- Email: [email protected]
Characterization of atmospheric aerosols using Synchroton radiation total reflection X-ray fluorescence and Fe K-edge total reflection X-ray fluorescence-X-ray absorption near-edge structure
Spectrochimica Acta Part B 63 (2008) 1489–1495
Contents lists available at ScienceDirect
Spectrochimica Acta Part B j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / s a b
Characterization of atmospheric aerosols using Synchroton radiation total reflection X-ray fluorescence and Fe K-edge total reflection X-ray fluorescence-X-ray absorption near-edge structure☆ U.E.A. Fittschen a,⁎, F. Meirer b, C. Streli b, P. Wobrauschek b, J. Thiele a, G. Falkenberg c, G. Pepponi d a
Department of Chemistry, University of Hamburg, Martin-Luther-King-Platz 6, 20146 Hamburg, Germany Atominstitut, Vienna University of Technology, Stadionallee 2, 1020 Wien, Austria c Hamburger Synchrotronstrahlungslabor at Deutsches Elektronen-Synchrotron DESY, Notkestr. 85, 22603 Hamburg, Germany d ITC-irst, Via Sommarive 18, 38050 Povo (Trento), Italy b
a r t i c l e
i n f o
Article history: Received 26 April 2008 Accepted 8 October 2008 Available online 22 October 2008 Keywords: SR-TXRF TXRF-XANES Aerosol analysis Iron
a b s t r a c t In this study a new procedure using Synchrotron total reflection X-ray fluorescence (SR-TXRF) to characterize elemental amounts in atmospheric aerosols down to particle sizes of 0.015 um is presented. The procedure was thoroughly evaluated regarding bounce off effects and blank values. Additionally the potential of total reflection X-ray fluorescence–X-ray absorption near edge structure (SR-TXRF-XANES) for speciation of FeII/III down to amounts of 34 pg in aerosols which were collected for 1 h is shown. The aerosols were collected in the city of Hamburg with a low pressure Berner impactor on Si carriers covered with silicone over time periods of 60 and 20 min each. The particles were collected in four and ten size fractions of 10.0–8.0 μm, 8.0–2.0 μm, 2.0– 0.13 μm 0.13–0.015 μm (aerodynamic particle size) and 15–30 nm, 30–60 nm, 60–130 nm, 130–250 nm, 250– 500 nm, 0.5–1 μm, 1–2 μm, 2–4 μm, 4–8 μm, 8–16 μm. Prior to the sampling “bounce off” effects on Silicone and Vaseline coated Si carriers were studied with total reflection X-ray fluorescence. According to the results silicone coated carriers were chosen for the analysis. Additionally, blank levels originating from the sampling device and the calibration procedure were studied. Blank levels of Fe corresponded to 1–10% of Fe in the aerosol samples. Blank levels stemming from the internal standard were found to be negligible. The results from the Synchroton radiation total reflection X-ray fluorescence analysis of the aerosols showed that 20 min of sampling time gave still enough sample material for elemental determination of most elements. For the determination of the oxidation state of Fe in the aerosols different Fe salts were prepared as a reference from suspensions in isopropanol. The results from the Fe K-edge Synchroton radiation total reflection X-ray fluorescence-X-ray absorption near-edge structure analysis of the aerosol samples showed that mainly Fe(III) was present in all particle size fractions. © 2008 Elsevier B.V. All rights reserved.
1. Introduction The importance of aerosols not only for human health but also for the cloud formation and the albedo of the earth have become apparent in recent years [1,2]. The particle size and their chemical load are two significant characteristics to determine their toxic potentials and give information to their origin and their geo potential. Until now characterization of aerosols representative for short time scales (1 h ☆ This paper was presented at the 12th Conference on Total Reflection X-ray Fluorescence Analysis and Related Methods held in Trento (Italy), 18-22 June 2007, and is published in the Special Issue of Spectrochimica Acta Part B, dedicated to that conference. ⁎ Corresponding author. Tel.: +49 40428386099; fax: +49 40428384381. E-mail addresses: ursula.fi[email protected] (U.E.A. Fittschen), [email protected] (F. Meirer), [email protected] (C. Streli), [email protected] (P. Wobrauschek), [email protected] (J. Thiele), [email protected] (G. Falkenberg), [email protected] (G. Pepponi). 0584-8547/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.sab.2008.10.016
and less) remains a challenging analytical aim. Size-resolved determination of major particulate matter components using conventional bulk analytical techniques typically fall short with time resolutions of 6–24 h. Fast bulk on-line methods are so far limited to non-metal components. Single particle on-line or off-line mass spectrometric methods can also detect metals [3]. Since only recently, these methods do allow for quantification on a mass per sampled air volume [4]. Recent studies on aerosol characterization showed that X-ray fluorescence using synchrotron radiation (SR) as excitation source – in total reflection geometry or using thin polymer foils as sample carriers to minimize background radiation – with its high sensitivity is an ideal method for this task [5–7]. The attribute of being non-destructive allows for further analysis of the same sample. Therefore in addition to the elemental amounts other characteristics like the species (chemical state, bonding etc.) can be determined. Only a few techniques are capable to determine the species of metals present in absolute amounts of e.g. 34–920 pg which are the
1490
U.E.A. Fittschen et al. / Spectrochimica Acta Part B 63 (2008) 1489–1495
