Article pubs.acs.org/est
Comparison of Daytime and Nighttime New Particle Growth at the HKUST Supersite in Hong Kong Hanyang Man,†,∥ Yujiao Zhu,† Fei Ji,† Xiaohong Yao,*,† Ngai Ting Lau,‡ Yongjie Li,#,‡ Berto P. Lee,‡ and Chak K. Chan*,‡,§ †
Key Lab of Marine Environmental Science and Ecology, Ministry of Education, Ocean University of China, Qingdao 266100, China Division of Environment and §Department of Chemical and Biomolecular Engineering, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon, Hong Kong, China ∥ School of Environment, Tsinghua University, Beijing 100084, China ‡
S Supporting Information *
ABSTRACT: Particles larger than 50−100 nm in diameter have been considered to be effective cloud condensation nuclei (CCN) under typical atmospheric conditions. We studied the growth of newly formed particles (NPs) in the atmosphere and the conditions for these particles to grow beyond 50 nm at a suburban coastal site in Hong Kong. Altogether, 17 new particle formation events each lasting over 1 h were observed in 17 days during 8 Mar−28 Apr and 1 Nov−30 Dec 2011. In 12 events, single-stage growth of NPs was observed in daytime when the median mobility diameter of NPs (Dp) increased up to ∼40 nm but did not increase further. In three events, two-stage particle growth to 61−97 nm was observed at nighttime. The second stage growth was preceded by a first-stage growth in daytime when the Dp reached 43 ± 4 nm. In all these 15 events, organics and sulfuric acid were major contributors to the first-stage growth in daytime. Ammonium nitrate unlikely contributed to the growth in daytime, but it was correlated with the second-stage growth of ∼40 nm NPs to CCN sizes at nighttime. The remaining two events apparently showed second-stage growth in late afternoon but were confirmed to be due to mixing of NPs with pre-existing particles. We conclude that daytime NP growth cannot reach CCN sizes at our site, but nighttime NP growth driven by organics and NH4NO3 can.
■
INTRODUCTION Nucleation and the subsequent growth of freshly nucleated particles in the atmosphere are conventionally referred as new particle formation (NPF) events.1 These NPF events are an important source of atmospheric particles, and they can rapidly increase the total particle number concentration by 1−2 orders of magnitude in a few hours.1,2 When the newly formed particles (NPs) grow in the atmosphere, they can activate as cloud condensation nuclei (CCN) and indirectly affect the climate.3−13 Previous studies showed that the threshold diameters for particles activating as CCN varied from about 50 to 100 nm under different water supersaturation (SS).3 For example, particles with the diameter less than 50 nm are hardly to activate as CCN unless SS >0.5%, while particles with diameter larger than 80 nm can activate as CCN at moderate SS (0.2%).14−16 A few studies use 50 nm as a CCN cutoff, although a larger diameter is needed in real atmospheric clouds.3,14,15 Hence, understanding the ability of NPs to grow beyond 50 nm is crucial to estimating their impacts on radiative forcing. Although the growth of NPs in various atmospheric environments has been well documented,1,2 the exact species © 2015 American Chemical Society
of condensing vapors dominating the growth of NPs have not been fully elucidated. It is largely agreed that organics play a crucial role in growing NPs to CCN sizes.4−13 Some researchers suggested that sulfuric acid can be a minor contributor. On the other hand, Bzdek et al.17 found that sulfate accounted for 29−46% of the total mass growth of particles of 20−25 nm diameter during NPF events. Ammonium nitrate is one of the important components in submicron atmospheric particles, and it may potentially also contribute in growing NPs to CCN sizes.4,7,15,17 Furthermore, in polluted atmospheres, NPs can mix with preexisting particles or freshly emitted plumes, and this complicates the analysis of the growth of NPs.1−3 In this paper, we analyzed the growth of NPs larger than 10 nm in a data set based on gas and particle measurements made at the HKUST supersite, which is located in a suburban coastal area in Hong Kong during the periods from 8 Mar to 28 Apr and from 1 Nov to 30 Dec in 2011. The growth events of NPs Received: December 16, 2014 Accepted: May 19, 2015 Published: May 19, 2015 7170
DOI: 10.1021/acs.est.5b02143 Environ. Sci. Technol. 2015, 49, 7170−7178
Article
Environmental Science & Technology were classified into two categories. In category 1, growth of NPs occurred only in daytime and they did not grow over 50 nm. In category 2, NPs exhibited a second stage growth from smaller than 50 nm to larger than 60 nm in late afternoon or nighttime. Semicontinuous measurements of chemical composition in PM2.5 and in different sized particles as well as gas measurements were used to interpret the different growth patterns of NPs. We will discuss the role of ammonium nitrate and mixing of air masses in the observed second stage growth to CCN particles.
