Measurement of the gross alpha activity of the fine fractions of road dust and near-roadway ambient particle matter

ABSTRACT

Traffic-related air pollution, including direct exhaust emissions and road dust (RD), impacts individuals living near busy roads. We recently conducted a study to investigate the sources and composition of tailpipe and non-tailpipe traffic emissions, where we collected and analyzed samples of ambient air fine particulate matter (PM_2.5) and fine RD (RD_2.5) at different distances from major roadways. We analyzed a subset of the samples, including those collected at the roadside and local background, for their alpha activity level. Subsequently, we investigated whether there is a distance-related decay in the alpha activity in RD_2.5 or PM_2.5 similar to those observed for traffic-related species in PM_2.5 and RD_2.5. We found that the alpha activity of ambient air PM_2.5 (Bq/mg) was more than an order of magnitude higher than the activity level of the corresponding RD_2.5 sample, suggesting that PM_2.5 may be more toxic than RD_2.5. Using mixed-effects regression models, we found that ambient PM_2.5 alpha activity was significantly higher during the cold months than during warm months, and that the background was higher than the roadside (though not significantly). In contrast, the RD_2.5 alpha activity was significantly higher at the background site compared to the roadside but was not significantly affected by season. In addition to sampling position, both Zn and elemental carbon (EC) were significant predictors of RD_2.5 alpha activity. In addition, the roadside RD_2.5 activity levels were found to be higher at highways as compared to secondary roads. While traffic-related emissions do not appear to be significant sources of either ambient PM_2.5 or RD_2.5 alpha activity, the RD_2.5 results suggest that traffic-related particles may contribute to RD_2.5 alpha-activity.

Implications: Many studies have reported the effects of traffic-related particulate matter (PM) on human health, and there is growing interest in the health effects of exposure to environmental PM alpha activity. This is the first study to report on the alpha activity of road dust (RD) or near-roadway ambient PM. We found that the alpha activity of ambient PM is twenty times higher than RD, suggesting that ambient PM may be more toxic. In PM and RD, the alpha activities were higher at background sites than at the roadside, indicating that traffic-related emissions are not a significant source of particulate radioactivity.

Introduction

Traffic-related emissions, which impact individuals living near and commuting on busy roads, are a significant source of air pollution (Hamra et al. 2015; Hankey, Lindsey, and Marshall 2017; Huang et al. 2018; Kim et al. 2005; Puett et al. 2014; Zhang and Batterman 2013). Exposure to traffic-related ambient fine particle matter (PM_2.5) or residential proximity to road have been associated with a wide range of adverse health effects, including respiratory and cardiovascular symptom exacerbation and birth outcomes (Boothe and Shendell 2008; Carlsten et al. 2011; Gan et al. 2011; Gauderman et al. 2005, 2007; Hoffman et al. 2012; Jerrett et al. 2008; McConnell et al. 2006, 2010; Van Hee et al. 2010; Wilhelm et al. 2012; Wilhelm and Ritz 2003; Williams et al. 2009; Zanobetti et al. 2010; Zeka et al. 2006).

Traffic-related air pollution is a complex mixture, which includes particulate matter (PM) originating from tailpipe emissions of vehicles and both direct and indirect non-tailpipe emissions. Tailpipe emissions contain elemental and organic carbon and various trace elements associated with incomplete combustion of fuel and oil additives. Non-tailpipe emissions include particles from wear of brakes, tires, engines and road surfaces, and resuspended road dust. Road dust itself commonly contains high concentrations of trace contaminants, which originate from tailpipe and non-tailpipe emissions, as well as crustal material (soil and sand), road salt, vegetation debris, etc. (Apeagyei, Bank, and Spengler 2011; Gunawardana et al. 2012; Hwang, Sahin, and Choi 2017). For example, Zn, Cu, Pb, Ni, Cr, and Cd primarily originate from non-tailpipe emissions; Si, Ca, and Al primarily originate from soil; Na and Cl mainly originate from ocean and road salt (Gunawardana et al. 2012; Ho et al. 2003; Johansson, Norman, and Burman 2009).