amounts of Fe found in aerosol samples in this study. Nonetheless speciation is important. The oxidation state of an element gives further important information towards toxicity e.g. Cr(III)/Cr(VI) and As(III)/As(V) or environmental activity e.g. N(III)/N(−III), Fe(II)/Fe(III). As total reflection geometry offers very high sensitivities for elemental determination in X-ray fluorescence analysis, this characteristics can be used to apply this set up for the speciation of minute amounts of metals in X-ray absorption experiments in the fluorescence mode [8]. X-ray absorption near edge structure in total reflection geometry (TXRF-XANES) studies on As in cucumber xylem sap [9] have shown that the oxidation state of As could be determined down to concentrations of 30 ng/mL. In aerosol science it was possible to distinguish between NH+4 and NO−3 in aerosol particulates of different size fractions by means of this method [10]. As elemental determination using SR-TXRF and speciation using TXRF-XANES need essentially the same experimental set up, both analyses can be performed at the same beamline with the same experimental configuration. In this study elements in atmospheric aerosols were determined with TXRF and the oxidation state of Fe was studied with XANES using the same SR-TXRF set up. Iron is the most abundant transition metal in atmospheric aerosols. It is transported from the great landmasses into the ocean. This transport is of utmost importance because Fe as a micro nutrient is essential for phytoplankton, the basis of marine life. As plankton is responsible for the Dimethyl sulphide (DMS) production in the oceans and DMS is known to drive cloud formation over the oceans [11] it is also relevant to geo climate. The Fe oxides from the earth crust contain mainly Fe(III). Many studies on redox cycles in the atmosphere indicate that the oxidation state of Fe differs by photo reduction in the atmosphere. Several studies on the oxidation state of Fe in rain water reported that Fe(II) was present in the samples in concentrations from up to 70% [12]. Additionally it has been proposed that it is reduced by sulphur or organic compounds [13]. As the uptake of Fe(II) is easy for marine organisms in contrast to the uptake of Fe(III) a change in the oxidation state is important. As it is proven that Fe undergoes reactions in the atmosphere [14], it is likely that the oxidation state changes also between the particle size fractions. Several studies deal with speciation of Fe in aerosol particulates [15–19]. Major concern for all speciation procedures is the alteration of the species during sampling and analysis. Oxidation of Fe(II) to Fe(III) is the most likely alteration process during sampling of the aerosols. Long storage on the filters in a continuous air stream can cause changes in the oxidation state. Although the use of online speciation techniques, like DOAS and aerosol mass spectrometry can prevent for unreasonable speciation results not all elemental species can be addressed using this methods by now. Online aerosol mass spectrometry has been used to determine oxidation state of Cr and S distinguishing between the mass spectra of their oxides [20,21]. Fe forms numerous different oxides some easily converted into each other [22] and rearrangements in the mass spectrometer itself can't be excluded [23]. Therefore most studies on speciation of Fe oxidation state uses methods like: wet extraction and photometric methods [15,24] or electrochemical methods [25], Moessbauer spectroscopy, which needs very high sample amounts (mg range) [26] or XANES where less sample amounts are needed [15–19]. The direct analysing techniques described above are not capable for analysing the small amounts analyzed in this study using SR-TXRFXANES. In consequence the sampling time used in those studies is minimum 24 h. Reduction of the sampling time to 1 h using SR-TXRFXANES therefore is a great advantage. Hoffmann et al. [15] and Zhang [17] collected aerosols over a very long time period of 5 months and 24 h respectively. In the study performed by Hoffman et al. [15] aerosols were collected on filters from a ventilation system without size fractionation. Electron micro probe analysis (EMPA) with energy dispersive detection was applied in addition to the X-ray absorption studies. The results give information about the chemical composition of single particles but
do not provide information about the bulk aerosol or the oxidation state of Fe. In the study from Hirabayashi et al. [16] no information towards sampling time are given and the statistic for the XANES analyses was very poor. All the same, the three studies found predominantly Fe(III) in the aerosol particulates and as well did A. Ohta et al. and B. J. Majestic et al. [18,19] using sampling time of 24 h. Liquid extraction followed by photometric or electrochemical analysis addresses only the species of a small fraction of the aerosol 0– 20% and inconsistencies occurring when results obtained from different methods are compared indicate that redox reaction influence the results from the wet analysis [19]. Beside Fe as micro nutrient the results for Pb will be discussed more detailed according to its ecotoxicity and human toxic potential as described in Ref. [27]. In the EU a limit value for Pb of 2 μg/m3 (24 h mean) was defined. Since the 1970s, abatement measures for Pb as additive in fuel, led to reduced levels in the atmospheric environment [28]. Nonetheless Pb is again object of public discussions according to its concentrations in electronic units and toys. As fine dust (PM10, Particulate matter b10 μm, aerodynamic particle size) und ultra fine dust (PM 2.5) have been related to severe health damage [29] this study focuses on elemental determination according to the particle size with special interest in small aerosol particles. Additionally a good size resolution was desired to enable the detection of Fe(II) which was expected to be enriched in specific particle size fractions according to its generation. A 12-stage round nozzle low pressure Berner impactor which enables sampling of particles down to 15 nm in aerodynamic particle size was used to collect aerosol particulates over sampling times up to 1 h. Other Berner type impactors have been used for aerosol analysis with TXRF [30–32], but none of these achieved comparable size resolution and a low particle cut off performance. The sampling time for the aerosols was kept as short as possible to avoid oxidation of Fe during sampling and to enable the detection of changes that may occur due to day and night time and different meteorological condition in the future. To achieve very low detection limits a direct analysis of the aerosols without digestion prior to the SR-TXRF analysis as described by Fittschen et al. [6] was applied. 