Outdoor dust has been widely recognized as an important medium and carrier of toxic substances, including metals (Adriano 2001; Duong and Lee 2011; Shi et al. 2011). Previous studies have noted that metals could enter the human body mainly via ingestion, dermal contact, and inhalation (Wei and Yang 2010; Du et al. 2013) and could cause adverse effects on organs and the central nervous system of humans, especially for children (Valko et al. 2006; Yongming et al. 2006; Zheng et al. 2010; Du et al. 2013). In recent years, with the rapid advances of industrialization and urbanization, metals in outdoor dust have many extensive sources, including atmospheric precipitation, various industrial process, coal combustion, and vehicle emission (Ahmed and Ishiga 2006; Qiang et al. 2015). High contents of heavy metals in outdoor dust in many cities have been observed around the world, e.g., Pb. The exceeding 1000 mg/kg mean concentration of Pb has been reported by many authors (Apeagyei et al. 2011; Kong et al. 2011; Okorie et al. 2012; Zhou et al. 2015). A previous study indicated that the concentrations of heavy metals in dust were closely linked to grain size of dust, and fine particles tended to contain higher concentration of heavy metals (Zhao et al. 2010). Different particles size might result in different human exposure to heavy metals via dust intake, because metals could be accumulated in the finer particles (Shi et al. 2011; Cao et al. 2012; Fang et al. 2013). Bioaccessibility of metal in dust is an important factor that directly impacts the risk assessment results. However, a number of previous studies were only focused on the total concentrations of metals, which were obtained by digesting with aqua regia acid or strong acids (Turner and Ip 2007; Lu et al. 2009; Shi et al. 2011; Okorie et al. 2012; Zhou et al. 2015). Therefore, there might be an overestimation of health risk, considering the fact that the measured total concentrations of metals are higher than the actual adsorption amounts of body (Oomen et al. 2002; Shi et al. 2011).

To the best of our knowledge, many previous studies focused on the heavy metals contamination in the mineral areas or industrial zone in China (Huang et al. 2015; Zhou et al. 2015), but little is known regarding heavy metals pollution in nonmineral urban areas. However, urban areas have higher population density, and thus need further study. In addition, previous studies only investigated the particle size associated heavy metal levels or bioaccessibilities of the metals in the bulk dust samples (Zhao et al. 2010; Huang et al. 2014; Pan et al. 2015; Wang et al. 2015). Little is known about the bioaccessibilities of heavy metals in the particle size associated fractions. Previous studies have documented that fine particles have higher surface area and contain higher levels of heavy metals (Mercier et al. 2011; Zhao et al. 2014). Furthermore, fine particles could be resuspended by wind erosion or traffic emission and have a higher tendency to intake by human via dermal contact and inhalation (Shi et al. 2011; Zhou et al. 2015). Consequently, it is important to understand bioaccessibilities of heavy metals in the fine particles.

Therefore, the primary objectives of this study were: (1) to investigate and compare the particle size associated concentrations of the selected metals in outdoor dust from three ring roads, parks and high spots in Chengdu city, China; (2) to determine the bioaccessibilities of metals in stomach and lung phases; and (3) to assess health risk to heavy metals via dust ingestion, inhalation, and dermal contact for residents in Chengdu, China. To our knowledge, this is the first report on the bioaccessibilities of the metals in the fine settled outdoor particles in China.

Materials and Methods

Study Site and Sample Collection

Chengdu is located in western area of Sichuan basin, which also is the capital of Sichuan province and the transportation hub of southwest China. Chengdu belongs to subtropical, humid, monsoon climate zone. The mean annual temperature is 16 °C, and the mean annual precipitation is 900–1300 mm. It covers approximately 12,000 square kilometers. At the end of 2014, it has a population of 14.4 million. As the center of technology, commerce trade in southwest China, Chengdu has a rapid economic development, and thus may suffer a severe contamination. At present, there are three main roads in urban area of Chengdu, namely first-ring road, second-ring road, and third-ring road, and the lengths of which are 19.4, 28.3, and 51.0 km, respectively.

