Hydroxymethanesulfonate and Sulfur(IV) in Fairbanks Winter During the ALPACA Study

Abstract

Hydroxymethanesulfonate (HMS) in fine aerosol particles has been reported at significant concentrations along with sulfate under extreme cold conditions (-35 °C) in Fairbanks, Alaska, a high latitude city. HMS, a component of S(IV) and an adduct of formaldehyde and sulfur dioxide, forms in liquid water. Previous studies may have overestimated HMS concentrations by grouping it with other S(IV) species. In this work, we further investigate HMS and the speciation of S(IV) through the Alaskan Layered Pollution and Chemical Analysis (ALPACA) intensive study in Fairbanks. We developed a method utilizing hydrogen peroxide to isolate HMS and found that approximately 50% of S(IV) is HMS for total suspended particulates and 70% for PM_2.5. The remaining unidentified S(IV) species are closely linked to HMS during cold polluted periods, showing strong increases in concentration relative to sulfate with decreasing temperature, a weak dependence on particle water, and similar particle size distributions, suggesting a common aqueous formation process. A portion of the unidentified S(IV) may originate from additional aldehyde-S(IV) adducts that are unstable in the water-based chemical analysis process, but further chemical characterization is needed. These results show the importance of organic S(IV) species in extreme cold environments that promote unique aqueous chemistry in supercooled liquid particles.

Figure 1

Comparison of filter collection with IC analysis by UNH and GT. UNH collected TSP and analyzed the filters in the field, GT measured PM_2.5 and analyzed the filters after 4 to 8 weeks of storage. Slope of 1 line (black) and orthogonal regression results are shown with the intercept forced through zero.

Figure 2

PILS online measurements of PM_2.5 sulfate and S(IV) for the ALPACA intensive study. Temperature is also shown along with periods used in contrasting case studies shaded in gray. Black bars indicate MOUDI sampling periods to characterize particle size distributions that are plotted in Figure 5.

Figure 3

Time series of sulfate, S(IV) and HMS determined by three different filter sampling systems. Gray shaded areas are periods analyzed in more detail.

Figure 4

HMS versus S(IV) from filter samples that collected particles over three different size ranges. The slopes are orthogonal regressions and give the HMS fraction of S(IV).

Figure 5

Selected size distributions from MOUDI measurements during the ALPCA study for sample ending dates shown (month/day; 01/26, 01/21 and 02/01). These sampling periods are shown in Figure 2 and correspond to periods of (i) low, (iii) moderate and (iii) high PM_2.5 mass concentration, given at the top of each column of plots. All species are plotted to zero (not stacked). The smallest size channel is from the MOUDI after filter with lower range arbitrarily set at 0.01 μm aerodynamic diameter. Three MOUDI measurements (each column is one MOUDI measurement) were selected for a range of PM_2.5 and S(IV) concentrations; (a) (d) relatively low, (b) (e) moderate and (c) (f) high. The lognormal fits for various modes and corresponding geometric mean aerodynamic diameters are also shown. More fit details, including geometric standard deviation and the uncertainty in fit parameters are given in Figure S6. Other S(IV) is equal to the difference in S(IV) and HMS.

Figure 6

Variation of sulfur species with surface temperature at the CTC site based on various measurement methods. (a) Sulfate concentration measured online with the PILS and with UNH TSP filters and GT PM_2.5 filters. (b) Ratio of S(IV) to sulfate, (c) ratio of HMS to sulfate and (d) ratio of other S(IV) to sulfate. The PILS did not speciate S(IV) with the H₂O₂ treatment so only filter data are shown in plots (c) and (d), which also shows the UNH data collected during the cold ((B) in Fig. 2) and warm ((C) in Fig. 2) events. Other S(IV) is equal to the difference in S(IV) and HMS.

Figure 7

PM_2.5 S(IV) concentrations as a function of ambient RH for all data measured during ALPACA with the PILS. Data are colored by ambient temperature and the various events of higher concentrations identified in Figure 1 are labeled as periods (A), (B), and (C).