1 Introduction

Fine particulate matter (PM2.5; aerodynamic diameter ≤ 2.5 µm) is responsible for the global burden of diseases and adverse effects on human health and surrounding environment such as respiratory problems, visibility, climate change (Cheng et al. 2016; IPCC 2013; Brauer et al. 2012). Health-related impacts are very distinguished for Asian environments (Tsiouri et al. 2015). The higher concentrations of particulate matter (PM) in metropolitan cities result from the higher consumption of energy (Patra et al. 2016a; Butler et al. 2008). Gurjar et al. (2010) have suggested the exposure to PM2.5 is a more severe problem in the urban area as compared to the rural area. PM2.5 pollution is defined by both quantitative and qualitative measurements of a number of activities conducted by different groups at particular places for short time intervals (Gautam et al. 2016a, b; Patra et al. 2016b; Gautam and Patra 2015). Moreover, most of the research work on PM2.5 has paid more attention to urban areas than to rural areas for the exploration of the sources and adverse effects (Gautam et al. 2016b; Lin et al. 2015a; Vela et al. 2015; Kumar et al. 2013). However, PM2.5 control strategies need to be taken or developed to minimize the impact of PM2.5 on the human and surrounding environment in the urban area. World Health Organization (WHO) (2014) has reported comprehensive data of PM2.5 concentrations of all possible and assessable locations of the world but need to be required more additional analysis in the present database.

The past literature has reported that higher exposure to PM2.5 is responsible for several types of health problems such as lung diseases (Chow et al. 2006), respiratory problems (Xia and Wang 2016; Chow et al. 2006), irregular heartbeat (Xie et al. 2016a; Kim et al. 2015), breathing problems (Yang et al. 2015; Labelle et al. 2015), eyes and itching problems (Salvi et al. 2016) and other associated problems in respiratory system (Xia and Wang 2016; Huang et al. 2016). Previous studies on topics such as source apportionment, chemical characterization, emission rate, emission inventory, chemical modeling, tempo-spatial variation, monitoring, dispersion analysis, health impact, exposure analysis, size distribution and statistical modeling of ambient PM2.5 have been carried out. All these past studies point for a holistic assessment of PM2.5 in China. Therefore, these works curtly synthesize the previously published research on PM2.5 at a different place of China, including monitoring networks and source area.

In China, the current status of air pollution is not acceptable because of exceedences against the national and international standards. The air quality status of different cities of China is more notable in the world (Zheng et al. 2016; Song et al. 2016; Hao and Wang 2005). A number of studies (quantitative and qualitative) have been reported the chemical composition and mass concentration of PM2.5 in China (Cheng et al. 2016; Yang et al. 2011). Because of the higher values of PM2.5 reducing the air quality, China introduced new national standards for PM2.5 in 2012. On the other hand, China has 1436 PM2.5 monitoring stations as compared with the 1500 monitoring stations in the USA (USEPA 2014). Data from these monitoring stations give the essential information for this study. The published literature (21 cities of China) on the PM2.5 level, distribution and composition in China was collected and reviewed. The main aim of the review is to introduce the status of the PM2.5 level, chemical composition and distribution in various cities of China, which will provide a good platform to the scientific community working on health impact related to PM2.5 and control strategies to minimize the exposure.

Approximately 190 published articles in the English language on source apportionment, health risk assessment, monitoring, and control of PM2.5 have been reviewed. We also included the literature from computer searches and bibliographic databases (e.g., Google scholar, Pub Med, Academia, Research Gate) in the analysis. We used the keywords such as PM2.5, air pollution, source apportionment, chemical composition, and China to search the suitable literature for this review article.

2 Chemical composition of PM2.5

The previous literature showed that heavy metal elements (i.e., Cd As, Pb, S, Si, K, Cl, Ca, Na, Fe, Al, Mg, Mn, Ba Br, Ba, Sb, and Ni) dominate the PM2.5 concentrations in urban atmosphere of China (Xie et al. 2016b; Cao et al. 2005; Dan et al. 2004). The high variations in elements composition could be ascribed to different sources or activities. All these elements are of great importance to assess the sources or related activities. Several studies have suggested the need of sensitive analytical techniques such as inductively coupled plasma mass spectrometry (ICPMS), particle elastic scattering analysis (PESA), particle-induced X-ray emission (PIXE), X-ray fluorescence (XRF), energy-dispersive X-ray fluorescence (EDXRF), scanning electron microscopy (SEM)–energy-dispersive X-ray spectroscopy (EDS), time-of-flight mass spectrometry (TOFMS), atomic absorption spectrophometer (AAS), and high-resolution continuum source atomic absorption spectrophometer (HR–CSGF–AAS) to identify the comprehensive details (toxic nature) of elemental information about PM2.5 in China’s atmosphere (Zhang et al. 2014d). Moreover, the hazardous nature of PM2.5 and its impact on human health, and measurement of water solubility of elements have been assessed in published articles (Jiang et al. 2014).

