Investigating three patterns of new particles growing to cloud condensation nuclei size in Beijing’s urban atmosphere

Investigating three patterns of new particles growing to cloud condensation nuclei size in Beijing’s urban atmosphere Liya Ma, Yujiao Zhu *, Mei Zheng, Yele Sun, Lei Huang, Xiaohuan Liu, Yang Gao, Yanjie Shen, Huiwang Gao 5 and Xiaohong Yao * Lab of Marine Environmental Science and Ecology, Ministry of Education, Ocean University of China, Qingdao 266100, 5 China Environment Research Institute, Shandong University, Qingdao, Shandong 266237, China State Key Joint Laboratory for Environmental Simulation and Pollution Control, College of Environmental Sciences and Engineering, Peking University, Beijing 100871, China State Key Laboratory of Atmospheric Boundary Layer Physics and Atmospheric Chemistry, Institute of Atmospheric Physics, 10 Chinese Academy of Sciences, Beijing 100029, China Laboratory for Marine Ecology and Environmental Sciences, Qingdao National Laboratory for Marine Science and Technology, Qingdao, China Correspondence to: Xiaohong Yao and Yujiao Zhu (*xhyao@ouc.edu.cn, zhuyujiao@sdu.edu.cn) Abstract. The growth of newly formed particles with diameters from ~10 nm to a larger size was investigated in Beijing’s 15 urban atmosphere on 10-23 December 2011, 12-27 April 2012 and through June-August 2014. The maximum geometric median diameter (Dpgmax) of newly formed particles in 11/27 new particle formation (NPF) events through June-August exceeded 75 nm, and the grown new particles may contribute to the population of cloud condensation nuclei. In contrast, no apparent growth in new particles with Dpgmax<20 nm was observed in all of the events occurring in December, in approximately half of the NPF events occurring in April and only 2/27 of the NPF events occurring in June-August. New particles observed 20 in the latter NPF events were too small to be activated as cloud condensation nuclei. Apparent new particle growth with Dpgmax50 nm was observed in the remaining NPF events. The 11/27 NPF events with Dpgmax exceeded 75 nm were thereby analyzed in-depth. They are clearly three particle growth pattern classifications: one-stage growth, which is characterized by a continuous increase in Dpgmax80 nm (4/11 NPF events), and two-stage growth-A and growth-B, which are characterized by either no apparent growth (two-stage growth-A) or a shrinkage of particles (two-stage growth-B) in the middle 2-4 hours of 25 the growth period (7/11 NPF events). Combining the observations of gaseous pollutants and the measured (or modeled) concentrations of particulate chemical species, the three growth patterns were discussed in terms of the spatial heterogeneity of NPF, the formation of secondary aerosols and the evaporation of semi-volatile particulates. Secondary organic species and NH4NO3 were argued to be two major contributors to the growth in new particles, but NH4NO3 likely contributed to growth only in the late afternoon and/or at nighttime. 30

NPF events were identified according to the definition by Dal Maso et al. (2005), and only NPF events with durations over one hour were analyzed in this study. In each NPF event, the net maximum increase in the nucleation mode particle number concentration (NMINP) was calculated according to Zhu et al. (2017). The nucleation mode was defined from 8 to 20 nm in this study.

