Environ Sci Pollut Res DOI 10.1007/s11356-013-2418-y

RESEARCH ARTICLE

Source identification of inorganic airborne particle fraction (PM10) at ultratrace levels by means of INAA short irradiation Pasquale Avino & Geraldo Capannesi & Alberto Rosada

Received: 9 October 2013 / Accepted: 28 November 2013 # Springer-Verlag Berlin Heidelberg 2013

Abstract Many studies have focused their attention on the determination of elements of toxicological and environmental interest in atmospheric particulate matter using analytical techniques requiring chemical treatments. The instrumental nuclear activation analysis technique allows achieving high sensitivity, good precision, and excellent limit of detection without pretreatment, also considering the problems related to the radioisotope characteristics (e.g., half-life time, interfering reactions, spectral interferences). In this paper, elements such as Al, As, Br, Cl, Cu, I, La, Mg, Mn, Na, Sb, Si, Ti, and V are studied in atmospheric PM10 sampled in downtown Rome: The relative radionuclides after activation of the sample are characterized by very short (ranging from 2.24 to 37.2 min) and short (ranging from 2.58 h to 2.70 days) half-lives. Furthermore, As, Br, La, Mn, and Sb were also determined for evaluating the aerosol characteristics. The results, elaborated considering the matrix effects and the interfering reaction contribution to the radioisotope formation (e.g., 28Al generated by both (n,γ) reaction from 27Al and (n,p) reaction from 28 Si), show interesting values of As (0.3–6.1 ng m−3), Cu (22– 313 ng m−3), Mn (17–125 ng m−3), V (7–63 ng m−3), higher than those determined in an area not influenced by

autovehicular traffic, and significant levels of I (1– 11 ng m−3) and Ti (25–659 ng m−3) in Rome PM10. The other elements show a pattern similar to the very few data present in the literature. It should be underlined the good correlation (r 2) of Al vs. Mg (0.915) and Al vs. La (0.726), indicating a same sources for these species as well as Br–Sb showing a little lower correlation (0.623). This last hypothesis is confirmed by the study of the enrichment factors: Sb and Br may be attributed to anthropogenic sources; Cu, Cl, and I show a mixed origin (natural and anthropogenic), whereas Al, Si, Ti, Mn, Na, Mg, and As are of crustal origin. For having more information, a statistical approach based on the principal component analysis and the canonical discriminant analysis has been performed: All the samples (except one) are grouped in a cluster, and elements such as As, Br, Cu, I, La, Mn, Sb, Ti, and Vare highly correlated, whereas Na and Cl and Mg and Al assemble in two different clusters. Finally, a comparison with other similar studies is reported showing interesting values for Al, As, Mg, Mn, and Ti.

Responsible editor: Gerhard Lammel

Introduction

Electronic supplementary material The online version of this article (doi:10.1007/s11356-013-2418-y) contains supplementary material, which is available to authorized users. P. Avino (*) DIPIA, INAIL Settore Ricerca, via IV Novembre 144, 00187 Rome, Italy e-mail: [email protected] G. Capannesi : A. Rosada UTFIST-CATNUC, R.C. Casaccia, ENEA, via Anguillarese 301, 00060 Rome, Italy

Keywords Airborne particle . PM10 . Short irradiation . INAA . Trace elements . Enrichment factor . Urban atmosphere . Statistical evaluation

Airborne particulate matter (PM) is a very complex matrix. In particular, PM10 plays an important role both in terms of human health and local climatology (Gobbi et al. 2013). Because of its high active surface area and metal presence, PM10 acts as powerful catalyst for the conversion reactions of sulfur and nitrogen oxides to sulfuric and nitric acid (World Health Organization 2000; Manigrasso and Avino 2009), respectively (in particular, Cu, Mn, or V, elements that will be investigated in this study, are able to play this role). They

