1 Introduction

Dust is the most pervasive and important environmental factor affecting human health (Duong and Lee 2011). Outdoor dust often contains elevated concentrations of a range of heavy metals (Tong and Lam 2000). Exposure to metal-contaminated dust through skin contact and hand-to-mouth contact can adversely affect human health, particularly through unintentional uptake by children in playgrounds and city streets (Saeedi et al. 2012). Nearly half of the world population lives in urban areas, and their health issues and living environment have become a major concern (Shi et al. 2008). Considerable attention has been paid to metal pollution in urban street- and roadside dust throughout the world, and studies have revealed that city dust contains a range of heavy metals, including Cd, Cu, Pb, and Zn (Ordóñez et al. 2003; Ferreira-Baptista and de Miguel 2005; Saeedi et al. 2012). In China, recent rapid urbanization, industrialization, and increased vehicular traffic have resulted in heavy metal accumulation in urban dust (Duzgoren-Aydin et al. 2006; Han et al. 2008; Li et al. 2012a, b). Most studies have shown that metal concentrations in dusts vary between different cities with level of development and environmental characteristics, but still lacking is a comprehensive characterization of the distribution of heavy metal levels in Chinese cities. A review from Wei and Yang (2010) of heavy metal contamination in urban soils, urban road dusts, and agricultural soils of China showed that Cd, Cr, Cu, Ni, Pb, and Zn are widespread in urban soils and urban road dusts. In China, the development of capital cities is much higher than other cities. The increases in anthropogenic Pb, Zn, Cu, Ni, and Cr in street dusts can most likely be attributed to rapid development (Ahmed and Ishiga 2006).

The primary aim of this paper is to present a characterization of the spatial distribution of heavy metals including Cd, Cr, Cu, Ni, Pb, and Zn in urban outdoor dust in capital cities in China and to discuss the causes of the differences in heavy metal levels between cities by summarizing and analyzing data for 20 provincial capitals from the published scientific literature.

2 The level and accumulation of heavy metals in urban dust in China

2.1 The level of heavy metals in urban dust

Properties of material in dust are highly variable in time and space, affecting the choices of sampling procedure, digestion procedure, and analysis method (Ferreira-Baptista and de Miguel 2005).

In the reviewed investigations (Table 1), dust samples were almost unanimously collected by brush; digested using HNO3–HF–HClO4; and analyzed by AAS, ICP-MS, ICP-AES, and/or XRF, all of which are acceptable. Although the particle size retained for analysis is different, direct comparisons among the results of different investigations are reasonable under such similar sampling, digestion, and analysis methods.

Table 1 The level of heavy metals in urban dust in China
Full size table

The concentration ranges of Cd, Cr, Cu, Ni, Pb, and Zn in urban dust in the studied Chinese cities were 0.47–11.0, 50.3–233, 49.5–261, 15.6–129, 50.4–375 and 196–2716 mg/kg, respectively. The highest levels of Cd, Cr, Cu, Ni, Pb, and Zn appeared in Shijiazhuang, Chongqing, Lanzhou, Chengdu, and Nanchang, respectively, with values 23.4, 4.63, 5.27, 8.27, 7.44 and 13.9-fold higher than the lowest values, respectively. The variations of Cd and Zn in city dust were considerably larger than that of other heavy metals. The data in Beijing from two investigations is similar in spite of different particle sizes being retained for analysis, while results of two investigations in Urumqi display some difference.

The geometric mean values of Ni and Cr in urban dust of China were similar to or less than the crustal levels, and those of Cd, Cu, Pb, and Zn were considerably higher than crustal levels. In particular, the geometric mean values of Cd, Pb, and Zn were respectively 10.2, 9.7, and 7.0 times crustal levels, demonstrating their heavy accumulation in urban dust in China. The following discussion will focus on the four elements with higher-than-crustal mean levels.

2.2 The spatial distributions of Cd, Cu, Pb, and Zn in urban dust

Spatial distributions of Cd, Cu, Pb, and Zn were mapped using ArcGIS v.9.1 (ESRI Co, Redlands, USA). The distribution of Cd showed no apparent spatial pattern, with a maximum value appearing in the north and a minimum in the northeast. A clear pattern of higher accumulation in southern cities than in northern ones was seen in the distributions of Cu, Pb, and Zn. The levels of Cu and Pb in cities in the southwest and east of China were higher than in the central and southern parts of China, and the highest level of Zn appeared in central China. A possible reason behind the spatial differences of heavy metal contents will be discussed in Sect. 3.

