Modes in the size distributions and neutralization extent of fog-processed ammonium salt aerosols observed at Canadian rural locations

Introduction Conclusions References


Introduction
Similar to clouds, fog plays important roles in the formation of secondary atmospheric aerosols.Fog events modify the size distribution, chemical composition and thus optical properties of preexisting atmospheric aerosols (Ondov and Wexler, 1998;Moore et al., 2004;Fahey et al., 2005;Sun et al., 2006;Aikaw et al., 2007;Herckes et al., 2007;Biswas et al., 2008;Collett Jr. et al., 2008;Dall'Ostol et al., 2009;Kaul et al., 2011;Figures Back Close Full Rehbein , et al., 2011;Yu et al., 2011).Fog droplets could efficiently scavenge atmospheric gaseous and particulate pollutants, followed by chemical reactions occurring in droplets (Pandis et al., 1990;Collett Jr. et al., 1999, 2008;Fahey et al., 2005;Biswas et al., 2008).When fog dissipates, fog droplets evolve into atmospheric aerosol particles with modified physical and chemical properties.Due to the high deposition velocity of large fog droplets (Herckes et al., 2007), it is difficult to characterize the size-distribution of fog-processed aerosols.Knowledge of size distribution and chemical composition of fog-processed aerosols are limited and factors determining these aerosol properties are poorly understood (Law and Stohl., 2007;Yu et al., 2011).Enhanced particle pollution was recently reported due to fog-processing events (Sun et al., 2006;Biswas et al., 2008;Yu et al., 2011).High number concentration of fog droplets was observed at sizes 5-6 µm of diameter in polluted ambient air environment (Quan et al., 2011).The small fog droplet size observed in Quan et al. (2011) could be caused by the high fog condensation nuclei (FCN) concentration, similar to the cloud formation under polluted environment (Zhang et al., 2006;Rosenfeld et al., 2008;Qian et al., 2009).These smaller fog droplets should have longer residence time than larger droplets (20-35 µm) as reported by Frank et al. (1998) and Herckes et al. (2007).It is not clear if fog-processing of aerosols enhance particulate pollution in relatively clean environments such as Canadian remote locations, what types of preexisting aerosols are precursors of FCN, and which factors determine physical and chemical properties of fog-processed aerosols.
The purpose of the present study is to identify fog processed aerosols from a suite of field measurements collected at seven rural inland sites and one coastal site in Canada (Zhang et al., 2008a, b) and to explore the variability in size distribution of the fog-processed ammonium salt aerosols with particular attention to the impacts of temperature, particle acidity condition, and preexisting aerosols serving as FCN.Sources and/or formation mechanisms of supermicron particle modes of ammonium salts are investigated in terms of gas-particle condensation, primary emissions, fog processing of aerosols, and heterogeneous reactions of gases with sea-salt aerosols.The results Figures

Back Close
Full are expected to improve our understanding on the size distribution of fog-processed ammonium salt aerosols.

Data
In this study, an eleven-stage MOUDI (Model 110) with 50 % cut-off points for the particle aerodynamic diameters: 18, 9.9, 6.2, 3.1, 1.8, 1.0, 0.54, 0.32, 0.18, 0.093, and 0.048 µm, were used for sampling at eight rural Canadian in eastern and central Canada (Fig. 1).Simultaneously, a PM 2.5 sampler equipped with a Na 2 CO 3 -coated and a citric-acid-coated denuders was also used to collect SO 2 , HNO 3 and NH 3 gases and PM 2.5 .Inorganic ions in particles were determined by an ion chromatograph while organics were not measured.In addition, SO 2 , NO x , NO y and O 3 analyzers were used to measure their concentrations in minutes.On-site meteorological data were recorded to support data analysis.Fog events were judged by on-site observed relative humidity and the record obtained from the nearest meteorological station to the sampling sites (http://www.wunderground.com/history).Detailed information about the sampling sites and chemical analysis can be found in Zhang et al. (2008a, b) and Yao and Zhang (2011).

