Introduction

Most modern vehicles are fitted with a three-way catalytic converter that uses platinum group elements (PGE) (mostly, Pt, Pd and Rh) as catalysts to minimize toxic gas emissions produced during gas combustion. Catalysts remove about 90% of total carbon monoxide, unburned hydrocarbons, and nitrogen oxides (NO x ) from exhausts and transform these pollutants to more innocuous carbon dioxide, nitrogen, and water (Barefoot 1997). However, PGE are emitted both in nanometer and micrometer particles on the surfaces of road dusts, roadside soils and plants during vehicle operation as evidenced by atmospheric particle analysis (Palacios et al. 2000; Hooda et al. 2007; Cicchella et al. 2008). The emission rates of Pt as high as 800 ng/km depend on speed, vehicle condition and exhausting temperatures (Moldovan et al. 2002). Although emission rates of PGE are relatively low, an estimated 500 million vehicles worldwide are equipped with catalysts (Barbante et al. 2001), resulted in increasing interest in their environmental behavior worldwide in the recent past. PGE contamination from vehicle emission is in metallic and oxide forms. They may be soluble, transported by water, deposited in sediments and soils, eventually they may enter the food chain and the living organisms. High levels of PGE in human body may cause asthma, nausea, hair loss, abortion, dermatitis and other health problems (Ravindra et al. 2004). Previous studies have found elevated PGE levels in urban airborne dusts (e.g. Rauch et al. 2005; Wichmann et al. 2007), road dusts and roadside soils (e.g. Lesniewska et al. 2004; Morcelli et al. 2005; Hooda et al. 2007; Cicchella et al. 2008), sediments (e.g. Rauch et al. 2004a; Whiteley and Murray 2005), and roadside vegetation (e.g. Zechmeister et al. 2006; Hooda et al. 2008). Bioaccumulation of PGE contaminants in aquatic (e.g. Zimmermann et al. 2004, 2005; Sures and Zimmermann 2007) and internal organs of birds has also been extensively studied (e.g. Jensen et al. 2002; Ek et al. 2004).

Automobile catalytic converters were introduced to China in 1993. As the biggest developing country in the world, the number of vehicles in China increases annually, with addition of about 7 million new vehicles each year in the recent past, and 10 million new vehicles in coming years. The rapid increase of vehicles equipped with catalysts may pose a problem of PGE contamination in urban areas. Thus, monitoring PGE contamination is important in populated regions, such as the Pearl River Delta region, South China. Although numerous studies of PGE contamination have been carried out in other countries, only a few similar studies with limited data have been reported in China (Kan and Tanner 2004; Fang et al. 2005; Kan and Tanner 2005; Wang and Qin 2007; Wang et al. 2005, 2007; Wang and Sun 2009; Pan et al. 2009). The aim of this study is to evaluate PGE contamination in road dusts and roadside soils by comparing with uncontaminated soils collected from three major cities, Shenzhen, Guangzhou, and Hong Kong in the Pearl River Delta region. A comparison is made for the PGE dataset between Hong Kong and mainland China that have different automobile catalyst systems.

Methodology

General information and sample collection

The Pearl River Delta region is one of the most populated areas in China (Fig. 1). Three major cities, Guangzhou, Shenzhen and Hong Kong, have populations of about 10, 9 and 7 million, respectively. They have registered vehicles of about 1, 1 and 0.6 million, respectively. All automobiles in Hong Kong are imported from overseas, but most of automobiles in Guangzhou and Shenzhen are made in China.

Fig. 1
figure 1

Geographic location of Pearl River Delta region, South China

Full size image

The sampling strategy was designed to ensure a representative coverage of areas of high traffic densities with no industrial input, i.e. the downtown road and its main intersections, as well as uncontaminated soils (Table 1). Sample collection was carried out in 2007 and 2008 during winter and spring time, approximately 2 weeks after the last rain. Road dust samples were hand-brushed and collected from the main street edge and the intersections located in the center of the city. Roadside soil samples were collected from the soil layer (about 5 mm thickness) of gardens along the main street about 3 m away from the street edge. Uncontaminated soil samples were collected from areas at least 200 m away from traffic and may represent samples with natural PGE background values.

Table 1 Sampling sites and their characteristics
Full size table

The collected samples were dried for 12 h at 105°C, homogenized and sieved to particle sizes of 150 mesh with a plastic sieve and stored in polypropylene bottles at room temperature for analysis.