J5.6 =
dN5.6 − 30 GR 5.6 − 30 + CoagS5.6 − 30N5.6 − 560 + N5.6 − 30 dt 24.4 + S losses
(1)
The coagulation loss for particles in the range 5.6−30 nm (CoagS5.6−30) was the sum of particle−particle inter- and heterocoagulation rates calculated as Yao et al.23 The growth loss is due to condensation growth (GR5.6−30) out of the 5.6− 30 nm size range during the calculation period. Slosses includes additional losses and is assumed to be zero in this study. Multiple log-normal distribution functions were used to fit the size distribution of atmospheric particles.7,29 In this study, Dp represents the median mobility diameter of the new particles mode, and Ni is the number concentration of particles at the median mobility diameter of mode i as measured by FMPS. The apparent growth rate (GR) of new particles is calculated as
■
EXPERIMENTAL SECTION Sampling Site and Instruments. Measurements were made at the HKUST Air Quality Research Supersite on the campus of the Hong Kong University of Science and Technology (HKUST). The site faces Port Shelter and Sai Kung, which is a clean rural area with a huge country park and little traffic and commercial development.18,19 The site includes a 10-m high automatic weather station, PM2.5 samplers outdoors, and an air-conditioned modular house, which houses a fast mobility particle sizer (FMPS, TSI Model 3091), a MARGA ambient air analyzer (ADI 2080 1S, Metrohm AG), a high-resolution aerosol mass spectrometer (HR-AMS), SO2, NOx, and O3 analyzers, and other instruments.18,19 The sampling lines for instruments are 3−4 m in length, and the sample inlets are 2 m above the roof of the house. HR-AMS measurements and their applications in closure analysis of hygroscopic tandem differential mobility analyzer data and cloud condensation activities have been reported at this site.20−22 Collocated FMPS and HR-AMS measurements were only available in Nov 2011, and the HR-AMS results are thereby used to analyze the NPF events during this period. The details of HR-AMS measurements are available in Li et al.20 The FMPS was used for measuring atmospheric particles of 5.6−560 nm in mobility diameter with a time resolution of 1 s at a flow rate of 10 L min−1, and the 1 min average was used for analysis. High time resolution sampling of FMPS allows investigation of rapid changes of ultrafine particles in the atmosphere and mixing of ultrafine particles from different sources.23,24 Details about FMPS can be found in the work of Jeong and Evans.25 Undersizing of FMPS has been reported for particles larger than 100 nm in diameter,26 but the size distribution of number concentration for ambient particles between 20 and 70 nm measured by FMPS and a scanning mobility particle sizer agreed well.25 The mass concentrations of SO42−, NO3−, and NH4+ were measured at 1 h resolution by the Monitor for Aerosols and Gases (MARGA). Although MARGA can simultaneously determine trace gases such as HCl, HNO2, SO2, HNO3, and NH3, the negative artifact for HNO3 due to absorption in the long sampling line has been reported to reach ∼80%, compared with the values measured by a sampler equipped with a denuder.27 Similar artifacts might also exist for other sticky gases such as HCl and HNO2, albeit to unknown extents. Thus, the gaseous data measured by the MARGA are not used for data interpretation. Calculation of Formation Rate, Growth Rate, and Condensational Sink. In this study, the formation rate of particles larger than 5.6 nm (J5.6), taking consideration of the coagulation and growth losses, was calculated by the following equation:28
