Ionic composition of TSP and PM2.5 during dust storms and air pollution episodes at Xi'an, China

  • Published on

  • View

  • Download

Embed Size (px)


<ul><li><p>in</p><p>Zhhign 71hinaCarlstitu</p><p>Department of Applied Chemistry, Faculty of Science and Technology, Keio University, Kohoku-ku, Yokohama 223-8522, Japan</p><p>a r t i c l e i n f o</p><p>Article history:Received 13 September 2008Received in revised form</p><p>species play an important role in determining the hygroscopicity ofatmospheric aerosols (Tang et al., 1995); and their tendency toabsorb water at high relative humidities (RHs) can exacerbatevisibility impairment (Dougle et al., 1996; Lee et al., 2005; Yuanet al., 2006). In addition, detailed information on particle-size</p><p>2002; Cao et al., 2005a; Shen et al., 2008). Although the air qualityhas improved recently (Zhang et al., 2002; Shen et al., 2008), highloadings of aerosols are still the major problem for the city. Duststorms (DSs) and pollution in the form of haze and smoke fromstraw burning and rework displays are considered as the mainreasons for the episodes of poor air quality in Xian and other citieslike Beijing in China (Cao et al., 2005a; Wang et al., 2006, 2007).Investigations of these sources are important because they alsocause high aerosol loadings at other cities in Asia.</p><p>* Corresponding author at: Department of Environmental Science and Engi-neering, Xian Jiaotong University, Xian 710049, China. Tel./fax: 86 29 82668385.</p><p>Contents lists availab</p><p>Atmospheric E</p><p>lse</p><p>Atmospheric Environment 43 (2009) 29112918E-mail address: (Z. Shen).1. Introduction</p><p>The composition of the atmospheric aerosol is important rela-tive to its effects on biogeochemical cycles, human health, andvisibility (Horvath, 1996; Yadav et al., 2003; Schichtel et al., 2001;Tsai and Cheng, 1999; Tsai et al., 2003; Lee et al., 2005; Ho et al.,2006a,b; Yuan et al., 2004, 2006). Prior studies showed that water-soluble ions accounted for one-third or more of the aerosol parti-cles mass in Chinese urban regions (He et al., 2001; Hu et al., 2002;Wang et al., 2002, 2006). Furthermore, the relative abundances ofwater-soluble ions, such as sulfate, nitrate, ammonium, and other</p><p>distributions of water-soluble ions has been obtained (Kadowaki,1976; Wall et al., 1988; Kerminen et al., 2001; Venkataraman et al.,2001; Tsai et al., 2003; Shen et al., 2008), and it has been shownthat the size distributions of the ions can vary signicantly withseason and geographic location (Fitzgerald, 1991).</p><p>Xian (area 1066 km2, population 5.1million), located on theGuanzhong Plain on the southern edge of the Loess Plateau, is thelargest city in northwestern China. Along with a rapid increase inmotor vehicle trafc and energy consumption during the past fewdecades, Xian is facing long-term air quality problems, with aerosolparticles (hereinafter aerosols) being a major concern (Zhang et al.,14 February 2009Accepted 5 March 2009</p><p>Keywords:TSPPM2.5Water-soluble ionsPollution episodes1352-2310/$ see front matter 2009 Elsevier Ltd.doi:10.1016/j.atmosenv.2009.03.005a b s t r a c t</p><p>TSP and PM2.5 samples were collected at Xian, China during dust storms (DSs) and several types ofpollution events, including haze, biomass burning, and rework displays. Aerosol mass concentrationswere up to 2 times higher during the particulate matter (PM) events than on normal days (NDs), and alltypes of PM led to decreased visibility. Water-soluble ions (Na, NH4, K, Mg2, Ca2, F, Cl, NO3,and SO4</p><p>2). were major aerosol components during the pollution episodes, but their concentrations werelower during DSs. NH4</p><p>, K, F, Cl, NO3, and SO42 were more abundant in PM2.5 than TSP but theopposite was true for Mg2 and Ca2. PM collected on hazy days was enriched with secondary species(NH4</p><p>, NO3, and SO42) while PM from straw combustion showed high K and Cl. Firework displays causedincreases in K and also enrichments of NO3 relative to SO42. During DSs, the concentrations ofsecondary aerosol components were low, but Ca2 was abundant. Ion balance calculations indicate thatPM from haze and straw combustion was acidic while the DSs samples were alkaline and the reworksPM was close to neutral. Ion ratios (SO4</p><p>2/K, NO3/SO42, and Cl/K) proved effective as indicators fordifferent pollution episodes.