Environ Sci Pollut Res

are also responsible for the dry deposition of acids on buildings and historical heritage. PM10 are generally irritating to the respiratory system. Their dangerous action, however, is mainly due to the harmful substances contained or adsorbed on them (Cecinato 1999; Monod et al. 2001; Buonanno et al. 2010, 2011). Several parameters can influence the toxicity of the PM10 that is present in the outdoor, indoor, or workplaces: the particle size distribution, the bulk chemical composition, and the contents of trace element, acid, and sulfate. An accurate knowledge on the airborne particulate matter nature and on the identification of their major sources and pathways through the environment is needed to identify mitigation actions (Almeida et al. 2013). According to some authors, an increase of 1 % of premature mortality (Michelozzi et al. 1998) is associated for every 10 μg m−3 of airborne PM10 concentration. In urban areas, more than 80 % of PM10 is formed by agglomerations of organic compounds, produced by condensation or sublimation of heavier gaseous compounds emitted from combustion processes or produced during photochemical smog (Cocheo 2000), while the inorganic fraction, although minor, is very important from the toxicological point of view. Thus, our attention was focused on the determination of the inorganic part. Indeed, alongside elements normally determined due to environmental and/or toxicological interest such as As, Cd, Cr, Fe, Hg, Mn, Ni, Pb, and Zn, some elements at ultratrace levels, whose origin and/or effects on health and environment are poorly known, are present onto airborne PM10: In particular, in the literature, there are few data on elements such as Cl, I, La, and Mg (Harrison and Sturges 1984; Ansari and Pandis 1999; Müller 1999). This deficiency results from a number of factors such as the limited availability of routine analytical techniques that allow to determine elements at ultratrace levels and the lack of knowledge of potential health (La is the only exception: its study has spread in the last decade) (Kroll et al. 2013) and environmental (few information on emission sources) effects. For aluminum, one of the most abundant elements in nature (Chen et al. 2011), the anthropogenic origin contribution is due to its wide use in the industry (Mason and Moore 1982; Lyle et al. 2005; Hetherington et al. 2007). Despite its natural abundance, aluminum has no known function in biology; it is not essential, whereas it is not a toxic element depending on both levels and target organ. Its limited geochemical mobility causes its absorption in the human body to be reduced (Dulka and Risby 1976). The chlorine origin in airborne PM10 is both natural (soil, marine aerosol, volcanic eruptions) and anthropogenic (production of a wide range of industrial and consumer products, chemical for water purification, reactions of gaseous hydrochloric acid with ammonia) (Koski et al. 1966; Greenwood and Earnshaw 1997; Wiberg et al. 2001; Seinfeld and Pandis

2006). In the literature, toxicity and environmental information are correlated especially to the reaction with water giving HCl and HClO, whereas very few data (as chloride) are present for Cl adsorbed onto particulate matter (Manigrasso et al. 2010). Iodine (the 47th element in abundance in the Earth's crust) is an essential trace element for human life, and its main role in animal biology is as a constituent of the thyroid hormones (T3 and T4). Iodine deficiency gives rise to hypothyroidism (Felig and Frohman 2001). In atmospheric PM10, iodine origin is essentially natural (soil, marine aerosol), whereas the anthropogenic contribution is very low, due to the burning of fossil fuels. Titanium, the ninth most abundant element in the Earth's crust (0.63 % by mass) (Barksdale 1968), is widely distributed and occurs primarily in minerals; about 95 % of Ti ore extracted from the Earth is destined for refinement into TiO2. Due to their high tensile strength/density ratio (Titanium 2006), high corrosion resistance (Lide 2005), fatigue resistance, high crack resistance (Moiseyev 2006), and ability to withstand moderately high temperatures without creeping, titanium alloys are used in aircraft, armor plating, naval ships, spacecraft, and missiles (Lide 2005; Krebs 2006). For these applications, titanium alloyed with aluminum, vanadium, and other elements is used for a variety of components. From a toxicological point of view, Ti is nontoxic even in large doses and does not play any natural role inside the human body (Emsley 2001). It does, however, have a tendency to bioaccumulate in tissues that contain silica. Although there is little information on the presence of such elements in particulate matter and its relative effects, it begins to be interesting to know the levels of these species. This paper reports data of Al, Cl, I, La, Mg, Na, Si, and Ti in airborne PM10 sampled in downtown Rome during an 18-month-long sampling campaign. As a result of neutron irradiation for a few minutes, the focus was on radionuclides characterized by short half-lives (ranging from 2.31 to 37.9 min). Furthermore, As, Br, Cu, Mn, Sb, and V are also determined for evaluating the toxicological aerosol characteristics. Among the different available analytical techniques, the instrumental neutron activation analysis (INAA) was used considering the analytical characteristics and the matrix investigated. In fact, the INAA allows: (a) to use minimum sample amount (fundamental aspect in the study of atmospheric PM); (b) to eliminate all the chemical–physical sample treatments (no positive/negative artifacts); (c) to perform bulk analysis; (d) to reach high selectivity and sensitivity; (e) to have very good precision and accuracy for a lot of elements; (f) to have wide dynamic range allowing element determination between milligrams per gram and picograms per gram (Campanella et al. 1998; Avino et al. 2000, 2006, 2008, 2010, 2011a, b, 2013; Fanizza et al. 2008; Seccaroni et al. 2008; Capannesi et al. 2008, 2009, 2012).

