC, the rate and quantum yield of this process have not been quantified, and glyoxal oxidation by photoactive IC has not been confirmed. To our knowledge, no direct evidence exists of photoenhanced reactions of fulvic acid with glyoxal, although photoactivated humic acid has been shown to promote SOA formation by limonene and isoprene in aerosols.16 Therefore, in the absence of detailed mechanistic and kinetic information, in order to qualitatively determine the effect of such chemistry, we posited potential photosensitizer reactions and analyzed the outcome they had when incorporated into the existing GAMMA mechanism. These tests were run on the simulations of chamber experiments in which the seed particles had a 50% fulvic acid composition: Volkamer et al.11 experiments 11, 17, 20, and 26. Photoactivation of fulvic acid (FA) was hypothesized to either lead to the direct generation of OH radicals in the aqueous aerosolphase or lead to the irreversible generation of a secondary organic aerosol product by reacting with glyoxal in the presence of light. In both cases, we assumed FA to be conserved during the reaction (i.e., it acted as a “photocatalyst”). FA + hv → OH + FA
(1)
GLYX + FA + hv → SOA + FA
(2)
Including either reaction 1 or 2 in GAMMA led to predictions of enhanced aerosol growth. Simulation results are shown in Figure 5. The generation of OH radicals via reaction 1 led to
Figure 5. Simulations of chamber results from the work of Volkamer et al.11 Experiment 11 using the photochemical box model GAMMA.20 In the baseline simulation, formation of organic acids via aqueousphase photooxidation is small, and SOA mass decreases with reaction time due to photolysis of gas-phase glyoxal leading to reemission of absorbed glyoxal. Including either reaction 1 or 2, both based on fulvic acid acting as a photosensitizer, could theoretically explain the observed efficient SOA growth.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected]. Phone: +1 (212) 854-2869. Fax: +1 (212) 854-3054. 11923
dx.doi.org/10.1021/es502020j | Environ. Sci. Technol. 2014, 48, 11919−11925
Environmental Science & Technology
Article
Notes
(17) Aregahegn, K. Z.; Noziere, B.; George, C. Organic Aerosol Formation Photoenhanced by the Formation of Secondary Photosensitizers in ageing Aerosols. Faraday Discuss. 2013, 165, 123−134. (18) George, C.; D’Anna, B.; Herrmann, H.; Weller, C.; Vaida, V.; Donaldson, D. J.; Bartels-Rausch, T.; Ammann, M. Emerging Areas in Atmospheric Photochemistry. Top. Curr. Chem. 2012, 339, 1−54. (19) Canonica, S.; Jans, U. R. S.; Stemmler, K.; Hoigne, J. Transformation Kinetics of Phenols in Water: Photosensitization. Environ. Sci. Technol. 1995, 29, 1822−1831. (20) McNeill, V. F.; Woo, J. L.; Kim, D. D.; Schwier, A. N.; Wannell, N. J.; Sumner, A. J.; Barakat, J. M. Aqueous-phase secondary organic aerosol and organosulfate formation in atmospheric aerosols: a modeling study. Environ. Sci. Technol. 2012, 46, 8075−8081. (21) Rossignol, S.; Aregahegn, K. Z.; Tinel, L.; Fine, L.; Noziere, B.; George, C. Glyoxal induced atmospheric photosensitized chemistry leading to organic aerosol growth. Environ. Sci. Technol. 2014, 48, 3218−3227. (22) Tan, Y.; Carlton, A. G.; Seitzinger, S. P.; Turpin, B. J. SOA from methylglyoxal in clouds and wet aerosols: Measurement and prediction of key products. Atmos. Environ. 2010, 44, 5218−5226. (23) Perri, M. J.; Lim, Y. B.; Seitzinger, S. P.; Turpin, B. J. Organosulfates from glycolaldehyde in aqueous aerosols and clouds: Laboratory studies. Atmos. Environ. 2010, 44, 2658−2664. (24) Compernolle, S.; Müller, J.-F. Henry’s law constants of diacids and hydroxypolyacids: recommended values. Atmos. Chem. Phys. 2014, 14, 2699−2712. (25) Wexler, A. S.; Clegg, S. L. Atmospheric aerosol models for systems including the ions H+, NH4+, Na+, SO42−, NO3−, Cl−, Br−, and H2O. J. Geophys. Res. 2002, 107 (D14), 4207 DOI: 10.1029/ 2001JD000451. (26) Paulot, F.; Crounse, J. D.; Kjaergaard, H. G.; Kürten, A., St; Clair, J. M.; Seinfeld, J. H.; Wennberg, P. O. Unexpected epoxide formation in the gas-phase photooxidation of isoprene. Science 2009, 325, 730−733. (27) Ervens, B.; Sorooshian, A.; Lim, Y. B.; Turpin, B. J. Key parameters controlling OH-initiated formation of secondary organic aerosol in the aqueous phase (aqSOA). J. Geophys. Res. Atmos. 2014, 119, 3997−4016. (28) Lelieveld, J.; Crutzen, P. J. The role of clouds in tropospheric photochemistry. J. Atmos. Chem. 1991, 12, 229−267. (29) Ervens, B.; Volkamer, R. Glyoxal processing by aerosol multiphase chemistry: towards a kinetic modeling framework of secondary organic aerosol formation in aqueous particles. Atmos. Chem. Phys. 2010, 10, 8219−8244. (30) Waxman, E. M.; Dzepina, K.; Ervens, B.; Lee-taylor, J.; Aumont, B.; Jimenez, J. L.; Madronich, S.; Volkamer, R. Secondary organic aerosol formation from semi- and intermediate-volatility organic compounds and glyoxal: Relevance of O/C as a tracer for aqueous multiphase chemistry. Geophys. Res. Lett. 2013, 40, 1−5. (31) Graber, E. R.; Rudich, Y. Atmospheric HULIS: How humic-like are they? A comprehensive and critical review. Atmos. Chem. Phys. 2006, 6, 729−753. (32) Shapiro, E. L.; Szprengiel, J.; Sareen, N.; Jen, C. N.; Giordano, M. R.; McNeill, V. F. Light-absorbing secondary organic material formed by glyoxal in aqueous aerosol mimics. Atmos. Chem. Phys. 2009, 9, 2289−2300. (33) Schwier, A. N.; Sareen, N.; Mitroo, D.; Shapiro, E. L.; McNeill, V. F. Glyoxal-methylglyoxal cross-reactions in secondary organic aerosol formation. Environ. Sci. Technol. 2010, 44, 6174−6182. (34) Anastasio, C.; McGregor, K. G. Chemistry of fog waters in California’s Central Valley: 1. In situ photoformation of hydroxyl radical and singlet molecular oxygen. Atmos. Environ. 2001, 35, 1079− 1089. (35) Zhao, R.; Lee, A. K. Y.; Abbatt, J. P. D. Investigation of Aqueous-Phase Photooxidation of Glyoxal and Methylglyoxal by Aerosol Chemical Ionization Mass Spectrometry: Observation of Hydroxyhydroperoxide Formation. J. Phys. Chem. A 2012, 116, 6253− 6263.