</p><p> 2009 Elsevier Ltd. All rights reserved.Ionic composition of TSP and PM2.5 durepisodes at Xian, China</p><p>Zhenxing Shen a,b,*, Junji Cao b, Richard Arimoto c,Suixin Liu b, Tomoaki Okuda e, Shunsuke Nakao e, SaDepartment of Environmental Science and Engineering, Xian Jiaotong University, Xiab SKLLQG, Institute of Earth Environment, Chinese Academy of Sciences, Xian 710075, CcCarlsbad Environmental Monitoring &amp; Research Center, New Mexico State University,dKey Laboratory of Regional Climate-Environment Research for Temperate East Asia, Ine</p><p>journal homepage: www.eAll rights reserved.g dust storms and air pollution</p><p>iwei Han d, Renjian Zhang d, Yuemei Han a,eru Tanaka e</p><p>0049, China</p><p>sbad, NM 88220, USAte of Atmospheric Physics, Chinese Academy of Sciences, Beijing 100029, China</p><p>le at ScienceDirect</p><p>nvironment</p><p>vier .com/locate/atmosenv</p></li><li><p>viroHaze, which is dened as a condition in which the atmo-spheric visibility is less than 10 km by China MeteorologicalAgency (,has been frequently observed in Xian. Most of the haze, which ismost common in winter and spring, results from excess PMemitted by anthropogenic sources and particles produced viagas-to-particle conversion (Watson, 2002). Dust storms areanother atmospheric phenomenon in Xian, and they are mostcommon in the dry springtime. Cao et al. (2005b) reported thata heavy DS in Xian led to extremely high PM2.5 levels(740 mg m3). Straw combustion in suburban areas, which is mostprevalent during the summer and fall harvests, also causesserious air pollution in Xian. The burning of agricultural resi-dues, mainly rice and wheat straw, is a common method forwaste disposal after the crops are harvested (Haleh and Lowell,2007). Although this source is sporadic, the combustion of strawmay release pollutants such as NOx, SO2, CO, and produce PMwith a very high content of K and Cl 4080%(w/w) (Jacobset al., 1997; Katsumi et al., 2000). Moreover, there are increasedrisks for bronchial asthma in children who live near the areaswhere the straw was burned (Jacobs et al., 1997; Katsumi et al.,2000). Firework displays are common during Chinese traditionalSpring Festivals, and they can also lead to the impairment of airquality. Fireworks contain chemicals such as potassium nitrate,potassium chlorate, potassium perchlorate, charcoal, sulfur,manganese, sodium oxalate, aluminum, and iron dust powder,strontium nitrate, and barium nitrate, etc. (Mclain, 1980). Thecomposition of PM produced from reworks includes tracemetals and organic compounds which may pose a threat tohuman health (Ravindra et al., 2001). Furthermore, reworkdisplays can generate ground-level ozone, which may put peopleat risk for respiratory disease (Attri et al., 2001).</p><p>Information on water-soluble ions in samples collected duringPM episodes is needed to: (1) create speciated emission inventoriesfor use in source-oriented models, (2) evaluate possible effects onhuman health, and (3) improve pollution control strategies. As thepotential effects of these pollution episodes on the environmentand human health have been recognized, several studies havefocused on one or two aspects of these four PM sources (Okadaet al., 2001; Schichtel et al., 2001; Yadav et al., 2003; Kang et al.,2004; Wang et al., 2006, 2007), but none has consider all foursimultaneously. Here, aerosol particles were collected to investi-gate the effects of the four PM sources on the major ion composi-tion of two types of PM, that is PM2.5 (particle size smaller than2.5 mm) and TSP (Total Suspended Particulates).