In situ chemical measurement of individual cloud residue particles at a 1 mountain site , South China

Qinhao Lin1,2, Guohua Zhang1, Long Peng1,2, Xinhui Bi1,*, Xinming Wang1, Fred J. 3 Brechtel3, Mei Li4, Duohong Chen5, Ping'an Peng1, Guoying Sheng1, Zhen Zhou4 4 5 1 State Key Laboratory of Organic Geochemistry and Guangdong Key Laboratory of 6 Environmental Protection and Resources Utilization, Guangzhou Institute of Geochemistry, 7 Chinese Academy of Sciences, Guangzhou, 510640, PR China 8 2 University of Chinese Academy of Sciences, Beijing, 100049, PR China 9 3 Brechtel Manufacturing Inc., Hayward, 94544, California, USA 10 4 Atmospheric Environment Institute of Safety and Pollution Control, Jinan University, 11 Guangzhou 510632, PR China 12 5 State Environmental Protection Key Laboratory of Regional Air Quality Monitoring, 13 Guangdong Environmental Monitoring Center, Guangzhou 510308, PR China 14 15 * Correspondence to: Xinhui Bi (bixh@gig.ac.cn) 16 Tel.: +86-20-85290195 17


Introduction
Aerosol-cloud interaction influences the thermodynamic and radiation balance of the atmosphere (IPCC, Boucher et al., 2013).Atmospheric aerosol particles can act as cloud condensation nuclei (CCN) and subsequently affect the chemical and physical properties of cloud droplets, which in turn influence global and regional climate change.The ability of atmospheric aerosol particles to act as CCN, particularly in terms of temporal and spatial variation, may usefully improve estimates of climate change.Anthropogenic particles have been observed to be enriched in the cloud droplets at Schmücke (Roth et al., 2016).
However, a lesser abundance of anthropogenic particles was found in the mixed-phase clouds during the Cloud and Aerosol Characterization Experiment (CLACE 6) (Kamphus et al., 2010).Therefore, it is crucial to assess how atmospheric aerosol particles contribute and respond to the chemical composition of cloud droplets in different regions.
The formation of CCN is dependent on the size and chemical composition of atmospheric aerosol particles at a given supersaturation (McFiggans et al., 2006).A change in the chemical composition of atmospheric aerosol particles during atmospheric aging processes can strongly alter their CCN ability.The presence of hydrophobic surface films lowers the CCN ability of atmospheric aerosols (Andreae and Rosenfeld, 2008).Elemental Atmos.Chem. Phys. Discuss., doi:10.5194/acp-2017-23, 2017 Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 14 February 2017 c Author(s) 2017.CC-BY 3.0 License.carbon (EC) particles, normally considered insoluble, show high CCN activity after mixing with sulfuric acid (Zhang et al., 2008).However, sulfate and nitrate, which are generally regarded as soluble materials, were found in particles ranging from high to low hygroscopicity (Herich et al., 2008).Furthermore, several cloud measurements have pointed to a lower Nf of sulfate in cloud droplets relative to ambient or interstitial particles (Twohy and Anderson, 2008;Pratt et al., 2010a).On the contrary, other study have reported a larger Nf of sulfate in cloud droplets (Roth et al., 2016).These discrepancies suggest that the influence of the mixing state of atmospheric aerosol particles on CCN activity remains unclear.
The combined technique of a counterflow virtual impactor (CVI) and Aerosol Mass Spectrometer (AMS) or single-particle measurement is widely used to characterize the chemical composition and mixing state of individual cloud/fog droplet residue particles.
These studies mainly focus on Europe (Drewnick et al., 2007;Kamphus et al., 2010;Roth et al., 2016;Schneider et al., 2016) and North America (Hayden et al., 2008;Berg et al., 2009;Pratt et al., 2010b;Zelenyuk et al., 2010).Over the past three decades, China has undergone rapid economic growth accompanied by increased aerosol emissions.Scientists have worked to increase our understanding of an emissions inventory and the temporal and spatial variation of atmospheric aerosols in China (Zhang et al., 2012b).However, few studies employ direct observation of the chemical composition and mixing state of cloud/fog droplets.Bi et al. (2016) used a ground-counterflow virtual impactor (GCVI) coupled with a real-time single particle aerosol mass spectrometer (SPAMS) to explore the chemical composition and mixing state of single fog residue particles in an urban area of South China at ground level.They found abundant anthropogenic emitted particles Atmos.Chem. Phys. Discuss., doi:10.5194/acp-2017-23, 2017 Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 14 February 2017 c Author(s) 2017.CC-BY 3.0 License.including soot or element carbon (EC) in fog droplets.Here, we present a study on the chemical composition and mixing state of individual cloud residue particles at a mountain site.The same experimental methods of Bi et al. (2016) were used in this study on the summit of South China's Nanling mountain region.The size distribution, chemical composition and mixing state of cloud residues during cloud events are discussed.
Moreover, the chemical compositions of ambient and interstitial particles were also compared with the cloud residues.The aim of this study is to assess the potential effects of anthropogenic aerosols from regional transportation on cloud formation and to investigate the dominant particle types in cloud droplets at a mountain site in South China.