GR =
ΔDp (2)
Δt
where ΔDp is the increased median mobility diameter of the new particles mode and Δt is the duration for the growth of new particles. The condensation sink (CS) was estimated based on the analysis by Kulmala et al.30,31 The CS reflects how rapidly condensable vapor molecules will condense on existing aerosol. Contribution of Sulfuric Acid to Growth of NPs. Gasphase sulfuric acid concentration can be estimated on the basis of global solar radiation (SR), SO2 concentration, and CS32,33 [H 2SO4 ] = k·
[SO2 ]SR CS
(3)
where k is a constant value 2.3 × 10−9 m2 W−1 s−1. The concentration of condensable vapor for particle growth from Dp0 to Dp1 is assumed to be constant and is expressed by30 2 ⎧ 2 ⎪ Dp1 − Dp0 ⎤ ⎡4 C = ρ⎨ +⎢ − 0.623⎥λ(Dp1 − Dp0) ⎪ ⎦ ⎣ 2 3 a ⎩
+ 0.623λ 2 ln
λ + Dp1 ⎫ ⎪ ⎬ /ΔtDm ⎪ λ + Dp0 ⎭
(4)
Here, the surface vapor pressure of the condensable material is assumed to be zero. C is the condensable vapor concentration (molecules cm−3), ρ is the particle density in g cm−3, a is mass accommodation coefficient (i.e., sticking probability), λ is the mean free path in nm, Δt (s) is the time during particle growth from Dp0 to Dp1, D (cm2 s−1) is the diffusion coefficient of the condensing vapor, and m is the molecular mass of the condensable vapor in g mol−1. The contribution of sulfuric acid vapor to the particle growth from Dp0 to Dp1 (R) can be expressed by30 R = ([H 2SO4 ]avg /C) × 100%
(5)
where [H2SO4]avg is the average concentration during the growth period. This proxy of sulfuric acid concentration has been verified to have a good correlation coefficient (0.81) with the measured value in a forest in Hyytiala. However, the use of this proxy may lead to uncertainties in the estimated contribution of sulfuric acid to the particle growth in other areas. 7171
DOI: 10.1021/acs.est.5b02143 Environ. Sci. Technol. 2015, 49, 7170−7178
Article
Environmental Science & Technology Table 1. Characteristics of NPF Events event order 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
formation rate (J5.6 particle cm−3 s−1)
type category category category category category category category category category category category category category category category category category
1a 1a 1a 1a 2a 2a 1a 2a 1a 2b 1a 1a 1b 1b 2b 1b 1b
condensation sinka (×10−2 s−1)
growth rate (nm h−1)
date
duration (h)
± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ± ±
5.8 4.3 4.5 3.6 4.6/7.1b 6.0/7.2b 4.1 5.7/9.5b 6.1 11/39b 7.8 2.4 −c 5.4 4.8/18.8b 6.9 6.0
16 Mar 2011 17 Mar 2011 8 Dec 2011 24 Dec 2011 25 Dec 2011 24 Apr 2011 23 Nov 2011 28 Mar 2011 10 Dec 2011 13 Apr 2011 25 Mar 2011 6 Apr 2011 2 Nov 2011 1 Nov 2011 4 Nov 2011 27 Nov 2011 29 Nov 2011
8 16 7 10 21 21 5 14 9 >4 3 3 >5 >6 >9 >5 >3
11.6 9.0 7.3 6.6 6.5 5.2 5.0 4.9 3.7 3.2 0.7 0.5 0.6 0.4 0.4 0.3 0.2
1.0 1.7 1.1 1.1 1.3 1.2 1.9 1.8 1.3 2.4 1.9 1.5 2.3 1.5 1.8 1.7 1.1
0.03 0.06 0.06 0.03 0.04 0.05 0.05 0.05 0.06 0.06 0.3 0.14 0.10 0.05 1.40 0.07 0.09
Condensation sink was averaged 1 h prior to the nucleation event. bRefers to the first-stage growth and the second-stage growth rates. c−, the rate cannot be estimated. a
Figure 1. Contour plot of category 1a NPF events (dN/dlogDp) ((a) 16 Mar 2011, (b) 17 Mar 2011, (c) 25 Mar 2011, (d) 6 Apr 2011, (e) 23 Nov 2011, (f) 8 Dec 2011, (g) 10 Dec 2011, (h) 24 Dec 2011).