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS This work was supported by NSF-ATM REU funding. REFERENCES
(1) Hallquist, M.; Wenger, J. C.; Baltensperger, U.; Rudich, Y.; Simpson, D.; Claeys, M.; Dommen, J.; Donahue, N. M.; George, C.; Goldstein, A. H.; et al. The formation, properties and impact of secondary organic aerosol: current and emerging issues. Atmos. Chem. Phys. 2009, 9, 5155−5236. (2) Volkamer, R.; Jimenez, J. L.; San Martini, F.; Dzepina, K.; Zhang, Q.; Salcedo, D.; Molina, L. T.; Worsnop, D. R.; Molina, M. J. Secondary organic aerosol formation from anthropogenic air pollution: Rapid and higher than expected. Geophys. Res. Lett. 2006, 33, L17811. (3) Whipple, E. B. Structure of glyoxal in water. J. Am. Chem. Soc. 1970, 92, 7183−7186. (4) Lim, H.-J.; Carlton, A. G.; Turpin, B. J. Isoprene Forms Secondary Organic Aerosol through Cloud Processing. Environ. Sci. Technol. 2005, 39, 4441−4446. (5) Kroll, J. H.; Ng, N. L.; Murphy, S. M.; Varutbangkul, V.; Flagan, R. C.; Seinfeld, J. H. Chamber studies of secondary organic aerosol growth by reactive uptake of simple carbonyl compounds. J. Geophys. Res. 2005, 110, D23207. (6) Galloway, M. M.; Chhabra, P. S.; Chan, a. W. H.; Surratt, J. D.; Flagan, R. C.; Seinfeld, J. H.; Keutsch, F. N. Glyoxal uptake on ammonium sulphate seed aerosol: reaction products and reversibility of uptake under dark and irradiated conditions. Atmos. Chem. Phys. 2009, 9, 3331−3345. (7) Schwartz, S. E. Mass-transport considerations pertinent to aqueous phase reactions of gases in liquid-water clouds; NATO ASI Series, Jaeschke, W., Ed.; Springer-Verlag: Berlin Heidelberg, 1986; Vol. G6, pp 425−471. (8) Betterton, E. A.; Hoffmann, M. R. Henry’s Law Constants of Some Environmentally Important Aldehydes. Environ. Sci. Technol. 1988, 22, 1415−1418. (9) Zhou, X.; Mopper, K. Apparent Partition Coefficients of 15 Carbonyl Compounds between Air and Seawater and between Air and Freshwater; Implications for Air-Sea Exchange. Environ. Sci. Technol. 1990, 24, 1864−1869. (10) Ip, H. S. S.; Huang, X. H. H.; Yu, J. Z. Effective Henry’s law constants of glyoxal, glyoxylic acid, and glycolic acid. Geophys. Res. Lett. 2009, 36 (L01802), 1−5. (11) Volkamer, R.; Ziemann, P. J.; Molina, M. J. Secondary Organic Aerosol Formation from Acetylene (C2H2): seed effect on SOA yields due to organic photochemistry in the aerosol aqueous phase. Atmos. Chem. Phys. 2009, 9, 1907−1928. (12) Galloway, M. M.; Loza, C. L.; Chhabra, P. S.; Chan, a. W. H.; Yee, L. D.; Seinfeld, J. H.; Keutsch, F. N. Analysis of photochemical and dark glyoxal uptake: Implications for SOA formation. Geophys. Res. Lett. 2011, 38 (L17811), 1−5. (13) Yu, G.; Bayer, A. R.; Galloway, M. M.; Korshavn, K. J.; Fry, C. G.; Keutsch, F. N. Glyoxal in aqueous ammonium sulfate solutions: products, kinetics and hydration effects. Environ. Sci. Technol. 2011, 45, 6336−6342. (14) Kampf, C. J.; Waxman, E. M.; Slowik, J. G.; Dommen, J.; Pfaffenberger, L.; Praplan, A. P.; Prévôt, A. S. H.; Baltensperger, U.; Hoffmann, T.; Volkamer, R. Effective Henry’s Law Partitioning and the Salting Constant of Glyoxal in Aerosols Containing Sulfate. Environ. Sci. Technol. 2013, 47, 4236−4244. (15) Yu, G.; Keutsch, F. N. Atmospheric organic-phase photosensitized epoxidation of alkenes by α-dicarbonyls. Atmos. Chem. Phys. Discuss. 2012, 12, 15115−15138. (16) Monge, M. E.; Rosenørn, T.; Favez, O.; Müller, M.; Adler, G.; Abo Riziq, A.; Rudich, Y.; Herrmann, H.; George, C.; D’Anna, B. Alternative pathway for atmospheric particles growth. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 6840−6844. 11924
dx.doi.org/10.1021/es502020j | Environ. Sci. Technol. 2014, 48, 11919−11925
Environmental Science & Technology
Article
(36) Bateman, A. P.; Nizkorodov, S. A.; Laskin, J.; Laskin, A. Photolytic processing of secondary organic aerosols dissolved in cloud droplets. Phys. Chem. Chem. Phys. 2011, 13, 12199−12212. (37) Epstein, S. A.; Blair, S. L.; Nizkorodov, S. A. Direct Photolysis of α-Pinene Ozonolysis Secondary Organic Aerosol: Effect on Particle Mass and Peroxide Content. Environ. Sci. Technol. 2014, DOI: 10.1021/es502350u. (38) Chen, Z. M.; Wang, H. L.; Zhu, L. H.; Wang, C. X.; Jie, C. Y.; Hua, W. Aqueous-phase ozonolysis of methacrolein and methyl vinyl ketone: a potentially important source of atmospheric aqueous oxidants. Atmos. Chem. Phys. 2008, 8, 2255−2265. (39) Renard, P.; Siekmann, F.; Gandolfo, A.; Socorro, J.; Salque, G.; Ravier, S.; Quivet, E.; Clément, J.-L.; Traikia, M.; Delort, A.-M.; et al. Radical mechanisms of methyl vinyl ketone oligomerization through aqueous phase OH-oxidation: on the paradoxical role of dissolved molecular oxygen. Atmos. Chem. Phys. 2013, 13, 6473−6491.