Measurement site
Measurements were carried out January 15-26, 2016.The sampling site was located in the Nanling Background Station (112° 53' 56" E, 24° 41' 56" N, 1,690 m a.s.l.) at the National Air Pollution Monitoring System in South China (Figure S1).This station is 200 km north of the metropolitan city Guangzhou and 350 km north of the South China Sea.This site is also surrounded by a national park forest (273 km 2 ) where there are hardly any emissions from anthropogenic activities.However, during the winter monsoon period, air pollution from northern China moves south to the southern coastal region and crosses the study region (Lee et al., 2005).
In this study, a GCVI inlet system (GCVI Model 1205, Brechtel Mfg.Inc.) was used to sample cloud droplets with a diameter greater than 8 µm.The sampled cloud droplets were passed through an evaporation chamber (air flow temperature at 40 °C), where the associated water was removed and the dry residue particles, considered CCN, remained.
The enrichment factor of the particles collected by the GCVI inlet was estimated to be 5.25 based on theoretical calculation (Shingler et al., 2012).Ambient particles were collected through an ambient inlet with a cut-off aerodynamic diameter (dva) of 2.5 μm when no cloud events were present.Additionally, interstitial particles were sampled through the ambient inlet during cloud events.The cloud droplet residues, interstitial particles or ambient particles were subsequently analyzed by a suite of aerosol measurement devices, including a SPAMS (Hexin Analytical Instrument Co., Ltd., Guangzhou, China), a scanning mobility particle sizer (SMPS) (MSP Cooperation) and an aethalometer (AE-33, Magee Scientific Inc.).Detailed information and parameter settings regarding the GCVI operation can be found in the work of Bi et al. (2016).Previous studies have found that the average size of cloud droplets in this region was approximately 10 μm, with a corresponding liquid water content of 0.11-0.15g m −3 (Deng et al., 2007).Therefore, it is reasonable to assume that particles larger than 8 µm are cloud droplets.Here, we focus on in situ observations of the size-resolved chemical composition and the mixing state of single cloud residue particles measured by the SPAMS.Meteorological parameters and PM2.5 values at this site were provided by Guangdong Environmental Monitoring Center.
A detailed operational principle of the SPAMS has been described elsewhere (Li et al., 2011).Briefly, aerosol particles are drawn into SPAMS through a critical orifice.The Atmos.Chem. Phys. Discuss., doi:10.5194/acp-2017-23, 2017 Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 14 February 2017 c Author(s) 2017.CC-BY 3.0 License.particles are focused and aerodynamically sized by two continuous diode Nd:YAG laser beams (532 nm).The particles are subsequently desorbed/ionized by a pulsed laser (266 nm) triggered exactly based on the velocity of the specific particle.The positive and negative ions generated are recorded with the corresponding size of each singe particle.
The velocity is related to vacuum dva using a calibration curve created from the measured velocities of a series of polystyrene latex spheres (Nanosphere Size Standards, Duke Scientific Corp., Palo Alto) with predefined sizes.Particles measured by SPAMS mostly fell within the size range of dva 0.2-2.0µm (Li et al., 2011).This makes it impossible to effectively detect particles that exceed such a size range (Figure S2).