■
RESULTS AND DISCUSSION
events in the latter 60-day period. The NPF events occurred over 13% and 17% of the sampling days in the two periods, and these percentages were 30−50% of those obtained in other urban or suburban/rural areas of China.34−36 All 17 events occurred under clear or partially cloudy conditions, e.g., as shown in Figures S1−6 (Supporting Information), and they started when ambient relative humidity (RH) was less than 70%. Half of sampling days had RH larger than 70% in daytime and this high RH did not favor NPF.1 In the 17 events, NPF rates, i.e., J5.6 in eq 1, varied broadly from 0.2 to 11.6 particle cm−3 s−1 (Table 1). NPF rates in the first 10 events (nos. 1−10) varied from 3.2 to 11.6 particle cm−3 s−1, which are consistent with the range of typical formation rates observed in the atmosphere (1−10 particle
Overview of NPF Events. In this study, only NPF events lasting over 1 h were analyzed. For the 2011 HKUST Supersite measurements, collocated FMPS and MARGA measurements were only available in March, April, November, and December. HR-AMS measurements were made only in November, and hence, both MARGA and HR-AMS data are used for interpreting NPF in November, but only MARGA data are available for interpreting NPF in the other three months. FMPS measurements were irregular in May, August, September, and October due to either poor weather conditions or instrument availability. Thus, our analysis focused on the data collected in the period of 8 Mar-28 Apr and 1 Nov−30 Dec of 2011. There were seven NPF events in the first 52-day period and 10 NPF 7172
DOI: 10.1021/acs.est.5b02143 Environ. Sci. Technol. 2015, 49, 7170−7178
Article
Environmental Science & Technology cm−3 s−1).1,37 NPF rates in the remaining seven events (nos. 11−17) were much lower at 0.2−0.7 particle cm−3 s−1. The CS, mixing ratio of SO2, or solar irradiance does not seem to explain the differences in observed formation rates when the first 10 and the next seven events are compared (Table 1, http://envf. ust.hk/dataview/gts). The presence of amines or extremely low volatility secondary organics could cause a large difference on nucleation rate,38−40 but amines and those secondary organics were not measured in this study. On the basis of whether NPs can grow over 50 nm or not, the 17 events are classified into two categories. Category 1 includes 12 events (nos. 1−4, 7, 9, 11−14, 16, and 17) in which NPs did not grow beyond 50 nm (Figure 1 and Figure S7, Supporting Information). There were eight events in which the growth of NPs was observed until their particle counts dropped to negligible values, and they are referred as category 1a (Figure 1). In the other four events (nos. 13−14 and 16−17), the NPs grew up to 30−40 nm but they were overwhelmed by the preexisting particles (Figure S7). They are referred to as category 1b. Category 2 includes five events (nos. 5−6, 8, 10, 15) in which NPs apparently exhibited a second phase of particle growth to beyond 60 nm (Figures 2−6). Two of these
Figure 3. NPF event occurred on 25 Dec 2011 ((a) contour plot of particles number concentration (dN/d log Dp); (b) time series of Dp and chemical components in PM2.5; (c) time series of N
