Assessment of Air Pollution around the Panzhihua VTi Magnetite Mine Region , Southwest China

This study investigated variations in air quality by evaluating trace gases, inhalable particulate matter (PM10), and associated trace elements at three sites in Panzhihua (a mining city located in Panxi Rift Valley, Southwest China) between January and December 2014. The concentrations of 19 trace elements in PM10 were determined through inductively coupled plasma mass spectrometry. Single particle morphology and chemical composition were determined through scanning electron microscopy with energy-dispersive X-ray analysis to identify their possible sources. Mean sulfur dioxide, nitrogen dioxide, and carbon monoxide concentrations were highest near the steel smelting district, whereas ozone concentrations were highest in the residential region. Annual mean concentrations of PM10 at three sites were 129.4, 165.5, and 187.2 μg m; all these exceed the annual mean (70.0 μg m) of the National Ambient Air Quality Standard. In addition, the concentrations of trace elements in PM10 exhibited significant spatial and seasonal variations at the three sites. The mean concentrations of trace elements in PM10 were in the order of Fe > Ti > Zn > Pb > Cu > Mn > Ba > V > Cr > Ni > Sr > Bi > Cd > As > Co > Sb > Sc > TI > U. The enrichment factor values of the trace elements suggested that anthropogenic activities were the dominant sources of As, Cd, Sb, Ti, TI, Zn, Cu, Pb, and Bi. Particle morphology and chemical composition analysis revealed five major particle types, namely aluminosilicate, Fe-containing, mineral, soot, and Ca-containing particles.


INTRODUCTION
Mining and steel smelting activities have been considered the primary cause of air quality deterioration worldwide (Allen et al., 2001;Csavina et al., 2012;Bhanu et al., 2014).These activities emit high concentrations of trace gaseous pollutants, particulate matter (PM), and toxic elements.These air pollutants play a crucial role in influencing human health (Garcia et al., 2011).In particular, many toxic elements are mostly concentrated in the fine particle size fraction (< 10 µm, PM 10 ) (Corriveau et al., 2011).Thus, these particles can be easily inhaled or ingested and can pose serious health hazards (Zheng et al., 2010).Moreover, mining cities in developing countries such as China have been adversely affected by air pollution in recent years because of rapid economic growth resulting in a tremendous increase in energy demand (Fang et al., 2010).Studies have reported high concentrations of sulfur dioxide (SO 2 ) (Gao et al., 2009;Lu et al., 2010;Francesco et al., 2016), nitrogen dioxide (NO 2 ) (Zhao et al., 2013;Francesco et al., 2016), ozone (O 3 ) (Tu et al., 2007;Francesco et al., 2016), and toxic metals in airborne particles (Kan et al., 2012;Lu et al., 2015), and a substantial part of the world's population is exposed to the highest pollution levels (Madaniyazi et al., 2016).Ambient air quality is a major concern in China.
The Panzhihua region in Sichuan Province, Southwest China, is a major ore belt of V-Ti magnetite, accounting for 11% of the world's reserves (Ren, 2003), with significant mining and steel smelting activities.Intensive mining, extraction, smelting, and related heavy traffic activities have affected the atmosphere, and the air quality of this region is seriously affected by atmospheric toxic metals (Xu et al., 1984;Xue et al., 2010), SO 2 , NO 2 , and carbon monoxide (CO) (Chen, 1985).In addition, the Panzhihua region is in the Panxi Rift Valley, and its urban area is surrounded by high mountains on three sides.Stable synoptic weather conditions with a high frequency of calm winds and temperature inversions trap air pollutants in the lower atmosphere, creating conditions that are detrimental effects on human health (Chen, 1985;Wang et al., 1985).In recent years, rapid urbanization and industrialization have aggravated atmospheric pollution in this region.Thus, assessing the quality of the environment around the mining area is necessary because of extensive concerns regarding health-and environment-related issues.
In the present study, we measured SO 2 , NO 2 , CO, O 3 , and PM 10 concentrations over a year period from January to December 2014 in three functional areas of the Panzhihua region.In particular, we focused on the concentration of trace elements (As, Ba, Bi, Cr, Cd, Co, Cu, Fe, Mn, Ni, Pb, Sb, Sc, Sr, Ti, TI, U, V, and Zn) in PM 10 .We analyzed the enrichment factor (EF) to distinguish the origins of trace elements.Single particle morphology and chemical composition were evaluated through scanning electron microscopy-energy-dispersive X-ray spectroscopy (SEM-EDX) to identify their origin.This preliminary research may help researchers and policy makers in preventing and managing air pollution.