There are currently not sufficient available measurements of ambient particle alpha activity available to allow epidemiological studies of the health effects of environmentally relevant exposure levels in populations. However, it is known that alpha activity causes DNA damage at the cellular level (Chauhan et al. 2012). In addition, exposure to alpha emitters at low dosage levels causes significantly more chromosome damage in liver cells than either beta or gamma radiation (Brooks 1975). A study by Little, Kennedy, and Mcgandy (1975) supports the hypothesis that alpha activity exposure from ²¹⁰Po associated with cigarette smoking may be a significant contributing factor in development of human lung cancer. Furthermore, recent studies have shown associations between a variety of human health effects and Particle Radioactivity (PR) measured as gross beta- and gamma-activities. These include association with total and cardiovascular mortality (Blomberg et al. 2018), effects on blood pressure (Nyhan et al. 2019), oxidative stress (Li et al. 2018), and lung and cardiac function (Vieira et al. 2019). In the absence of anthropogenic release, PR is attributed to predominantly to radionuclides of terrestrial origin, specifically radon progeny (Camacho et al. 2009; Ceballos et al. 2016; Cabello et al. 2018). Radon is a noble gas, a decay product of ²³⁸U and ²³²Th, which are both naturally occurring radioactive elements found in earth’s crust. After formation, radon diffuses from the soil into ambient air, where it decays to radioactive progeny. The progeny quickly form charged clusters that attach to airborne accumulation mode particulate matter. Other natural sources of particle radioactivity include cosmogenic radionuclides such as ⁷Be, and terrestrial radionuclides such as ⁴⁰K, from resuspension of soil (Blomberg et al. 2020; Winkler et al. 1998).

In other publications from our current study, which was designed to investigate the sources and composition of tailpipe and non-tailpipe traffic emissions, we report that trace elements associated with traffic emissions show distance-related decreases in concentration for both ambient particles and road dust (Huang et al. 2020; Moreira et al. 2020; Silva et al. 2020). In this paper, we measure the gross alpha radioactivity of ambient fine particles and road dust at different distances from the road, and investigate whether there are decay patterns for PM gross alpha activity that are similar to those of the traffic-related elements we have reported previously (Huang et al. 2020; Silva et al. 2020). To the best of our knowledge, this is the first study to investigate the alpha activity of near roadway ambient PM or RD.

Methods

1. Study design

We conducted a field study from June 2018 to December 2019 in the Greater Boston metropolitan area to examine the relationships between road proximity and traffic-related composition of the coarse and fine fractions of particle matter in ambient air (PM) and road dust (RD), at different distances from major roadways. On each sampling day, RD and ambient PM samples were collected at three distances from the road, one at the roadside (0–25 m), one at an intermediate distance (50–200 m) and one at a background site (500–1000 m) away from the roadside. Major roadways sampled included multi-lane divided state and interstate highways (with or without limited access via onramps and exit ramps) and busy state secondary and connecting roads. Background and intermediate samples were collected the same day at locations on adjacent roadways within the target distances from the roadside site and were almost entirely on residential roads. For this paper, we measured the gross alpha radioactivity in 52 samples of fine ambient PM (PM_2.5) and the fine fraction of RD (RD_2.5) collected at paired roadside and background sites for 20 major roads in the greater Boston area Samples were collected twice at six of the sites, during cold and warm seasons. Sampling locations are shown in the Supplementary Material, Figure S1; road parameters and site locations are also included in the Supplementary Material, Table S1.

• (2) Sample Collection

For this study, we designed a Mobile Particle Concentrator Platform (MCMP). The MPCP was equipped with fine (0.2 to 2.5 µm) and coarse (2.5 to 10 µm) particle concentrators and a Road Dust Aerosolizer sampler (RDA), which was designed and constructed to simultaneously re-suspend road dust and separate it into fine and coarse fractions. Methods are described in detail elsewhere in this special issue (Martins et al. 2020) and summarized below.