2.3 Dust heavy metals in different city functional areas

Several studies reported heavy metal concentrations in urban dust for different city functional areas, but the reported functional areas were not the same for all cities. We chose four major functional areas, including industrial areas (IA), commercial areas (CA), traffic areas (TA), and residential-educational areas (REA), to study the distribution of heavy metals in the urban dust of different functional areas within cities. Urban dust heavy metal data for four functional areas of seven cities—Urumqi (Liu et al. 2009), Shenyang (Li et al. 2008), Xi’an (Han et al. 2008), Taiyuan (Cao 2012), Nanjing (Hu et al. 2011), Guiyang (Li 2013)—were collected, and the ratios of their geometric means to the crustal levels were calculated (Table 2). CA and REA had more Cd in their dust than did IA and TA. No significant difference was detected for Cu in different functional areas. Pb and Zn in IA were significantly higher than in the other three functional areas. Pb levels were higher in REA than in CA and IA, but Zn levels showed no obvious difference between those three functional areas.

Table 2 The ratios of the geometric means of heavy metals to the crustal levels
Full size table

For the urban area as a whole, as shown by Duong and Lee (2011), heavy metals in dust have a variety of sources that vary between different functional areas. Industrial activity is among the main factors affecting heavy metal levels in urban street dust, and Qin (2008) showed that the concentrations of Cu, Pb, and Zn in industrial dust in the city of Lanzhou were 368, 152, and 582 mg/kg, respectively, and industrial activity contributed 61.5 % of the heavy metals in urban dust.

The heavy metal levels in urban dust in TA was mainly affected by traffic activity, including abrasion of tires, brakes, and other auto parts; tail pipe emissions; and re-raised and sedimented road dust (Apeagyei et al. 2011). Zn release mainly results from the abrasion of tires, to which zinc oxide is added as a catalyzer during the vulcanizing process (Adachi and Yoshiaki 2004). Cu comes from brake abrasion (Apeagyei et al. 2011), and Pb may come from lead tire weights and tail pipe emissions (Caravanos et al. 2006).

Composition of resedimented road dust is more complex, containing harmful heavy metals from varied sources. Quantitative estimates calculated by Kumar et al. (2001) indicate that road dust contributes 41 % to resedimented dust, vehicular emissions 15 %, marine aerosols 15 %, metal industries 6 %, and coal combustion 6 %. Thus the resedimentation of dust is a common source of multiple heavy metals in TA.

Heavy metals in urban dust in CA come from soil, atmospheric deposition, and transportation activity. Heavy metal levels in CA are also related to peeling of urban paints (Turner and Sogo 2012), goods abrasion, and human-mediated transport of dust-containing heavy metals. Moreover, the contribution of pedestrian traffic to transporting and redistributing dust and heavy metals cannot be ignored (Hunt et al. 2006).

The sources of heavy metals in CA and in REA partially overlap, but REA have a unique source that is almost absent in other areas—indoor dust. Some indoor articles and materials may contain heavy metals (Lisiewicz et al. 2000), and construction abrasion, furniture, and cosmetics can increase heavy metal content in indoor dust as well. The metals in indoor dust may travel outside through windows and balconies or through dumping of trash. The levels of Pb, Cd, and other elements in indoor dust have been reported to be higher than in outside dust and soil (Liggans and Nriagu 1998; Tong and Lam 2000). Our work shows that the dust levels of Pb and Cd in REA are higher than in CA and TA, which may reflect the influence of indoor dust.

3 The cause of heavy metal level differences between cities

3.1 Natural background

Soil is one of the sources of dust (Hunt et al. 2006). It is an interesting question to examine whether heavy metal concentrations in dust are related to their background values in soils. Background values of the heavy metals were collected from literature (Ref Yang et al. 2010; Li et al. 2008, 2010, 2011; Wei et al. 2009; Zhao 2008; Qin 2008; Han et al. 2006; Chen et al. 2013; Shi et al. 2008; Zhang and Zhang 2007; Tang 2010 in Table 1) and from Background Value of Soil Elements in China (The National Environmental Protection Agency 1990). Correlation analysis conducted with SPSS 13.0 showed that the concentrations of Cd (P = 0.001), Cr (P = 0.016), and Cu (P = 0.004) in dust were significantly correlated with their background values in soils. Zn, Pb, and Ni concentrations in soil showed no significant correlation with those in dust, which might be influenced by more factors, e.g., industry types, traffic activity, climatic conditions, etc., than the corresponding soils.

3.2 Industrial activity

Besides element background values in soils, the heavy metal variation between the cities could be attributed to city history, industrial types and distributions, energy types and consumption, population, vehicles, and many other factors (Ordóñez et al. 2003). Industry related to metallurgical activity, metal-processing, and transportation of raw materials is the most important factor influencing the metal level in dust. Different types of industry may release different heavy metals (Long et al. 2010; Liu et al. 2012). The most common element released by the iron and steel industry is Zn, and the automobile industry may increase the levels of Cu and Zn in urban dust (Meng et al. 2007). Machinery and electrofacing may be related to the release of Cd, and the thermal power industry may be connected to the release of many elements, including Pb. The distribution pattern of higher accumulations of Cu, Pb, and Zn in southern cities than in northern ones may be partly due to relatively more smelting and processing of metals in southern cities.