Instrumentation and reagents

The instrument used in this study is an ELAN DRC-e ICP-MS in the State Key Lab of Ore Deposit Geochemistry, Institute of Geochemistry, Chinese Academy of Sciences, Guiyang. The sensitivity of the instrument was adjusted to >40,000 cps for 1 ng/ml of 115In.

HCl and HNO3 were purified by sub-boiling distillation. Water was obtained from 18 MΩ cm grade Millipore purification system. Te solution (about 2 mg/ml) and SnCl2 solution (20%, w/v) were purified by Te-coprecipitation (Qi et al. 2004). The Carius tubes and Os distillation system are same as that described by Qi and Zhou (2008). Spike solutions with enriched stable isotopes, 194Pt, 105Pd, 101Ru, 193Ir, and 190Os, were prepared from pure metals (US Services Inc., Oxbow, NJ). An ICP multi-element standard solution of 100 μg/ml of Pt, Pd, Rh, Ru, Ir, and Au (AccuStandard, USA) was used and diluted as needed for calibrating the spike solutions and the concentrations of mono-isotopic element, Rh and Au.

Analytical procedure

In this study, powdered samples (5,000g for road dust, 10,000g for roadside soil and uncontaminated soil) were accurately weighted and ignited in a porcelain crucible at 650°C to remove the organic materials, and then transferred to a 75 ml Carius tube. Appropriate amount of enriched isotope spike solution containing 194Pt, 105Pd, 101Ru, 193Ir, and 190Os was accurately added and mixed with 15–30 ml aqua regia depending on the sample mass. The sealed Carius tube was placed in a custom-made high-pressure autoclave filled with water to prevent explosion of the tube when heated to about 250°C (Qi and Zhou 2008). After 10 h, the Carius tube was cooled and the contents were transferred to a 50-ml centrifuge tube. After centrifuging, the upper solution was transferred to a distillation system for Os distillation. The collected Os solution was used for measuring the Os concentration by isotope dilution and the ratios of 187Os/188Os. After distillation, the remaining solution was used to pre-concentrate PGE by Te-coprecipitation and all the interference elements are removed using a cation exchange resin and P507 extraction chromatography resin combined in the same column (Qi et al. 2004). Platinum, Pd, Ru, and Ir were measured by isotope dilution, whilst 194Pt was used as the internal standard to calculate the concentration of the mono-isotope elements Rh and Au. The total procedural blanks were lower than 0.0003 ng/g for Os, 0.003 ng/g for Ru, Rh and Ir, 0.04 ng/g for Pd, and 0.02 ng/g for Pt (Table 2).

Table 2 Blank (ng) and analytical results (ng/g) of reference materials, BCR-723
Full size table

Quality control

An urban dust reference material, BCR-723 (0.5 g), was used for quality control in this study and digested both directly and after ignition at 650°C to investigate the loss of PGE during this process. Analytical results are shown in Table 2. The results for direct digestion and after ignition agree well, indicating no lost of PGE during the ignition treatment. The results of Pt, Pd and Rh are in agreement with the certified values and Ru, Ir and Os agree well with reported values (Meisel et al. 2003).

The recovery of Au was only about 80% when Te-coprecipitation technique was used for pre-concentration. In this study, the solution was heated to boil for about 1 h after the addition of SnCl2 to increase the recoveries of PGE and Au (Amosse 1998), and the standard solution was prepared same as the procedure of sample preparation described above. Thus, the recoveries of Au for the standard and samples are nearly identical.

Results and discussion

The background values

To identify the elevated PGE contents through anthropogenic contributions in environmental samples, a local natural background level is required to establish as a reference point. Therefore, the local uncontaminated soils can be considered as the background values. As listed in Table 3, the background values have PGE abundances generally similar to those reported for the Upper Continental Crust (Wedepohl 1995; Peucker-Ehrenbrink and Jahn 2001) and the crust levels in eastern part of China (Chi and Yan 2006).

Table 3 PGE abundance (ng/g) for upper crust and local background values
Full size table