2. Experimental 2.1. Chemicals Standard solutions containing 1 g/L of cobalt (Merck, Darmstadt, Germany), subboiled HNO3 prepared from HNO3 (65%) from Merck and bi-distilled water were used in this work. As reference for the Fe SR-TXRF-XANES experiments following substances were used: Fe2O3, Fe3O4, Fe(II)oxalate, Fe(II)sulphide, Fe(II)chloride, Fe(II)sulphate, Fe (III)sulphate, Fe(II)ammonium sulphate and Fe(III)ammonium sulphate (Merck). These reference compounds were suspended in isopropanol (Merck) and pipetted on the reflectors. 2.2. Instruments For aerosol sampling a modified 12-stage low-pressure Berner impactor covering the size range from 16 to 0.015 μm (aerodynamic particle diameter, Haucke KG, Gmunden, Austria) was used at a flow rate of 1.5 m3/h. The impactor flow was controlled several times using a calibrated flowmeter (Gilibrator, Gilian) [33]. The ring-shaped steel impaction plates (with outer and inner diameter 70 and 30 mm respectively) were modified such that Si-wafer supports could be inserted on the plate. The supports were embedded in cavities in such a way that the level of the impaction side of the plate remained unchanged. The cut off were not recalibrated. The cut offs given here rely on those given by the manufacturer and may be changed slightly by the modification. Nonetheless, Particle sizes determined by SEM
U.E.A. Fittschen et al. / Spectrochimica Acta Part B 63 (2008) 1489–1495
analysis, which are not presented here, indicate that mean particle sizes collected on the different stages remained mainly unchanged (estimated deviation 20%). Due to the construction of this type of impactor, the particulate matter is deposited in small spots on ringshaped targets being placed on stagnation plates. The impactor was cleaned after and before each sampling with pure water in an ultra sonic bath for 1 h. Silicon wafers were chosen as material for the impaction plates, because they show very good performance in TXRF in comparison with quartz and glassy carbon carriers as a result of their low surface roughness [34]. The silicon carriers were cut in pieces of 24.7 × 15.0 mm from a silicon wafer (diameter 200 mm) with the aid of a diamond saw. To make sure that the aerosols were collected exclusively in the middle of the silicone carriers, the other parts were protected by cellulose nitrate filters from Sartorius (pore diameter: 3 μm, filter diameter: 40 mm) that were removed after the sampling. To study “bounce off” effects during sampling on the silicon reflectors these were covered with medical Vaseline or with a silicone layer. Therefore every reflector was treated with 10 μL of 10% silicone solution in isopropanol from SERVA (Heidelberg, Germany). For calibration one droplet containing 160 pg of cobalt was spotted with an HP DeskJet 500C onto the aerosol as described in Ref. [6]. 2.3. Measurements Particles were collected with a Berner impactor on silicon carriers on a university building in 10 m height in Hamburg twice for 1 h and once for 20 min. The particles on the silicon carriers from stages containing particles with aerodynamic particle diameter from 16– 8 μm, 8–2 μm, 2 –0.130 μm and 0.130–0.015 μm were analysed with both Fe K-edge TXRF-XANES and SR-TXRF. In a second experiment particles were collected as described above but in ten size fractions with aerodynamic particle diameter from 16–8 μm, 8–4 μm, 4–2 μm, 2–1 μm, 1–0.5 μm, 0.5–0.25 μm, 0.25–0.130 μm, 0.130–0.060 μm, 0.060–0.030 μm, 0.030–0.015 μm. The samples were kept under Argon atmosphere and analysed by Fe K-edge TXRF-XANES. The TXRF measurements for the study off “bounce off” effects and blank levels were performed on an EXTRAII TXRF device (Seifert, Ahrensburg). The SR-TXRF measurements were performed at the setup for TXRF comprising a vacuum chamber at HASYLAB bending magnet beamline L [7], recently equipped with a sample changer capable to carry eight samples. A peltier-cooled silicon drift detector with an active area of 50 mm2 (VORTEX, SII NanoTechnology, Northridge, USA) was used. Aerosol samples were excited with synchrotron-radiation monochromized to 18.5 keV with the aid of a multilayer monochromator and an integration time of 100 s was used. The sample was mounted vertical in front of the side-looking detector and the beam dimensions were set to 200 μm × 1400 μm (height × width). Vertical line scans with 200 μm step width have been performed in order to localize the collected aerosol and the applied picodroplet containing the Co standard. To assure an accurate quantification the single spectra of each line scan have been summed to make sure that the total amounts of aerosol and standard were measured respectively. The Kα-lines were used for the determination of all elements investigated, except for Pb, for which L-lines were used. The amounts
1491
Fig. 1. Lateral distribution of elements on a “clean” Si-wafer that was imprinted with one drop of a Co standard solution obtained by means of a SR-TXRF line scan (50 um beam heights, 50 um step width). The constant amounts of Fe and Cu indicate that these blank values are not from the printed solution.