25
NMINP= N8-20 nm (t1)-N8-20 nm (t0) (1) N8-20 nm represents the sum of particle number concentrations with diameters from 8 nm to 20 nm, t0 and t1 represent the time of an NPF event to be initially observed and the time when the N8-20 nm arrives at the maximum value, respectively.
The growth rate (GR) and shrinkage rate (SR) of new particles are determined by the slope of the fitted geometric median diameter of new particles (Dpg) with time (Whitby et al., 1978;Yao et al., 2010;Zhu et al., 2014;Man et al., 2015). In an NPF 30 event or in each growth period of one NPF event, the maximum value of Dpg is defined as Dpgmax.
The units for [H2SO4] and [SO2] are molecule cm -3 , and the unit for UVB (ultraviolet B) is W m -2 . UVB occupies 5% of the ultraviolet radiation that reaches the Earth's surface (https://en.wikipedia.org/wiki/Ultraviolet#cite_note-Skin_Cancer_Foundation-23). Thus, UVB values were obtained by multiplying the downward ultraviolet radiation at the surface by 5% in this study, and the ultraviolet radiation was downloaded from https://cds.climate.copernicus.eu/. The contribution of sulfuric acid vapor to particle growth was calculated based on the method reported by Kulmala et al. (2001).
When Dpg of the grown new particles just reached 50 nm, the survival probability ratio (SPR) of grown new particles at Dpg = 50 nm was estimated as: where σ represents the standard deviation of the median diameter in the fitted log-normal distribution of grown new particles.
N50+3σ refers to the integral value of the number concentration of new particles with the diameter from 50 nm to 50 + 3σ nm.
We further defined another technical term, i.e., two times the SPR (2*SPR). The final SPR was between SPR and 2*SPR since some new particles with diameter from 50 -3σ nm to 50 nm may eventually grow over 50 nm with an increase in Dpg. However, 15 how many new particles with diameter from 50 -3σ nm to 50 nm can grow over 50 nm varies case by case. The same can be said for the final SPR. Similar definitions are applied for the SPR of grown new particles with Dpg reaching over 70 nm.
The amount of chemical species required to grow new particles from Dpg1 to Dpg2 (Massrequried) is roughly estimated as below (Man et al., 2015;Zhu et al., 2014): Massrequried = 4/3 [(Dpg2/2) 3 -(Dpg1/2) 3 ]*N ̅ *  (4) 20 N ̅ represents the average integral value of new particle number concentrations with the particle size from Dpg1 to Dpg2;  is the density, which is assumed as 1.5 µg m -3 for SOA and 1.7 µg m -3 for NH4NO3, respectively. Local standard time was used to describe the NPF events in this paper.

Overview of NPF events in three campaigns
A total of 46 NPF events were observed during three campaigns in Beijing, and the occurrence frequencies of NPF events decreased clearly in the rainy season (  (Wu et al., 2007).
The NMINP varied largely from event to event in the five months, but the monthly averages were generally closer to each

Season-dependent growth patterns of newly formed particles
When the growth behaviors of newly formed particles were studied, three growth patterns, i.e., Classes I, II and III, were identified . Class I was characterized by no apparent particle growth. For example, the fitted Dpg of new particles was almost constant at 11 nm for ~10 hours on 25 April 2012 until the new particle signal dropped to a negligible level ( Table 1, Fig. S3a). Class II was characterized by the fitted Dpg of new particles growing from 10±2 nm to 20-50 nm, e.g.,

5
in Fig. S3b-e. Class II can be further subclassified into four scenarios. In Scenario 1, the new particle growth lasted for a few hours with Dpg increasing to 27-48 nm and then stopping (Fig. S3b). The stop lasted for a few hours until the new particle signal dropped to a negligible level; in Scenario 2, new particles grew with Dpg approaching 32-45 nm. Afterwards, the signal of the new particles was apparently replaced by another signal of the new particles with an obviously smaller diameter (Fig.   S3c). The phenomenon could also be explained by the spatial heterogeneity of NPF, as discussed later, and may not represent 10 two NPF events occurring in one day. In Scenario 3, new particles grew with Dpg increasing to 20-50 nm, and the new particle signal was then overwhelmed by aged plumes. In the half or one hour transient period from new particle signals to aged plume signals, Dpg rapidly increased by dozens of nanometers ( 2-4 and Figs. S4-6). Class III can be further classified into three growth patterns, which will be detailed later in this study.
Overall, the Dpgmax of grown new particles increased from Class I to III.
In December, all of these observed NPF events (three NPF events at the rooftop site plus seven NPF events at the street 20 site) were subject to Class I ( 0.2%. The new particle growth behaviors in Class III NPF events were thus deeply studied, and the survival probabilities of the particles with Dpgmax at 50 nm and 70 nm are also estimated below.

Three growth patterns of newly formed particles reaching CCN size
When the observational results in June, July and August 2014 were examined separately, the occurrence frequencies of Class III NPF events in the three months were very close to each other, i.e., 4, 3 and 4 in June, July and August, respectively.