1. Ambient PM_2.5

We used a modified Harvard Ambient Fine Particle Concentrator (HAFPC), originally a 5,500 LPM three-stage fine particle concentrator that has been described in detail in previous studies (Lawrence et al. 2004; Sioutas et al. 1997). The HAFPC was modified to require less power to be accommodated by the mobile platform. The modified system used two parallel two-stage concentrators whose outputs were combined, followed by separate collection of samples on Teflon and Quartz fiber filters, with flows of 45 LPM for each. For this study, we analyzed the samples collected on Teflon. Samples were collected for one hour at a total intake flowrate of 2,200 LPM; ambient concentrations were calculated using the concentrator enrichment factor as described elsewhere (Martins et al. 2020).

• (b) Road dust fine fraction (RD_2.5)

We developed a novel Road Dust Aerosolizer (RDA) sampler to re-suspend and collect PM_2.5 from the road surface (RD_2.5). The RDA sampler aerosolizes road dust at a very high flow rate from the surface of the roadway and simulates road PM resuspension into the air, which results in a more realistic measurement of PM composition and size, as described in our methods paper published in this special issue (Martins et al. 2020). RD was sampled for approximately 5 minutes, during which 12 m2 of road surface was vacuumed. The RD_2.5 fraction was collected using a 30LPM cascade sampler (Demokritou et al. 2004). Like the HAFPC above, the RDA sampler collected samples on both Teflon and Quartz fiber filters, and the Teflon filters were analyzed for this study.

• (3) Sample radioactivity analysis

The analysis of PM gross alpha activity is described in detail elsewhere (Kang et al. 2020; Liu et al. 2020) and summarized here. We measured PM alpha activity in 52 samples of ambient PM_2.5 and RD_2.5 collected at paired roadside and background sites for 20 major roads in the greater Boston area using a low background gas proportional counter (LB4200, Canberra Industries, Inc., Meriden, CT) with a counting time of 600 min per sample. The instrument was calibrated using a 0.0518 μCi NIST traceable ²¹⁰Po source, and measurements were blank corrected. The background level was below 0.1 cpm and the minimum detectable activity (MDA) level for alpha averaged 0.071 Bq/mg. Most of the measured alpha activities in samples were above the detection limit.

The total activity on a filter at any time after collection is a function of radionuclide type and concentration as well as filter loading. After 1.5–2 years of storage, the total alpha activity is likely dominated by ²¹⁰Po decay, which is the decay progeny of ²¹⁰Pb (a beta-emitter) whose half-life is the longest of the radon decay radionuclides (22.3 years). Therefore, long-lived alpha activities in the air at the time of sampling can be estimated using the following equation modified based on Sheets and Thompson’s Equation (1)(1) $\mathrm{C}=\frac{A_t\cdot e^{\lambda t}}{M}$(1) :

(1) $\mathrm{C}=\frac{A_t\cdot e^{\lambda t}}{M}$(1)

where $A_t$ is the alpha activity on the filter at time of counting (Becquerel, Bq), $\lambda$ is the decay constant for ²¹⁰Pb (8.51 × 10⁻⁵ day⁻¹), t is the time duration from the end of PM sampling to the start of alpha counting (day), and M is the mass loading of the sample collected on the filter (mg).

Although we measured both alpha- and beta-activity in these samples, the beta-activity levels were low, below the MDA (average of 0.170 Bg/mg) in 70% of PM_2.5 and 93% of the RD_2.5 samples. Background correction of measurements using blank filters generates some negative counts. For this reason, in this paper, we focus primarily on the alpha activity of the samples. Gross beta activity measurements for these samples are included in the Supplementary Materials, Figure S2.