</p><p>2. Experimental</p><p>2.1. Aerosol sample collection</p><p>The TSP and PM2.5 samples were collected simultaneously on47-mm diameter cellulose nitrate membranes (0.8 mm pore size,AAWP04700, Millipore Corp.) from the roof of a 15-m high building.The sampling site is located in the southeastern part of downtownof Xian, only about 100m from the South Second Ring Road (Fig. 1).North and east of the sampling site are residential areas and thecampus of Xian Jiaotong University, while to the south andwest arethe South Second Ring and Xingqin Roads, where the trafc isalways heavy. During the sampling periods, meteorological data,including ambient temperature, relative humidity (RH), dew point,wind speed, wind direction, visibility, and rainfall were recorded bythe Shaanxi meteorological agency.</p><p>Twenty-four hour TSP and PM2.5 (10:00 am to 10:00 am, localtime) samples were obtained simultaneously every third day from</p><p>Z. Shen et al. / Atmospheric En29121 October 2006 to 30 September 2007 using a dual-channel low-volume sampler (Tokyo Dylec Corp., Japan) operating at a ow rateof 5 l min1. A total of 227 aerosol samples was collected, and theseinclude 14 pairs of haze samples, 5 sets of DSs samples, 2 pairs ofstraw combustion samples, 5 PM2.5 rework burning samples (noTSP samples were collected for the latter due to a failure of thevolume measurements for the TSP channel), and 90 pairs ofsamples collected on normal days (NDs). Of the fourteen hazesamples, twowere collected during autumn; onewas in spring, andothers were in winter. DSs samples were collected in springexclusively; the straw combustion lters were sampled in earlyJune, and during that time the peasants had just nished harvestingthe wheat and burning the straw residues. Firework displaysamples were collected during the Chinese Spring Festival(sampling from 17 to 21 in Feb 2007). Field blank lters were alsocollected each month by exposing lters in the sampler but notdrawing air through them; these were used to account for anyartifacts introduced during the sample handling procedure. Theexposed lters were stored in a refrigerator atw4 C until analysisto minimize the evaporation of volatile components.</p><p>Aerosol mass loadings were determined gravimetrically withthe use of a Sartorius MC5 electronic microbalance (1 mg sensi-tivity, Sartorius, Gottingen, Germany). Before weighing, the lterswere equilibrated for 24 h at 2023 C and a relative humiditybetween 35% and 45% (USEPA, 1997). Each lter was weighed atleast three times before and after sampling following the 24 hequilibration period. The mean net mass for each lter wasobtained by subtracting the pre-deployment weight from theaverage of the post-sampling readings.</p><p>2.2. Water-soluble ions analysis</p><p>One-fourth of each lter sample was used to determine theaerosol ion mass concentrations. Six anions (SO4</p><p>2, NO3, Cl, andF) and ve cations (Na, NH4, K, Mg2, and Ca2) in aqueousextracts of the lters were determined by an ion chromatography(IC, Dionex 500, Dionex Corp, Sunnyvale, CA). To extract the water-soluble species from the cellulose nitrate lters, one-fourth of eachlter was put into a separate 20 ml vial containing 10 ml distilled-deionized water (a resistivity of 18 MU), and then placed in ultra-sonic water bath for 60 min. The samples were shaken two timesusing a mechanical shaker for 1 h each time to completely extractthe ionic compounds. The extracts were ltered with a 0.45 mmpore size microporous membrane, and the ltrates were stored at4 C in a clean tube before analysis.