11925
dx.doi.org/10.1021/es502020j | Environ. Sci. Technol. 2014, 48, 11919−11925
Model Analysis of Secondary Organic Aerosol Formation by Glyoxal in Laboratory Studies: The Case for Photoenhanced Chemistry Andrew J. Sumner, Joseph L. Woo, and V. Faye McNeill* Department of Chemical Engineering, Columbia University, New York, New York 10027, United States ABSTRACT: The reactive uptake of glyoxal by atmospheric aerosols is believed to be a significant source of secondary organic aerosol (SOA). Several recent laboratory studies have been performed with the goal of characterizing this process, but questions remain regarding the effects of photochemistry on SOA growth. We applied GAMMA (McNeill et al. Environ. Sci. Technol. 2012, 46, 8075−8081), a photochemical box model with coupled gas-phase and detailed aqueous aerosol-phase chemistry, to simulate aerosol chamber studies of SOA formation by the uptake of glyoxal by wet aerosol under dark and irradiated conditions (Kroll et al. J. Geophys. Res. 2005, 110 (D23), 1−10; Volkamer et al. Atmos. Chem. Phys. 2009, 9, 1907−1928; Galloway et al. Atmos. Chem. Phys. 2009, 9, 3331− 306 3345 and Geophys. Res. Lett. 2011, 38, L17811). We find close agreement between simulated SOA growth and the results of experiments conducted under dark conditions using values of the effective Henry’s Law constant of 1.3−5.5 × 107 M atm−1. While irradiated conditions led to the production of some organic acids, organosulfates, and other oxidation products via well-established photochemical mechanisms, these additional product species contribute negligible aerosol mass compared to the dark uptake of glyoxal. Simulated results for irradiated experiments therefore fell short of the reported SOA mass yield by up to 92%. This suggests a significant light-dependent SOA formation mechanism that is not currently accounted for by known bulk photochemistry, consistent with recent laboratory observations of SOA production via photosensitizer chemistry.
■
INTRODUCTION Secondary organic aerosol (SOA) represents a significant fraction of tropospheric particulate matter, though its formation is not yet fully understood.1 Glyoxal, a gas-phase oxidation product of biogenic and anthropogenic volatile organic compounds (VOCs), has been suggested as a direct contributor to SOA formation.2 Despite its high vapor pressure, glyoxal readily partitions to aerosol water. Glyoxal is highly watersoluble, and once in the aqueous phase, it spontaneously undergoes two successive hydrations and may self-oligomerize or form dimers.3 Due to these properties, the uptake of glyoxal to aerosols can lead to SOA formation, even in the absence of the in-particle photochemistry which drives the formation of SOA by glyoxal in cloudwater.4 Kroll and co-workers observed very high uptake of glyoxal onto ammonium sulfate seed under dark conditions in the Caltech dual 28 m3 chamber.5 They showed for the first time that glyoxal uptake by deliquesced ammonium sulfate aerosols at constant relative humidity and inorganic aerosol composition depends linearly on gas phase glyoxal concentration. In other words, in that study, dark uptake of glyoxal to aqueous sulfate aerosols could be described accurately using Henry’s Law. Galloway and others later revisited glyoxal uptake onto ammonium sulfate seed conditions in the Caltech chamber.6 Along with demonstrating reversible SOA formation under dark conditions, they reported the irreversible formation of a small amount (∼0.5 wt %) of nitrogen-containing organic species, identified as imidazole carboxaldehyde and related compounds. The effective Henry’s Law constant, H*, inferred from laboratory studies such as those described above, is often © 2014 American Chemical Society
applied in simulations of glyoxal uptake to aerosols to describe the reversible component of uptake. H* describes enhanced uptake from the gas phase (beyond what would be predicted using the physical Henry’s Law constant) due to reversible aqueous-phase reactions7 and depends on a number of factors including aerosol composition, relative humidity, and pH. Reported experimental values for H* for glyoxal range from ∼105 M atm−1 for bulk solutions8−10 to 108 M atm−1 or higher in aerosol chamber studies.5,6,11,12 Ip et al. showed via a laboratory study of H* of glyoxal in bulk aqueous solutions that sulfate and, to a lesser degree, chloride, have a salting-in effect.10 Yu and co-workers showed that the sulfate ion induces a shift in the hydration equilibrium of glyoxal toward the hydrated gem-diol form, consistent with the observations of Ip et al.13 Kampf et al. systematically investigated glyoxal uptake to aerosols of varying ionic strength and proposed a modified Setschenow relationship between H* and the concentration of ammonium sulfate in the aerosol.14 Laboratory chamber studies have produced conflicting evidence regarding the effect of photochemistry on SOA growth via glyoxal uptake to seed aerosols. Volkamer and coworkers studied the uptake of glyoxal generated via acetylene oxidation to aerosols of varying compositions, including some containing fulvic acid and amino acids.11 Their experiments were performed in a 7 m3 indoor smog chamber at UC Riverside. They found that aerosol growth increased greatly Received: Revised: Accepted: Published: 11919
April 24, 2014 September 10, 2014 September 16, 2014 September 16, 2014 dx.doi.org/10.1021/es502020j | Environ. Sci. Technol. 2014, 48, 11919−11925
Environmental Science & Technology
Article
aerosol-phase chemistry.20 Allowing for differences in H* due to varying seed aerosol conditions, we find it necessary to invoke newly proposed mechanisms for SOA formation involving photosensitizer chemistry16,17,21 in order to achieve closure with the irradiated experiments, particularly when fulvic acid is present in the seed aerosol.