5
The 11 NPF events can be further classified into three particle growth patterns, i.e., one-stage particle growth, two-stage particle growth-A and two-stage particle growth-B, as detailed below: The one-stage particle growth pattern occurred in 4/11 NPF events, in which the fitted Dpg of newly formed particles continuously increased from 11 nm to 80-100 nm in 6-17 hours on 18 June, 12-13 July and 25 August 2014 ( Fig. 2, Fig. S4).
The new particle growth stopped at ~24:00 in 3/4 NPF events, while the new particle growth stopped as early as ~16:00 in the 10 last event.
The two-stage particle growth-A pattern also included 4/11 NPF events, in which the Dpg of newly formed particles increased from 9-22 nm to 23-69 nm in the daytime, then oscillated for several hours, and eventually restarted the increase at nighttime (Fig. 3, Fig. S5). In 2/4 events, the Dpg of newly formed particles stopped the increase for 2~3 hours in the middle period and then restarted the increase up to 75 nm at 22:00. In the other 2/4 events, the Dpg stopped the increase for ~4 hours 15 in the middle period and then restarted the increase up to 110-115 nm at 1:00 the next day.
A total of 3/11 NPF events experienced the two-stage particle growth-B, in which the Dpg of newly formed particles increased from 10-19 nm to 36-79 nm, then decreased down to 24-50 nm in the next 2-4 hours, and eventually restarted the increase with the Dpg up to 84-120 nm (Fig. 4, Fig. S6). In two events, the decrease in newly formed particles occurred at approximately 18:00, e.g., at 18:00-21:22 on 23 June and 17:50-20:30 on 26 July. However, the shrinkage occurred as early 20 as 15:20-17:20 on 11 June.

One-stage new particle growth to CCN size
Among four one-stage growth NPF events, newly formed particles took the shortest time to reach the maximum size on 18 June 2014 (Fig. 2a) of Ox (NO2+O3) largely increased from ~ 60 ppb to ~ 130 ppb during the growth period, supporting the photochemical formation of secondary species to drive particle growth. Based on the observed mixing ratio of SO2, sulfuric acid was estimated to contribute < 2% to the period particle growth (Fig. 2b). Wiedensohler et al. (2009), Ehn et al. (2014 and  proposed that the photochemical formation of oxidized organic compounds played a more important role than the sulfuric acid vapor in growing new particles with diameters >10 nm in the daytime on basis of field measurements, laboratory experiments 5 and modeling results. Assuming that the particle growth in this period was completely driven by secondary organic compounds, the required amount of SOA was estimated as 13 μg m −3 . The observed concentration of OM (including SOA and POA) in PM1.0 increased by 15 μg m -3 during the five-hour growth period. The SOA seemingly acted as a major contributor to particle growth in this period, as supported by the decrease in the hygroscopicity parameter of 50 nm atmospheric particles from ~ 0.3 to ~ 0.1 during the same event independently reported by Wu et al. (2016). Almost constant concentrations of NO3and NH4 + 10 observed at 11:00-14:00, implied that no ammonium nitrate was freshly formed and contributed to the particle growth before 14:00 (Fig. 2c). From 14:00 to 16:00, the concentrations of NO3and NH4 + largely increased. Assuming the increase in NO3due to formation of NH4NO3, the net increase in NH4NO3 was 10 μg m -3 . Thus, formation of NH4NO3 may also play an important role in growing new particles after 14:00. Zhu et al. (2014) and Man et al (2015) reported that ammonium nitrate can be an important contributor to the growth in new particles (from 40-50 nm to a larger size at night). The new particles 15 stopped growth after 15:54 until the new particle signal gradually disappeared at ~20:00. The observed concentrations of OM and NO3had no increasing trends during the four hours, although they largely oscillated.
Another example subject to one-stage growth occurred on the event on 25 August 2014, and newly formed particles took the longest time to reach Dpgmax (Fig 3e). RH was lower than 50%, and the ambient air temperature varied from 24℃ to 31℃ during the growth period (Fig. 2h), also indicating dry and hot conditions during the particle growth period. The NPF event 20 was observed from 07:50 on 25 August 2014 to 08:00 the next day (Fig. 2e). The new particle signal was unstable in the initial three hours due to the spatial heterogeneity of NPF.
The Dpg of newly formed particles started an increase from 12 nm at 10:48 to 80 nm at 24:00 with a particle growth rate . Thus, the modeled concentrations of SOA were considered as a semiquantitative estimation to argue their contributions to growing new particles >10 nm. The modeled concentrations of NO3and NH4 + were almost constant at 11:00-18:00 (Fig.   2g), suggesting that no NH4NO3 was newly formed to drive particle growth.