were calculated from fundamental parameters [35].The same set up was utilized to carry out K-edge XANES measurements for Fe in the fluorescence mode. The exciting energy was tuned with a Si(111) double crystal monochromator from 6950 eV to 7600 eV in varying steps (10 eV far above the edge to 0.5 eV at the absorption edge) across the iron K-edge at 7111 eV. The fluorescence spectra were recorded using the silicon drift detector described above. The acquisition time for the spectra of a scan was set from 2 to 10 s depending on the sample mass to get a reasonable signal to noise ratio. For each specimen at least three repetitive scans were performed to improve the measurement statistic. To check the energy calibration the absorption by an elemental Fe foil was recorded in transmission mode simultaneously for each scan. The first inflection point (i.e. the first maximum of the derivative spectrum) of the Fe metal foil scan was assumed to be 7111 eV (Fe–K edge). The energy scale of each XANES scan (standards and samples) was corrected with respect to the Fe–K edge. 3. Results and discussion Prior to the aerosol analysis blank values originating from the internal standard and originating from the sampling device were studied. Additionally the loss of aerosol particles according to the “bounce off” effect was investigated.
Table 1 Elemental amount of Fe in aerosol particulates collected for 1 h under class ISO 5 clean bench with an low pressure berner impactor on Si-wafer plates coated with silicone Fraction size
Fe pg/m3
Fe pg
16–8 μm 8–2 μm 2–0.13 μm 0.13–0.015 μm
415 283 410 n. d.
5±1 17 ± 4 14 ± 3 n.d.
Elemental amounts of lead were below the detection limits. n. d.: not detected.
Fig. 2. Low pressure Berner impactor with a modified impaction plate carrying four Siwafer plates which were directly used as sample carrier in SR-TXRF analysis.
1492
U.E.A. Fittschen et al. / Spectrochimica Acta Part B 63 (2008) 1489–1495
Table 2 Elemental amounts of S, Fe and Zn in aerosol particulates (2–0.060 μm aerodynamic diameter) collected for 2 h on Si-wafer plates coated with Vaseline and coated with silicone
S Fe Zn
Silicone coated ng
Vaseline coated ng
5±1 0.10 ± 0.01 0.037 ± 0.009
1.1 ± 0.51 0.053 ± 0.009 0.033 ± 0.008
3.1. Blank values Some studies report severe blank values of Fe, Cr, Ni and Mn stemming from a slightly different impactor itself [31,32]. To estimate contamination arising from this source aerosol were collected in a clean bench (class ISO 5) for 1 h. The results are given in Table 1. Chromium, Mn, Pb and Ni could not be detected. Iron was present in absolute amounts from “not detected” to 17 pg depending on the particle size fraction. That corresponds to 1–10% of the concentrations of Fe found in atmospheric aerosol samples collected under similar conditions. For calibration one droplet of a cobalt standard solution (1 g/L) containing 160 pg cobalt was spotted as reference onto the aerosol particulates with an ink-jet printer [6,36]. Blank values stemming from this procedure were determined prior by spotting one droplet of cobalt solution (1 g/L) with a printer on a “clean” reflector followed by a line scan with SR-TXRF (50um beam heights, 50um step width). The results show that the blank values are of the same order of magnitude as the blank values for non-spotted areas on the reflectors themselves (Fig. 1). 3.2. Bounce off “Bounce off” effects were monitored by collecting aerosols on Siwafer plates with different surface coatings. For this experiment an impaction plate was constructed which can carry four Si-wafer plates
simultaneously (Fig. 2). For the aerosol sampling less modified impaction plates having a cavity to carry one Si-wafer were used. The coating of the surface with Vaseline is said to prevent “bounce off” to a high extent [32]. A few other authors report that silicon coating was used to prevent “bounce off” effects [37,38] but didn't examine its performance. Because the Vaseline was sometimes contaminated with Ca or Fe and the preparation of thin and homogeneous layers of Vaseline was not always satisfactory silicone layers were tested to prevent “bounce off” effects. As silicone solutions are commonly used to generate smooth hydrophobic surfaces in TXRF, the preparation is a routine and contamination known to be negligible. The coatings form thin elastic layers and seemed therefore suitable to prevent for “bounce off” effects during aerosol impaction. The loss of aerosols depending on the surface coating was studied using Si-wafer plates covered with Vaseline as well as coatings with silicone. The results are given in Table 2. For S and Fe the amounts collected on Silicone coated surface are higher than on Vaseline. This can be explained by the high background in the spectra taken from samples collected on Vaseline. Amounts of Zn were more or less the same. The same set up was used to determine the error of the results when different sample spots from the same particle size fraction are analyzed and found to be between 5 and 18%. For this experiment only Si-wafer plates covered with silicone were used. 3.3. Aerosol analysis During the first campaign atmospheric aerosols were collected for 20 and 60 min in the city centre of Hamburg. Fe K-edge TXRF-XANES on Fe were performed prior to the TXRF analysis, because the addition of the reference element necessary for quantification in SR-TXRF could change the Fe species. The results from the SR-TXRF analysis of the collected aerosols are presented in Fig. 3. A spectrum of the aerosols collected on stage 4, size fraction 2–0.130 μm, is shown in Fig. 4. It can be seen that 20 min of sampling time gave still enough sample material for elemental
Fig. 3. Amount of elements determined in atmospheric aerosol from three different samplings: 60 min day time (a) 20 min day time (b) and 60 min night time (c) in four size fractions 10.0–8.0 μm, 8.0–2.0 μm, 2.0–0.13 μm and 0.13–0.015 μm.