Definition of cloud events
To reliably identify the presence of cloud events, an upper-limit visibility threshold of 5 km and a lower-limit relative humidity (RH) threshold of 95% were set in the GCVI software (Bi et al., 2016).Three long-time cloud events occurred during the periods of 16:00 (local time) 15 January -07:00 17 January (cloud I), 20:00 18 January -12:00 19 January (cloud II) and 17:00 19 January -13:00 23 January (cloud III), as marked in Figure 1.In addition, a cloud event occurred during 14:40 -15:00 17 January, but we did not complete an analysis due to the short duration of this cloud event.The average values of cloud droplet concentrations integrated by the SMPS were 218 cm -3 , 284 cm -3 and 272 cm -3 for cloud I, cloud II and cloud III, respectively (Figure S2).Note that during cloud events, RH was close to 100%, as illustrated in Figure 1.Hazy days associated with low visibility were almost completely excluded in this study due to the low level of PM2.5 (~ 12.7 μg m −3 ).A rainfall detector of the GCVI system was also used to exclude rainy droplet Atmos.Chem. Phys. Discuss., doi:10.5194/acp-2017-23, 2017 Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 14 February 2017 c Author(s) 2017.CC-BY 3.0 License.contamination.When cloud events occurred without precipitation, sampling was automatically triggered by the GCVI control software.

Particle classification
During the study period, a total of 73996 sampled particles including 49322 ambient, 23611 cloud residues and 1063 interstitial particles with bipolar mass spectra were chemically analyzed in the size range of dva 0.2-1.9µm.The sampled particles were first classified into 101 clusters using an Adaptive Resonance Theory neural network (ART-2a) with a vigilance factor of 0.75, a learning rate of 0.05, and 20 iterations (Song et al., 1999).By manually combining similar clusters, aged EC, Potassium-rich (K-rich), Amine, Dust, Fe, Pb, Organic carbon (OC), and Sodium-rich (Na-rich), eight major particle types with distinct chemical patterns were obtained, which represented ~99.9% of the population of the detected particles.The remaining particles were grouped together as "Other".
Assuming that number of individual particles follows Poisson distribution, standard errors for number fraction of particle type were estimated (Pratt et al., 2010a).

Back trajectories and meteorological conditions
Back air trajectories in this study were calculated using the Hybrid Single Particle Lagrangian Integrated Trajectory (HYSPLIT Model).During the study period, the station was mainly affected by southwesterly or northerly air masses (Figure 2).The southwesterly air masses, accompanied by warm and moist airflows, occurred during 15-17 and 19-22 January, which promoted cloud formation (Figure 1).Conversely, the northerly air masses, Atmos.Chem.Phys. Discuss., doi:10.5194/acp-2017-23, 2017 Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 14 February 2017 c Author(s) 2017.CC-BY 3.0 License.associated with cool and dry airstreams, occurred during 18 and 23-26 January and led to a decrease in temperature and relative humidity.Meteorological conditions were unstable, with high southwesterly flow (~ 6.5 m s −1 ) during 15-17 and 20-22 January (Figure 1).The level of PM2.5 remained low with a value of approximately 3 μg m −3 for this time period.
A high level of PM2.5 (~20 μg m −3 ) was observed during 18 January when the northerly flow dominated.Similarly, the average PM2.5 value reached 24 μg m −3 during 24-26 January when the local northerly and southwesterly flows occurred alternately.However, the particles still originated from northerly air masses for this period (Figure 2).During 23-24 January, a big freeze associated with a violent northerly flow and a wind speed that exceeded the upper-limit speed (~12 m/s) of a wind speed sensor resulted in a sharp decrease in temperature (Figure 1).