Environ Sci Pollut Res

Experimental part Sampling site For PM10 sampling, a SWAM 5a Dual 9 Channel Monitor (FAI Instruments, Fonte Nuova, Italy) operating at 16.7 L min−1 was used. The instrument is an automatic sampling and mass measurement system. The PM10 samples are accumulated on filtering membranes, and their mass is determined by the β-attenuation technique. Particulate matter was collected on polymethylpentane ringed, 2.0-mm pore size, 37mm Teflon membranes (Gelman, type R2PJ). The samples were stored in a box under controlled conditions (atmosphere and temperature). The sampling campaign was performed in downtown Rome: The location is a wide area characterized by strong anthropogenic pollution (e.g., heating, combustion processes). Furthermore, Rome, as a big city with a very large district, has also a large area considered “downtown.” Avenues at very high autovehicular traffic surround the area where the sampling was performed, during daytime and nighttime. The sampler was located at the area center and positioned at ground level; the campaign was 19 months long from April 2010 to November 2011: The sampling was 24-h long for each filter, and 38 samples were collected (2 per month, difference of 15 days between two samplings) for understanding the element trends. All the treatments of sample storage and handling were carried out at the ISPESL's Primary Reference Laboratory according to the EU regulations (European Standard 123241 1998). Meteorological data were also collected for interpreting eventual pollution episodes. INAA analysis Samples, blank, and standards, put in nuclear-grade polyethene cylinders (Kartell, Milan, Italy), were irradiated at a neutron flux of 1.25×1013 n×cm−2 ×s−1 for 20 min in the pneumatic channel “Rabbit” of the nuclear reactor Triga Mark II of the ENEA-Casaccia Laboratories. The flux stability (>99.8 %) was tested irradiating Au standard as monitor. For the analysis, primary and secondary standards were used. Primary standards (Carlo Erba, Milano, Italy) were Al, As, Br, Cl, Cu, I, Mg, Mn, Na, and V, whereas, as secondary standards, three standard reference materials (SRMs) such as SRM 2709 (S. Joaquim Soil) and SRM 98a (Plastic Clay) from the National Institute of Standards and Technology (NIST), and a SRM GRX-4 (Soil) from the US Geochemical Survey (USGS) were involved in this study. Actually, it should be better to use a reference material with a matrix as similar as possible to the samples investigated. We used three SRMs, i.e., SRM 2709, SRM 98a, and SRM GRX-4, along with 10 primary standards, for overcoming this issue. In this way, we can assume that the measurements performed have

good quality assurance and quality control (QA/QC), and they have high reproducibility and precision, as certified by several International proficiency tests in different matrices, to which our laboratory routinely takes part. After irradiation, γ-ray spectrometry measurements of different durations were carried out using a Ge(HP) Canberra detector (Meriden, CT, USA) (full width at half maximum 1.68 keV at 1,332 keV) (Fig. 1 of the Supplementary Material shows the efficiency calibration of the detector) connected to a multichannel analyzer equipped with software packages (Canberra Genie 2k) for a γ-spectra analysis. A first measurement series was performed 6 min after the end of irradiation with measurement times of 600 s for determining 28Al, 38Cl, 66Cu, 128I, 27Mg, 29Al(Si), 51Ti, and 52V (Erdtmann and Soyka 1988) (Fig. 2 of the Supplementary Material). The second series was performed 300 min after the end of irradiation with measurement times of 7,200 s for determining 76As, 82Br, 140La, 56Mn, 24Na, and 122Sb (Erdtmann and Soyka 1988). Table 1 reports all the nuclear data and limits of detection (LODs) for the elements studied in this work.

Results and discussion Analytical methodology validation In this study, a fundamental task regards the QA/QC validation due to both the investigated matrix complexity and the elements determined. In fact, the INAA determination of elements with both short half-life and low mass number (i.e., Al, Cl, Mg, and V) shows a methodological aspect that must be considered: the no-negligible fast component of the neutron flux in the Rabbit position, where the cadmium h i fluxþfast fluxÞ ratio Cd ¼ ðthermal fast , is 2.9. This means that flux it needs to consider carefully the possible interfering reactions of type (n,p), (n,α), and (n,2n ) (Table 2). The reproducibility of the INAA measurements was tested on the pure standards of Al, As, Br, Cl, Cu, I, Mg, Mn, Na, and V, irradiated in the Rabbit. For QA/QC, secondary standards were used such as NIST 2709, NIST 98a, and USGS GRX-4. Furthermore, blank filter samples were also processed: For the calculation of the element level, the background activity of each radionuclide was subtracted. Table 3 shows the results obtained in irradiating and analyzing the three standards in the same analytical conditions of the real samples. First of all, it should be noted that the very low levels of some elements (