Platinum, palladium and rhodium

The PGE concentrations of road dusts and roadside soils from Guangzhou, Shenzhen and Hong Kong are given in Table 4. The road dusts and roadside soils have Pt, Pd and Rh contents significantly higher than the background values. The road dusts contain the highest PGE concentrations (Table 4), suggesting inputs from automobile catalyst system. Pt, Pd and Rh concentrations of the road dusts from Hong Kong range from 12.1 to 187, 12.4 to 287, 1.03 to 29.4, respectively. Those from Shenzhen range from 14.1 to 178, 33.8 to 514, 4.9 to 53 ng/g, respectively, those from Guangzhou range from 4.66 to 48.2, 13.0 to 554, 2.14 to 14.5 ng/g, respectively. These values are comparable with those reported for other countries (Lesniewska et al. 2004). The large range of Pt, Pd and Rh contents in the dust samples from these three cities may be caused by different traffic densities and sample spots from different distance away from the road as reported previously (Jarvis et al. 2001; Morcelli et al. 2005; Whiteley and Murray 2005). All the soil samples have Pt, Pd and Rh much lower than the dust samples. This may also reflect the longer transportation distance of PGE to the soil samples rather than the dust samples. Pan et al. (2009) reported Pt, Pd and Rh concentrations in soil samples from five cities in China (Beijing, Guangzhou, Hong Kong, Macao and Qingdao). The highest values are 160 ng/g Pt, 107 ng/g Pd and 34.5 ng/g Rh for the Hong Kong soils which are higher than the soil samples but lower than the dust samples in this work. Wang and Sun (2009) also reported Pt and Pd concentrations (2.9 and 2.8 ng/g) in roadside soils from Xuzhou, China. Their values are lower than those for the Pearl River Delta region reported here.

Table 4 The concentration of PGE (ng/g) in road dusts, roadside soils collected from Guangzhou, Shenzhen and Hong Kong
Full size table

Automobile catalysts mainly contain Pt, Pd and Rh, and may release into the environment through exhausting system. Thus, these three elements show positive correlations for the roadside dusts and soils samples. Our data indicate that there are clearly positive correlations between Pt and Pd, Pt and Rh and Pd and Rh (Fig. 2) for the samples from three cities in the Pearl River Delta region, again consistent with automobile catalyst sources. The Pt/Pd ratios of the dust samples from Shenzhen (0.35–2.17, with an average of 0.93) are comparable with that in Hong Kong (0.3–1.98, with an average of 0.94; Table 4). The Pt/Pd ratios of the dust samples from Guangzhou (0.21–1.35, with an average of 0.51) are lower than those from Shenzhen and Hong Kong. The Pt/Pd ratios in our samples have smaller range than those reported previously (0.3–7.3) (Pan et al. 2009). The Pt/Rh ratios of the dust samples from Hong Kong (4.4–14.9, with an average of 8.51) are clearly higher than those from Shenzhen (2.87–10.8, average 3.96) and Guangzhou (1.49–5.62, average 3.37) (Table 4), but similar to those reported previously by Pan et al. (2009). The Pt/Rh and Pt/Pd ratios of our samples are similar to those reported previously for urban dusts and soils (Gomez et al. 2002; Lesniewska et al. 2004), within the ranges of Pt/Rh 5–16 and Pt/Pd 1–2.5 (Ely et al. 2001). They are within the range in commonly used gasoline catalysts (Pt/Rh = 5) (Palacios et al. 2000). The different Pt/Pd and Pt/Rh ratios for the samples from Guangzhou, Shenzhen and Hong Kong appear to reflect different automobile catalyst systems used in Hong Kong and mainland China.

Fig. 2
figure 2

Plots of Pt versus Pd, Pt versus Rh, Pd versus Rh and Pt/Rh versus Rh for the samples from Pearl River Delta region

Full size image

Ruthenium and iridium

Until present, other than Pt, Pd and Rh, only limited studies have been reported for automobile catalyst-related Ru and Ir in environmental samples (Muller and Heumann 2000; Ely et al. 2001; Fritsche and Meisel 2004; Rauch et al. 2004a, b, 2006). Iridium has been used in some new diesel catalysts together with Pt (Palacios et al. 2000). Muller and Heumann (2000) reported elevated Ir and Ru concentrations in roadside soils and tunnel dust samples. They found that Ru concentrations in the samples are related to the distance from roads. Rauch et al. (2004a) reported elevated Ir and Ru contents in sediments in an urban lake near Boston, and suggested that Ir is present as catalyst component or impurity and Ru is present as catalyst impurity. Ely et al. (2001) also reported elevated Ir concentrations in soils, and suggested the Ir is probably present as an impurity in automobile catalysts.

The Ir and Ru concentrations of our samples are clearly higher than the background values and Upper Continental Crust (Wedepohl 1995; Peucker-Ehrenbrink and Jahn 2001; Chi and Yan 2006; Table 4; Fig. 3). There are clearly positive correlations between Pt and Ir, Pt and Ru, and Ir and Ru in our samples, except no correlations between Pt and Ru, and Ir and Ru in Guangzhou (Fig. 3). We suggest that the elevated Ir and Ru in soil and dust samples from the Pearl River Delta region are related to emission of automobile catalyst in which these elements occur as impurities.