U.E.A. Fittschen et al. / Spectrochimica Acta Part B 63 (2008) 1489–1495
1493
Fig. 4. SR-TXRF spectrum obtained from aerosol particulates size fraction 2–0.130 μm at 18.5 keV excitation energy and a life time of 100 s.
determination of most elements for the size fraction 16–8 μm, 8–2 μm and 2–0.130 μm, but not enough for the determination of elements in the smallest size fraction. For the first three stages the TXRF results show significant differences between the elemental compositions, as could be expected according to the studies of Ref. [5]. The results from the first and the last 60 min sampling vary in the composition which indicates a variation of aerosol composition in the atmosphere between the two sample periods. As the quantification is based on single measurements overall uncertainty of the determination was estimated from the errors of the single methods and regarding the representativeness of the taken sample to be 20%. 3.4. Iron and lead in aerosols The concentration of Fe and Pb in the particles varied depending on the particle size from 3–24 ng/m3 and 0.045–4 ng/m3, respectively (Table 3). The highest amounts were found in the 8 to 2 μm particle size fraction for Fe and in the 2 to 0.130 μm fraction for Pb. The lowest amounts for both elements were found in the smallest particles. Summation of Pb concentrations over all stages gives bulk elemental concentration of 5 ng/m3 for sampling period “a” and 2 ng/m3 for the sampling period “c”. These concentrations are in the same order of magnitude as the concentrations (1.6–90 ng/m3) determined during a large-scale campaign on post-abatement Pb-levels in several urban and rural sites in Germany. In this campaign the aerosols were collected over a time period of 24 h [39]. Absolute amounts from 34–920 pg of Fe and 3–260 pg of Pb per sample were found. Detection limits for Fe were 100–400 fg (100 s life time). To estimate detectable levels of Fe in the aerosol particulates,
blank values stemming from the impactor have to be taken into account. Those were below the detection limits for the smallest particles but 5–17 pg for the other size fractions (Table 1). Therefore detectable levels of Fe in the aerosol correspond to ca. 0.8–3.2 pg/m3 (hourly sampling) for the smallest size fraction and to 283–415 pg/m3 (hourly sampling) for the other size fractions. Detection limits for Pb were found to vary from 200–500 fg depending on the individual sample. These values correspond to detectable levels of approx. 1.5– 8 pg/m3 (hourly sampling). The detection limits found here would allow for an hourly monitoring of Pb in the four mentioned size fractions. If a better fractionation of the particles is necessary the sampling time would have to be elongated approximately to 2 or 3 h. To learn more about the redox characteristics of Fe present in atmospheric particulate matter the oxidation state of Fe was analysed by Fe K-edge TXRF-XANES with respect to the particle size. As several studies reported, Fe is photo chemically reduced in the atmosphere it seemed possible that Fe(II) was enriched in specific particle size fractions according to its generation.
Table 3 Elemental amounts of lead in atmospheric aerosols in four size fractions: 16–8 μm, 8– 2 μm, 2–0.130 μm and 0.130–0.015 μm from three sampling for a (60 min), b (20 min) and c (60 min) in the city of Hamburg 16–8 μm Iron a Iron b Iron c Lead a Lead b Lead c
8–2 μm
2–0.13 μm
0.13–0.015 μm
pg/m3
pg
pg/m3
pg
pg/m3
pg
pg/m3
pg
7000 12,000 4000 80 n. d. 50
800 400 500 11 n. d. 6
24,000 19,000 22,000 300 n. d. 200
800 200 700 10 n. d. 7
15,000 5000 3000 4300 200 1600
900 70 200 300 3 100
n. d. n. d. 3000 n. d. n. d. 200
n. d. n. d. 30 n. d. n. d. 3
n. d.: not detected.
Fig. 5. Fe K-edge XANES spectra from different Fe species: Fe-metal, Fe(II)sulphide, Fe(II) oxalate, Fe(II)sulphate, Fe(II)chloride, Fe(II+III)oxide, AmmoniumFe(II)sulfate, AmmoniumFe(III)sulfate and Fe(III)oxide.