Study Area and Sampling Description
The Panzhihua region is a crucial mining and industrial base in Southwest China.Approximately 76 types of mineral resources are found in this region.Furthermore, Fe and steel smelting, V-Ti processing, and chemical and coal-fired power plants are the major industries in this region.The Panzhihua Fe-Ti-V oxide mine is the largest known V-Ti magnetite ore body in the country, providing 20% Fe, 64% V, and 53% Ti supply of China.The economy of the Panzhihua region depends almost entirely on mining activities that have been developed during the past half-century.In 2009, the population of this region is approximately 1.16 million and its urbanization ratio was 60%.However, air quality degradation has become a serious problem because of rapid urbanization and mining processes.
The Panzhihua region has a typical subtropical monsoon dry-hot valley climate, with a large temperature difference between day and night.The annual mean temperature is approximately 19.7-20.5°Cwith long hours of sunshine (2300-2900 h a -1 ).The wet precipitation is 760-1200 mm, with 90% of the total rainfall occurring predominantly from June to October.The Jinsha River (upper stream of the Yangtze River) traverses the city from west to east, and industrial enterprises are located on both sides of the river and surrounded by high mountain ranges.Many residential buildings are present in and around the mining and industrial districts where the living environment is very poor.

SEM-EDX Measurements
Individual aerosol particles in the PM 10 samples were analyzed through SEM by using an FEI Quanta 250 FEG (20 kV) equipped with an EDX spectrometer (EDX-Oxford INCAx-max 20) at the State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation (Chengdu University of Technology, China).A single portion of quartz filters (10 × 10 mm in size) was cut from the central part of the membrane and mounted on SEM aluminum stubs by using double-sided carbon-coated tape and observed using randomly selected fields of view.The chemical composition was determined through EDX analysis (spot size: 5, working distance: 10 mm); the measurement time of EDX analysis for each particle was 60 s.

Total Element Concentration
For trace metal analysis, PM samples were extracted from filter using a mixed acid of ultra-pure hydrogen fluoride and nitric acid (3:2) and the microwave digestion method and analyzed for 19 trace metals (As, Ba, Bi, Cr, Cd, Co, Cu, Fe, Mn, Ni, Pb, Sb, Sc, Sr, Ti, TI, U, V, and Zn) by using a PE 6000 inductively coupled plasma mass spectrometer (Perkin Elmer Corp., Norwalk, USA) at the Applied Nuclear Technology in Geosciences Key Laboratory of Sichuan Province (Chengdu University of Technology).All samples were determined in triplicate and a difference lower than 1% was considered acceptable.

Pollution Assessment
EF is often used to identify a natural and anthropogenic source of aerosol components.In addition, EF indicates the enrichment degree of a particular element in particles compared with the relative abundance of that element in the crustal material.EF is mathematically defined as follows (Waheed et al., 2010): EF = (X/R) air/(X/R) crust, where R is a reference element.In this study, Sc was selected as the reference element because of its low volatility, lack of an anthropogenic source, and reliable quantitative determination (Bilos et al., 2001).The average composition of the Earth's crust was referenced from Zhao et al. (1985).If the EF was > 10, anthropogenic activity was considered to be the predominant source.Furthermore, if the EF was < 1 or 1 < EF < 10, natural or mixed sources were considered to be the predominant source, respectively.