under irradiated conditions, especially when fulvic acid was present in the seed aerosol. They attributed this enhancement to organic photochemistry driven by the OH radical. Quantitatively, the rate of SOA formation under irradiated conditions was found to be a factor of ∼500 times higher compared to dark conditions, a fast enough rate to compete with gas-phase glyoxal photolysis as a sink for glyoxal. That group recently performed a followup study of photochemically enhanced SOA formation by glyoxal in ammonium sulfate/ fulvic acid aerosol, in a different reaction chamber. They again saw rapid growth during periods of irradiation. These quick increases in SOA growth coincided directly with UV exposure and appeared significantly stronger in experiments with ammonium sulfate/fulvic acid seed than experiments with only ammonium acid seed.14 In contrast to the observations of Volkamer and co-workers, in the 2009 study of Galloway et al., overall organic growth was reduced under irradiated conditions as compared to dark conditions, although an OH source was not present.6 Glyoxal oxidation products as well as glycolic acid sulfate were observed, indicative of active radical chemistry; however, increased temperature and gradual particle dehydration during the irradiated experiments may have caused glyoxal to repartition to the gas phase. Glyoxal photolysis in the gas phase is also proposed as a reason for less growth, but reemission of glyoxal species adsorbed on the chamber walls likely compensated at least partially for this loss. Following up on these experiments, they later studied glyoxal SOA formation in the same chamber under photochemical conditions in the presence of OH as well as dark conditions.12 They found that the chemical composition, as measured by an Aerodyne Aerosol Mass Spectrometer (AMS), of glyoxal SOA formed on ammonium sulfate seed aerosol during photochemical experiments with OH present was identical to that formed under dark conditions. They suggested that the identity of the seed aerosol played an important, although poorly understood, role in determining photochemical SOA formation from glyoxal. Recently, experimental evidence has emerged that may shed some light on the apparent discrepancies among these chamber observations. Several studies indicate that light-absorbing organic species, when photochemically excited, may contribute to the formation of SOA.15−18 The photoexcited molecule may either directly oxidize volatile organic material via electron transfer, or it may generate oxidizing radicals in situ.19 Aregahegn and co-workers found evidence that imidazole carboxaldehyde (IC), a known product of dark reactions between glyoxal and ammonia in the aerosol phase,6 can oxidize VOCs such as limonene, isoprene, and toluene under UV irradiation.17 The ability of a VOC to be oxidized by a particular photosensitizer depends on the relative energy levels of the molecular orbitals of the VOC and the photosensitizer. The control experiments of Aregahegn et al.17 suggest that IC may not be capable of oxidizing glyoxalthat is, in a controlled chamber experiment where glyoxal is the only VOC present, the IC-glyoxal reaction would not lead to autocatalytic glyoxal SOA formation. Other conjugated molecules such as fulvic acid, present in the seed aerosols of some of the experiments performed by Volkamer et al.11 and Kampf et al.,14 could also act as photosensitizers. In this study, we analyze the experimental results of Kroll et al.,5 Volkamer et al.,11 and Galloway et al.6,12 using GAMMA (Gas Aerosol Model for Mechanism Analysis), a photochemical box model with coupled gas-phase and detailed aqueous
■
METHODS Details of the model GAMMA and its application to modeling aqueous SOA formation chemistry in ammonium sulfate aerosols can be found in the work of McNeill et al.20 In brief, the aqueous phase processing of glyoxal includes glyoxal self-oligomerization and hydration, glyoxal oxidation reactions with OH to eventually form organic acids (succinic, glycolic, glyoxylic, formic, oxalic) and photochemical formation of organosulfates.4,20,22,23 Henry’s Law equilibration is allowed for oxalic, formic, glycolic, glyoxylic, pyruvic, and acetic acids. Succinic and malonic acids have exceedingly high Henry’s Law constants (>109 M atm−1)24 and therefore are assumed to remain in the particle phase. Ionized acids (e.g., oxalate) are assumed to remain in the particle phase. The sum of all organic acids and their ionized forms are reported as “organic acids” in this study. Relevant initial concentrations and experimental parameters (T, initial glyoxal concentrations, seed concentration and size, RH, hv, OH concentration) were extracted from the literature5,6,11,12 and organized for input to GAMMA. Aerosol liquid water content and the activity coefficient of H3O+ were calculated using E-AIM.25 All of the experiments analyzed were performed with ammonium sulfate (AS) seed except for a subset of the Volkamer et al.11 experiments (11, 17, 20, and 26), which were performed using seed particles composed of 50% AS and 50% fulvic acid. For the purposes of those simulations, fulvic acid was specified in E-AIM, following the work of Kampf et al.14, assuming a molecular weight of 150 g mol−1 and a molar volume of 4.39 cm3 mol−1. In order to account for the variability in H* across experiments with different experimental conditions (T, RH, initial glyoxal concentration, seed aerosol), we aimed to use the results of dark experiments at a given set of conditions to determine H*, then applied that value to model irradiated experiments under similar conditions. Experiment 3 from Galloway et al.6 and experiment 4 from Galloway et al.12 were coupled in this way (the average H* value was used for the remainder of the Galloway et al.12 experiments). Similarly, dark experiments from Kroll et al.5 and Volkamer et al.11 were grouped to calibrate effective Henry’s Law constants and then coupled with irradiated experiments in the work of Volkamer et al.11 The H* values obtained in this manner were consistent with the recommendations of Kampf et al.14 More details are provided in the Results and Discussion section. Information regarding the absolute actinic flux of the aerosol chamber light sources was not available for the experiments analyzed here. In order to investigate the effects of gas-phase glyoxal photolysis further, chamber simulations were run with varying photolysis strength factors. Photolysis strength is scaled to the Caltech dual chamber running with 50% of the UV lights. This condition was calibrated by Paulot et al. to correspond to a gas-phase photolysis rate constant for HOOH of 2.15 × 10−6 s−1.26 In other words, the Caltech chamber running with 50% lights corresponds to Photolysis Factor = 1. Experiments 4 through 7 of Galloway et al.12 were performed with 10% lights in that chamber, corresponding to a Photolysis 11920
dx.doi.org/10.1021/es502020j | Environ. Sci. Technol. 2014, 48, 11919−11925
Environmental Science & Technology
Article
Table 1. Comparison of Experimental Results and GAMMA Output for Dark Experimentsa average H* (2.7 × 107 M atm−1)
varying H*
experimental experiment
[Glyx] (ppb)
RH (%)
T (K)
ΔV (μm3 cm−3)
H* (107 M atm−1)
ΔV (μm3 cm−3)
% error
ΔV (μm3/cm3)
% error
Kroll2005_1 Kroll2005_2 Kroll2005_3 Kroll2005_4 Kroll2005_5 Volkamer2009_15 Volkamer2009_16 Volkamer2009_32 Volkamer2009_34 Galloway2009_2 Galloway2009_3
55 153 126 110 158 189 167 100 173 67 182
47.5 54.5 53.5 44.5 54.5 60 49 50 52 56 56
293 293 293 293 293 298 298 298 298 293 293
3.43 11.12 8.58 4.5 12.6 20.9 42 11.00 11.6 10.7 40.4
2.0 1.6 1.6 1.3 1.7 3.2 3.0 4.4 2.3 5.0 5.5
3.38 11.29 8.65 4.5 12.3 20.4 42 11.04 11.3 10.7 39.7
−1.55 1.45 0.82 0.91 −2.27 2.61 −0.18 0.40 −2.73 0.404 −1.76
4.49 17.4 14.1 9.14 19.0 17.4 38.4 6.98 13.1 6.12 21.1
30.9 56.4 64.7 103 50.4 −16.8 −8.6 −36.5 13.3 −42.8 −47.7
a
Experiment numbers correspond to experiment numbers reported in the work of Kroll et al.,5 Volkamer et al.,11 and Galloway et al.6 Results are reported for GAMMA simulations where H* was varied to match experimental data and when an average value of H* was used.