Hypothesis 1: Fingerprint of fog-processed aerosols
Among the total of 192 MOUDI samples, ten samples (about 5 % of total data samples) had one or two supermicron particle modes of ammonium salts (colored lines in Fig. 2).However, supermicron particle modes of ammonium salts were absent in nonfog samples (dashed lines in Fig. 2).The difference between fog samples and non-fog samples was statistically significant.Our hypothesis is that the supermicron modes of ammonium salts were the result of fog processing of ammoniated sulfate and/or nitrate.These ten samples having supermicron particle modes of completely-neutralized Figures

Back Close
Full or incompletely-neutralized ammonium salts were not likely related to cloud processing of aerosols due to the two considerations: (1) cloud-processed aerosols usually have a fingerprint of the 0.6-1.0µm mode (Ondov andWexler, 1998, Yao et al., 2003;Huang et al., 2006); and (2) partially cloud or overcast weather conditions occurred in over 60 % days in a year based on the weather records at airports of eastern Canada, yet only 5 % samples were found to have supermicron ammonium salt aerosols.If cloud processing indeed led to the supermicron modes of ammonium salt aerosols, there should be a lot more samples than the 10 samples identified.Gas-particle condensation alone usually leads to a particle mode at 0.2 ± 0.1 µm in mass spectra and this formation mechanism does not likely lead to the supermicron modes of ammonium salt aerosols.In another 10 % of samples which might have gone through foggy days, fog-processed aerosols cannot be clearly identified due to the absence of the supermicron modes of ammonium salts.One explanation could be due to the high efficiency of fog in removing particles from ambient air at that time and the amount of fog-processed aerosols was too small to be detectable.Thus, only the 10 samples were used to examine the hypothesis in the present study.

Hypothesis 2: Impact of acidity on fog-processed aerosols
The ambient acidity condition can affect chemical composition, and possibly, size distribution, of fog-processed aerosols.For example, if ambient condition is acidic, fogprocessed aerosols could theoretically be acidic and contain less nitrate, chloride and semi-volatile weak organic acids, and vice versa.Strong acidic, less acidic and neutral fog droplets have been reported under different ambient conditions (Kelmm et al., 1994;Collett Jr. et al., 1999;Moore et al., 2004;Biswas et al., 2008;Str äter et al., 2010;Watanabe et al., 2010;Yao et al., 2011).These studies also showed that NO nitrate through evaporation of HNO 3 gas.The hypothesis that the ambient acidity conditions affect chemical composition and size distributions of fog-processed aerosols will be examined here.The relative acidity (RA) is calculated using all observed ion species (in their equivalent concentrations, Kerminen et al., 2001): Considering that analytical errors of ionic concentrations were about 5 %, RA ≤ 0.9 was thereby considered a threshold to judge the presence of acidic aerosols in this study.
The missing RA in this study is due to concentrations of the major ion being close to the detection limit when the relative analytical error could be large.

Hypothesis 3: Impact of ambient temperature on fog-processed aerosols
Because most physical and chemical processes strongly depend on T , its potential impact on size and composition of fog-processed aerosols is worth examining.Freezing fog could occur under T < 0 • C and frozen fog could occur under T < −35 • C. Frozen fog was also reported with T in the −1 to −12 • C range, e.g., in the inland areas of the Pacific Northwest (http://en.wikipedia.org/wiki/Fog).Corbin et al. (2012) recently reported that combustion particles could be important ice nuclei in ambient air.In both freezing and frozen fog, ice crystals could be present.When fog dissipates, ice crystals might still exist, depending on ambient conditions.In addition, freezing T could lower rates of chemical reactions and favor semi-volatile species partitioning more in the particle phase (Seinfeld and Pandis, 2006).Thus, size distributions of aerosol particles produced from freezing or frozen fog could be different than those from T > 0 • C.This hypothesis will be examined in this study.
Based on the knowledge discussed above, the presentation of the data analysis is categorized into T > 0 • C and T < 0  et al., 2003, 2007;Huang and Yu, 2008;Lan et al., 2011), but the two supermicron modes of ammoniated sulfate and nitrate aerosols (ASNA) were rarely reported.The mass ratio of NO − 3 /SO − 4 at the 0.4-0.5 µm mode was lower than at the two supermicron modes (Fig. 3c), indicating that coexistence of ASNA in the submicron mode with ASNA in the two supermicron modes was thermodynamically unfavorable.
The 1-2 µm mode aerosols were once reported (Dall'Ostol et al., 2009;Nie et al., 2010) and could be associated with fog processing (Ondov and Wexler, 1998); but no direct evidence is available to support this assumption.This can only be demonstrated by excluding all the other possible formation routes as discussed below.(1) The sum of ([Ca 2+ ]+[Na + ]+[K + ]) in equivalent concentration at the 1.6-1.7 µm mode accounted for only about 10 % of the sum of ([ ), heterogeneous reactions of acidic gases with crustal, sea-salt and biomass burning aerosols cannot explain most of (SO 2− 4 + NO − 3 ) at this size range.The same was true for the 8.2-9.0 µm mode in which the sum of ( (2) Theoretically, it was less likely for hygroscopic growth of ASNA at the 0.4-0.5 µm mode, together with uptake of SO 2 (gas), HNO 3 (gas) and NH 3 (gas), to the 1.6-1.7 µm mode of ASNA (Kerminen and Wexler, 1995).In the central and eastern Canada, organic carbon only accounted for one-fourth to one-third of the sum of ([SO ) in mass concentration (Jeong et al., 2011) and should play a minor in particle growth.However, it was practically impossible to grow ASNA from 0.4-0.5 µm to 8.2-9.0 µm because it needed an impractical growth factor of 20.In fact, the hygroscopic growth factor of (NH 4 ) 2 SO 4 was only 2 at relative humidity of 98 % (Matsumura and Hayashi, 2007).In order to reach the growth factor of 20, a 0.4-0.5 µm ammonium salt aerosol particle needed to be coated by ammonium nitrate which mass should be about three orders of magnitude higher than that of the 0.4-0.5 µm particle.Supersatuation weather condition was thereby needed for the growth, which was present in fog events.Thus, fog processing of SO 2− 4 and NO − 3 seems to be the only possible path leading to the two supermicron models of ASNA.Introduction