5
The modeled concentrations of SOA increased rapidly from 0.56 μg m -3 at 18:00 to 1.24 μg m -3 at 22:00 (Fig. 2g), with the Dpg of newly formed particles increasing from 47 nm to 70 nm. A total of 3.3 μg m −3 of SOA was required to drive the particle growth in this period if SOA were the only contributor. The modeled net increase in particulate ammonium nitrate was 3.6 μg m −3 from 18:00 to 22:00 (Fig. 2g). Assuming that the new particle growth from 18:00 to 22:00 was completely driven by ammonium nitrate, the required amount was estimated to be 3.8 μg m −3 . Ammonium nitrate may yield an important 10 contribution to the particle growth in this period since its net increase apparently satisfied the required amount, to some extent.
The Dpg of newly formed particles increased from 70 nm to ~80 nm from 22:00 to 24:00 when the modeled concentrations of all species decreased due to the dilution effect. Afterwards, the new particles stopped growing until their signal gradually disappeared at 08:00 the next day. The modeled concentrations of SOA and ammonium nitrate were almost constant after 1:00 the next day, consistent with the lack of apparent growth in these large new particles.

15
During the two NPF events on 12 and 13 July, sulfuric acid vapor was estimated as a minor contributor to particle growth ( Fig. S4). The modeled results suggested that both ammonium nitrate and SOA were important contributors to particle growth, but ammonium nitrate contributed to growth only at nighttime (Fig. S4c, Fig. S4g). However, the concentrations of chemical species in different sized nanometer particles need to confirm this.  Fig. 3a-d), the NPF events started to be observed at 09:00 and lasted for 18 h, with the RH generally lower than 40%. Apparent growth also cannot be observed for newly formed particles from 09:00 to 10:30. The Dpg of newly formed particles increased from ~10 nm at 10:30 to 35 nm at 15:20, with a GR of 5.2 nm h -1 . Using the observed mixing ratio of SO2, sulfuric acid vapor was estimated to contribute 3% to the first stage particle 25 growth (Fig. 3b). The constant concentrations of NO3observed during the period implied that no ammonium nitrate was freshly formed and contributed to the period of particle growth (Fig. 3c). Assuming that the period particle growth was completely driven by SOA, the required amount of SOA was estimated as 0.64 μg m −3 . The observed OM dramatically varied at that period (Fig. 3c).

Two-stage new particle growth-A to CCN size
After 15:20, the Dpg of newly formed particles stopped the growth and fluctuated at approximately 35 nm for 30 approximately two hours. The first stage particle growth apparently encountered an upper limit. Compared with the https://doi.org/10.5194/acp-2019-1151 Preprint. Discussion started: 4 May 2020 c Author(s) 2020. CC BY 4.0 License.
concentrations observed before and after the two-hour period, the largely decreased number concentrations of newly formed particles implied spatial heterogeneity of NPF on that day, i.e., much weaker atmospheric nucleation generated newly formed particles in the upwind atmosphere at a certain spatial range, and the grown new particles at a lower number concentration were transported and observed at rooftop site at 15:20-17:40. The slightly decreased mixing ratios of Ox during this time, which were unexpected on the basis of a sharp increase in the observed Ox after the period, implying the reduced photochemical 5 reaction activities in the upwind atmosphere at certain spatial ranges. The photochemical reaction activities during the period may be too weak to generate sufficient amounts of secondary organic and inorganic precursors in supporting the growth of new particles >35 nm to a larger size and subsequently lead to the growth encountering the upper limit, liking the diagram in the graph abstract.
After 17:40, the Dpg of newly formed particles started the second stage increase from 32 nm to 75 nm at 22:30, with a GR 10 of 9.7 nm h -1 , which nearly doubled the growth rate observed during the first growth stage. The observed mixing ratio of Ox increased from 66 ppb at 17:20 to ~ 90 ppb at 21:20, supporting the secondary formation of chemical species to drive particle growth (Fig. 3b). The observed concentrations of OM and NO3rapidly increased from 18:00 to 21:00 while the former was one order of magnitude larger than the latter. The required amount of NH4NO3 for the period particle growth was estimated as Following the analysis mentioned above, freshly formed SOA were argued to dominantly drive the first-stage particle 20 growth on 6 August 2014 (Fig. 3e), on 12 and 15 August 2014 (Fig. S5). On the other hand, newly formed SOA and NH4NO3 were likely to be two major contributors to second-stage particle growth. Again, large uncertainties on modelled concentrations may exist due to lack of direct measurements of chemical species in different sized nanometer particles.

