Article pubs.acs.org/JPCA
Heterogeneous Glyoxal Oxidation: A Potential Source of Secondary Organic Aerosol B. M. Connelly,*,† D. O. De Haan,‡ and M. A. Tolbert† †
Cooperative Institute for Research in the Environmental Sciences and Department of Chemistry and Biochemistry, University of Colorado, Boulder, Colorado 80309, United States ‡ Department of Chemistry and Biochemistry, University of San Diego, San Diego, California 92110, United States S Supporting Information *
ABSTRACT: Laboratory studies are described that suggest reactive uptake of glyoxal on particulate containing HNO3 could contribute to the formation of secondary organic aerosol (SOA) in the upper troposphere (UT). Using a Knudsen cell flow reactor, glyoxal is observed to react on supercooled H2O/HNO3 surfaces to form condensed-phase glyoxylic acid. This product was verified by derivatization and GC−MS analysis. The reactive uptake coefficient, γ, of glyoxal varies only slightly with the pressure of nitric acid, from γ = 0.5 to 3.0 × 10−3 for nitric acid pressures between 10−8 and 10−6 Torr. The data do not show any dependence on temperature (181−201 K) or pressure of glyoxal (10−7 to 10−5 Torr). Using the determined reactive uptake kinetics in a simple model shows that glyoxal uptake to supercooled H2O/HNO3 may account for 4−53% of the total organic mass fraction of aerosol in the UT.
■
We have utilized a Knudsen continuous flow reactor to form supercooled liquid H2O/HNO3 films on ice and organic films under UT relevant temperatures (181−211 K) and pressures of nitric acid (10−8 to 10−6 Torr HNO3). We then studied glyoxal uptake on these H2O/HNO3 films and determined the uptake coefficient, γ, of glyoxal to the liquid layer under relevant conditions. Additionally, in situ and ex situ analysis was performed to identify condensed products. The experimentally determined heterogeneous kinetics and existing literature data on glyoxal gas-phase kinetics can be combined to estimate SOA formation by glyoxal uptake in the UT. From there, the significance of the SOA formation can be determined by comparing the estimated SOA to field observations. Although SOA formation by glyoxal uptake has been investigated in the lower troposphere, this study is the first to investigate SOA formation by glyoxal uptake under UT conditions. Because the concentration of nitric acid and the majority of the glyoxal precursors are increased by anthropogenic activities,17,18 formation of SOA by reactive uptake of glyoxal to aerosol may have implications for future global climate change and intercontinental transport of pollution.6
INTRODUCTION Recent field studies have found that organics compromise a significant fraction of aerosol mass throughout the troposphere.1−6 Organic mass fractions up to 60% have even been found at high altitudes (1 h) to vapor-phase glyoxal (∼10−5 Torr). After exposure, the chamber and film were pumped at ∼10−7 Torr for more than 1 hour. Both the Q-MS and FTIR-RAS indicated the surface contained only organic material, free of both nitric acid and liquid water. After the organic films were synthesized, we then prepared the Knudsen reactor for glyoxal uptake experiments on the films. The organic films were brought to a steady state with the chamber atmosphere set to atmospherically relevant pressures of nitric acid and water vapor. The organic film (now with a fresh liquid layer of supercooled H2O/HNO3) was then isolated from the chamber, and then a desired pressure of glyoxal introduced. Once the chamber was at a steady state, the cup was retracted and a decrease in glyoxal signal was observed. Equation 1 was used to determine the uptake coefficient. The values for S0 and S are averaged from at least 30 data points using the same procedure described previously for the ice films. Glyoxal uptake experiments on organic films coated with H2O/ HNO3 were performed as a function of the relative humidity with respect to ice (RHice), in addition to temperature (181− 201 K), pressure of nitric acid (10−8 to 10−6 Torr), and pressure of glyoxal (10−7 to 10−5). The RHice ranged between the onset of H2O/HNO3 uptake to the organic film (∼60% RHice) until ice nucleation was observed in the FTIR-RAS spectra. The RHice