Gaseous Pollutants
Table 1 lists the monthly variations of daily concentrations of SO 2 , NO 2 , CO, 8-h peak O 3 , and PM 2.5 to PM 10 ratios (PM 2.5 /PM 10 ) measured at the three sites.These data were collected from the Panzhihua AQMS in 2014.The mean concentration of SO 2 was 45, 66, and 107 µg m -3 at XSG, HMK, and NNP, respectively.The annual mean concentration of SO 2 at HMK and NNP exceeded the Grade-II value (60 µg m -3 ) of the National Ambient Air Quality Standard (NAAQS).The mean concentration of NO 2 was 32, 32, and 64 µg m -3 at XSG, HMK, and NNP, respectively.Similar to the SO 2 concentration, the annual mean concentration of NO 2 at NNP exceeded the Grade-II value (60 µg m -3 ) of the NAAQS.The mean concentration of CO was 1.9, 1.8, and 2.1 mg m -3 at XSG, HMK, and NNP, respectively.The annual mean concentration of CO at all sites was lower than the Grade-II value (4.0 mg m -3 ) of the NAAQS.Furthermore, the annual mean concentration of O 3 was 104, 104, and 84 µg m -3 at XSG, HMK, and NNP, respectively.The annual mean concentration of O 3 at all sites was lower than the Grade-II 8-h peak O 3 standard (160µg m -3 ) of the NAAQS.The annual mean PM 2.5 /PM 10 ratios were 0.5, 0.5, and 0.6 at XSG, HMK, and NNP, respectively, indicating a large fine fraction of PM 10 in Panzhihua City.The SO 2 , NO 2 , and CO concentrations followed a bimodal distribution over the three sampling sites, which may be linked to local meteorological conditions, biomass burning, and heat emission.However, the SO 2 concentration at NNP exhibited no significant seasonal differences, indicating that the SO 2 concentration in this region is likely homogeneous and mainly due to heavy-duty truck emissions and coal combustion during steel smelting.O 3 demonstrated considerable seasonal variation, with a relatively high magnitude in summer and relatively low magnitude in winter.The seasonal variation of O 3 differed markedly from those of other gaseous and PM pollutants.The concentrations of gaseous pollutants were significantly higher at industrial sites, indicating that local industrial activities apparently release gaseous pollutants.

PM 10 Mass Concentrations
The monthly mean PM 10 values were calculated from daily concentration data and are presented in Fig. 2.During the monitoring period, the daily concentration of PM 10 at XSG, HMK, and NNP varied from 45.1 to 250.6 µg m -3 , 63.1 to 369.8 µg m -3 , and 98.2 to 321.7 µg m -3 , respectively, with the highest concentration at NNP, followed by HMK and XSG sites.The average mass concentration of PM 10 was approximately 129.4, 165.5, and 187.2 µg m -3 at XSG, HMK, and NNP sites, respectively.The corresponding average PM 10 concentrations at these three sites were 1.3, 1.7, and 1.9 times higher than the China NAAQS (GB3095-2012) annual mean of 100 µg m -3 .Furthermore, approximately 35%, 60%, and 72% PM 10 samples exceeded the 24-h NAAQS value of 150 µg m -3 at XSG, HMK, and NNP sites, respectively.The mass concentrations of PM 10 exhibited significant seasonal variations, with markedly higher concentrations observed in winter (November-January) and spring (February-April) and lower values in summer (May-July) and autumn (August-October).Particles are easily scavenged by rain in the rainy season (Csavina et al., 2012); therefore, PM 10 levels may be low during the rainy season (May-September).