Two-stage new particle growth-B to CCN size
Among the three two-stage growth-B NPF events, the longest shrinkage in grown new particles occurred on 23 June 2014 25 (Fig. 4a) and lasted approximately 4 hours. On that day, our analysis implied the first stage particle growth to be driven by SOA. In detail, the Dpg of newly formed particles increased from 17 nm at 11:20 to 79 nm at 17:20, with a GR of 10 nm h -1 .
From 11:20 to 17:20, the mixing ratio of Ox increased from 74 ppb to 122 ppb (Fig. 4b). A net increase in the observed OM was 14.6 μg m −3 during this period (Fig. 4c) while the required amount of SOA was estimated as 4.4 μg m −3 . SOA were very likely to be the major contributor to particle growth in this period. As independently reported by Wu et al. (2016), the hand, the estimated sulfuric acid and observed NO3plus NH4 + yielded either a small percentage or negligible contribution to particle growth in this period.
The Dpg of newly formed particles stopped growth at 79 nm from 17:20 through 18:00 and then decreased from 79 nm to 52 nm at 21:22, with a decreasing rate of 8 nm h -1 . During this period of shrinkage, the observed mixing ratio of Ox largely decreased from 130 ppb to 80 ppb (Fig. 4b). Repartition of the semivolatile species in gas and particle phases was hypothesized 5 to cause the evaporation of semivolatile particulate species to the gas phase. Unfortunately, the observed OM fluctuated a lot at 20:00-22:00 when local signals of OM apparently overwhelmed its regional signals (Fig. 4c). The shrinkage may also be argued as spatial heterogeneity of NPF, but size-segregated number concentration modeling needs to confirm this.
After 21:22, the Dpg restarted to increase from ~50 nm to 90 nm over 4 hours. The formation of NH4NO3 likely yielded an important contribution to the second stage of particle growth, i.e., a net observed increase of 4.5 µg m -3 versus the required amount of 8.1 µg m -3 . SOA may also contribute to the second stage of particle growth on basis of a net increase in OM by 18 µg m -3 (Fig. 4c). After the second stage of growth, the Dpg of new particles experienced small oscillations at ~90 nm until the signal was overwhelmed completely by aged plumes.
Following a similar analysis on 23 June, reduced photochemical reaction activities were also argued to cause the shrinkage in newly formed particles on 11 June and 26 July 2014 (Fig. S6). The observed and modeled results for the two days 15 implied that NH4NO3 played an important role in new particle growth only at night. In the daytime, SOA likely acted as the major contributor to particle growth.