At least a week before sampling period, the weather of Chengdu was cloudy. A total of 27 dust samples were collected from 27 different sampling sites from March 30 to April 3, 2015, including 19 (4 on the first-ring road, five on the second-ring road, 10 on the third-ring road) road dust samples. The other eight came from four parks (XinHua Park, People’s Park, HuanHuaxi Park, Garden Park) and four high spots. Four high spots were selected on the roofs of four different buildings, and the height of each was required to reach more than 18 floors at least. Approximately 200 g of settled dust on the pavement and road was sampled using a vacuum cleaner (Dyson, DC 34, England) and then sealed in a polypropylene bag and labeled. All samples were transported to the laboratory and stored at 4 °C until pretreatment.

Grain Size Fractionation

Leaves, little stones, and other debris were removed by hand. Each sample was sieved with five nylon sieves with different mesh numbers and classified into five grain size fractions of <63, 63–125, 125–250, 250–500, 500–1000 μm. Then, each sorted fraction was weighed and kept in 50-ml polyethylene tubes.

Sample Extraction

In this study, the processes of extraction were divided into two stages: the “stomach” phase extraction and the “lung” phase extraction. The “stomach” phase extraction was performed using the method developed by Le Bot (2010) and Glorennec (2012), including bioaccessible metal digestion and residual metal digestion.

The “stomach” phase extraction was described as follow. Briefly, (1) 0.05 g of dust sample was digested with 20 ml of 1.4 % hydrochloric acid (Darmstadt, Germany) with pH = 1.5 in a 50 ml polyethylene tube at 37 °C for 1 h. After that, the digested solution was centrifuged at 4000 rpm for 10 min, and 10 ml of supernatant was transferred as bioaccessible part in the stomach phase for analysis; (2) the remaining portion continued to be digested with 3.75 ml of aqua regia (3:1 HCl/HNO3) in Teflon tube at 60 °C for 30 min, and then 95 °C for 140 min. After cooling down, the solution was transferred to polyethylene tube as residual part in the stomach phase for analysis. Therefore, the total concentrations of each metal in dust can be described as the summation of contents determined in the above two parts.

The procedures employed in “lung” phase extraction was described by Huang et al. (2014). Small particles (<10 μm) could enter the nose and mouth during normal breathing (Shi et al. 2011). Therefore, in the case of “lung” phase extraction, only grain size fractions <63 μm in each bulk sample were extracted. Briefly, (1) the simulated lung phase solution with pH = 7.4 was prepared according to Huang et al. (2014), and the details are shown in Table S1 (Online Resource); (2) 0.5-g dust sample was digested in a polyethylene tube with 10 ml of simulated lung phase solution at 37 °C for 30 min; (3) the obtained suspension was centrifuged (4000 rpm, 8 min), and 1 ml supernatant liquid was transferred for further ICP-MS analysis.

Chemicals and Reagents

Hydrochloric acid (HCl) and nitric acid (HNO3) used in stomach digestion process were trace metal grade and purchased from Merck (Darmstadt, Germany) and Fisher (Fisher Scientific, St. Louis, MO, USA), and chemicals used to prepare simulated lung solution (listed in Table S1 in Online Resource) were purchased from Sigma Aldrich (America). The standards of metals applied in the determination were obtained from Perkin Elmer (PE, USA) and the internal standard Rh was purchased from O2Si (Charleston, SC). Standard reference material NIST 2584 was obtained from National Institute of Standards & Technology of the US (Gaithersburg, MD). Milli-Q water was used throughout the study.

Instrumental Analysis

The concentrations of 11 selected metals (Cr, Mn, Co, Ni, Cu, Zn, As, Sr, Cd, Sb, and Pb) were analyzed by ICP-MS (NEXION 300X, PE) with an automatic sampler (SC2 DX, ESI). The running parameters of ICP-MS are listed below: RF power of 1200 W, plasma gas flow rate 15 l/min, nebulizer gas flow rate 0.94 l/min, analog stage voltage −1900 V, pulse stage voltage 950 V, scan mode peak hoping, MCA channels 1, dwell time 50 ms, 141 integration time 1000 ms, and readings were taken as triplicates.

