Article pubs.acs.org/est
Complex Anthropogenic Sources of Platinum Group Elements in Aerosols on Cape Cod, USA Indra S. Sen,*,†,‡ Bernhard Peucker-Ehrenbrink,† and Nicholas Geboy§ †
Department of Marine Chemistry and Geochemistry, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts 02543, United States ‡ School of Earth, Ocean and Climate Sciences, Indian Institute of Technology − Bhubaneswar, Bhubaneswar, OD 751013, India § Eastern Energy Resources Science Center, U.S. Geological Survey, Reston, Virginia 20192, United States ABSTRACT: Platinum group elements (PGE) of anthropogenic origin have been reported in rainwater, snow, roadside soil and vegetation, industrial waste, and urban airborne particles around the world. As recent studies have shown that PGE are bioavailable in the environment and pose health risks at chronic levels, the extent of PGE pollution is of global concern. In this study, we report PGE concentrations and osmium isotope (187Os/188Os) ratios of airborne particles (particulate matter, PM10) collected in Woods Hole, a small coastal village on Cape Cod, Massachusetts, U.S.A. The sampling site is more than 100 km away from the nearest urban centers (Boston, Providence) and has no large industrial emission center within a 30 km radius. The study reveals that, although PGE concentrations in rural airborne particulate matter are orders of magnitude lower than in urban aerosols, 69% of the total osmium is of anthropogenic origin. Anthropogenic PGE signatures in airborne particles are thus not restricted to large cities with high traffic flows and substantial industries; they can also be found in rural environments. We further conclude that the combination of Pt/Rh concentration ratios and 187Os/188Os composition can be used to trace PGE sources. The Pt/Rh and 187 Os/188Os composition of Woods Hole aerosols indicate that the anthropogenic PGE fraction is primarily sourced from ore smelting processes, with possible minor contributions from fossil fuel burning and automobile catalyst-derived materials. Our results further substantiate the use of 187Os/188Os in source apportionment studies on continental scales.
■
performed in urban environments such as Mexico City,18 Boston,20 Beijing,19 Buenos Aires,21 and Frankfurt,22 with only few studies focusing on rural areas. 3,22,24−26 In fact, investigations of PGE concentrations in rural airborne particles in the United States, where catalytic converters were first introduced, are still missing. Inadequate knowledge of PGE concentrations in rural aerosols limits our understanding of the magnitude of PGE pollution in urban environments, as PGE background levels are not well constrained. Here, we report on PGE concentrations and 187Os/188Os of airborne particles collected in Woods Hole, a small coastal village on Cape Cod, Massachussetts, U.S.A. In addition, we further investigate the potential of 187Os/188Os to support a source apportionment study of continental scale. The objective of this study is to address the source and magnitude of regional PGE pollution based on elemental and radiogenic isotope data.