where ice nucleation occurred varied depending on temperature, with the films at 181 K having the highest value of ∼155% RHice, whereas the films at 201 K nucleated ice at 125% RHice. After uptake, the organic/H2O/ HNO3 films could be again pumped free of water and nitric acid, returning them to their composition prior to the uptake experiment (but with added thickness). After the FTIR-RAS and MS confirmed that the film contained only the organic material, additional uptake experiments could be performed. In addition to the FTIR- RAS, the products of the uptake were analyzed ex situ using gas chromatography/mass spectrometry (GC/MS) with derivatization. N-Methyl-N(trimethylsilyl)trifluoroacetamide with 1% trimethylchlorosilane (MSTFA + 1% TMCS) dissolved in an equal volume of pyridine was used as the derivatization agent. After an uptake experiment was performed, the chamber was flushed with dry nitrogen and the gold substrate removed. Then 200 μL of the derivatization agent was added directly to the gold surface and removed using a glass capillary after the solid products dissolved. Following 15 min reaction time a 5 μL sample was injected into the GC−MS for analysis. A Finnigan Trace GC Ultra equipped with a Zebron ZB-5mS capillary column (95% dimethylpolysiloxane and 5% phenylarylene, 30 m × 0.25 mm × 0.25 μm) was used for separations and mass spectra were obtained with a Polaris Q ion trap detector (ThermoFinnigan, San Jose, CA). The helium carrier gas flow rate was 1.0 mL min−1. The initial column temperature was 85 °C, which was held for 5 min, then ramped to 320 at 30 °C min−1, and then
held for 1 min. The EI-Ion trap mass spectrometer collected ions from 30 to 650 m/z.
■
RESULTS AND DISCUSSION Glyoxal Uptake Kinetics. Glyoxal uptake was first probed on the bare gold substrate and pure ice surfaces. Figure 3, traces
Figure 3. Mass spectrometer signal at 58 m/z versus time for glyoxal uptake experiments. (A) Bare gold substrate. (B) Pure ice film. (C) Uptake of glyoxal to supercooled HNO3/H2O layer on an ice film exposed to 10−6 Torr nitric acid at 190 K.
A and B, shows that glyoxal uptake does not occur on either of these surfaces. In contrast, Figure 3C shows glyoxal uptake to a supercooled H2O/HNO3 film on ice. Figure 3C shows a slight recovery of the signal of ∼2% after 5 min; however, the recovery was undetectable in other uptake experiments, even in those lasting up to 2 h. Uptake without recovery can indicate irreversible uptake, constant absorption and desorption flux, or both occurring simultaneously. To shed light on which regime was being observed, we isolated the surface from the chamber atmosphere after uptake and pumped out the remaining glyoxal. When the cup was retracted, desorption of glyoxal was not observed. Therefore, it is likely that the uptake is largely irreversible, though adsorption/desorption during uptake cannot be ruled out. The data from the uptake experiments is summarized in Table 1. Glyoxal uptake was observed on ice and organic surfaces only in the presence of a supercooled H2O/HNO3 layer, within the nitric acid pressure range studied (10−8 to 10−6 Torr). The uptake of glyoxal did not vary with the pressure of glyoxal (10−7 to 10−5 Torr) or temperature (181−201 K). The open triangles in Figure 4A are uptake for films where the initial FTIR-RAS indicated only ice and supercooled H2O/HNO3. These films were observed at nitric acid pressures below ∼10−6 Torr. A slight increase in γ with HNO3 pressure could possible be read from this data, but the data are almost equally well fit assuming no dependence on the nitric acid pressure. 6182
dx.doi.org/10.1021/jp211502e | J. Phys. Chem. A 2012, 116, 6180−6187
The Journal of Physical Chemistry A
Article