The chemical characterization of cloud droplet residues
Figure 3 shows the average positive and negative mass spectra of nine particle types.The aged EC particles were identified by EC cluster ions (e.g., m/z ±12C +/-, ±36C3 +/-, ±48C4 +/-, ±60C5 +/-, …) and a strong sulfate ion signal (m/z -97HSO4 − ) and some organic markers (m/z 27, 37).The aged EC particle type was the largest fraction (49.3% by number) of the cloud residues (Figure 4).In addition, the number fraction (Nf) of aged EC particles in the cloud residues significantly decreased while size increased (Figure S3).The K-rich particles exhibited the highest peak at m/z 39K + , mainly combined with sulfate and nitrate (m/z -46NO2 − , -62NO3 − ).The K-rich particles presumably resulted from biomass/biofuel burning (Moffet et al., 2008;Zhang et al., 2013).The K-rich particle type, the second largest contributor, accounted for 33.9% by number of the cloud residues (Figure 4).Aged EC and K-rich particles mainly originated from combustion processes (Andreae and Rosenfeld 2008;Bond et al., 2013).The Nanling mountain sampling site does not contain any sources of anthropogenic emissions; thus, the abundant aged EC and K-rich particles in cloud residues are expected to come from regional transportation.
The Amine particles represented 3.8% by number of the cloud residues (Figure 4), mainly within the size fraction of 0.7-1.9µm (Figure S3).A recent study also showed a low fraction (<10% by number) of amine species in the cloud residues (Roth et al., 2016).It has been reported that in-cloud/fog processing could enhance amine species (Rehbein et al., 2011;Zhang et al., 2012a).However, this possibility was not supported by the observations of Bi et al. (2016), who did not detect amine-containing particles in the fog residues.In this study, the Nf of the Amine particles varied from 0.2% to 15.1% of the cloud residues dependent on air mass history (see Sect. 3.4).
The Dust particles presented significant ions at m/z 40Ca + , 56CaO + /Fe + , 96Ca2O + and -76SiO3 − , with an internal mixture of sulfate and nitrate.This type contributed 2.9% by number of the cloud residues (Figure 4).Dust/mineral aerosol accounted for approximately 35% of the total aerosol mass in China (Zhang et al., 2012b).Approximately 12% by number of fog contained dust particles at ground level in South China (Bi et al., 2016).At Mt. Taishan in northern China, a high concentration of Ca 2+ in cloud/fog water was mainly attributed to a sandstorm event (Wang et al., 2011).In this study, a low fraction (2.9% by number) of dust cloud residue suggests that dust particles did not play a significant role in cloud formation in South China or that they occupied larger CCN (Tang et al., 2016), which cannot be detected by the SPAMS.
The Fe and Pb particles had their typical ions at m/z 56Fe + and 208Pb + , respectively, and were associated with sulfate and nitrate.The Fe and Pb particles made up 4.1% and 0.5% by number of the cloud residues, respectively (Figure 4).The presence of Fe in the cloud droplets might play an important role in aqueous-phase SO2 catalytic oxidation in cloud processing (Harris et al., 2014), thus accelerating the sulfate content of Fe-containing particles in cloud processing.
The Na-rich particles were mainly composed of ion peaks at m/z 23Na + and 39K + in the positive mass spectra, and inorganic soluble nitrate and sulfate species in the negative mass spectra.Moffet et al. (2008) attributed Na-rich particles to varied sources of industrial emissions or sea salt particles and dry lake beds.The OC particles presented dominant intense OC signals (e.g., m/z 27C2H3 + , 37C3H + , 43C2H3O + and 51C4H3 + ) and abundant sulfate.The Na-rich and OC types contributed 3.0% and 2.4% by number to the cloud residues, respectively (Figure 4).Internally mixed EC with metal signatures was observed in the Other particles.However, Other particles contributed only 0.1% by number to the cloud residues, which suggests their minor contribution to cloud formation (Figure 4).