Factor of 0.2. Note that, since we do not directly simulate OH formation via photolysis of CH3ONO in the gas phase but rather use the authors’ reported gas-phase OH concentrations in our simulations, the photolysis factor primarily impacts the glyoxal photolysis rate.
■
RESULTS AND DISCUSSION Table 1 displays our simulation results for dark experiments and compares them to the experimental data. The effective Henry’s law constant for glyoxal was varied in these simulations in order to achieve agreement with the experimental data to within 3%. The inferred experimental values of H* vary from 1.3 to 5.5 × 107 M atm−1. These values are consistent with the compilation of Kampf et al.,14 which included much of the same experimental data. We note that the ionic content of the aerosols in these studies places them in what those authors referred to as the “kinetically limited” region of the Setschenow plot, suggesting that there should not be significant variation of H* with changes in ionic strength under these conditions (and making it inappropriate to apply their parametrization of H* for this study). However, the simulation results are sensitive to these variations in H*: when an average value of 2.7 × 107 M atm−1 is applied to simulate all of the experiments, errors of up to 103% are introduced (Table 1). This is demonstrated in Figure 1, which shows predicted vs measured volume growth when H* is allowed to vary (blue boxes) and when the average value is used (red dots). The errors are nearly random in sign, with a small bias toward overestimation of particle growth: the average error is 15%, but average absolute error is 36%. Predicted SOA growth fell significantly short of the experimental values for most of the irradiated experiments regardless of the presence of OH precursor, and for all of the experiments for which fulvic acid was present in the seed aerosol (see Table 2 and Figure 2). Using GAMMA as described in the work of McNeill et al.,20 with no photosensitizer chemistry, we predict that SOA growth is dominated by Henry’s Law uptake of glyoxal from the gas phase to the aqueous aerosol phase even under irradiated conditions; that is, significant photochemical SOA formation such as what was observed experimentally, was not predicted. To illustrate this, we can discuss as an example a chamber experiment (Galloway et al.12, experiment 4) performed under irradiated conditions with no source of hydroxyl radicals. Using the value of H* inferred from dark reactions under the nearest corresponding
Figure 1. Volume growth predicted by GAMMA vs measured volume growth for the dark glyoxal uptake experiments listed in Table 1. Results obtained by varying H* to find the best agreement with each experimental data set are indicated by the box symbols, and predictions using the average value of H* obtained from that exercise (2.7 × 107 M atm−1) are shown as red dots.
conditions (5.5 × 107 M atm−1, Galloway et al.6, experiment 3), GAMMA underestimates the experimentally observed SOA formation by 68%. The predicted concentrations of organic species in the aqueous aerosol phase for experiment 4 of Galloway et al.12 are shown in Figure 3. Organic acids (such as succinic and oxalic acids) formaldehyde and trace organosulfates are present in the aqueous phase under irradiated conditions, but their concentrations are several orders of magnitude less than glyoxal and its hydrates and oligomers (shown together on the figure as “Glyoxal (aq)”). In contrast to cloudwater conditions such as those studied by Lim et al.,4 in the absence of additional in-particle sources of OH, the high organic concentrations in aerosol lead to an OH-depleted chemistry regime, decreasing the relative importance of OH photoxidation for SOA production in aerosol water. This is consistent with the results of McNeill et al.20 and the recent simulations of Ervens et al., who also predicted OH-limited conditions during glyoxal SOA formation in aqueous aerosol.27 We predict similar results for the chamber studies performed by Volkamer et al.11 For four of the experiments from that study, volumetric aerosol growth was predicted as a function of the photolysis strength (Figure 4). Photolysis intensity was varied 11921
dx.doi.org/10.1021/es502020j | Environ. Sci. Technol. 2014, 48, 11919−11925
Environmental Science & Technology
Article
Table 2. Comparison of Experimental and Modeled Results for Irradiated Experiments, Using Gamma without the Addition of FA Photocatalysisa EXPERIMENTAL
CALCULATED
Experiment
[Glyx] ppb
RH (%)
T (K)
ΔV (μm3/cm3)
H* (107 M atm‑1)
ΔV (μm3/cm3)
% Error
Galloway2011_4 Galloway2011_5 Galloway2011_6 Galloway2011_7 Volkamer2009_7 Volkamer2009_9 Volkamer2009_13 Volkamer2009_20* Volkamer2009_11* Volkamer2009_17* Volkamer2009_26*
131 38 59 253 143 117 160 70 100 162 121
60 76 76 67 88 49 46 86 54 49 67
293 292 291 292 298 298 298 298 298 298 298
33 5 5 96 10.6 1.5 3.1 11.4 5.5 10 22.8
5.5 2.7 2.7 2.7 4.0 4.4 2.1 4.0 4.4 2.1 3.5
10.6 1.32 1.95 7.33 10.7 1.53 1.73 3.2 2.18 1.99 7.13
−68 −74 −61 −92 1.19 2.0 −44 −72.0 −60 −80 −68.7
a
Experiment numbers correspond to experiment numbers reported in the work of Volkamer et al.11 and Galloway et al.12 Experiments in which the seed aerosol contained fulvic acid are marked with an asterisk (*). Note that H* was assigned for each simulation based on dark experiments at the nearest corresponding conditions (see text for details).