In case of analytical instrument, European countries are frequently using X-ray diffraction (XRD), scanning electron microscopy–energy-dispersive spectroscopy (SEM–EDS), inductively coupled plasma mass spectrometry (ICL–MS), and energy dispersive X-ray fluorescence (EDXRF), or aerosol mass spectrometer (AMS) and other related physicochemical characterization instruments to assess the accurate physical and chemical properties of fine PM. Similarly, Chinese environment (urban/rural) exhaustively needs to assess the chemical composition of fine PM and their effects on health and to comparatively improve the air quality of working environments (Cheng et al. 2016).

Different levels of PM2.5 concentrations originate from activities around rural (Zhu et al. 2012), urban (Zhang and Cao 2015), industrial areas, including mining (Zhang and Cao 2015; Hu et al. 2014; Hu and Jiang 2013) and marine activities (Wang et al. 2006). The composition of PM2.5 has been well reported to be the mixture of organic, inorganic, water-soluble ions, elemental carbon, crustal material, and hydrocarbons (Wang et al. 2006; Zheng et al. 2005; Yao et al. 2002; Wei et al. 1999). Some elements like barium (Ba), bromine (Br), nickel (Ni), sulfur (S), and magnesium (Mg) are observed higher in composition than other elements in PM2.5 in China (Wu et al. 2014; Chao and Wong 2002). On the other hand, several studies (Pui et al. 2014; Huang et al. 2012; Kan et al. 2007) organized in China indicate potential hazards (cardiovascular diseases, low birth weight, urinary effects, eyes irritation, and other health problems) due to higher amount of different elements with varying chemical composition. The assessment of PM2.5 related studies in different cities of China, based on average concentration and chemical composition, is presented in Fig. 1.

Fig. 1
figure 1

PM2.5 concentrations (µg m−3) and chemical composition (%) in different Chinese cities. Reported data are taken from Guangzhou EPB (Guangzhou), Shenzhen EPB (Shenzhen), Hong Kong (Huang et al. 2014b), Hangzhou (Cheng 2014), Qingdao (Cao et al. 2012), Nanjing (Chen et al. 2015a), Wuhan (Zhang et al. 2015), Xiamen (Cao et al. 2012), Xi’an (Wang et al. 2015b), Chongqing (Yang et al. 2011), Jinchang (Cao et al. 2012), Chengdu (Chen et al. 2015b), Yulin (Cao et al. 2012), Suzhou (Cheng 2013), Tianjin (Xu et al. 2015), Shanghai (Shanghai EPB), Harbin (Jia 2014), Beijing (Liu et al. 2015a), Jinan (Gu et al. 2014), Zhengzhou (Geng et al. 2013), Changchun (Cao et al. 2012)

Full size image

The presented database covers most part of China where the average concentration of PM2.5 was reported to vary in a broad range of 31–175 µg m−3. The higher amount of SNA, organic matter, and elemental carbon can be identified in fine PM, besides other unidentified substances. Transportation, household activities, vehicular movement, and industrial sector are possible PM2.5 sources in China (Liu et al. 2015a). However, most Chinese cities, especially Beijing, Yulin, Qingdao, and Zhengzhou, are mostly dependent on transportation and on solid fuels for household activities (Liu et al. 2015a). The variation of concentration occurs due to the number of residents or working area (urban/rural) (Smith et al. 2013). A number of studies have been done on chemical characterization and monitoring practice of particle in China, however, quantitative assessments for aerosol, its distribution formation and relationship between personal exposure, household exposure and ambient exposure are still needed.

3 Source apportionment

Source apportionment (SA) is used to investigate emission from different sources or the impact of fine PM at different monitoring sites (Balachandran et al. 2013). SA techniques are basically the statistical methods (i.e., monitoring data analyzed using basic statistics or numerical technique to identify the emission sources) including PCA (principal component analysis), nonparametric wind regression, back trajectory analysis, FA-MR (factor analysis with multiple regression), PMF (positive matrix factorization), and CMB (chemical mass balance) (Balachandran et al. 2013). PMF is useful to assess factors without source information of receptor site, but some techniques such as CMB models require unique profiles of each source (Gao et al. 2013). Nowadays, PMF is popular because of its convenient and realistic performance. However, the SA was underestimated by highly time-resolved data of PM2.5 compared to Vedantham et al. (2014). To solve this problem, the results should be validated by characterization of PM2.5 at local spatial and temporal patterns, local emissions, and meteorological parameters.