Calculation of Enrichment Factor and Heavy Metal Loads

Enrichment factor (EF) is an index measuring the impact of human activity on environment, and calculated using the Eq. (1) (Chen et al. 2015). In this study, Mn is regarded as reference element, and the conservativeness of Mn in dust has received confirmation in previous study (Yongming et al. 2006). Background values of 11 heavy metals of soil in Chengdu, China were derived from China National Environmental Monitoring Center (CNEMC 1990).

$${\text{EF}} = \left( {C_{i - s} /C_{n - s} } \right)/\left( {C_{i - b} /C_{n - b} } \right)$$
(1)

where C i–s refers to the concentration of studied element in dust; C n–s means the concentration of reference element in sample; C i–b is the background value of the selected element; C n–b is the background value of the reference element.

Heavy metal loads are used to compare the contributions of different particle sizes to overall heavy metal pollution of dust. Metal loads on each particle size fraction (GSFLoad) were calculated for each individual sample using Eq. (2) (Sutherland 2003).

$${\text{GSF}}_{\text{Load}} = \frac{{C_{i} \times {\text{GS}}_{i} }}{{\mathop \sum \nolimits_{i = 1}^{m} C_{i} \times {\text{GS}}_{i} }}$$
(2)

where C i is the concentration of metal in different grain sizes of individual sample, mg/kg; GS i represents the mass percentage of each grain size fraction in one sample; m refers to the number of grain size fractions in each sample.

Exposure Assessment

As many studies indicated, there were three main exposure routes for adults and children to heavy metals via outdoor dust, including inhalation, ingestion, and dermal contact (Wei and Yang 2010; Du et al. 2013), and exposure is closely linked to chemical concentration, frequency, duration of exposure, and the different exposed age stages (Reis et al. 2014). In this study, the models adopted to evaluate the three exposure pathways for adults and children were based on the method provided by US Environmental Protection Agency (EPA 1989, 1996). The calculation equations of daily heavy metals dose via ingestion (D ingestion), inhalation (D inhalation) and dermal contact (D dermal) are listed below.

$$D_{\text{ingestion}} = \frac{{C \times {\text{Ing}}R \times {\text{EF}} \times {\text{ED}} \times 10^{ - 6} }}{{{\text{BW}} \times {\text{AT}}}}$$
(3)
$$D_{\text{inhalation}} = \frac{{C \times {\text{In}}\;hR \times {\text{EF}} \times {\text{ED}}}}{{{\text{PEF}} \times {\text{BW}} \times {\text{AT}}}}$$
(4)
$$D_{\text{dermal}} = \frac{{C \times {\text{SA}} \times {\text{SL}} \times {\text{ABS}} \times {\text{EF}} \times {\text{ED}} \times 10^{ - 6} }}{{{\text{BW}} \times {\text{AT}}}}$$
(5)

where C is the concentration (mg/kg) of heavy metal in outdoor dust. In this study, the bioaccessible concentration of each heavy metal was employed: IngR and InhR mean the ingestion rate (mg/day) and respiration rate (m3/day), respectively; EF refers to exposure frequency (day/year); ED is exposure duration (year); PEF is a particle emission factor (m3/kg); SA is exposed skin area (cm2); SL is skin adherence factor (mg/cm2/day); ABS is dimensionless dermal absorption factor; BW is average body weight (kg); AT is averaging time (day). The specific values of these parameters are summarized in Table S2 (Online Resource). To obtain a better understanding of daily human exposure amount of heavy metals in outdoor dust for children and adults, two exposure scenarios (A and B) were employed. Similar to the scenarios used by Gan et al. (2014, 2015), the median concentration of heavy metal, mean ingestion rate, and mean respiration rate were applied in Scenario A as an estimation of mean exposure level, and 95th percentage concentration of heavy metal, high ingestion rate, and respiration rate were used in Scenario B on behalf of a high exposure level.