Statistical analysis of factors related to new particle growth
The growth rate of newly formed particles is mainly determined by concentrations of condensable vapors such as sulfuric acid, organics in various volatilities, nitric acid and ammonia (Zhang et al., 2012;Ehn et al., 2014;Man et al., 2015;Lee et al., 20 2019). The Dpgmax are, however, determined by the total amount of vapors condensed on grown new particles, which may or may not have a positive correlation with the concentrations of these vapors (Zhu et al., 2019). When the values of Dpgmax are plotted against those of GR in Fig. 5a (two variables during the first growth period were used for plotting if two-stage particle growth existed), the values of Dpgmax largely scattered with r=0.23. When three circled points were excluded, the Dpgmax had a significant correlation with GR, but r value is still as low as 0.48 (Fig. 5a). GR alone is not sufficient to characterize the growth 25 of newly formed particles by considering their potential impacts on the climate, and the Dpgmax and GR should be alternatively used.
As abovementioned, SOA and NH4NO3 are likely two major contributors to the new particle growth in different periods with the contributions of sulfuric acid only in a few percentages. Fig. 5b shows the net hourly increases in OM and NH4NO3 against the hourly required masses for particle growth by assuming the density of 1.5 g m -3 for OM and 1.7 g m -3 for NH4NO3.
Both OM and NH4NO3 generally increase with increasing required masses and reasonably satisfy the required masses, but https://doi.org/10.5194/acp-2019-1151 Preprint. Discussion started: 4 May 2020 c Author(s) 2020. CC BY 4.0 License. they scatter largely in Fig. 5b. There is still a challenge to accurately quantify those contributors to the growth of newly formed particles.
The generation of OM and HNO3 are strongly related to oxidation reactions at daytime. Thus, we further plot Dpgmax and GR against Ox (Ox=NO2+O3) during the particle growth period at daytime. Fig. 5c shows a good correlation between Dpgmax and Ox (the hourly average value when reaches Dpgmax) with r=0.80 and P<0.01. The slope further suggests that an increase of 10 5 ppb in Ox likely causes an increase of 5 nm in Dpgmax. The values of Ox in Class I NPF events were significantly smaller than those in Class II and Class II with P<0.05, and the lower Ox could be one of factors for no apparent particle growth in Class I.
In addition, there was no significant difference of Ox between Class II and Class III. Including Ox, other factors should also affect the particle growth in Class I, II and III NPF events. Fig. 5d shows a significant correlation between GR andOx (the average value during the whole growth period) with r=0.67 and P<0.01. The decreased r value implies the response of GR to 10 the increase in Ox to be highly variable.
Oxidation products of biogenic VOCs, such as highly oxygenated molecules (HOMs), have been reportedly overwhelmed to determine the condensation growth of newly formed particles in the smaller size range because of their low volatilities (Ehn et al., 2014;Lee et al., 2019). In this study, the clear seasonal boundary of the Class I and Class II+III NPF events, e.g., 100% of Class I events in winter versus 7% and 93% of Class I and Class II+III events in summer, also points towards the importance 15 of oxidation products of biogenic VOCs in growing particles from ~10 nm to larger size. In the summertime, theoretically increased emissions of biogenic VOCs and enhanced photochemical reactions indicating by Ox are expected to generate more HOMs for the growth of particles from ~10 nm to larger size. In spring, approximately half of NPF events are subject to Class I. However, there were no Class III then in. The seasonal transience may further imply that the generated amount of oxidation products of biogenic VOC not only determines the growth of new particles from ~10 nm to larger size, but also to CCN size.

20
Unfortunately, we had no direct measurements of HOMs in small sized nanoparticles. In fact, it is still a common challenge to measure them for research community as reviewed by Lee et al (2019).
When the estimated CS were compared, the averaged value was 1.8±2.0×10 -2 s -1 , 2.1±1.5×10 -2 s -1 and 2.0±1.2×10 -2 s -1 for Class I, Class II and Class III, respectively. No significant difference exists between any two of them. Therefore, CS alone cannot explain the obtained three classes of particle growth patterns.  According to the observed mixing ratio of SO2, the sulfuric acid vapor generally yielded minor contributions to new 15 particles growth. The observed and modeled concentrations of particulate chemical species suggested that the growth in newly formed particles in the daytime was mainly caused by OM/SOA. At night and late afternoon, the increased amount of NH4NO3 can reasonably account for the required amount to support new particle growth in most Class III NPF events. Besides, organics were also an important contributor to nighttime new particle growth in most Class III NPF events. However, direct measurements of these chemicals in different sized nanometer particles need to confirm this.

20
Regarding climates impacts of NPF events, the final SPR50 and final SPR70 are essential to be quantified. In Class III NPF events, the final SPR50 and final SPR70 varied from 42% to 200% and from 31% to 138%, respectively, implying that a significant fraction of new particles can grow to CCN size. However, the final SPR70 in one stage growth NPF events were significantly smaller than those of in two-stage growth-A plus growth-B NPF events with P<0.05. The difference is worthy of further investigation. However, no significant difference of final SPR70 between two-stage growth-A and growth-B NPF events.

25
The percentage values of SPR50 and final SPR70 larger than 100% were due to spatial heterogeneity of NPF on a regional scale.
The uncertainties on SPR50 and SPR70 associated with spatial heterogeneity of NPF cannot be removed on basis of observations alone while they may be reduced, to some extent, by a combination of observations and modeling results of newly formed particles in future. Our observations indicated that spatial heterogeneity always occurred in each NPF event to some extent and may be caused by varying photochemical reaction activities. When photochemical reaction activities are reduced, the