Conclusions References
Tables Figures

Back Close
Full Frank et al. (1998) reported that fog droplets sometimes had bi-modal size distribution in number concentrations and sometimes exhibited a uni-modal pattern, although the causes were not well understood.Bi-modal ASNA observed in this study were possibly from evolution of bi-modal fog droplets.However, the FCN associated with the 8.2-9.0 µm mode of ASNA should be different from those associated with the 1.6-1.7 µm mode of ASNA because of different chemical composition between the two modes (Fig. 3a, c).
It was obvious, in the submicron size, that the concentrations of ASNA overwhelmed those of biomass burning, crustal and sea-salt aerosols.Thus, the preexisting ASNA from 1 µm down to 0.05 µm could be one of the major sources of FCN.Yao et al. (2011) reported that ASNA were usually internally mixed with organics in eastern Canada.Although the mass concentration of organics in submicron particles was significantly lower than the sum of those of inorganic ions (Jeong et al., 2011), organics were reported to enhance the efficiency of inorganic aerosols as fog or cloud condensation nuclei (Ming and Russell, 2004;Asa-Awuku et al., 2011;Gierlus et al., 2012).Aerosols <0.05 µm needs an impractically high supersaturation condition to be activated (Dusek et al., 2006) and could not play a significant role in forming FCN.The sum of ([K + ]+[Na + ]) in equivalent concentration was one order of magnitude smaller than that of NH + 4 .The role of K + and Na + was thereby negligible for the 8-9 µm mode aerosols.The concentrations of Ca 2+ and NH + 4 in the particles >3 µm were at the same order of magnitude.Preexisting Ca 2+ -contained aerosols could be an important source of FCN under supersatutaion conditions (Koehler et al., 2009;Gierlus et al., 2012) and contributed to the observed 8-9 µm mode aerosols.Moore et al. (2004) reported that the mean ratio of Ca 2+ in large fog droplets (about 30 µm) to small fog droplets (about 10 µm) was 4.4 at Whiteface Mountain, New York site, but the ratios for SO 2− 4 , NO − 3 in cold seasons, as was the case for most of the samples collected in cold seasons (Zhang et al., 2008a).However, in this particular sample, the concentration of the 1.6 µm mode NO − 3 is higher than that of the 0.5 µm mode (Fig. 3a), indicating that fog increased the concentration of particulate NO − 3 in the coarse mode in the atmosphere.This also implies that the new formation of nitrate in fog droplets was likely more than the removal of nitrate through deposition of fog droplets and interstitial aerosols.The elevated concentrations of gas precursors of NO − 3 support this hypothesis (Table 1).It is also noted that the SO 2− 4 concentrations at the 1.7 µm and 9.0 µm modes were much lower than those at the 0.4 µm mode, suggesting that the new formation of ammonium salt aerosol was mostly ammonium nitrate and that the new formation of SO 2− 4 in fog droplets was slower than the deposition of SO 2− 4 in fog droplets and interstitial coarse particles.
We further examined size distributions and composition of ions in the samples collected on 17-18 November 2004 at SPR.Three modes of SO 2− 4 were observed at 0.3, 1.0 and 5.0 µm while NO − 3 had only two modes at 1.0 µm and 5.0 µm (Fig. S2a, b).The 0.3 µm mode was between the upper limit of the condensation mode and the lower limit of primary aerosols (Ondov and Wexler, 1998).Following the analysis presented earlier, formation of ASNA at the 1.0 µm and 5.0 µm modes was also probably ascribed to fog-processing of aerosols.In this sample, Ca 4 between the 1.0 µm mode and the 5.0 µm mode aerosols also suggested that the two modes of aerosols were likely evolved from different types of fog droplets.
The concentration of SO 2− 4 in the 1.0 µm aerosols was slightly higher than that in the 0.3 µm mode aerosols, suggesting that the formation rate of SO 2− 4 in fog droplets was slightly higher than its deposition rate.The concentrations of NO − 3 or NH + 4 in the 1.0 µm mode aerosols were evidently higher than those in the 0.3 µm mode aerosols, indicated that fog probably increased concentrations of those species in the atmosphere.