Based on calculated values according to the Eqs. 3–5 listed above, the corresponding hazard quotients (HQs) for noncarcinogenic risk and cancer risk (CR) could be acquired using Eqs. 6 and 7. Among analyzed heavy metals in this study, only Cr, As, and Pb are carcinogenic.

$${\text{HQ}} = \frac{D}{{{\text{Rf}}D_{0} }}$$
(6)
$${\text{CR}} = D \times {\text{SF}}$$
(7)

where RfD 0 refers to the reference dose, SF means the slope factor. In this study, the values of RfD 0 and SF for each metal were obtained from regional screening levels (EPA 2010), except for Pb, which was derived from the Guidelines for Drinking Water (WHO 1993). While for exposure estimate of the dermal contact pathway, RfD ABS and SFABS were calculated by Eqs. 8 and 9, which were suggested by U.S. EPA (2004). The values of RfD 0 and SF are shown in Table S3 (Online Resource).

$${\text{RfD}}_{\text{ABS}} = {\text{RfD}}_{0} \times {\text{ABS}}_{\text{GI}}$$
(8)
$${\text{SF}}_{\text{ABS}} = \frac{\text{SF}}{{{\text{ABS}}_{\text{GI}} }}$$
(9)

where ABSGI is gastrointestinal absorption value, the reference data of ABSGI of heavy metals was obtained from the US EPA (2004). Due to the absence of ABSGI values of Co, Sr, Zn, and Pb, RfD 0 and SF also were used in the exposure evaluation of dermal contact pathway instead of RfDABS and SFABS.

Hazard index (HI) was expressed as the sum of individual HQ value of each metal. If HI values >1, indicating there might be a potential health risk. On the contrary, if HI values <1, suggesting there might be a relatively low no-carcinogenic risk. For carcinogens, the values of CR for different metals between 1 × 10−6 and 1 × 10−4 can be tolerable. A value higher than 1 × 10−4 means that 1 in 10,000 people may develop any type of cancer from lifetime exposure to carcinogenic hazards (Lim et al. 2008).

Quality Control and Quality Assurance

Procedural blanks were employed with each batch of 20 samples during digestion, and the standard reference material NIST 2584 was only processed by aqua regia due to the absence of related bioaccessibility data. The analysis results of metals in NIST 2584 are listed in Table S4 (Online Resource). In addition, three samples were selected randomly by triplicate and spiked mixed metal standard, and the mean recoveries and standard deviation values of the three samples are shown in Table S5 (Online Resource). During detection process, a reagent blank, 0.5- and 50-ng/ml standard solutions at each interval of 20 samples were analyzed to check instrumental blank and memory effect, accuracy, and precision. The detailed results are given in Table S6 (Online Resource). The limit of quantification (LOQ) was calculated according to Gomez-Taylor et al. (2003). The LOQ values also are summarized in Table S5 (Online Resource). It’s worth mentioning that the determination values lower than the limit of detection (LOD) were treated as 0.

Statistical Analysis

Statistical analysis of data was conducted using SPSS 21.0. Kolmogorov–Smirnov test was used to estimate the normal distribution of data. For the data of nonnormal distribution, differences among each heavy metal level in different sampling sites were performed by Mann–Whitney U test. Conversely, it was done by one-way ANOVA. In addition, bi-variables correlation test (Spearman) was implemented to analyze relationship between EF values of diverse metals and different grain sizes (<63, 63–125, 125–250, 250–500, 500–1000 μm).

Result and Discussion

Concentrations of Heavy Metals in Dust

The statistic total concentrations of the heavy metals in the 27 dust samples are shown in Fig. 1. Generally, the mean concentrations of the metals contained in the dust from Chengdu were in descending order: Zn, Mn, Cu, Cr, Sr, Pb, Ni, As, Co, Sb, Cd. The concentrations of the 11 metals except for Co varied considerably, especially for Cu, Zn. The concentrations of Cu and Zn ranged from 40.0 to 5770 and from 212 to 7200 mg/kg, respectively, which is far higher than corresponding values from street dust investigated by Qiao et al. (2013) in Chengdu, indicating the inherent heterogeneity of the outdoor particles. It also exceeded other cities globally, such as Newcastle in England (Okorie et al. 2012), Punjab in Pakistan (Mohmand et al. 2015), Guang Zhou (Huang et al. 2014), and Beijing (Li et al. 2015) in China, implying the existence of heterogeneity in dust from different cities. The differences of climate, population density, and economic development degree between different cities might lead to this result (Acosta et al. 2015).