1494
U.E.A. Fittschen et al. / Spectrochimica Acta Part B 63 (2008) 1489–1495
verification of these results is necessary. In future studies LA-ICPTOFMS will be used to determine Pb in aerosols for comparison. The TXRF-XANES of Fe in the aerosol samples gave the result that Fe was present in the oxidation state of three and very likely in form of Fe(III)oxide in all collected aerosols. This corresponds to other studies on the oxidation state of Fe in aerosols particulates [15–19] though in rain and cloud samples high amounts of Fe(II) are reported [12]. As oxidation of Fe(II) by air flow should be drastically reduced due to the short sampling times (1 h) the fact, that the aerosols analyzed here were composed of Fe(III) may be due to the location, especially the height, where the aerosols were collected. Nonetheless, it can't be excluded that Fe(II) present in the aerosols is oxidized during sampling. Acknowledgements
Fig. 6. Fe-XANES spectra from different aerosol particle size fractions: particles size from 15–30 nm, 30–60 nm, 60–130 nm, 130–250 nm, 250–500n, 0.5–1 μm, 1–2 μm, 2– 4 μm, 4–8 μm, 8–16 μm, in addition to the aerosols one Fe-XANES curve from Fe(II) sulfate and one from Fe(III)oxide are shown.
Different Fe salts were analyzed as reference (Fig. 5). The references were prepared on silicon reflectors from low concentrated suspension of different Fe salts in isopropanol to avoid oxidation which more easily occurs in aqueous solutions. The curves were energy calibrated using the absorption edge of metallic Fe as reference. The curves were fitted using IFFEFIT Athena software [40– 42]. The shape of the XANES-curves matches very well with those that can be found in literature [43]. This shows that the procedure used here is suitable to prepare reference samples from the different Fe species for TXRF-XANES analysis. According to the XANES analysis of the aerosol collected during the first campaign, Fe was present in the oxidation state (III) in all four particle size fractions. Nonetheless it could not be excluded that small amounts of Fe(II) were present but could not be detected because of the dominance of the Fe(III) species. Therefore in a second campaign aerosol particles were collected from ten size fractions. The samples were kept under Ar atmosphere to prevent oxidation and determined by Fe K-edge TXRF-XANES (Fig. 6). It can be seen from the curves, that all fraction contain Fe predominantly in the oxidation state of three. Because the amounts of Fe on the collection plates were very low around 500–900 pg all repetitive energy scans of the same sample were merged to improve statistics. Repetitive scans of the same sample showed no changes of the XAFS confirming that no oxidation of the samples took place during the measurements (which were all performed in vacuum). 4. Conclusions Using SR-TXRF elemental amounts in atmospheric aerosols were determined down to particle sizes of 15 nm aerodynamic particle diameter and sampling times of 20 min. The blank values for Fe stemming from the impactor were 10–100 times lower then the amounts found in the aerosols. Nonetheless this subject has to be studied more detailed in the future. To address the sources of post-abatement levels of Pb in West Europe, information about the variation of Pb in the atmosphere on short time scales and information on Pb concentration according to the particle sizes are necessary. The detection limits found here will allow for a large-scale study of Pb in the atmosphere and offer the possibility of further speciation e.g. by Pb L-edge XANES. The Pb level found here 2–5 ng/m3 in particles from 10–0.015 μm are similar to those found in other studies [39]. Nonetheless a
The authors would like to thank Prof. Gerhard Lammel from the Max Plank Institute of Meteorology for the allowance to use his impactor. This work was supported by the European Community — Research Infrastructure Action under the FP6 “Structuring the European Research Area” Programme — Integrating Activity on Synchrotron and Free Electron Laser Science (I-03064EC) and the Austrian Science Foundation: FWF (Fonds zur Förderung der wissenschaftlichen Forschung), Projekt: P18299. This study was done within the frame of CRP (G4.20.02./1271) “Unification of nuclear spectrometries: integrated techniques as a new tool for material research.” References [1] K. Katsouyanni, G. Touloumi, E. Samoli, A. Gryparis, A. Le Tertre, Y. Monopolis, G. Rossi, D. Zmirou, F. Ballester, A. Boumghar, H.R. Anderson, B. Wojtyniak, A. Paldy, R. Braunstein, J. Pekkanen, C. Schindler, J. Schwartz, Confounding and effect modification in the short-term effects of ambient particles on total mortality: results from 29 European cities within the APHEA2 project, Epidemiology 12 (2001) 521–531. [2] J.E. Penner, M. Andreae, H. Annegarn, L. Barrie, J. Feichter, D. Hegg, A. Jayaraman, R. Teaitch, D. Murphy, J. Nganga, G. Pitari, Aerosols, their direct and indirect effects, in: Climate Change 2001 — the scientific basis, Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change (J.T. Houghton et al., eds.), Cambridge University Press, Cambridge, UK, pp. 289–348. [3] S.H. Pastor, J.O. Allen, L.S. Hughes, P. Bhave, G.R. Cass, K.A. Prather, Ambient single particle analysis in Riverside, California by aerosol time-of-flight mass spectrometry during the SCOS97-NARSTO, Atmos. Environ. 