• (4) Other Data

Other data used in this paper include the concentrations of trace elements as well as elemental and organic carbon and their thermally evolved fractions, measured during this study and reported elsewhere in this special issue (Huang et al. 2020; Moreira et al. 2020; Silva et al. 2020). We analyzed our collected ambient PM_2.5 and RD_2.5 samples for trace elemental composition using X-Ray Fluorescence (Silva et al. 2020; Huang et al. 2020, respectively). We analyzed elemental carbon (EC) and organic carbon (OC) using thermal optical reflectance (TOR) (Moreira et al. 2020, in this special issue). We used the IMPROVE TOR method (Chow et al. 2004; Chow et al. 2007), which distinguishes EC/OC as thermally evolved fractions, including 3 EC fractions (EC1, EC2, and EC3), 4 OC fractions (OC1, OC2, OC3, and OC4), a pyrolyzed carbon fraction (OP), and an optical EC fraction (Kang, Koutrakis, and Suh 2010). The different characteristics of the carbonaceous species composing these thermally evolved fractions may offer insight into their origins (Habre et al. 2014; Kim et al. 2005, 2011, 2004; Lim et al. 2012; Sahu et al. 2011; Yu, Xu, and Yang 2002).

• (5) Statistical analysis

All statistical analyses were performed using SAS (version 9.4, SAS Institute, Cary, NC). Differences between background and roadside samples were tested using t-tests and paired t-tests, where appropriate. Because our study featured repeated measurements (i.e., multiple distances from each major road and measurements at the same sites during different seasons), we used a linear mixed-effect regression model to estimate the relationships between the alpha activity and study variables (e.g., season, proximity to roadway, road type), including PM composition. Models for PM_2.5 and RD_2.5 were run separately, with a random intercept for each major road. Slope coefficients for each parameter and species included were treated as fixed effects. Parameters investigated included the dichotomous variables reflecting distance from road (roadside versus local background), season (warm versus cold), and road type (A3 versus combined A1 and A2). Species investigated included the trace elements and carbon fractions measured for each sample. We considered results significant when p <.05 for the t-tests and mixed-effects model.

Results and discussion

Results of alpha activity measurements are shown in Figures 1 and 2 for ambient PM_2.5 and RD_2.5, respectively. The alpha-activity of PM_2.5 and RD_2.5 is reported in Becquerels per milligram of mass (Bq/mg). The alpha activity of PM_2.5 is also shown in Bq/m³ (Figure 1(b)). The one-hour alpha activity concentrations measured in this study are consistent with those reported in a recent study in greater Boston, where the 24-hr average during 2005–2006 was 0.0019 ± 0.0011 Bq/m³ (range, ND to 0.0076 Bq/m³) (Liu et al. 2020).

Figure 1. Ambient PM_2.5 alpha activity (Bq/mg PM and Bq/m³) and uncertainty measured in samples collected at the roadside and local background positions at major roadways in the greater Boston area

Figure 2. Road dust (RD_2.5) alpha activity (Bq/mg RD) and uncertainty measured in samples collected at the roadside and local background positions at major roadways in the greater Boston area

Overall, ambient PM_2.5 alpha activity per unit mass was more than an order of magnitude higher than RD_2.5 as shown in Table 1. The ratio of PM_2.5 to RD_2.5 alpha activity was higher for roadside samples compared to the background, but this difference was not statistically significant (paired sample t-test, p-value 0.20).

Table 1. Ratio of ambient PM_2.5 to RD_2.5 alpha activity (Bq/mg)

Position	Air/Dust Alpha Activity Ratio
	Average (SD)*	Minimum	Maximum
Roadside	23.5 (11.0)	1.7	40.8
Background	19.7 (8.7)	9.0	51.7
Overall	21.6 (10.0)	1.7	51.7

Notes. *Ratios are all statistically significant; paired sample t-test for PM_2.5 and RD_2.5 roadside, background and overall yield p-values ≪.001.

Both near road and background ambient PM_2.5 alpha activity levels were significantly higher during the cold season (October–January) than during the warm season (June–September), averaging 0.236 ± 0.112 and 0.172 ± 0.052 Bq/mg, respectively (t-test, p-value 0.013). PM_2.5 alpha activity levels were also lower at the roadside sites than background sites during the cold season (0.183 ± 0.068 and 0.226 ± 0.109 Bq/mg, respectively; p-value 0.025), but similar during the warm season. There were no significant differences in PM_2.5 alpha activity levels observed in roadside samples for highways (combined A1 and A2 roads) compared to secondary (A3) roads. This suggests that traffic-related emissions to not contribute significantly to the PM_2.5 alpha activity.