Under T > 0 • C and acidic conditions
One sample collected at the coastal site (KEJ) and two samples collected at one inland site (CHA) meets the conditions of T > 0 • C and acidic particles.Size distributions of ionic species and RA in the sample collected at Kejimkujik (KEJ) during 10-11 November 2002 were shown in Fig. 4a, b, and fog occurred from 6 p.m. on November to 12 a.m. on 11 November and was also considered as a regional event based on the record from ground weather stations.Since the site is situated at a coastal area, nonsea-salt-SO were observed at 0.3, 1.5 and 3.7 µm (Fig. S3a).The RA values between 0.3-2 µm were less than 0.8, indicating that the nss-SO 2− 4 was incompletely neutralized.Under such acidic condition, the hygroscopic growth together with SO 2 oxidation to nss-SO 2− 4 cannot grow the 0.3 µm mode nss-SO 2− 4 to the 1.5 µm mode nss-SO 2− 4 according to a theoretical analysis by Kerminen and Wexler (1995).In addition, the sum of Ca 4 at the 3-5 µm mode than the 1-2 µm mode.However, the 3.7 µm mode nss-SO 2− 4 was close to the detection limit, much lower than the concentration of the 1.5 µm mode.It is thus concluded that fog processing of nss-SO 2− 4 most probably occurred, leading to the 1.5 µm mode of nss-SO 2− 4 .It is interesting to note that the concentration of nss-SO 2− 4 at the 1.5 µm mode was less than that at the 0.3 µm mode.This might be due to the low conversion rate of SO 2 to SO 2− 4 under acidic condition in fog droplets.However, it was also possible that fog-processed nss-SO 2− 4 were diluted, leading to a lower concentration of nss-SO 2− 4 at the 1.5 µm mode.
In acidic environment, incompletely-neutralized sulfate aerosols inhibited the formation of ammonium nitrate (Seinfeld and Pandis, 2006) and the NO − 3 can exist only as metal salts.Therefore, the concentration of NO − 3 in the aerosols between 0.3-2 µm was negligible because of lack of metal ions in this size range.The NO − 3 dominated at the 2.9 µm mode (Fig. S3b).However, Na + and Cl − dominated at the 3.7 µm mode (Fig. S3c).The maximum surface area concentration of sea-salt aerosols is usually located at a smaller size compared to the size of the maximum mass concentration of sea-salt aerosols.Thus, NO − 3 formed through surface reactions between acidic gases and sea-salt aerosols would peak at smaller sizes than that of the maximum sea-salt mass concentrations.The similar observations were reported by Zhuang et al. (1999) and Zhao and Gao (2008).
The two samples collected at Chalk River (CHA) on 14-15 June 2004 and 16-20 June 2004 are shown in Fig. 5. Fog occurred from 10 p.m. on 14 June to 5 a.m. on 15 June from 9 a.m. to 11 a.m. on 16 June, and from 1 a.m. to 4 a.m. on 19 June, all of which appeared to be local fog events since no fog was recorded at a station 20 km away from the site.For the sample collected during 14-15 June 2004, the RAs in the particles less than 3.1 µm varied from 0.80 to 0.87 except one outlier (with a value of 1.34) at the size bin of 0.32-0.54(Fig. 5b), which was caused by suspected high concentration of Ca 2+ (Fig. 5a) because the rest of the size bins had extremely low Introduction