INTRODUCTION Human activities such as mining, construction, fossil fuel burning, biomass burning, and human apportionment of primary production have disturbed the global cycle of elements at the Earth’s surface.1,2 The disturbance is most pronounced for platinum group elements (PGE: Os, Ir, Ru, Rh, Pt, Pd), for which the total anthropogenic mobilization at the Earth surface is orders of magnitude higher than the total natural mobilization.2 This human alteration to the biogeochemical cycles of PGE is mostly attributed to the introduction of exhaust catalyst by the automobile industry.3−5 Platinum, Pd, and Rh based automobile exhaust catalysts have been used over the last 30 years to convert noxious gas emissions into more benign forms. Although these exhaust catalysts have significantly reduced the emission of nitrogen oxides, carbon monoxide, and short-chained hydrocarbons, they have resulted in the release of Pt, Pd, and Rh during vehicle operation.4,5 In addition to Pt, Pd, and Rh, emission of Os, a trace impurity in the PGE catalysts, from catalytic converters has also been reported.6,7 As PGE are bioavailable in the environment and pose health risks, PGE emission from catalytic converters are of global concern.8−11 Abrasion of PGE particles from exhaust catalysts not only causes elevated PGE concentrations in roadside environments,9,12−17 but also their presence is widely reported in airborne particulate matter.3,18−23 However, most of the previous studies have been © 2013 American Chemical Society
■
METHODS Sample Preparation. Aerosols were collected using a high volume (∼1000 L/min) atmospheric aerosol sampler (ThermoFisher Mass Flow Controlled PM10 Sampler) between April
Received: Revised: Accepted: Published: 10188
April 15, 2013 August 1, 2013 August 5, 2013 August 5, 2013 dx.doi.org/10.1021/es4016348 | Environ. Sci. Technol. 2013, 47, 10188−10196
Environmental Science & Technology
Article
187−189 (8 s integration), 188−190−192 (4 s), 190−192−194 (4 s), and 192−194−196 (2 s). Mass/charge ratio of 192 was cycled through all the three MICs to monitor the yields of the ion counters. Platinum interferences on 190Os and 192Os were corrected using blank-corrected count rates at mass/charge ratios of 194 and 196, while blank-corrected count rates at the mass/charge ratio of 185 were used to correct for Re interferences on 187Os. To assess our data quality, we repeatedly measured 5 pg solutions of Leoben (Austria) Os Standard solution (LOsSt). The external reproducibility of 187 Os/188Os for a 5 pg solution of LOsSt (187Os/188Os = 0.10690 ± 0.00020) was ∼1.5% (2 S.D., n = 5), and an internal precision of ∼0.3% (2 S.D., n = 5) was achieved. Platinum, Ir, and Rh were analyzed using a single-collector, magnetic sector ICPMS fitted with a Scott-type double-pass spray chamber. Iridium concentrations were calculated on the basis of the 191/193 mass/charge ratios, whereas the 198/194 and 198/195 ratios were used to quantify Pt concentrations. As Pt concentrations were determined using 198/194 and 198/ 195 ratios, and both concentrations were in close agreement (1).37−40 For example, chromite and PGE ore deposits have unradiogenic 187Os/188Os values ranging from 0.1 to 0.2,37,38,40 whereas base metal sulfide deposits can have unradiogenic to radiogenic 187Os/188Os values.39 Anthropogenic PGE sources have 187Os/188Os signatures that extend from unradiogenic to radiogenic ratios. Therefore, PGE source identification based only on 187Os/188Os is challenging. PGE source identification can also be achieved using PGE concentration ratios such as Pt/Pd and Pt/Rh.3,21,22 However, PGE ratios are not very robust indicators, as catalyst-derived material and smelter-derived materials have overlapping ranges.3 For example, industrial waste, incinerated ash, and catalyst-derived materials occupy a common field in Pt/Ir vs Rh/Ir space (Figure 3). Figure 3, however, emphasizes that aerosols and the commercially available PGE follow different trends in Pt/Ir vs Rh/Ir space. Aerosols are characterized by a steeper positive slope than anthropogenic sources (exhaust catalyst, sewage sludge, incinerated ash). Different proportions of natural and anthropogenic sources (exhaust catalyst, sewage sludge, incinerated ash) cannot explain the steep trend. As the Pt/Rh ratio controls the slope of the line in Pt/Ir vs Rh/Ir space, we surmise that aerosols and catalyst-derived materials