RD_2.5 alpha activity levels were higher at the background than at the roadside, 0.012 ± 0.006 and 0.009 ± 0.004 Bq/mg, respectively. During the cold season, this difference was even more pronounced, with levels of 0.014 ± 0.006 and 0.008 ± 0.004 Bq/mg for background and roadside, respectively. During the warm season, the background (0.011 ± 0.006 Bq/mg) and roadside (0.009 ± 0.003 Bq/mg) RD_2.5 alpha activities were more similar. These differences were statistically significant both overall (paired sample t-test, p-value 0.015) and during the cold season (paired sample t-test, p-value 0.005). This suggests that traffic-related emissions are not a significant source of RD_2.5 alpha activity. Although it did not reach statistical significance, the RD_2.5 alpha-activity in roadside samples was observed to be higher in the warm season than the cold season, 0.010 ± 0.003 and 0.008 ± 0.004 Bq/mg, respectively.

RD_2.5 roadside alpha-activity levels were significantly higher for highways (combined A1 and A2 roads, 0.010 ± 0.003 Bq/mg) than for busy state secondary and connecting roads (A3, 0.007 ± 0.004 Bq/mg) where traffic density and speed are generally lower (Table S1, supplementary material); t-test p-value 0.022. At background sites, alpha activity levels were similar for highways and secondary roads.

The results of the mixed-effects regression analysis for PM_2.5 and RD_2.5 are presented in Table 2. The only significant predictor of the PM_2.5 alpha activity levels was season, with higher activities observed during the cold months. Position was borderline significant as a predictor, with lower activities at the roadside compared to the background. None of the trace elemental species or carbon fractions were significant predictors for the ambient PM_2.5 alpha-activity. This is in contrast with previous findings from our group, which show strong associations of alpha activity with both PM_2.5 and Sulfur (S), a known regional and secondary PM_2.5 component (Kang et al. 2020; Liu et al. 2020). However, these previous studies used large datasets. Our results agree with a previous study from our group, which found that particle radioactivity in Boston was highest in February (Blomberg et al. 2020). This suggests that the alpha activity measured in our study is similarly due to long-lived radon decay products.

Table 2. Fixed effect estimates for ambient PM_2.5 and road dust RD_2.5 alpha activity models

Model	Fixed Effect	Slope	SE	p-value
Ambient PM2.5	Season Cold (Warm = 0)*	0.07434	0.02344	.0037
	Position Background (Roadside = 0)	0.04469	0.02269	.0589
Road Dust (RD2.5)	Zinc*	0.00338	0.00100	.0020
	EC1*	0.00019	0.00005	.0004
	Position Background (Roadside = 0)*	0.00431	0.00136	.0040

Notes. *Statistically significant predictors of alpha activity level in Bq/mg at p < .05.

The RD_2.5 alpha activity is significantly predicted by Zn, EC1 and position, with activity levels higher at the background than roadside sites. The Zn and EC1 both have positive slope estimates, indicating that higher concentrations of Zn and EC1 are both associated with higher alpha activity. In our previous papers on the concentrations of trace elements and carbon fractions in RD_2.5 published in this special issue, both Zn and EC1 showed a decay with distance from the road (Huang et al. 2020; Moreira et al. 2020). These species are both related to traffic. EC1 is associated with diesel combustion in vehicles and engines (Kim et al. 2005, 2004) and Zn is associated with tire wear, brake wear, and engine oil (Garg et al. 2000; Hjortenkrans, Bergback, and Haggerud 2003; Lough et al. 2005). The positive associations observed between traffic-related species and alpha activity in RD_2.5, together with roadside activity levels that are higher at highways than beside secondary roads, may suggest that traffic-related emissions may contribute to RD_2.5 activity, though not a significant source. However, as the concentrations of road dust gross alpha activity are significantly higher in PM_2.5 than in RD_2.5, deposition of long-lived radon progeny attached to ambient PM_2.5 is likely to be a contributor.