37 (2003) 239–258. [4] J.L. Jimenez, J.T. Jayne, Q. Shi, C.E. Kolb, D.R. Worsnop, I. Yourshaw, J.H. Seinfeld, R.C. Flagan, X. Zhang, K.A. Smith, J.W. Morris, P. Davidovits, Ambient aerosol sampling using the Aerodyne Aerosol Mass Spectrometer, J. Geophys. Res. 108 (2003) 8425, doi:10.1029/2001JD001213. [5] N. Bukowiecki, M. Hill, R. Gehrig, C.N. Zeicky, P. Lienemann, F. Hedegüs, G. Falkenberg, E. Weingartner, U. Baltensperger, Trace metals in ambient air: hourly size-segregated mass concentrations determined by synchrotron-XRF, Environ. Sci. Technol. 39 (2005) 5754–5762. [6] U.E.A. Fittschen, S. Hauschild, M.A. Amberger, G. Lammel, C. Streli, S. Förster, P. Wobrauschek, C. Jokubonis, G. Pepponi, G. Falkenberg, J.A.C. Broekaert, A new technique for the deposition of standard solutions in total reflection X-ray fluorescence spectrometry (TXRF) using pico-droplets generated by inkjet printers and its applicability for aerosol analysis with SR-TXRF, Spectrochim. Acta Part B 61 (2006) 1098–1104. [7] C. Streli, G. Pepponi, P. Wobrauschek, C. Jokubonis, G. Falkenberg, G. Zaray, A new SR-TXRF vacuum chamber for ultra trace analysis at HASYLAB, beamline L, X-Ray Spectrom 34 (2005) 451–455. [8] G. Falkenberg, G. Pepponi, C. Streli, P. Wobrauschek, Comparison of conventional and total reflection excitaion geometry for fluorscence X-ray absorption spectroscopy on droplet samples, Spectrochim. Acta Part B 58 (2003) 2239–2244. [9] F. Meirer, G. Pepponi, C. Streli, P. Wobrauschek, V.G. Mihucz, G. Zaray, V. Czech, J.A.C. Broekaert, U.E.A. Fittschen, G. Falkenberg, Arsenic Speciation in cucumber (Cucumis sativus L.) xylem sap by K-edge TXRF-XANES, X-ray Spectrom. 36 (2007) 408–412. [10] S. Török, J. Osan, B. Beckhoff, G. Ulm, Ultratrace speciation of nitrogen compounds in aerosols collected on silicon wafer surfaces by means of TXRF-NEXAFS, Powder Diffr. 19 (1) (2004) 81–86. [11] L. Spokes, T.D. Jickells, B. Lim, Solubilisation of aerosol trace metals by cloud processing: a laboratory study, Geochim. Cosmochim. Acta 58 (1994) 3281–3287. [12] G. Zhuang, Z. Yi, G.T. Wallace, Iron(II) in rainwater, snow, and surface seawater from coastal environment, Mar. Chem. 50 (1995) 41–50. [13] Y. Ertel, S.O. Pehkonen, M.R. Hoffmann, Redox chemistry of Iron in fog and stratus clouds, J. Geophys. Res. 98 (1993) 18423–18434. [14] M.O. Andrea, P.J. Crutzen, Atmospheric aerosols: biogeochemical sources and role in atmospheric chemistry, Science 276 (1997) 1052–1056.
U.E.A. Fittschen et al. / Spectrochimica Acta Part B 63 (2008) 1489–1495 [15] P. Hoffmann, A.N. Dedik, J. Ensling, S. Weinbruch, S. Weber, T. Sinner, P. Gütlich, H.M. Ortner, Speciation of iron in atmospheric aerosol samples, J. Aerosol Sci. 27 (1996) 325–337. [16] M. Hirabayashi, M. Matsuo, Characterization of iron in airborn particulate matter by X-ray absorbtion fine structure technique, Anal. Sci. 17 (2001) i1581–i1584. [17] G. Zhang, Atmospheric particulate matter studied by Mössbauer spectroscopy and XAFS, Hyperfine Interact. 151/152 (2003) 299–306. [18] A. Ohta, H. Tsuno, H. Kagi, Y. Kanai, M. Nomura, R. Zhang, S. Tershima, N. Imai, Chemical composition and XANES speciation of Fe, Mn and Zn from aerosols collected in China and Japan during dust events, Geochem. J. 40 (2006) 363–376. [19] B.J. Majestic, J.J. Schauer, M.M. Shafer, Application of synchrotron radiation for measurement of iron red-ox speciation in atmospherically processed aerosols, Atmos. Chem. Phys. Discuss. 7 (2007) 1357–1390. [20] K.R. Neubauer, M.V. Johnston, A.S. Wexler, Chromium speciation in aerosols by rapid single-particle mass spectrometry, Int. J. Mass Specrom. Ion Process. 151 (1995) 77–87. [21] K.R. Neubauer, S.T. Sum, M.V. Johnston, Sulfur speciation in individual aerosol particles, J. Geophys. Res. 101 (1996) 18701–18707. [22] A. Navrotzky, L. Mazeina, J. Majzlan, Size-driven structural and thermodynamic complexity in Iron oxides, Science 319 (2008) 1635–1638. [23] D.M. Murphy, D.S. Thomson, M.J. Mahoney, In situ measurement of organics, meteoritic material, mercury and other elements in aerosols at 5 to 19 kilometers, Science 282 (1998) 1664–1669. [24] R.J. Kieber, K. Williams, J.D. Willey, S. Skrabal, G.B. Avery, Iron speciation in coastal rainwater: concentration and deposition to seawater, Mar. Chem. 73 (2001) 83–95. [25] E. Viollier, P.W. Inglett, K. Hunter, A.N. Roychoudhury, P. Van Capellen, The ferrozine method revisited: Fe(II)/Fe(III) determination in natural waters, Appl. Geochem. 15 (2000) 785–790. [26] H.M. Ortner, Characterization of individual aerosol particles with special reference to speciation techniques in Handbook of elemental speciation: techniques and methodology, Whiley-VCH Weinheim (2003), 505–526. R. Cornelis, H. Crews, J. Caruso, K.Heumann. [27] T.C. Hutchinson, K.M. Meema (eds), lead, mercury, cadmium in the environment, SCOPE vol. 31, New York, Wiley. [28] R.M. Harrison, W.T. Sturges, The measurement and interpretation of Br/Pb ratios in airborn particles, Atmos. environ. 17 (1983) 311–318. [29] U. Lahl, W. Steven, Pneumology 59 (2005) 704–714.