have different Pt/Rh concentration ratios. It is worth emphasizing that we have limited understanding of how PGE fractionate during ore smelting, fossil fuel combustion, and atmospheric transport and deposition. For example, Ir/Os ratios maintain a linear relationship with 187Os/188Os (Figure 2b), where the natural end-member has radiogenic 187Os/188Os signature and high Ir/Os ratio (≥1). Since unmodified natural sources of PGE with Ir/Os ≫1 are not known,41 the high Ir/ Os ratio plausibly indicates that Os and Ir are differentially mobilized and dispersed during industrial processing, atmospheric transport, and deposition. The volatility of OsO4 may contribute to, or even control, this differential mobilization. Figure 4 indicates that the combination of Pt/Rh ratios and 187 Os/188Os systematics can be used as a tracer of PGE sources. As natural aerosols are derived from the deflation of soil, PGE concentrations in soil should serve as a useful proxy for natural sources. However, data on globally averaged PGE composition of soils free of anthropogenic PGE are not available. We therefore take the composition of the average eroding continental crust, as derived from data on loess, as a suitable proxy for natural PGE composition of aerosols.42,43 In addition to soil dust, extraterrestrial matter44,45 and volcanic aerosols46,47 could contribute PGE to the natural environment. Extraterrestrial matter and volcanic aerosols have unradiogenic Os48,49 and an average Pt/Rh ratio of 7.1450 and 3−10,47 respectively. However, their annual fluxes are miniscule when compared to soil formation and soil erosion fluxes. For example, it has been estimated that wind and water erosion removes 75 000 Tg of soil annually,51 compared to 0.03 Tg of extraterrestrial matter accumulating annually on the Earth’s surface.44 Furthermore, Chen et al.52 have shown that subaerial volcanism and extraterrestrial matter cannot be the primary source of natural Os in the environment. We use the Pt/Rh (3.9)43 and the 187Os/188Os composition (1.4)32 of average upper continental crust as the best available analog for the natural PGE source. PGE compositions of urban and rural 10191
dx.doi.org/10.1021/es4016348 | Environ. Sci. Technol. 2013, 47, 10188−10196
Environmental Science & Technology
Article
aerosols are very different from the natural source. Figure 4 shows that aerosols in cities with heavy traffic volume such as Boston,20 Buenos Aires,21 and Mexico City18 have an average Pt/Rh ratio and 187Os/188Os of 4.2 (0.9−13.7) and 0.13−2.7 (1 S.D., n = 66), respectively. In contrast, rural (background) sites in Europe and USA (this study) have an average Pt/Rh ratio of 2.0 (0.6−5.0), with 187Os/188Os ranging from 0.13 to 1.1. Similar results were obtained in Frankfurt22 (Germany), Sweden,53 and Italy.54 It is thus evident that the aerosols have a much lower Pt/Rh ratio when compared to the natural end-member. Among the anthropogenic sources, automobile exhaust catalyst and industrial waste (sewage sludge and incinerated ash) are considered major sources of PGE pollution.9,12−15,18−21,33,55 These anthropogenic sources have Pt/ Rh values greater than 5 and 187Os/188Os values of 0.1 to 0.2. Other possible anthropogenic sources may include PGE released from fossil fuel (coal and petroleum) burning. PGE composition of coal and petroleum is heterogeneous, and estimates of globally average coal and petroleum compositions are not available. If PGE concentration data in the older literature can be trusted, such data reveal that coal has a Pt/Rh ratio of 0.7 ± 0.3 (1 S.D, n = 122).56 It is therefore plausible that PGE released in the process of fossil fuel burning could
Figure 4. Pt/Rh vs 187Os/188Os of Cape Cod aerosols [this study] and the Rauch et al.20 study that compares potential anthropogenic sources. Fossil fuels, catalytic converters, and smelting of ore minerals plot near the three corners of Pt/Rh vs 187Os/188Os space. Cape Cod aerosols mostly fall along the mixing line between natural and smelting sources. Smelting seems to be the major source of PGE pollution, with minor contributions from fossil fuel and catalytic converters. The data references are as follows: upper continental crust,32,43 automobile catalyst derived material,9,70 fossil fuels,34−36 and smelting source.40,58 Note that the isotope ratios of the upper continental crust, fossil fuel, and smelting sources can vary substantially, and only typical values are shown.