Fig. 1
figure 1

Concentrations of investigated metals contained in outdoor dust from Chengdu, China (top and bottom of each box represent 75th and 25th percentiles, respectively; top and bottom of each asterisk represent maximum and minimum, respectively; line across inside of each box represents median; small square represents the mean value)

Full size image

According to different sampling sites, the mean level and standard deviation of heavy metals in the dust from the roads, parks, and the high spots are listed in Table S7 (Online Resource). The results showed that dust samples from high spots exhibited highest heavy metal pollution level (p > 0.05) compared with those from the roads and the parks, with the exception of Ni, Cu, and Pb. The average concentration of Zn from the high spots was up to 2440 mg/kg. This could partially explain in that heavy metals in dust from building roofs have extensive sources, e.g., vehicle emission, crustal materials, industrial activities, and coal combustion (Kong et al. 2011). Discrete analysis results are shown in Table S8 (Online Resource). No significant differences occurred in the concentrations of Cr, Ni, and Zn in the high spots, roads, and the parks. The concentrations of Mn, Co, As, Sr, and Cd displayed significantly differences among the high spots, roads, and the parks. Therefore, the sources of metals in Chengdu need further research.

Heavy Metals in Size-Fractionated Outdoor Dust

The mean concentrations of heavy metals in different dust sizes are presented in Table S9 (Online Resource). Among 11 investigated metal elements, the highest concentrations were determined in the particles <63 μm for most metals, including Co (14.3 mg/kg), Zn (1150 mg/kg), As (20.1 mg/kg), Sr (154 mg/kg), Cd (2.12 mg/kg), and Sb (10.3 mg/kg), which are consistent with previous findings. However, the highest contents of Ni (55.0 mg/kg) and Cu (656 mg/kg) occurred in particles with median sizes (125–250 and 250–500 μm). As stated by Han (2008) and Liu (2015), medium-sized coarse particles have lower mobility compared with fine particles and are not easy to be transported by external conditions, such as wind and movement of foot. Therefore, in the case of some metals, medium-sized particles had relatively higher concentrations than fine particles. Different from other ten metals, the mean Pb level in the coarsest fraction (500–1000 μm) was relatively higher than those in the other size fractions; this finding was in line with a previous study (Zhao et al. 2010).The average heavy metal loads of five grain size fractions are shown in Fig. 2. The mean loads of each grain size fraction of dust of 11 determined metals displayed similar distribution. Briefly, the contribution of median sizes (63–125, 125––250, 250–500 μm) dusts accounted for more than 70.0 % of overall heavy metal loads. However, the finest and coarsest fractions only contributed approximately 15.0 and 10.0 %, respectively, which differed from previous study that the maximum loading percentage of the particle fractions occurred in the particles with a grain size of <100 μm (Qiang et al. 2015). These findings suggested that the median size outdoor dusts in Chengdu made greater contribution to overall metal contamination. Therefore, more attention should be attached to removal of medium-size particles in Chengdu.

Fig. 2
figure 2

Average heavy metal loads of different particle sizes

Full size image

Enrichment Factor

The enrichment factors (EFs) and background values of 11 metals in Chengdu are presented in Table S7 (Online Resource). The EFs of Pb in the dust from the roads, Cr, Cu, and Sb in the dust from the high spots, and Cu, Zn, and Sb in the dust from the parks ranged from 2.12 to 4.29. Based on the five metal contamination categories classified by Han et al. (2006), they were classified as the second degree, revealing a moderate contamination degree. The EF values of Cu, Zn, Cd, and Sb in the dust from the roads, as well as Pb in the dust from the parks were smaller than 20 belonging to the third degree, showing significant metal contamination. Severely, EFs of Zn and Cd in the dust from the high spots and Cd in the dust from the parks were larger than 20, implying a very high metal contamination level. The contamination of Cu, Zn, Cd, and Pb in the dust from the roads and the parks might mainly be originated from deterioration of vehicles, the use of organic reagents on vehicles, such as lubricants, vehicle emissions, and atmospheric deposits, whereas the related sources of Zn, Cd, Cr, Cu, and Sb in the dust from high spots might be from vehicle emission, crustal materials, industrial activities, and coal combustion (Kong et al. 2011; Qiang et al. 2015).