Our field study was not designed to investigate the alpha activity of RD_2.5 or ambient near-road PM_2.5, but rather to investigate direct and indirect traffic-related PM emissions, as reported in several papers in this special issue (Huang et al. 2020; Martins et al. 2020; Moreira et al. 2020; Silva et al. 2020). However, for this paper, we measured gross alpha activity in a subset of samples collected from the original study. Our current study, to the best of our knowledge, is the first to investigate the alpha activity of either road dust or near-roadway ambient PM_2.5. With the relatively small sample size, we were unable to include many variables in our analysis. However, we found that the alpha activity of ambient PM_2.5 is more than an order of magnitude higher than that of RD_2.5 (Table 1), suggesting that PM_2.5 may be more toxic than RD_2.5. In addition, we observed that both EC1 and Zn, traffic-related species, are both significant positive predictors of alpha activity in RD_2.5, which may suggest that there could be some traffic-related contribution.

Conclusion

In the first study investigating the alpha activity of near roadway traffic-related PM and RD, we found that ambient PM_2.5 alpha activity levels in the greater Boston area were more than 20 times higher than their corresponding samples of RD_2.5. This suggests that ambient PM_2.5 may be considerably more toxic than road or soil dust.

Using linear mixed effects regression models, we found that PM_2.5 alpha-activity is influenced significantly by season, with lower levels during the warm season, consistent with observations from previous studies. Distance from roadway is a borderline significant predictor of PM_2.5 alpha activity, with higher levels at background locations than roadside. Our results indicate that traffic-related sources do not directly contribute significantly to PM_2.5 alpha activity. In contrast to previous studies, neither S nor PM_2.5 concentrations were associated with PM_2.5 alpha activity; however, this may be due to the relatively small sample size we analyzed.

Alpha-activity of RD_2.5 was also higher at the background locations than the roadside, but in contrast to PM_2.5, was not significantly affected by the season. We found that the roadside alpha activity of RD_2.5 was higher at highways than secondary roads. Significant predictors for RD_2.5 alpha activity using mixed-effects models included sample location (roadside or background), Zn, and EC1. Together, these results may suggest that although traffic-related emissions are not a significant source of RD_2.5 alpha activity, they may contribute and thus may impact the hazards of RD_2.5 exposure.

Acknowledgment

Research described in this article was conducted under contract to the Health Effects Institute (HEI), an organization jointly funded by the United States Environmental Protection Agency (U.S. EPA) (Assistance Award No. CR-83467701) and certain motor vehicle and engine manufacturers. The contents of this article do not necessarily reflect the views of HEI, or its sponsors, nor do they necessarily reflect the views and policies of the EPA or motor vehicle and engine manufacturers.

This publication was made possible by U.S. EPA grant RD-83587201. Its contents are solely the responsibility of the grantee and do not necessarily represent the official views of the U.S. EPA. Further, U.S. EPA does not endorse the purchase of any commercial products or services mentioned in the publication.

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary material

Supplemental data for this article can be accessed on the publisher’s website.

Additional information

Funding

This work was supported by the Environmental Protection Agency [CR-83467701]; Health Effects Institute [CR-83467701].

Notes on contributors

Joy Lawrence

Joy Lawrence is a research associate in the Exposure, Epidemiology and Risk program, Department of Environmental Health, Harvard T.H.Chan School of Public Health, Boston, MA.

Marco Martins

Marco Martins is a research associate in the Exposure, Epidemiology and Risk program, Department of Environmental Health, Harvard T.H.Chan School of Public Health, Boston, MA.

Man Liu

Man Liu is a doctoral student in the Exposure, Epidemiology and Risk program, Department of Environmental Health, Harvard T.H.Chan School of Public Health, Boston, MA.

Petros Koutrakis

Petros Koutrakis is a Professor of Environmental Sciences in the Department of Environmental Health at the Harvard T.H.Chan School of Public Health, Boston, MA.