Fig. 3
figure 3

Plots of Pt versus Ir and Pt versus Ru for the samples from Pearl River Delta region

Full size image

Osmium

During Earth’s mantle–crust differentiation, Os is compatible and Re is incompatible. As a result, crustal materials have very low Os abundance and high Re/Os ratios which yield extremely radiogenic Os isotopic signatures with high 187Os/188Os ratio in upper crust materials (Koide et al. 1991; Esser and Turekian 1993). The 187Os/188Os ratio of primitive mantle is 0.1290 ± 0.0009 (Meisel et al. 1996). This value is significantly lower than the crustal value. A small anthropogenic input of Os that is originally derived from mantle may reduce the 187Os/188Os ratios. Elevated Os concentrations have been found in environmental samples and are thought to be the result of emission from automobile catalysts in which Os occurs as an impurity (Fritsche and Meisel 2004; Poirier and Gariepy 2005; Rauch et al. 2005, 2006). In this study, almost all samples have higher Os contents and lower 187Os/188Os ratios than the background values (Fig. 4; Table 4), indicating an anthropogenic source of Os that originally derived from mantle. Our results support the interpretation that Os is an impurity in the automobile catalysts.

Fig. 4
figure 4

Plots of Pt versus Os, Pt versus Au and 187Os/188Os versus Os for the samples from Pearl River Delta region

Full size image

When Os is emitted from the high temperature automobile engine, it may be oxidized to volatile OsO4, and therefore may be transferred further than other PGE particles (Poirier and Gariepy 2005). Samples from Hong Kong, Shenzhen and Guangzhou show no correlation between Pt and Os in both soil and dust samples (Fig. 4). We suggest that this is because volatile OsO4 has been transported farther away from its source than other PGE.

Gold

Few studies are available on Au contamination by the emissions from automobile catalysts. Eisler (2004) reported up to 440 ng/g Au in atmospheric dusts collected from heavy traffic roads in Germany. Hooda et al. (2007) reported relatively high Au concentrations (up to 18 ng/g) in soils which are higher than the crustal average (5 ng/g), suggesting an anthropogenic source. It should be noted that Au is an effective catalyst and recent advances in technology allow its use in automobile catalytic converters, offering a cost-effective alternative to PGE to oxidate carbon monoxide (Haruta et al. 1993; Haruta 1997a, b; Mellor et al. 2002). The significant feature of gold-based three-way catalyst in gasoline and diesel applications is the low temperature oxidation of carbon monoxide, where the catalysts display activity as low as −70°C (Mellor et al. 2002).

The contents of Au in dusts from Hong Kong (42–453 ng/g, average 209 ng/g) and Shenzhen (148–1345 ng/g, average 471 ng/g) are much higher than soils (average 19.4 and 14.7 ng/g) and the background values (average 0.98 ng/g). Dust samples from Hong Kong and Shenzhen show clearly positive correlation between Au and Pt (Fig. 4). Because these samples were collected from the commercial streets and highways with no industrial input, the anomalies are likely related to contamination from automobile catalysts. The contents of Au in all the samples from Guangzhou (average 12.2 ng/g) are much lower than those from Shenzhen and Hong Kong and show no correlations between Pt and Au (Fig. 4). The similarity between Shenzhen and Hong Kong may have caused by the frequent shuttle traffic between these two cities. The difference between Guangzhou and Hong Kong may reflect the different automobile catalyst systems between mainland China and Hong Kong.

Conclusions

This study presents the results of PGE and Au in road dusts and roadside soils collected from three major cities in Pearl River Delta region, South China. The proposed low procedure blank and effective Carius tube technique together have improved the precision of PGE and Au determination in environmental samples, especially for the uncontaminated soil samples for establishing the background values.

The background values of PGE and Au in the study area were established. The highest levels of Pt, Rh and Pd were found in road dusts and are due to contamination from automobile catalyst. Similar Pt/Pd ratios for the samples from Hong Kong, Guangzhou and Shenzhen reflect similar Pt/Pd ratios in the automobile catalysts used in these cities. The Pt/Rh ratios of dust samples from Hong Kong are higher than those from Guangzhou and Shenzhen, reflecting the different automobile catalyst system between Hong Kong and mainland China. The elevated levels of Ru, Ir and Os in road dust and roadside soil samples are possibly related to the impurity of these elements in the automobile catalysts. Anomalous high concentrations of Au in the dust samples from Shenzhen and Hong Kong may also be related to emission from automobile catalyst systems.