Bi-variables correlation test results between EFs of selected metals and different particle sizes (<63, 63–125, 125–250, 250–500, 500–1000 μm) are listed in Table S10 (Online Resource). The results demonstrated that the EFs of Cu, Sb, and Cd with particle size displayed negative correlation, implying that Cu, Sb, and Cd were more inclined to accumulate in fine particles. On the contrary, the EFs of Co and Sr with particle size displayed positive correlation. Co and Sr have higher EF values in coarser particle size fractions, indicating Co and Sr tended to enrich in coarse particles. But for Cr, Ni, Zn, As, and Pb, no correlation was found between the two variables.

Bioaccessibilities of Heavy Metals in “Stomach” and “Lung” Solutions

The bioaccessible concentrations and bioaccessibilities of determined metals in “stomach” and “lung” solutions are shown in Tables S11 and S12 (Online Resource), respectively. In this study, bioaccessibilities of the investigated metals in the two phases varied greatly, especially for As, Cu, and Cr. Cd (68.6 %), Sr (64.9 %), and Zn (60.5 %) exhibited higher bioaccesiibilities compared with the other eight metals in “stomach” phase with medians greater than 50 %. In contrast with the other metals, Sr (0.61 %) displayed relatively higher bioaccessibilities in “lung” phase with mean higher than 0.5 %. Despite using the same digestion method, the mean bioaccessibility values in the simulated stomach fluid in this study were higher than those reported by Le Bot (2010) and Huang (2014), with the exception of Cr, in which the values in the lung phase solution were relatively lower. This might be attributed to the effect of other factors on bioaccessibility, such as pollution sources, sample matrix, chemical species, etc. (Bi et al. 2015; Patinha et al. 2015). The statistics results revealed that statistics distinctions were observed between simulated gastric solution and lung solution (p < 0.05). The mean bioaccessibilities of 11 investigated metals in the gastric solution were significantly higher compared with those in the composite lung simulating serum, especially for Zn, Cd, and Pb. Concretely, the bioaccessibilities of Zn, Cd, and Pb in the stomach phase ranged from 49.0 to 82.0 %, from 47.0 to 79.0 %, and from 24.0 to 69.0 %, respectively, whereas the maximum bioaccessibilities of the corresponding metals in the lung fluid were only 0.83, 0.56, and 0.01 %. In the case of the remaining metals, similar results were obtained. This might be attributed to the existence of inorganic ligands (e.g., Cl−, HCO3 −) and higher pH in simulating lung solution. In the composite lung serum, metals may undergo complexation, readsorption to altered particle surface sites, or precipitation as insoluble compounds (Turner and Ip 2007; Turner et al. 2009; Turner and Radford 2010; Huang et al. 2014), which could result in the reduction of mobilization of metals, and thus reducing the bioaccessibilities of metals in lung solution.