Mixing state of secondary species in cloud residues
Particles that coated with inorganic species (e.g., sulfate, nitrate and ammonium) can facilitate water uptake to growth into cloud droplets (Andreae and Rosenfeld, 2008).
Number fractions of sulfate were found to be highly related to the K-rich (91%), OC (100%), aged EC (98%), Pb (74%), Fe (93%) and Amine (99%) cloud residues, as shown in Figure 5. Lower number fractions of sulfate were observed in the Na-rich (41%) and Dust (42%) cloud residues.In contrast, nitrate contributed 89% and 88% by number to the Na-rich and Dust cloud residues, respectively.The heterogeneous and/or aqueous chemistry of HNO3 in the Na-rich and dust particles may lead to the preferential enrichment of nitrate (Li and Shao, 2009).The detection of nitrate in the cloud residues was thought to be the form of ammonium nitrate by estimating the ratio of m/z 30 to m/z 46 in AMS data (Drewnick et al., 2007;Hayden et al., 2008).However, low portions of ammonium in the Na-rich (23% by number) and Dust (15% by number) cloud residues suggest that ammonium nitrate is not a predominant form of nitrate in these cloud residue particle types.
Note that the evaporation chamber of the GCVI may lead to a reduction of ammonium nitrate in the cloud residues (Hayden et al., 2008;Prabhakar et al., 2014).However, this effect would be insignificant because the dry carrier air of the GCVI was set at 40 °C.A volatility study found that the temperature to evaporate ammonium nitrate particles reached at least 75 °C (Bi et al., 2015).We found that nitrate accounted for only 46% by number of the aged EC cloud residues, which is significantly less than the contribution of sulfate.
Previous studies found that aged EC (soot) fog/cloud residues are mainly internally mixed with sulfate (Pratt et al., 2010a;Harris et al., 2014;Bi et al., 2016).The presence of abundant sulfate in aged EC cloud residues was considered to be a good CCN species before activation, rather than formed by in-cloud processing (Bi et al., 2016;Roth et al., 2016).High portions (75-86% by number) of ammonium were observed for the OC and EC cloud residues, suggesting that ammonium plays a key role in cloud processes for the two cloud residue types.Organics (e.g., amine and oxalate) have previously been measured in cloud residues (Sellegri et al., 2003;Sorooshian et al., 2007b;Pratt et al., 2010a).Amine and oxalate particles with mixtures of inorganic salts could enhance water uptake behavior (Sorooshian et al., 2008;Wu et al., 2011).Enrichment of TMA (93% by number) in the Amine cloud residues is expected to promote water uptake in sub-and supersaturated regimes (Sorooshian et al., 2007a).A total of 3,410 oxalate-containing (m/z, -89HC2O4 − ) particles represented 14.4% of the cloud residues by number, which was mainly associated with the K-rich cloud residues including 2,144 oxalate particles.The oxalate in the K-rich cloud residues is likely attributed to biomass burning, which facilitates the CCN ability of biomass-burning particles due to the hygroscopic property of oxalate (Pratt et al., 2010a).
Relative high portions (~30% by number) of oxalate in the metal (Pb, Fe) cloud residues might be the form of metal oxalate complexes from reactions of in-cloud formation oxalate with metals (Furukawa and Takahashi, 2011).Oxalate can readily partition into the particle phase to form amine salts (Pratt et al., 2009).This may result in 33% by number to the Amine residues containing oxalate.As mentioned above, the Nf of the cloud residue types significantly changed as the air mass origin varied from northerly to southwesterly.To further investigate the influence of air mass history, we selected cloud residues that had arrived from a northerly air mass on 19 January and compared these to cloud residues originating from a southwesterly air mass during the periods of 16-17 and 21-22 January.The detected number of cloud residues for the northerly and southwesterly air masses are given in Table 1.Note that southwesterly air mass accompanied by high relative humidity (>90%) (Figure S4) may have triggered particles activated to CCN prior to their arrival to the sampling site.