Figure 5. 187Os/188Os (in parentheses following sample identifiers) plotted against average air mass back trajectories. The figure shows that the backward air mass trajectories from the southwest of Cape Cod are characterized by unradiogenic Os, whereas the air-mass coming from the north have more radiogenic signatures. 10192
dx.doi.org/10.1021/es4016348 | Environ. Sci. Technol. 2013, 47, 10188−10196
Environmental Science & Technology
Article
masses coming from the North American continental interior. Sample 52408 (Figure 5, lower left) is derived from an air mass originating near Canada’s Hudson Bay and passing through the urban center of Montreal, whereas samples 61908 and 61008 (Figure 5, lower right) have passed over the industrialized Ohio River Valley. In a study conducted concurrently and at the same location with ours, Kolker et al.28 evaluated major and trace metal compositions of smaller aerosol particles (PM2.5) during sampling intervals identical to ours. Those results are suggestive of air masses from different geographic areas carrying distinct chemical signatures. Specifically, aerosols coming from offshore are enriched in Na, Mg, and Sr (sea salt contributions), while aerosols derived from the industrialized Ohio River Valley are characterized by elevated concentrations of elements commonly associated with fossil fuel combustion, such as Mo, V, Sb, and Ni. We do not see a stark contrast in the 187Os/188Os between these two geographic areas in the PM10 size fraction. This finding is not unexpected, because Os is one of the least abundant elements in seawater52 (approximately 10 fg/g). Therefore, anthropogenic and/or continental signatures would overwhelm any seawater Os signature even if the aerosols were derived from marine air masses. It is possible, however, to observe differences in the 187Os/188Os values of aerosols that have interacted with air masses from distinct geographic sources. For example, particles that have air mass back trajectories from the South and West of Cape Cod are characterized by unradiogenic Os, whereas air masses coming from the North have more radiogenic signatures. Regional differences in the 187Os/188Os of the eroding continental crust may contribute to this finding, as areas in NE Canada are characterized by Precambrian shield areas that are likely more radiogenic than those found in the central and eastern U.S. Therefore, 187Os/188Os can make unique contributions to source apportionment studies. At the very least, fossil fuel combustion can be distinguished from geogenic or ore processing sources. However, a larger data set is needed to more fully investigate the relationship between source area, PGE concentration, 187Os/188Os, and other major and trace elements in aerosols. Our results demonstrate that anthropogenic PGE are not restricted to urban centers with high traffic volumes and important industries but can also be seen in rural environments. In general, PGE in aerosols are a complex mixture of various anthropogenic and natural sources, showing significant variations in composition as well as elevated environmental concentrations. Considering the recent investigations of PGE toxicity, environmental bioavailability, and concentrations in biological specimens, it is evident that elevated PGE concentration may pose previously unrecognized health risks.10,11,65−67 The presence of anthropogenic PGE in rainwater, snow, roadside soil and vegetation, and the urban and rural airborne particles around the world indicates global scale PGE pollution. As the demand for PGE metals continues to increase, exemplified by a 26% increase in Pt production between 2000 and 2010,68 global PGE contamination of the critical zone is a growing concern that requires continuing observation and study.
have low Pt/Rh ratios. However, fossil fuel has radiogenic 187 Os/188Os signatures (>1) and hence will not contribute unradiogenic Os to aerosols.34−36 An additional source with low Pt/Rh ratio and unradiogenic Os is therefore required to explain our observations. Smelting of certain PGE ores could release anthropogenic PGE with low Pt/Rh concentration ratios57 and unradiogenic Os to the environment. Most of the world’s PGE deposits are related to magmatic activity (e.g., Bushveld, Stillwater, Sudbury, and Duluth),58 and magmatic differentiation can significantly fractionate PGE.59,60 For example, fractionation of Ni−Cu rich sulfide liquid that has already concentrated PGE from silicate magma would produce two classes of sulfide ore deposits. The deposit that is associated with the liquid fraction will have high Pt/Rh ratios (>10), while the fractional crystallization products of the same liquid would produce ore deposits with low (