Figure 4. Aerosol growth vs photolysis scaling factor for selected experiments from the work of Volkamer et al.11 See the text for more details.
Figure 2. Volume growth predicted by GAMMA vs measured volume growth for the irradiated experiments listed in Table 2. The black line indicating 1:1 agreement is shown as a guide to the eye.
consistent with the observations of Galloway et al.6 Photolysis destroys gas-phase glyoxal, generating CO, HO2, and HCHO, and leading to the off-gassing of glyoxal from the aqueous aerosol phase following Henry’s Law. This effect increases with increasing light intensity. Besides reducing SOA formation due to Henry’s Law partitioning, glyoxal photolysis also results in the generation of a small amount of OH, but the products of the resulting aqueous-phase glyoxal oxidation do not significantly contribute to SOA mass. Note that, since calibrated photolysis intensity was not available for the Volkamer et al.11 experiments, the values reported in Table 2 are the upper bound predicted volume growth (obtained with a photolysis factor of 0.1−0.25). Galloway et al.12 suggested that their chamber walls served as a glyoxal reservoir and therefore gas-phase depletion of glyoxal due to uptake or photolysis was not likely. Assuming a fixed gas-phase glyoxal concentration in GAMMA, but allowing for OH production via photochemistry in the chamber, we still arrive at a 64% underprediction in ΔV for their experiment 4 (as compared to 68% when gas-phase glyoxal was not held fixed). Based on our analysis as described above, we can eliminate variations in H*, differences in photolysis intensity, and a glyoxal reservoir on chamber walls as potential explanations for discrepancies in growth phenomena observed across laboratory
Figure 3. Aqueous aerosol composition predicted by GAMMA experiment 4 of the work of Galloway et al.12
in order to evaluate its potential impact on the experimental observations. We find that, for each experiment simulated, predicted SOA formation decreased with increasing photolysis intensity in the model due to gas-phase glyoxal photolysis, 11922
dx.doi.org/10.1021/es502020j | Environ. Sci. Technol. 2014, 48, 11919−11925
Environmental Science & Technology
Article
significant formation of the organic acids, malonic acid and succinic acid, by reacting with glyoxal, which in turn led to further Henry’s Law uptake of glyoxal. We found that, assuming that reaction 1 takes place in the aerosol bulk, a photochemical rate constant of ∼10−4 s−1 was sufficient to match the growth observed by Volkamer et al. This corresponds to an OH production rate of ∼10−4 M s−1, similar to the OH production rate calculated by Ervens et al. as being required to overcome OH limitations in box model simulations of SOA formation by glyoxal in aqueous aerosol particles.27 The rate constant that we infer is somewhat higher than the photolysis rate for aqueous hydrogen peroxide in our model (3.77 × 10−6 s−1),28 suggesting that this pathway likely does not account for all the missing photochemical SOA formation. Reaction 2 was also able to account for the observations made by Volkamer et al. with a rate constant ranging from 5 × 10−5 to 10−4 M−1 s−1. In the absence of mechanistic information and measured rate constants, it is difficult to distinguish via modeling whether the proposed photoenhanced chemistry is a surface or bulk process. However, since FA and IC are both surface-active, VOCs are more readily available at the surface, and UV intensity is higher at the surface than in the aerosol interior, it is likely that surface processes are important. This is consistent with assumptions made in modeling studies by Volkamer and co-workers.29,30 Besides the application to this analysis, fulvicand humic-like “brown carbon” substances are common components of atmospheric aerosols31 and the potential impact of their photoenhanced chemistry on ambient SOA formation deserves further study. We note that GAMMA also underestimates particle volume growth in the absence of fulvic acid (e.g., experiment 13 of Volkamer et al.11). Fulvic acid was not used in the experiments of Galloway et al.,12 and they did not observe the dramatic photoenhancement of SOA formation seen by Volkamer et al.;11 however, our detailed analysis reveals a consistent unknown photochemical source of SOA in their experiments (Table 2). While some error is likely attributable to variations in H*, we would expect that error to be randomly distributed or biased positive (see Figure 1) rather than showing the consistent underprediction demonstrated in Table 2. As Aregahegn et al.17 proposed, light-absorbing glyoxal SOA products such as IC could be participating in photoenhanced chemistry. No evidence exists from the experiments of Aregahegn et al. for IC’s ability to oxidize glyoxal either directly or via the oxidation of radicals in situ.19 However, a variety of other light-absorbing, fulvic-like compounds are known to be formed by glyoxal in ammonium sulfate solutions which could potentially act as photosensitizers.32,33 This chemistry, which involves the participation of ammonium, would be enhanced at lower relative humidities (i.e., higher solute concentrations), perhaps explaining the absence of this effect in experiment 7 of the work of Volkamer et al.11 Finally, additional unknown pathways of SOA formation under irradiated conditions may exist besides the photoenhanced chemistry analyzed here−for example, unaccountedfor in-particle sources of OH, such as photolysis of organic peroxides34−38 which would open up the possibility of photochemically driven oligomerization.22,39
studies. The systematic underprediction of SOA growth we see under irradiated conditions reveals a light-dependent formation mechanism that is not accounted for in GAMMA. We therefore explored the potential for photoenhanced chemistry in the aerosol phase. Candidate photosensitizers include fulvic acid, present in the seed aerosols of Volkamer et al.,11 and imidazole carboxaldehyde (IC), one of the products of glyoxal reacting with ammonium sulfate solutions in the dark.6 While Aregahegn and co-workers17 reported photoenhanced SOA production by IC, the rate and quantum yield of this process have not been quantified, and glyoxal oxidation by photoactive IC has not been confirmed. To our knowledge, no direct evidence exists of photoenhanced reactions of fulvic acid with glyoxal, although photoactivated humic acid has been shown to promote SOA formation by limonene and isoprene in aerosols.16 Therefore, in the absence of detailed mechanistic and kinetic information, in order to qualitatively determine the effect of such chemistry, we posited potential photosensitizer reactions and analyzed the outcome they had when incorporated into the existing GAMMA mechanism. These tests were run on the simulations of chamber experiments in which the seed particles had a 50% fulvic acid composition: Volkamer et al.11 experiments 11, 17, 20, and 26. Photoactivation of fulvic acid (FA) was hypothesized to either lead to the direct generation of OH radicals in the aqueous aerosolphase or lead to the irreversible generation of a secondary organic aerosol product by reacting with glyoxal in the presence of light. In both cases, we assumed FA to be conserved during the reaction (i.e., it acted as a “photocatalyst”). FA + hv → OH + FA
(1)
GLYX + FA + hv → SOA + FA
(2)
Including either reaction 1 or 2 in GAMMA led to predictions of enhanced aerosol growth. Simulation results are shown in Figure 5. The generation of OH radicals via reaction 1 led to
Figure 5. Simulations of chamber results from the work of Volkamer et al.11 Experiment 11 using the photochemical box model GAMMA.20 In the baseline simulation, formation of organic acids via aqueousphase photooxidation is small, and SOA mass decreases with reaction time due to photolysis of gas-phase glyoxal leading to reemission of absorbed glyoxal. Including either reaction 1 or 2, both based on fulvic acid acting as a photosensitizer, could theoretically explain the observed efficient SOA growth.