Conclusions References
Tables Figures

Back Close
Full concentration (e.g., less than one tenth of the size bin of 0.32-0.54µm).The small RAs suggest that ammoniated sulfate aerosols were incompletely neutralized.SO 2− 4 and NH + 4 were two dominant ionic species and NO − 3 was negligible in the acidic aerosols.Three modes of SO 2− 4 were obtained at 0.4, 1.4 and 6.0 µm, respectively (Fig. S4a).The incompletely neutralized ammoniated sulfate aerosols at the 1.4 µm mode were probably ascribed to fog-processing events as discussed above.Compared to SO 2− 4 in the submicron mode, the supermicron models of SO 2− 4 had evidently lower concentrations under acidic conditions.The size distributions and chemical composition in the sample collected on 16-20 June 2004 were almost the same as those collected on 14-15 June 2004, except without an outlier of Ca 2+ at the at the size bin of 0.32-0.54µm (Fig. 5c, d and Fig. S4b).These two samples were expected to have gone through similar fog conditions.

At a coastal site under T < 0 • C condition
Two samples collected at the coastal site (KEJ) likely encountered fog, one sample was collected during 8-9 November 2002 (Fig. 6a, b) and another during 9-10 November 2002 (Fig. 6c, d).The ambient temperature ranged from −2.6 to −8.8 • C during the fist sample period (Table 1) and fog occurred from 5 a.m. to 8 a.m. on 8 November by judging from the measured RH.This fog seemed to be a local event.In this sample, nss-SO 2− 4 and NH + 4 were two dominant ionic species in the particle <3 µm where the RA varied from 0.7 to 0.9, suggesting that the ammoniated sulfate aerosols were incompletely neutralized.The low NH 3 concentration of 0.2 ppb (Table 1) favored formation of incompletely-neutralized sulfate aerosols and inhibited formation of ammonium nitrate.
Four modes of nss-SO 2− 4 were obtained at 0.2, 0.6, 2.9 and 8.0 µm, respectively during the first sample (Fig. S5a), with the dominant mode at 2.9 µm.The two submicron modes can be explained by known mechanism such as condensational growth and cloud processing (Ondov and Wexler, 1998;Zhuang et al., 1999;Yao et al., 2003;Lan Introduction