3.3 Population and vehicle

A comparison study of two cities of a large urban area with population of 2.3 million and a small one with population of 0.3 million (Charlesworth et al. 2003) shows that the levels of some hazard elements in large urban areas are higher than those in smaller ones, and comparison of the populations of the cities with heavy metal concentrations similarly shows that in general, the larger the population, the higher the heavy metal concentration in street dust. This is particularly true of Cd and Pb (Fig. 1).

Fig. 1
figure 1

Spatial distributions of Cd, Cu, Pb, and Zn in urban dust of China

Full size image

Data for population density, area, and public transport vehicle stock (tram, bus, and taxi) of these cities during the sampling year or the year before or after the sampling year were obtained from China City Statistical Yearbook (The National Bureau of statistics, the city social economy investigation division 2004, 2006–2010, 2012). These factors’ possible effects on heavy metal (Cd, Cu, Pb, and Zn) levels in urban dust were evaluated. The results are shown in Figs. 2 and 3. Correlation analysis conducted with SPSS 13.0 showed that the concentrations of Cd (P = 0.000) and Zn (P = 0.015) in dust were significantly correlated with population density; the correlation coefficients (R) were 0.825 and 0.548, respectively. No significant correlation was detected between Cu and Pb concentration and city population. The larger the population, the more complex the urban public facilities are. Urban public facilities (e.g. walls, lamp posts, railings), furniture, and other consumer products (e.g. cosmetics) all likely contain Cd, Cu, Pb, and Zn (Lisiewicz et al. 2000; Turner and Sogo 2012). This work showed the effects of population on Cd and Zn in the urban dust, but not on Pb and Cu (Table 3).

Fig. 2
figure 2

Correlation between heavy metal levels in urban dust and population density (R relation ratio, N number of the data)

Full size image
Fig. 3
figure 3

Correlation between heavy metal levels in urban dust and public vehicular density (per km2)

Full size image
Table 3 Comparison between urban dust heavy metal levels in China and in selected cities throughout the world
Full size table

Cd also significantly correlated with the public vehicle stock. The concentrations of Cu, Pb, and Zn showed little correlation with vehicle stock. Traffic activity is one source of heavy metal in dust; the heavy metal released by traffic activity is also related to traffic flow.

From the analysis shown in the figures above, it could be inferred that population density has a stronger effect on Cd and Zn in urban dust than on Cu and Pb, while the effect on Cu, Pb, and Zn in urban dust by vehicular density is unclear, and may be confused by other factors.

4 Comparison of heavy metal levels in urban dust between China and cities in other countries

Compared with urban dust data from thirteen cities in different countries, the 20 provincial capitals in China were lower in Cd and Cu, comparable in Zn, and lower than most of the cities in Pb except for Birmingham, Dhaka, and Ottawa. Birmingham, Dhaka, and Ottawa are the only cities that have significantly lower heavy metal concentrations than the Chinese average.

Taking the geometric mean of Cd, Cu, Pb, and Zn concentrations in street dust of the 13 cities to represent the world average, means of Cd, Cu, and Zn levels of Chinese cities are close to the world average, and Pb is significantly lower than the world average. According to literature reports and development features of each city, this finding might be related to industry type as well as range and type of paint used in cities. Further discussion on this interesting question could be carried out in a subsequent study.

5 Conclusions

An analysis of heavy metal levels in urban dust in 20 capital cities in China shows that the geometric mean values of Ni and Cr are less than or close to the crustal levels, and those of Cd, Cu, Pb, and Zn are significantly greater than the crustal levels.

The spatial distributions of Cu, Pb, and Zn in urban dust are similar: all exhibit a pattern in which the metal levels in cities located in the south of China are higher than in the northern parts.

Commercial areas and residential-education areas accumulate more Cd in dust than industrial areas and traffic areas, and industrial areas and residential-education areas accumulate more Pb than commercial areas and traffic areas. Zn levels in the dust of industrial areas are significantly higher than in other areas, but Cu levels exhibit no significant difference between functional areas.

A positive correlation exists between Cd and Zn in urban dust and population density, but no obvious correlation was detected between Pb and Cu and population density. Vehicular density (per km2) has little effect on Cu, Pb, and Zn concentrations in dust.

Urban dust Pb of cities in China is lower than the world average calculated with data of thirteen cities in different countries. Cd, Cu, and Zn levels of China are close to the world average values.