■
AUTHOR INFORMATION
Corresponding Author
*E-mail: [email protected]. Phone: +1 (212) 854-2869. Fax: +1 (212) 854-3054. 11923
dx.doi.org/10.1021/es502020j | Environ. Sci. Technol. 2014, 48, 11919−11925
Environmental Science & Technology
Article
Notes
(17) Aregahegn, K. Z.; Noziere, B.; George, C. Organic Aerosol Formation Photoenhanced by the Formation of Secondary Photosensitizers in ageing Aerosols. Faraday Discuss. 2013, 165, 123−134. (18) George, C.; D’Anna, B.; Herrmann, H.; Weller, C.; Vaida, V.; Donaldson, D. J.; Bartels-Rausch, T.; Ammann, M. Emerging Areas in Atmospheric Photochemistry. Top. Curr. Chem. 2012, 339, 1−54. (19) Canonica, S.; Jans, U. R. S.; Stemmler, K.; Hoigne, J. Transformation Kinetics of Phenols in Water: Photosensitization. Environ. Sci. Technol. 1995, 29, 1822−1831. (20) McNeill, V. F.; Woo, J. L.; Kim, D. D.; Schwier, A. N.; Wannell, N. J.; Sumner, A. J.; Barakat, J. M. Aqueous-phase secondary organic aerosol and organosulfate formation in atmospheric aerosols: a modeling study. Environ. Sci. Technol. 2012, 46, 8075−8081. (21) Rossignol, S.; Aregahegn, K. Z.; Tinel, L.; Fine, L.; Noziere, B.; George, C. Glyoxal induced atmospheric photosensitized chemistry leading to organic aerosol growth. Environ. Sci. Technol. 2014, 48, 3218−3227. (22) Tan, Y.; Carlton, A. G.; Seitzinger, S. P.; Turpin, B. J. SOA from methylglyoxal in clouds and wet aerosols: Measurement and prediction of key products. Atmos. Environ. 2010, 44, 5218−5226. (23) Perri, M. J.; Lim, Y. B.; Seitzinger, S. P.; Turpin, B. J. Organosulfates from glycolaldehyde in aqueous aerosols and clouds: Laboratory studies. Atmos. Environ. 2010, 44, 2658−2664. (24) Compernolle, S.; Müller, J.-F. Henry’s law constants of diacids and hydroxypolyacids: recommended values. Atmos. Chem. Phys. 2014, 14, 2699−2712. (25) Wexler, A. S.; Clegg, S. L. Atmospheric aerosol models for systems including the ions H+, NH4+, Na+, SO42−, NO3−, Cl−, Br−, and H2O. J. Geophys. Res. 2002, 107 (D14), 4207 DOI: 10.1029/ 2001JD000451. (26) Paulot, F.; Crounse, J. D.; Kjaergaard, H. G.; Kürten, A., St; Clair, J. M.; Seinfeld, J. H.; Wennberg, P. O. Unexpected epoxide formation in the gas-phase photooxidation of isoprene. Science 2009, 325, 730−733. (27) Ervens, B.; Sorooshian, A.; Lim, Y. B.; Turpin, B. J. Key parameters controlling OH-initiated formation of secondary organic aerosol in the aqueous phase (aqSOA). J. Geophys. Res. Atmos. 2014, 119, 3997−4016. (28) Lelieveld, J.; Crutzen, P. J. The role of clouds in tropospheric photochemistry. J. Atmos. Chem. 1991, 12, 229−267. (29) Ervens, B.; Volkamer, R. Glyoxal processing by aerosol multiphase chemistry: towards a kinetic modeling framework of secondary organic aerosol formation in aqueous particles. Atmos. Chem. Phys. 2010, 10, 8219−8244. (30) Waxman, E. M.; Dzepina, K.; Ervens, B.; Lee-taylor, J.; Aumont, B.; Jimenez, J. L.; Madronich, S.; Volkamer, R. Secondary organic aerosol formation from semi- and intermediate-volatility organic compounds and glyoxal: Relevance of O/C as a tracer for aqueous multiphase chemistry. Geophys. Res. Lett. 2013, 40, 1−5. (31) Graber, E. R.; Rudich, Y. Atmospheric HULIS: How humic-like are they? A comprehensive and critical review. Atmos. Chem. Phys. 2006, 6, 729−753. (32) Shapiro, E. L.; Szprengiel, J.; Sareen, N.; Jen, C. N.; Giordano, M. R.; McNeill, V. F. Light-absorbing secondary organic material formed by glyoxal in aqueous aerosol mimics. Atmos. Chem. Phys. 2009, 9, 2289−2300. (33) Schwier, A. N.; Sareen, N.; Mitroo, D.; Shapiro, E. L.; McNeill, V. F. Glyoxal-methylglyoxal cross-reactions in secondary organic aerosol formation. Environ. Sci. Technol. 2010, 44, 6174−6182. (34) Anastasio, C.; McGregor, K. G. Chemistry of fog waters in California’s Central Valley: 1. In situ photoformation of hydroxyl radical and singlet molecular oxygen. Atmos. Environ. 2001, 35, 1079− 1089. (35) Zhao, R.; Lee, A. K. Y.; Abbatt, J. P. D. Investigation of Aqueous-Phase Photooxidation of Glyoxal and Methylglyoxal by Aerosol Chemical Ionization Mass Spectrometry: Observation of Hydroxyhydroperoxide Formation. J. Phys. Chem. A 2012, 116, 6253− 6263.