Conclusions References
Tables Figures

Back Close
Full 4 .Under the acidic condition found in this sample, the above mentioned mechanism cannot explain the dominant mode at 2.9 µm of nss-SO 2− 4 associated with NH + 4 , and the most likely process behind this was fog-processing.Preexisting incompletely-neutralized ammonium sulfate and their mixture with organics in the submicron size could be the major FCN which eventually led to the 2.9 µm mode of nss-SO 2− 4 .Fog processing of aerosols apparently increased the concentration of nss-SO 2− 4 in the atmosphere as evident by the higher nss-SO 2− 4 concentration at the 2.9 µm mode than those at the submicron sizes.Also in this sample, NO − 3 had a dominant mode at 8 µm and a minor condensation mode at 0.2 µm (Fig. S5b).As mentioned earlier, nss-SO 2− 4 also had a minor mode at 8 µm in which nss-SO 2− 4 was apparently associated with NH + 4 .The 8 µm mode of NO − 3 was apparently associated with Na + because 1) the Na + also had a unique mode at 8 µm; and 2) the equivalent ratio of ([NO − 3 ]+[Cl − ])/[Na + ]) was in the range of 1.18-1.22,close to the ratio of [Cl − ])/[Na + ] in sea water, which is 1.17.The fresh/aged sea-salt aerosols could be the major source of the FCN which eventually led to the 8 µm mode aerosols.It was reported that sea spray bubble aerosol could also contain organics and the later could enhance efficiency of CCN (Russell et al., 2009;Moore et al., 2011).On the other hand, the primary sea-salt aerosols usually had a mode at 2-5 µm in mass concentration worldwide.The re-suspended road dust was not likely the major source of the 8 µm mode of Na + because of the negligible Ca 2+ (Fig. 6a) and the wind speed of about 1-2 m s −1 did not favor re-suspension of road dust.Fog droplets derived from sea spray aerosols could scavenge ammoniated sulfate preexisting in the atmosphere and uptake SO 2 and NH 3 gases, leading to nss-SO 4 at the 0.7 µm and 4.5 µm modes were basically the same.Assuming that the 0.7 µm mode ASNA grows into the 4.5 µm mode ASNA, a growth factor of 6.5 is needed and can only be reached under supersatuation conditions.Moreover, fog also appeared to significantly lower concentrations of ASNA in the atmosphere by inferring a substantial decrease of ASNA's concentration at the 4.5 µm mode, comparing their concentrations to the submicron particles.
No fog occurred during the period of 12:40 to 18:10 local time on 6 March 2002 at Egbert (EGB); however, regional fog lasted for four hours before sampling when the ambient temperature was at −4 • C. A mode at 3.1-3.5 µm was observed for ASNA in this sample (Fig. 7c, d and Fig. S8ab).Crustal and sea-salt ions accounted for only 20-30 % of the sum of ([SO Very high concentration of HNO 3 gas (1.7 ppb, Table 1) was observed during the sampling period.The value was even higher than the maximum daily averaged concentration of HNO 3 gas observed in winter in urban areas of North America cities, e.g., Toronto (Godri et al., 2009) and New York (Ren et al., 2006), even under high NO 2 concentration conditions.Photochemical formation of HNO 3 in the gas-phase under such cold conditions and low NO 2 concentrations at EGB should not be able to produce the observed high concentration of HNO 3 gas.Besides, the mass ratio of NO − Introduction

Conclusions References
Tables Figures

Back Close
Full this occasion, fog droplets were probably acidic.This hypothesis was also supported by the RA (0.88) at the 1.8-3.2µm size bin.RAs at other supermicron size bins were in the range of 1.0 and 1.03, which could be explained by the released of HNO 3 that decreased the particle acidity.Thus, the post process of fog processing during this sampling period was a source of HNO 3 gas.
Comparing samples discussed in this section with those in Sect.3.1, it is found that fog processed aerosols under T < 0 • C condition were generally distributed at larger size (e.g., 2-5 µm) than those under T > 0 • C condition (e.g., 1-2 µm).It is speculated that ice could be present in fog-processed aerosols under T<0 • C condition, leading to the 2-5 µm mode.It is noticed that one sample collected on 9-10 December 2004 at SPR with T in the range of −2.9 to 1.7 • C had similar size distribution and composition of fog-processed SO 2− 4 (Fig. 8a, b and Fig. S9) to some of the samples collected under T > 0 • C and acidic conditions.In this sample, fog occurred from 5 p.m. on 9 December to 8 a.m. on 10 December and appeared to be a local event.Ice was not likely present in this particular sample since the temperature was only slightly below zero during part of the sampling period.

Conclusions
Fog-processed aerosols observed at Canadian inland and coastal rural sites were identified and factors determining their size distributions and chemical composition were investigated.The supermicron modes of ASNA were identified as the fingerprint of fogprocessing.Consistent with previous studies, fog processing could lead ammonium salt aerosols to a mode of 1-2 µm.This study further identified that fog processing could also lead ammonium aerosols at modes of 2-5 µm and 5-10 µm.The ammonium salt was in the form of sulfate under acidic condition and in the form of sulfate and nitrate under other conditions.
Temperature was found to be an important parameter in determining size distributions of fog-processed aerosols.When T > 0 • C, size distributions and chemical Introduction