Trace Element Concentrations
The concentrations of 19 trace elements (As, Ba, Bi, Cr, Cd, Co, Cu, Fe, Mn, Ni, Pb, Sb, Sc, Sr, Ti, TI, U, V, and Zn) present in PM 10 in the four seasons at the three sampling sites are listed in Table 2.We observed that the concentration of the crustal elements Ti and Fe (Lim et al., 2010) was considerably higher than that of any other toxic metals in PM 10 .This result is consistent with that reported by Xu et al. (2012), in that the predominant elements in airborne PM were often from crustal soils.In addition, the concentration of the anthropogenic metals Zn, Pb, Cu, and Mn was higher among the measured trace metals in PM 10 .This finding is in agreement with that of Bilos et al. (2001), who reported that industrial metallurgic processes result in an increase in Cu, Pb, and Zn concentrations.The concentrations of most trace elements (except for As) were significantly higher at the industrial sites than at the residential area, suggesting that industrial operations have a crucial effect on air quality.Similar to the PM 10 mass concentration, the mass concentrations of most elements exhibited seasonal variations, with the highest and lowest concentrations in spring and summer, respectively.The seasonal variations of most elements might be partly associated with differences in meteorological conditions.The rainout and washout scavenging of wet deposition effectively removes metalbearing particles from the atmosphere in summer.Related studies have reported that approximately 40%-90% of heavy metals can be removed by wet deposition (Yang et al., 2009).For example, a study reported that in Qingdao, approximately 99% of Pb and 17% of Cu in PM 10 were removed by rainfall (Li et al., 2003).
The trace elements As, Cd, Co, Cr, Ni, Pb, Mn, and V are carcinogenic heavy metals registered in the U.S. Agency for Toxic Substances and Disease Registry (Tao et al., 2014).The annual mean concentrations of As, Cd, Ni, and Mn in the two industrial sites exceeded the WHO standard by 1.1-2.5 times (WHO, 2000; As: 6.6 ng m -3 , Cd: 5 ng m -3 , Ni: 20 ng m -3 , Mn: 150 ng m -3 ).The Pb concentration was comparable to the limit of 500 ng m −3 , and the V Fig. 2. 24-h average PM 10 concentration at three sampling sites in 2014 (µg m -3 ) concentration was lower than the limit of 1000 ng m -3 .

EF Analysis
The EFs of the elemental species at the three sampling sites are presented in Fig. 3.The EFs of As, V, Ni, Cr, Co, Sr, Mn, U, and Ba were less than 10, indicating that their predominant source was crustal elements.However, the aforementioned elements can also have other sources.Studies have reported that As in atmospheric particles are coal combustion indicators (Tian et al., 2010), and the metal elements Ba, Cr, Cu, Co, Sb, Sr, and V have been identified as markers of vehicle emissions (Lim et al., 2010).The EFs of Cd, Ti, TI, Cu, Zn, Pb, and Sb ranged from 10 to 100, indicating that the anthropogenic emission was predominant.The EF of Bi was > 100, indicating that it was highly enriched.We observed that the EFs of most toxic metals were much higher at NNP [Fig.3(c)] and HMK [Fig.3(b)] and relatively lower at XSG [Fig.3(a)], confirming that anthropogenic sources contributed significantly to the input of these metals.