The authors declare no competing financial interest.
■ ■
ACKNOWLEDGMENTS This work was supported by NSF-ATM REU funding. REFERENCES
(1) Hallquist, M.; Wenger, J. C.; Baltensperger, U.; Rudich, Y.; Simpson, D.; Claeys, M.; Dommen, J.; Donahue, N. M.; George, C.; Goldstein, A. H.; et al. The formation, properties and impact of secondary organic aerosol: current and emerging issues. Atmos. Chem. Phys. 2009, 9, 5155−5236. (2) Volkamer, R.; Jimenez, J. L.; San Martini, F.; Dzepina, K.; Zhang, Q.; Salcedo, D.; Molina, L. T.; Worsnop, D. R.; Molina, M. J. Secondary organic aerosol formation from anthropogenic air pollution: Rapid and higher than expected. Geophys. Res. Lett. 2006, 33, L17811. (3) Whipple, E. B. Structure of glyoxal in water. J. Am. Chem. Soc. 1970, 92, 7183−7186. (4) Lim, H.-J.; Carlton, A. G.; Turpin, B. J. Isoprene Forms Secondary Organic Aerosol through Cloud Processing. Environ. Sci. Technol. 2005, 39, 4441−4446. (5) Kroll, J. H.; Ng, N. L.; Murphy, S. M.; Varutbangkul, V.; Flagan, R. C.; Seinfeld, J. H. Chamber studies of secondary organic aerosol growth by reactive uptake of simple carbonyl compounds. J. Geophys. Res. 2005, 110, D23207. (6) Galloway, M. M.; Chhabra, P. S.; Chan, a. W. H.; Surratt, J. D.; Flagan, R. C.; Seinfeld, J. H.; Keutsch, F. N. Glyoxal uptake on ammonium sulphate seed aerosol: reaction products and reversibility of uptake under dark and irradiated conditions. Atmos. Chem. Phys. 2009, 9, 3331−3345. (7) Schwartz, S. E. Mass-transport considerations pertinent to aqueous phase reactions of gases in liquid-water clouds; NATO ASI Series, Jaeschke, W., Ed.; Springer-Verlag: Berlin Heidelberg, 1986; Vol. G6, pp 425−471. (8) Betterton, E. A.; Hoffmann, M. R. Henry’s Law Constants of Some Environmentally Important Aldehydes. Environ. Sci. Technol. 1988, 22, 1415−1418. (9) Zhou, X.; Mopper, K. Apparent Partition Coefficients of 15 Carbonyl Compounds between Air and Seawater and between Air and Freshwater; Implications for Air-Sea Exchange. Environ. Sci. Technol. 1990, 24, 1864−1869. (10) Ip, H. S. S.; Huang, X. H. H.; Yu, J. Z. Effective Henry’s law constants of glyoxal, glyoxylic acid, and glycolic acid. Geophys. Res. Lett. 2009, 36 (L01802), 1−5. (11) Volkamer, R.; Ziemann, P. J.; Molina, M. J. Secondary Organic Aerosol Formation from Acetylene (C2H2): seed effect on SOA yields due to organic photochemistry in the aerosol aqueous phase. Atmos. Chem. Phys. 2009, 9, 1907−1928. (12) Galloway, M. M.; Loza, C. L.; Chhabra, P. S.; Chan, a. W. H.; Yee, L. D.; Seinfeld, J. H.; Keutsch, F. N. Analysis of photochemical and dark glyoxal uptake: Implications for SOA formation. Geophys. Res. Lett. 2011, 38 (L17811), 1−5. (13) Yu, G.; Bayer, A. R.; Galloway, M. M.; Korshavn, K. J.; Fry, C. G.; Keutsch, F. N. Glyoxal in aqueous ammonium sulfate solutions: products, kinetics and hydration effects. Environ. Sci. Technol. 2011, 45, 6336−6342. (14) Kampf, C. J.; Waxman, E. M.; Slowik, J. G.; Dommen, J.; Pfaffenberger, L.; Praplan, A. P.; Prévôt, A. S. H.; Baltensperger, U.; Hoffmann, T.; Volkamer, R. Effective Henry’s Law Partitioning and the Salting Constant of Glyoxal in Aerosols Containing Sulfate. Environ. Sci. Technol. 2013, 47, 4236−4244. (15) Yu, G.; Keutsch, F. N. Atmospheric organic-phase photosensitized epoxidation of alkenes by α-dicarbonyls. Atmos. Chem. Phys. Discuss. 2012, 12, 15115−15138. (16) Monge, M. E.; Rosenørn, T.; Favez, O.; Müller, M.; Adler, G.; Abo Riziq, A.; Rudich, Y.; Herrmann, H.; George, C.; D’Anna, B. Alternative pathway for atmospheric particles growth. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 6840−6844. 11924
dx.doi.org/10.1021/es502020j | Environ. Sci. Technol. 2014, 48, 11919−11925
Environmental Science & Technology
Article
(36) Bateman, A. P.; Nizkorodov, S. A.; Laskin, J.; Laskin, A. Photolytic processing of secondary organic aerosols dissolved in cloud droplets. Phys. Chem. Chem. Phys. 2011, 13, 12199−12212. (37) Epstein, S. A.; Blair, S. L.; Nizkorodov, S. A. Direct Photolysis of α-Pinene Ozonolysis Secondary Organic Aerosol: Effect on Particle Mass and Peroxide Content. Environ. Sci. Technol. 2014, DOI: 10.1021/es502350u. (38) Chen, Z. M.; Wang, H. L.; Zhu, L. H.; Wang, C. X.; Jie, C. Y.; Hua, W. Aqueous-phase ozonolysis of methacrolein and methyl vinyl ketone: a potentially important source of atmospheric aqueous oxidants. Atmos. Chem. Phys. 2008, 8, 2255−2265. (39) Renard, P.; Siekmann, F.; Gandolfo, A.; Socorro, J.; Salque, G.; Ravier, S.; Quivet, E.; Clément, J.-L.; Traikia, M.; Delort, A.-M.; et al. Radical mechanisms of methyl vinyl ketone oligomerization through aqueous phase OH-oxidation: on the paradoxical role of dissolved molecular oxygen. Atmos. Chem. Phys. 2013, 13, 6473−6491.
11925
dx.doi.org/10.1021/es502020j | Environ. Sci. Technol. 2014, 48, 11919−11925