Health Risk Assessment for Exposure to Heavy Metals in Outdoor Dust

For noncarcinogenic heavy metals, the percentage contributions of each exposure pathway for adults and children are shown in Fig. 3. The HQ and CR values for children and adults exposure to metals via different dust exposure pathways are listed in Table S13 (Online Resource). The HI and CR values of selected metals via ingestion, dermal, and inhalation for adults and children are summarized in Table 1. As indicated in Fig. 3 and Table S14 (Online Resource), for 11 studied metals, regardless age group, ingestion was the most important exposure pathway with percentage greater than 65.0 %. The maximum value was up to 99.0 %. Inhalation only occupied less than 0.01 % and could be ignored. As shown in Table 1, for noncarcinogenic heavy metals, in the scenario A, all HI values were lower than 1, indicating low noncarcinogenic risk. However, in the scenario B, HI value of Cr for children was higher than 1 (1.64), suggesting a noncarcinogenic risk might happen to children exposed to outdoor dust in Chengdu. In the case of the scenario A, HI values of non-carcinogenic heavy metals for adults were in the following order: Cr > Pb > Co > As > Mn > Sb > Cd > Zn > Cu > Ni > Sr. For children, similar order was found, namely Cr > Pb > Co > As > Sb > Mn > Cd > Zn > Cu > Ni > Sr. Based on the current study, Cr was main contributor to total noncarcinogenic risk for both children and adults. HI values for children were close to ten times higher than that for adults, and children seemed to suffer more severe risks in outdoor environment. For carcinogenic heavy metals (Cr, As, and Pb), similar conclusion was observed that carcinogenic risks for children were one to two times higher than that for adults. Considering the tolerable range (1 × 10−6 to 1 × 10−4) of cancer risk, Cr in the scenario B and Pb in both A and B scenarios exceeded the safe range with values greater than 1 × 10−4 for adults and children, which implied a potential carcinogenic risk.

Fig. 3
figure 3

Percentage composition of different exposure pathways for children and adults in Chengdu

Full size image
Table 1 Total HI values and CR values of investigated metals for children and adults exposure to metals via different dust exposure routes
Full size table

The total HI and total CR were the summation of HI values of individual noncarcinogenic metal and the summation of CR values of the three carcinogenic metals, respectively. Based on the scenario B, the total HI value was greater than the threshold value (1) for children (3.31), indicating children in Chengdu might suffer noncarcinogenic risk when exposed to outdoor dust. Likewise, in the A and B scenarios, the total CR values for adults and children were higher than 1 × 10−4, suggesting the potential carcinogenic risk might occur for Chengdu residents through outdoor dust intake.

Conclusions

Eleven trace metals were determined in the size fractionated outdoor dust samples from the roads, parks, and the high spots in Chengdu, China, and related human exposure and health risk to metals via dust intake were evaluated. The concentrations of 11 metals except for Co varied considerably. Dust samples from the high spots exhibited the highest heavy metal pollution level compared with the roads and the parks, with the exception of Ni, Cu, and Pb. The statistics results showed no significant differences in the concentrations of Cr, Ni, and Zn among the high spots, roads, and parks. However, the concentrations of Mn, Co, As, Sr, and Cd displayed significantly differences among the high spots, roads, and the parks.

The mean loads of each grain size fraction of dust of 11 determined metals displayed similar distribution. The contribution of median size (63–125, 125–250, 250–500 μm) dust accounted for more than 70.0 % of overall heavy metal loads, and more attention should be attached to removal of medium-size particles in Chengdu. Cu, Sb, and Cd were inclined to enrich in small particles, and Co and Sr tended to accumulate in coarse particles.

Cd exhibited the highest bioaccessibility with a median of 68.6 % in the stomach phase, followed by Sr (64.9 %) and Zn (60.5 %). Sr (0.61 %) displayed relatively higher bioaccessibility in the lung phase with mean higher than 0.50 %. The mean bioaccessibilities of 11 investigated metals in the gastric solution were much higher than the corresponding values in the composite lung simulated serum, especially for Zn, Cd, and Pb.

The health risk assessment results showed that for both children and adults, Cr was main contributor to total noncarcinogenic risk. HI values for children were far higher than that for adults, implying that children seemed to suffer more severe risks in outdoor environment. Ingestion was the most important exposure pathways for both children and adults. Children in Chengdu might suffer non-carcinogenic risk when exposed to outdoor dust contained high metal levels. The total CR values for adults and children were higher than 1 × 10−4, suggesting a potential carcinogenic risk might occur for Chengdu residents through outdoor dust intake. However, considering the fact that only particles <10 μm can reach the lung, so there may be an overestimation of health risk resulting from heavy metals via inhalation, which would magnify the final risk assessment results.