3.1 Positive matrix factorization (PMF)

Since information about the chemical characterization of fine PM necessary for researchers to make control strategies requires to be very accurate, it is tough to assess the source of air pollutants using traditional SA methods. However, Paatero and Tapper (1994) developed the PMF technique that is now widely used in China and elsewhere to investigate the source distribution and identification of PM2.5 (Liu et al. 2015b; Cheng et al. 2015; Gu et al. 2014). Numbers of studies have been reported with results on different proportions of composition (i.e., SNA, OC, EC, heavy metals) of PM2.5 with PMF (Shi et al. 2016; Wang et al. 2016; Geng et al. 2013). Table 1 comprehensively summarizes related PMF studies in China.

Table 1 Comprehensive details of recent SA results of PM2.5 (μg m−3) by using PMF in China
Full size table

Table 1 indicates the wide use of PMF methods to assess the source and chemical characterization of PM2.5 in different cities of China. However, source marker data (tracer methods) is needed to avoid the pre-assumptions in field measurement and overcome limited analysis in particle-based PMF analysis (Xie et al. 2013, 2016a, b). Similarly, the combination of PMF and AMS with radiocarbon is very useful to understand the atmospheric oxidation. Chemical transport CAMx model with PMF is rare in source apportionment analysis (Bove et al. 2014). However, it should be combined or integrated into future research on emission inventories. The results of this approach will demonstrate the role of atmospheric dynamics with respect to PM2.5 composition. Similarly, the outputs from the combination of PMF with cluster analysis would be useful to investigate the role of meteorological parameters on PM2.5 sources (Masiol et al. 2014).

3.2 PCA

A huge dataset can be easily handled by using of PCA, which is considered as most successful methods to explore the different set of variables and potential results of the input data. The method includes mathematical relationships to investigate the association of the different source of air pollutant and health effects. The recent results of SA on PM2.5 by PCA in different cities of China are summarized in Table 2.

Table 2 Summary of the recent studies on PCA of PM2.5 in China
Full size table

The outcomes of PCA method are very useful to understand the relationship between the source of air pollutant and the exposure analysis. However, it provides an only qualitative assessment of source variation, with some limitation on utility in health studies. Some advanced receptor models such as PCA (absolute PCA), CMB (chemical mass balance), and PMF (positive matrix factorization) are now being used to understand the source contribution (i.e., specific concentration increments) and are of help to the researcher who is working on chemical exposure analysis of fine PM.

3.3 Other SA methods and summary

Figure 2 summarizes prominent source apportionment studies and its key findings in China. In addition to the use of coal for electricity generation, goal is an important fuel for cooking in North China. Vehicular traffic is another considerable source of PM2.5 in most part of China. Fine particles could not settle immediately after their formation. Due to rapid urbanization and industrialization with higher energy consumption in China after 1985s, air pollution has become a major problem for human health. Moreover, a comprehensive summary of relevant past SA studies is presented in Table 3.

Fig. 2
figure 2

A systemic diagram of recently used source appointment methods of PM2.5. Reported data are taken from different previous studies such as PMF (Shi et al. 2016; Masiol et al. 2014; Gu et al. 2014; Gao et al. 2014; Tao et al. 2013), PCA (Qin et al. 2016; Meng et al. 2015; Zhao et al. 2014), CMB (Zhang et al. 2014a; Gao et al. 2013), statistical/trajectories (Li et al. 2015a), CMAQ; Community Multiscale Air Quality (Wang et al. 2014; Ying et al. 2014a, b), PSCF; Potential Source Contribution Function (Zhang et al. 2014b), Radiocarbon (Liu et al. 2013, 2014), elements (Yu et al. 2014; Yang et al. 2014) and EC (Huang et al. 2014a; Zhang et al. 2014c)

Full size image
Table 3 Summary of SA studies on PM2.5 at different places in China
Full size table

Chinese scientists started their work on fine PM monitoring and measurement in the 2000s, while standards were promulgated in 2012s in China (Yang et al. 2011; Gu et al. 2010; Feng et al. 2005). In China, the national standard is 15 µg m−3 for annual and 35 µg m−3 for 24 h average (Gautam et al. 2016c). Several studies have been conducted to assess the source of PM2.5 in different places of China (Liang et al. 2016; Huang et al. 2012; Yang et al. 2011; Cheng et al. 2011). Coal burning, vehicle movement, industrial pollution, and secondary aerosol formation are the major source of fine PM in megacities of China. However, dust storm affects the northern part of China every year (early winter to spring). Zheng et al. (2005) reported that 36% of the PM2.5 mass is contributed by dust only, which has not yet well controlled. Most of the measurement and modeling technique of USA are being applied in China for air pollution control. However, a technique is very much required to explain data collected from the nation-wide network. This technique has been applied in the USA (i.e., IMPROVE 2011), but it is a challenging area for research in China.