Conclusions References
Tables Figures

Back Close
Full composition of fog-processed aerosols also strongly rely on acidity condition.(1) Under the neutrality condition, ASNA had two modes at the 1-2 µm and 5-10 µm, respectively.ASNA at the 1-2 µm mode was likely the result of preexisting submicron aerosols being activated as FCN which eventually evolved into aerosols.In addition to preexisting submicron aerosols, Ca 2+ -conatined supermicron aerosols could also be activated as FCN, contributing to the 5-10 µm mode of ASNA.Moreover, fog apparently increased concentrations of ASNA in the atmosphere under such condition.
(2) Under the acidic condition, fog-processing of ammoniated sulfate, which was incompletely neutralized, had a mode at 1-2 µm.Preexisting submicron aerosols could be the major contributor of FCN while concentrations of Ca 2+ -conatined aerosols were too low to be important.
The concentrations of SO 2− 4 and NH + 4 at the 1-2 µm were lower than those at submicron size in these acidic samples, suggesting that fog probably played a role in lowering concentrations of these species.
When T < 0 • C, fog-processed nss-SO 2− 4 at a coastal site exhibited a bi-modal size distribution (the dominant mode at 2.8-2.9 µm and a minor mode at 8.0 µm).The nss-SO 2− 4 at the 2.8-2.9 µm mode was incompletely neutralized and was apparently evolved from fog droplets being activated from submicron nss-SO 2− 4 aerosols.The nss-SO 2− 4 at the 8.0 µm mode was completely neutralized and was evolved from fog droplets being activated from fresh/aged sea-salt aerosols.Fog-processed nss-SO 2− 4 at inland sites had a unique mode at 3-5 µm or 1-2 µm where nss-SO 2− 4 sometimes was completely neutralized and sometimes not.At low T , Ca 2+ -contained supermicron aerosols were likely not FCN and the 5-10 µm mode of fog-processed aerosols were absent.
Fog processing was found to modify submicron NH 4 NO 3 aerosols and release a large quantity of HNO 3 gas under acidic conditions.This process substantially lowered the residence time of reactive nitrogen in the atmosphere because of the much higher deposition rate of HNO 3 gas than that of submicron NH 4 NO 3 aerosols, and is worth further investigation.Full  Full Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | droplets, regardless of acidity conditions because of a large quality of liquid water.However, when those fog droplets in low pH evaporated, the formed aerosols could be acidic and contain much less amount of Discussion Paper | Discussion Paper | Discussion Paper | • C conditions.In the T > 0 • C regime, data were further subcategorized into neutrality fog-processed aerosols and acidic fog-processed Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | unity.The higher concentration of Ca 2+ at about 30 µm fog droplets should be due to the direct activation of Ca 2+ aerosols, while any other atmospheric processes would favor higher Ca 2+ at about 10 µm fog droplets.When large fog droplets evaporated into aerosols, the formed aerosols should be large in size and Introduction small number concentrations of fog droplets.The above mechanism explains the observed 1.6-1.7 µm and 8.2-9.0 µm mode aerosols collected on 16-17 November 2004 at SPR.The concentrations of coarse particle NO − 3 are generally much lower than fine particle NO − 2+ -contained aerosols played a negligible role as FCN because of a much lower concentration of Ca 2+ .The same can be said for K + -contained and Na + -contained aerosols.FCN could be mainly activated from preexisting ammoniated sulfate aerosols and/or their mixture with organics in the submicron size.Condensation and coagulation possibly grew the FCN into bi-modal pattern fog droplets and eventually evaporated into bi-modal aerosols.The different at the size range were unimportant.A significant high concentration of sea-salt aerosols was present at 3-5 µm (Fig.4a).Heterogeneous formation of nss-SO 2− 4 or sulfuric acid gas condensation process should have produced higher Discussion Paper | Discussion Paper | Discussion Paper | other metal ions accounted for only a small fraction of the nss-SO 2− heterogeneous reactions between acidic gases and crustal, sea-salt and biomass burning aerosols cannot explain the observed ( mode.The ASNA at this mode was most likely the result of fog processing of aerosols as supported by the discussions below.

Fig. 6 .
Fig. 6.Size distributions of ionic concentrations and RA during the period from 17:50 on 8 November to 10:20 on 9 November 2002 and from 11:10 on 9 November to 10:20 on 10 November 2002 at KEJ (a and b represents the sample collected on 8-9 November; c and d represents the sample collected on 9-10 November).

Fig. 7 .Fig. 8 .
Fig. 7. Size distribution of ionic concentrations, RA and mass ratio of NO − 3 /SO 2− 4 during the period from 11:00 on 19 February 2003-10:30 on 20 February 2003 at ALG and during the period from 11:40 on 16 November to 9:50 on 17 November 2004 at EGB (ab represented the sample collected at ALG; cd represented the sample collected at EGB).