Comparison of cloud residues in different air mass sources
The K-rich type was found to contribute 23.9% to the cloud residues in the northerly air mass, which was significantly lower than its contribution to the southwesterly air mass (51.5%), as summarized in Table 1.The considerable increase of K-rich cloud residues suggests a major influence of regional biomass-burning activities.Biomass-burning emissions from Southeast Asia, including Myanmar, Vietnam, Laos and Thailand, where abundant fire dots are observed (Figure 2), could have been transported to the sampling site under a southwesterly air mass (Duncan et al., 2003).In contrast, the aged EC type represented only 23.7% of the cloud residues under the influence of a southwesterly air Atmos.Chem. Phys. Discuss., doi:10.5194/acp-2017-23, 2017 Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 14 February 2017 c Author(s) 2017.CC-BY 3.0 License.mass, which was significantly lower than observations for the northerly air mass (59.9%).
This result suggests that the northern air mass has a greater influence on the presence of aged EC cloud residues.
In addition, an obvious increase in Nf of the Amine type was observed in the southwesterly air mass (15.1%) compared to the northerly air mass (0.2%).This implies that the sources or formation mechanisms of amine in cloud residues varied in different air masses.The southwesterly air mass arrived from as far as the Bay of Bengal and then travelled through Southeast Asia before reaching South China (Figure 2).The potential gas amine emissions from ocean (Facchini et al., 2008) and livestock areas (90 million animals, data was available at the website http://faostat3.fao.org) in Southeast Asia might promote the enrichment of amine particles.Note that after the activation of amine particles, the partitioning of the gas amine on cloud droplets may further contribute to the enhanced Amine cloud residues (Rehbein et al., 2011), especially for air masses delivered via routes with high relative humidity, as mentioned above (Figure S4).In contrast, northerly air mass accompanied with dry airstreams may inadequately induce the partitioning of gas amines into the particle phase (Rehbein et al., 2011).