</p><p>Cation (Na, NH4, K, Mg2, and Ca2) concentrations weredetermined with the use of a CS12A column (Dionex Corp.), using20 mMmethanesulfonate as an eluent. Anions (SO4</p><p>2, NO3, Cl, andF) were separated by an AS11-HC column (Dionex Corp.), using20 mM KOH as the eluent. The method detection limits (MDLs)were 4.6 mg L1 for Na, 4.0 mg L1 for NH4, 10.0 mg L1 for K, Mg2</p><p>and Ca2, 0.5 mg L1 for F and Cl, 15 mg L1 for NO3, and 20 mg L1</p><p>for SO42. Standard Reference Materials produced by the National</p><p>Research Center for Certied Reference Materials, China wereanalyzed for quality assurance purposes. Blank values were sub-tracted from sample concentrations.</p><p>3. Results and discussion</p><p>3.1. Mass concentrations of aerosol particles during four PMepisodes</p><p>TSP and PM2.5 mass concentrations for the DSs, pollution events,and NDs are given in Fig. 2. The average TSP mass concentrationduring haze, DSs, and straw combustion were 677.3 mg m3,</p><p>nment 43 (2009) 29112918631.2 mg m3, and 622.6 mg m3, respectively, and all were more</p></li><li><p>of</p><p>Z. Shen et al. / Atmospheric Envirothan double what was measured on NDs. These high mass loadingsexceed the Chinese Class 2 TSP standard (24-h mean value of300 mg m3) and the yearly standard, that is, a mean value of200 mg m3. Compared with the TSP mass, the variation of PM2.5concentrations for DSs and three pollution episodes showed somedifferences, exhibiting a decreasing order as follows: strawcombustion (404.1 mg m3) &gt; haze (351.2 mg m3) &gt; reworks(298.8 mgm3)&gt;DSs (137.6 mgm3). Among the four PM types, DSsshowed a relatively low PM2.5 mass, closest to the NDs mean valueof 130.0 mg m3, while the PM2.5 concentrations for haze, strawburning, and reworks increased by 2- to 3-fold over NDs. ThePM2.5 concentrations for these pollution events also exceeded the24-h American air quality standard of 35 mg m3 and yearly meanvalue of 15 mg m3 (EPA, 2006). China currently has no National</p><p>Fig. 1. LocationPM2.5 standard, but this comparison illustrates the seriousness ofthe ne particle pollution in Xian.</p><p>Previous studies have shown that aerosol loadings are one of thefactors most responsible for visibility reduction (Sequeira and Lai,1998; Watson, 2002; Lee et al., 2005). One would expect the highPM loadings at Xian to impair visibility, and so we examined thedata for relationships between aerosol mass and visibility. The</p><p>Co</p><p>nc</p><p>en</p><p>tra</p><p>tio</p><p>n (g</p><p>m-3)</p><p>Haze DSs Straw Fireworks NDs0</p><p>100</p><p>200</p><p>300</p><p>400</p><p>500</p><p>600</p><p>700 TSPPM2.5</p><p>Fig. 2. TSP and PM2.5 mass concentration for dust storms (DSs), pollution events, andnormal days (NDs).meteorological records show that the mean ambient visibilities onNDs averaged 13.1 km, but they were reduced to 10.4 km for DSsdays, 5.3 km for haze events, 8.5 km for straw combustion, and11.8 km for rework shows, respectively. Furthermore, the reduc-tion in visibility was strongly related to PM2.5 mass (the correlationcoefcient R 0.71, not shown), but the relationship betweenvisibility and TSP was much weaker (R 0.34). These preliminaryresults suggest that ne-particles are especially important forvisibility reduction, but there are additional factors (such asmeteorological variable) that need to be evaluated, and a detailedanalysis of PM and visibility is beyond the scope of this paper.</p><p>The ratios of PM2.5 to TSP mass concentrations for DSs, haze,straw, and NDs were 21.0%, 51.8%, 64.9%, and 49.0%, respectively,and therefore the dust storms had the greatest proportion of coarse</p><p>sampling site.nment 43 (2009) 29112918 2913particles. Wang et al. (2006) reported a similar result; in theirstudies, which were conducted in Beijing, the PM2.5/TSPmass ratiosduring dust storm events were 22%. For the pollution events, PM2.5composed more than half of TSP mass, and this demonstrates themajor contribution of ne-particles to...</p></li></ul>