Particles Morphology Characteristics
SEM-EDX is a powerful tool for understanding the elemental composition, morphology, and density of particles, and it provides more definitive insight into the origin of particles than other methods do (Samara et al., 2016).The morphological properties of PM 10 collected from the three sites are presented in Fig. S1.We observed that the morphology and particle number concentration of atmospheric PM varied considerably because of variations in the emission sources.Aerosol particles were more abundant at the industrial sites [Figs.S1(b) and S1(c)] than at the urban residential site [Fig.S1(a)], indicating that industrial activities have a crucial effect on air quality.On the basis of particulate morphology and elemental composition, we identified five clusters of particles, namely aluminosilicate, Fe-containing, mineral, soot, and Ca-containing particles.
Aluminosilicate particles (Fig. S2) are primarily composed of Si, Al, and Ca; Si, Al, K, and Si; or Si, Al, and Fe (Kurth et al., 2014), which were originally associated with either combustion (Slezakova et al., 2008;Pipal et al., 2011) or crustal processes (Martinez et al., 2008).These particles are abundant and have diverse morphological characteristics.In the current analysis, we observed that the predominant proportion of aluminosilicate particles was irregularly shaped with sharp edges [Figs.S2(a) and S2(b)], indicating their crustal or suspended dust origins (Chow et al., 1995;Slezakova et al., 2008).The concentrations of these particles were higher at XSG.Some aluminosilicate particles had a spheroidal morphology [Fig.S2(c)], indicating that their origin is related to high-temperature combustion processes, with major abundance at the two industrial sites (HMK and NNP).
Fe-containing particles (Fig. S3) consisted of S, Fe, and Si [Figs.S3(c) and S3(d)] and had a typical spherical shape [Fig.S3(b)], indicating their anthropogenic origin particularly from Fe and steel smelting activities and other hightemperature combustion processes in the region.The concentration of Fe-containing particles was higher at the two industrial sites.We observed angular-shaped Fe-bearing particles [Fig.S3(a)], indicating that these particles are typically mechanically generated by crushing and grinding and often dominated by mining activities (Slezakova et al., 2008).Ca-containing particles [Fig.S4(a)] have a high concentration of Ca and low concentration of Si, indicating that these particles can be related to crustal sources and were mainly from limestone mining in the region.In addition, Ca has been associated with suspended dust.Cement contains approximately 90% calcium carbonate (Tas, 2007), and cement manufacturing industries are developed in the sampling region.
The presence of soot particles is illustrated in Fig. S4(b).The fluffy and amorphous deposition demonstrated the presence of carbonaceous spherules, possibly originating from fuel combustion associated with traffic emissions and biomass burning (Chow et al., 1995;Slezakova et al., 2008).These particles were the most abundant at NNP.We identified carbonaceous species mixed with other trace combustion particulates such as Al, Na, Ca, Si, and K and, in certain cases, with crustal species.Aerosol EDX spectra [Fig.S4(d)] indicated an association of K [Fig.S4(d)] with C and O particles, invariably suggesting the contribution from local biomass burning.The presence of Cl in carbonaceous aerosols with low Si and K concentrations indicated the contribution of farming and burning practices, which are mainly associated with regional livelihoods (Murari et al., 2016).
Mineral particles (Fig. S5) contain a high concentration of metals such as Fe, Mn [Fig.S5 V,and Cr,[Figs. S5(d) and 8(e)] and are one of the most critical groups from a toxicological viewpoint.We observed that these mineral particles mostly had an irregular shape but occasionally had a regular shape and lumpy structure [Fig.S5(c)].These particles may have originated from different sources depending on the metal concentration.

CONCLUSIONS
The study results revealed that the annual mean concentrations of SO 2 and NO 2 at NNP exceeded the Grade-II value (60 µg m −3 ) of the NAAQS.The mean PM 10 concentration during the study period was higher than the annual national standard.Furthermore, approximately 35%, 60%, and 72% of PM 10 samples exceeded the 24-h NAAQS value of 150 µg m −3 at XSG, HMK, and NNP, respectively.Spatial and seasonal variations in mass concentrations of PM 10 and trace elements were affected by distribution of pollution sources and meteorological conditions.EFs at the three sites indicated that trace elements in inhalable particles can be divided into two categories: elements from the Earth's crust and anthropogenic sources.For the elements As, Cd, Sb, Ti, TI, Zn, Cu, Pb, and Bi, anthropogenic activities were the dominant sources.The SEM-EDX analysis indicated that the morphology and particle mass concentration of atmospheric PM varied considerably because of variations in emission sources.The concentration of aerosol particles was markedly higher at the industrial sites than at the urban residential site.On the basis of the particulate morphology and elemental composition, we identified five clusters of particles: aluminosilicate, Fe-containing, mineral, soot, and Ca-containing particles.The complex composition of anthropogenic particles included S, O, C, Si, F, K, Ca, AI, Na, Mg, CI, Cr, Ti, Fe, V, Mn, and Zn.

Fig. 1 .
Fig. 1.Map of the study area showing three sampling sites.

Table 1 .
Monthly mean level of gaseous pollutants (SO 2 , NO 2 , O 3 in µg m -3 , CO in mg m -3 ) and ration of particulate matter PM 2.5 /PM 10 NO 2 CO O 3 PM 2.5 /PM 10 SO 2 NO 2 CO O 3 PM 2.5 /PM 10 SO 2 NO 2 CO O 3 PM 2.5 /PM 10 Data were provided by Air Quality Monitoring Network of Panzhihua, 2014.

Table 2 .
Concentration of trace elements (ng m -3