4 Future perspectives and conclusions

A number of key research challenges and associated future direction on the assessment and mitigation of PM2.5 can be foreseen. These may include:

4.1 Measurement and methodology

The assessment or monitoring process should be accurate and following the international guidelines, in order to minimize the error during data collection. Therefore, methods should be developed according to location and climate condition to get more accurate details such as chemical composition and quantitative data, for making successful strategies or policies.

4.2 Chemical analysis

Organic and related microorganisms which are present in very high concentrations in PM2.5 are yet to be identified. To assess the tempo-spatial variation and source identification of PM2.5, tracer methods using ions, elements, radiocarbon, and EC need to be improved.

4.3 Particle formation

Studies on the assessment of secondary aerosol formation in rural areas of China are lacking. More studies should be carried out on the secondary aerosol formation to identify their impact on primary users during cooking and heating especially in the rural area, China.

4.4 Monitoring network and standards

According to MEPPRC (2015), China has developed huge monitoring network with 1436 monitoring stations in 367 cities to understand the air quality status and concentration level of PM2.5 to develop the air quality management. Similarly, China has minimized the concentration level of PM2.5 in different cities by awareness through free mobile APPs, online facility to monitor concentration from static and mobile sources, ban on certain vehicles and machines run by diesel, promotion of environment-friendly transport for public, and launching of programs for adopting new cleaner fuel technology for cooking and heating. However, most of these initiatives are confined to urban or township area. Promotion and implementation of these policies and installation of PM2.5 monitoring network in rural areas are therefore needed for effective implementation of strategies or policies for air quality management across the country.

PM2.5 concentrations in China are reported well in the literature and scientific reports (MEPPRC 2015; Zhang and Cao 2015). There are several current and upcoming issues (i.e., source, effects, and control) very common with PM2.5 exposure in China (Zhang and Cao 2015), and they are indicating attention on PM2.5 related issues. A systematic approach to the act of research (academic), anthropogenic activities, private agency, and government authority can minimize the pollution level in China (Fig. 3).

Fig. 3
figure 3

A systemic approach to address the PM2.5 emission and exposure

Full size image

The outcomes from academic and research could provide correct information about the sources and effects of PM2.5 which will help in improving the regulations/standards and controlling the PM2.5 concentration profile. The actions of government and the private agency are very important to develop new regulations/standards commensurate with the outcomes of research and academic that will protect human health and help with air quality management. The industry should be bounded by regulations and follow the national standards suggested by the Federal Government and local authorities. Figure 3 suggests that each stakeholder can support others to reduce the level of PM2.5. The general public will move forward to minimize the pollution level by the use of less of diesel vehicles, adoption of the cleaner fuel, increased use of public transport, and support for green technologies.

The current study is a combination of recently published articles from 2012 to 2016 with the PM2.5 title and source apportionment-related study. Several measurements such as concentration variation with respect to time and space and chemical composition are taken in the various parts of the world, especially in China, where the PM2.5 concentration is reported higher as compared to other Asian countries.

The chemical composition of PM2.5, secondary aerosol formation, tertiary aerosol formation, microbes, intervention, climate change, personal and ambient exposure relationship, and exposure apportionment are emerging subject and should be addressed by future research. To assess and control the PM2.5 concentrations, cost-effective and updated instruments for quantitative measurement and new control or intervention studies are highly needed.

Source apportionment, exposure, and emission studies are very necessary to improve or develop the policy to minimize PM2.5 in the surrounding atmosphere. China has conducted to a large extent the source apportionment studies to understand the social consumption, energy structure profile, emission inventories in urban and township area. Identification of PM2.5 source or exposure apportionment in rural area of China should now be rigorously carried out in order to identify the PM2.5 emission source (traditional stove, open biomass burning), fuel types (cow dung, brushwood, forest wood, etc.), emission duration (food type and kitchen use), in order to better understand the PM2.5 level in rural areas. The outcomes of the suggested studies will improve the policy strategies to control the emission and reduce the exposure to PM2.5.

Qualitative and quantitative information on PM2.5 has been found from the past literature. However, further understanding on the PM2.5 issue should be supported by the utilization of multidisciplinary approaches involving: (1) dispersion, deposition and suspension dynamics, (2) the role of meteorology, chemistry, and terrain, and (iii) assessment through remote sensing and computational program.