also present, presumably on the remaining ice surface not yet covered by NAT crystals. Example spectra are provided in the Supporting Information. For these films, glyoxal uptake was observed when gaseous HNO3 (and hence a supercooled liquid H2O/HNO3 film) was present. In contrast, when the nitric acid vapor was removed from the chamber, the supercooled liquid H2O/HNO3 film quickly evaporated, and glyoxal uptake was not observed on the ice/NAT surface left behind. This again suggests that glyoxal uptake requires an aqueous surface film. The calculated γ values varied from 10−3 to 10−2 for experiments performed on ice/NAT with supercooled liquid HNO3/H2O films. These values are also shown in Table 1. The variability of the γ values may be explained by an ice surface partially covered by NAT crystals, with a supercooled H2O/ HNO3 liquid film covering the remainder. Because the nucleation and growth of NAT on the ice surface is kinetically controlled, the surface coverage varies between experiments. However, because only the liquid film takes up glyoxal, and because the surface area of liquid film coverage is both variable and unknown, it is not possible to determine γ values with accuracy. The uptake experiments performed on the organic/H2O/ HNO3 films are shown in Figure 4B as a function of RHice, nitric acid pressure, and temperature. In Figure 4B the circle, square, and diamond plots represent experiments performed at 7 × 10−8, 4 × 10−7, and 2 × 10−6 Torr HNO3, respectively. The shading represents different temperatures for experiments performed at 7 × 10−8 Torr. As stated previously, the data does not reflect a dependence on temperature. Only at the highest pressure of nitric acid (2 × 10−6 Torr) does γ appear to be dependent on RHice. For the experiments performed between 10−7 and 10−8 Torr, linear regression analysis reveals that within error, the trend line may be flat. Here, the most significant dependence is an increase in γ with increased nitric acid pressure. For comparison, the average γ for the uptake to organic/H2O/HNO3 films at each pressure of nitric acid is shown as solid symbols in Figure 4A. These γ values are similar to the range of γ values for the uptake on the ice films. A linear regression analysis for all the data taken together shows a slight trend in γ with respect to nitric acid pressure. For the range studied here, an increase of a factor of 100 in HNO3 pressure yields only an increase in γ of a factor of 3. There are many chemical and physical components that may contribute to the kinetics of reactive uptake. For a reaction in the supercooled H2O/HNO3 bulk, the uptake could be influenced by the nitric acid concentration, hydronium ion activity, aqueous diffusion rate, and the activation barrier for the reaction. The reaction could occur directly on the surface, as well as in the bulk. There may be multiple uptake and reaction mechanisms, as it is possible that organic intermediates, organic nitrates, or organic nitrites may be present in quantities below the limit of detection of the FTIR-RAS during the uptake process. The experiments performed do not allow us to determine the detailed reaction mechanism; therefore, we cannot conclude with confidence a definitive trend relating γ and nitric acid concentration. On the other hand, we can confidently conclude that reactive uptake of glyoxal will occur when a supercooled H2O/HNO3 layer is present. Product Analysis. After glyoxal uptake, the chamber was pumped down to less than ∼3 × 10−7 Torr with the cup closed for over an hour. When the cup was retracted, the Q-MS did not indicate desorption of glyoxal from the film. Once the cup was retracted, both HNO3 and H2O evaporated/sublimated
Table 1. Summary of Glyoxal Uptake Experiments film type ice/H2O/ HNO3
ice/NAT
ice/NAT/ H2O/HNO3
organic/ H2O/HNO3
a
HNO3, ×10−7 Torr
RHice
temp, K
0.14
100
191
0.26 0.56 0.56 0.56 0.99 1.12 1.43 2.60 trace trace trace 2.8
100 100 100 100 100 100 100 100 100 100 100 100
191 181 181 181 201 191 201 191 181 191 201 181
2.8 15 22 40 50 50 100 100 0.70
100 100 100 100 100 100 100 100 55
181 201 181 191 201 201 201 201 181
0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.70 0.70 4.00 4.00 4.00 4.00 20.00 20.00 20.00 20.00
57 65 68 73 74 78 79 83 92 95 100 103 105 107 111 113 116 129 129 149 66 92 100 114 72 100 107 115
191 201 191 191 201 191 201 191 181 191 201 201 191 181 201 191 201 201 181 181 201 201 201 201 201 201 201 201
glyoxal, ×10−6 Torr 46 1.3 3.5 1.37 0.9 0.4 3 4.4 3.5 6.2 4.2 8.6 1.3 3.4 3.9 4.1 7.0 72 78 6.5 70 4.1 2.8 1.3 4.2 3.4 0.93 3.4 3.8 4 0.3 4 1.6 3.4 2.5 4.9 4 3.4 3.3 3.1 4 1.7 3.4 3.1 3 3.5 3.1 2.7 2.5 2.4
γ × 10−3 0.71 1.8 1.26 1.37 1.43 0.893 1.37 1.94 1.83