Comparison of cloud residues with ambient and interstitial particles
A direct comparison between cloud residues and ambient particles was limited because of their differences in air mass origins.During the sampling period, the cloud events occurred once the southwesterly air masses were dominant.Therefore, a comparison between cloud residues and ambient particles cannot be addressed under the influence of southwesterly air masses.Here, we chose five hours before and after the beginning of the cloud II period Atmos.Chem. Phys. Discuss., doi:10.5194/acp-2017-23, 2017 Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 14 February 2017 c Author(s) 2017.CC-BY 3.0 License. in order to compare cloud residues and ambient particles with similar northerly air mass origins, as discussed in Sect 3.4.The time, detected number and Nf of ambient particles for this comparison are listed in Table 1.From 10:00 21 January to 13:00 23 January, the particles were manually switched in an hourly cycle between the CVI and ambient inlets during the cloud III period to provide information on cloud residues and interstitial particles.
The number and Nf of particle types in the cloud residues and interstitial particles are given in Table 2.Note that air mass origin shifted from southwesterly to northerly during 22-23 January.
Table 1 shows that the contribution of K-rich particles in cloud residues slightly decreased relative to ambient particles (23.9% versus 30.7%), which may be due to the small size distribution of K-rich in ambient particles (Figure S5).A slight increase in the aged EC cloud residues was attributed to the decreased K-rich cloud residues.The remaining particle types showed no clear differences between cloud residues and ambient particles.A comparison of cloud residues and ambient particles may yield information on particle's CCN activity due to the in-cloud processing effect.Rather, differences between cloud residues and interstitial particles can better reflect whether particles become activated (Zelenyuk et al., 2010).However, few studies have focused on this issue, in part because interstitial particles show a smaller size than that detected by single-particle mass spectrometry (Roth et al., 2016).In comparing the cloud residues with the interstitial particles, a significant change in Nf was found for the aged EC and K-rich type.A higher Nf of K-rich particles and a lower Nf of EC particles were found for the cloud residues relative to the interstitial particles (Table 2).Aged EC particles may require very high Atmos.Chem.Phys. Discuss., doi:10.5194/acp-2017-23, 2017 Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 14 February 2017 c Author(s) 2017.CC-BY 3.0 License.supersaturation to grow into cloud droplets and thus only form interstitial hydrated aerosol (Hallberg et al., 1994).
Nitrate intensity was found to be enhanced for the cloud residues relative to the ambient particles, as shown in Figure 7. Drewnick et al. (2007) suggested that, high nitrate, rather than sulfate, content in pre-existing particles preferentially acted as cloud droplets.Hayden et al. ( 2008) argued that a high nitrate content in cloud residues resulted from the uptake of HNO3 gas during the cloud process and estimated that the absorption of HNO3 gas has the increment of 100-200 nm nitrate cloud residues.However, this is not likely to be the dominant source of 300-500 nm nitrate cloud residues in this study (Figure S6).The enhancement of nitrate in cloud residues may be explained by pre-existing particles before activation, rather than in-cloud nitrate formation.Interestingly, we observed a decrease in nitrate intensity in cloud residues (Figure 8) and a large size distribution of nitratecontaining cloud residues compared with the interstitial particles (Figure S7).This result suggests that particle size, rather than nitrate content, plays a more important role in the activation of particles into cloud droplets.
Sulfate intensity increased for the aged EC and OC cloud residues, while it decreased for the Dust, Na-rich cloud residues compared with both ambient and interstitial particles.
Although the in-cloud addition of sulfate occurred by an aqueous Fe-catalyzed reaction (Harris et al., 2013), sulfate was observed to diminish in the Fe cloud residues relative to ambient particles.Compared with interstitial particles, sulfate enhanced in the Fe cloud residues.In a similar comparison of cloud residues with interstitial particles, ambient particles were observed for the K-rich type.Previous studies also showed that the mass or number fraction of sulfate in the cloud residues changed between ambient and interstitial Atmos.Chem. Phys. Discuss., doi:10.5194/acp-2017-23, 2017 Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 14 February 2017 c Author(s) 2017.CC-BY 3.0 License.particles (Pratt et al., 2010a;Hao et al., 2013;Schneider et al., 2016).However, the reason for this discrepancy remains unclear.
No remarkable change in organic signals between cloud residues and ambient interstitial particles was obtained for the different particle types.The in-cloud process was an important pathway for the production of amine particles (Rehbein et al., 2011;Zhang et al., 2012a).In this study, no significant enhancement of the Amine cloud residues was obtained relative to the ambient particles (Table 1).Bi et al. (2016) considered that the absence of amine species in fog residues may be partially affected by droplet evaporation in the GCVI.
We did find a high fraction of the Amine cloud residues when the southwesterly air mass prevailed, as discussed in Sect 3.4.Therefore, the effect of amine volatilization in the GCVI on the reduction of the Amine cloud residues is likely an unimportant factor in this study.
A lack of gas-phase amines may be the cause of few amine particles detected in the ambient particles and cloud residues (Rehbein et al., 2011).

Comparison with previous studies on cloud/fog residues
Our finding can be compared with previous observations of cloud residues in various environments including mountain sites (Kamphus et al., 2010;Roth et al., 2016) and aircraft measurement (Zelenyuk et al., 2010).In these studies, cloud residues showed a larger size distribution relative to ambient and/or interstitial particles, although in-cloud processes may modify the size distribution of cloud residues.Cloud residues also exhibited discrepancies in particle types.The aged EC particles in the stratocumulus cloud residues were negligible at an altitude of 2-3 km over Alaska (Zelenyuk et al., 2010).In another study, Pratt et al. (2010a) observed the abundant soot (~19% by number) and biomass Atmos.Chem. Phys. Discuss., doi:10.5194/acp-2017-23, 2017 Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 14 February 2017 c Author(s) 2017.CC-BY 3.0 License.burning (~43% by number) mix-phase cloud residues at an altitude of 5-7 km over Wyoming.High Nf of soot (~30%) and biomass burning (~25%) orographic cloud residues were also observed at a mountain site Schmücke (937 m a.s.l.) in central Germany (Roth et al., 2016).However, at Jungfraujoch station (3580 m a.s.l.) in Europe, the K-rich (biomass burning) particles was only found to contribute 3% of the mix-phase cloud droplets and the aged EC cloud residue was insignificant (< 1% by number) (Kamphus et al., 2010).At a ground site in Guangzhou city, aged EC particles contributed up to 67.7% of fog residues by number (Bi et al., 2016).In this study, aged EC and K-rich particles dominated the cloud residues.We also found no distinct change in the Nf of aged EC and K-rich particles in cloud residues relative to ambient particles, which was consistent with the previous observation of the mix-phase cloud condition (Pratt et al., 2010a).However, Roth et al. (2016) reported a higher Nf of aged soot particles in orographic cloud residues rather than ambient particles, but no clear difference between cloud residues and ambient particles for the biomass burning particle type.This disagreement between studies may suggest that the ability of particle types to form cloud droplets strongly varies depending on geographic location rather than cloud type and altitude.

Conclusions
This study presented an in situ observation of individual cloud residues, interstitials and ambient particles at a mountain site in South China.We found that the largest fraction of cloud residues was the aged EC type (49.3%), followed by K-rich particles (33.9%).The remarkable change in Nf of the cloud residue types influenced by varied air masses highlights the important role of regional transportation in the observed cloud residue Atmos.Chem.Phys.Discuss., doi:10.5194/acp-2017-23,2017   Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 14 February 2017 c Author(s) 2017.CC-BY 3.0 License.

Figure 6
Figure 6 displays hourly average unscaled counts and Nf values of nine types of cloud residues and ambient particles.During 18-19 January, the cloud residues and ambient particles showed similar chemical characteristics and were dominated by aged EC particles.A lack of significant variation in the Nf of particle types for this period suggests that the original particles did not change.Based on a backward trajectory, air masses changed from northerly on 18 January to southwesterly on 19 January (Figure2), consistent with variation Atmos.Chem.Phys.Discuss., doi:10.5194/acp-2017-23,2017   Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 14 February 2017 c Author(s) 2017.CC-BY 3.0 License.

Figure 1 :
Figure 1: The hourly average variations in meteorological conditions (temperature, relative humidity, visibility, pressure, wind speed and direction) and PM2.5.

Figure 2 :
Figure 2: HYSPLIT back trajectories (72 h) for air masses at 1,800 m during the whole sampling period.The black and red lines refer to northerly and southwesterly air masses, respectively.The yellow rots represented the fire dots during the study periods.The fire dots are available at https://earthdata.nasa.gov/.

Figure 3 :
Figure 3: Averaged positive and negative mass spectra for the 9 particle types (Aged EC, K-rich, Amine, Dust, Fe, Pb, Na-rich, OC, and Other) of the 73996 sampled particles during the whole sampling period.RPA in the vertical axis refers to relative peak area.m/z in the horizontal axis represents mass-to-charge ratio.

Figure 4 :
Figure 4: Number fraction of the cloud residual types during the whole sampling period.

Figure 5 :Figure 6 :
Figure 5: Mixing state of secondary markers with the cloud residue particle types.

Figure 7 :Figure 8 :
Figure 7: Difference between mass spectra for the cloud residues and ambient particles.

Table 2 .
Particle number and number fraction of cloud residues and interstitial particles 644 using a manually switched way during the cloud III period.Atmos.Chem.Phys.Discuss., doi:10.5194/acp-2017-23,2017 Manuscript under review for journal Atmos.Chem.Phys.Discussion started: 14 February 2017 c Author(s) 2017.CC-BY 3.0 License.