1495
[30] B. Schneider, The determination of atmospheric trace metal concentrations by collection of aerosol particles on sample holders for total-reflection X-ray fluorescence, Spectrochim. Acta Part B 44 (1989) 519–524. [31] M. Schmeling, D. Klockow, Sample collection and preparation for analysis of airborne particulate matter by total reflection X-ray fluorescence spectrometry, Anal. Chim. Acta 346 (1997) 121–126. [32] R. Klockenkämper, H. Bayer, A.v. Bohlen, M. Schmeling, D. Klockow, Collection of airborne particulate matter for a subsequent analysis by Total Reflection X-ray Fluorescence, Anal. Sci. 11 (1995) 495–498. [33] G. Lammel, D.G. Baumgardner, U.E.A. Fittschen, B. Peschel, Evolution of anthropogenic aerosols in the coastal town of Salina Cruz, Mexico: part III sizesegregated elemental composition analyzed by total-reflection X-ray fluorescence spectrometry, Int. J. Environ. Anal. Chem. 87 (2007) 659–672. [34] F. Esaka, K. Watanabe, T. Onodera, T. Taguchi, M. Magara, S. Usuda, The use of Si carriers for aerosol particle collection and subsequent elemental analysis by totalreflection X-ray fluorescence spectrometry, Spectrochim. Acta Part B 58 (2003) 2145–2155. [35] D.K.G. de Boer, Fundamental parameters for X-ray fluorescence analysis, Spectrochim. Acta Part B 44 (1989) 1171–1190. [36] U.E.A. Fittschen, N.H. Bings, S. Hauschild, S. Foerster, A.F. Kiera, E. Karavani, A. Froemsdorf, J. Thiele, G. Falkenberg, Picoliter droplet dried residues as standards for direct analysis techniques, Anal. Chem. 80 (2008) 1967–1977. [37] J. Injuk, R. Van Grieken, Optimisation of total-reflection X-ray fluorescence for aerosol analysis, Spectrochim. Acta Part B 50 (1995) 1787–1803. [38] N. Streit, E. Weingartner, C. Zellweger, M. Schwikowski, H.W. Gaeggeler, U. Baltensperger, Characterization of size-fractionated aerosol from the Jungfraujoch (3580 m asl)using total reflection X-ray fluorescence (TXRF), Int. Environ. Anal. Chem. 76 (1) (2000) 1–16. [39] G. Lammel, A. Röhrl, H. Schreiber, Atmospheric lead and bromine in Germany, Environ. Sci. Pollut. Res. 9 (2002) 397–404. [40] http://cars9.uchicago.edu/ifeffit/, The IFEFFIT homepage. 2007. [41] M. Newville, IFEFFIT: interactive XAFS analysis and FEFF fitting, J. Synchrotron Radiat. 8 (2001) 322–324. [42] M. Newville, EXAFS analysis using FEFF and IFEFFIT, J. Synchrotron Radiat. 8 (2001) 96–100. [43] http://cars9.uchicago.edu/xafs/xas_fun/xas_fundamentals.pdf.
Contents lists available at ScienceDirect
Spectrochimica Acta Part B j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / s a b
Characterization of atmospheric aerosols using Synchroton radiation total reflection X-ray fluorescence and Fe K-edge total reflection X-ray fluorescence-X-ray absorption near-edge structure☆ U.E.A. Fittschen a,⁎, F. Meirer b, C. Streli b, P. Wobrauschek b, J. Thiele a, G. Falkenberg c, G. Pepponi d a
Department of Chemistry, University of Hamburg, Martin-Luther-King-Platz 6, 20146 Hamburg, Germany Atominstitut, Vienna University of Technology, Stadionallee 2, 1020 Wien, Austria c Hamburger Synchrotronstrahlungslabor at Deutsches Elektronen-Synchrotron DESY, Notkestr. 85, 22603 Hamburg, Germany d ITC-irst, Via Sommarive 18, 38050 Povo (Trento), Italy b
a r t i c l e
i n f o
Article history: Received 26 April 2008 Accepted 8 October 2008 Available online 22 October 2008 Keywords: SR-TXRF TXRF-XANES Aerosol analysis Iron
a b s t r a c t In this study a new procedure using Synchrotron total reflection X-ray fluorescence (SR-TXRF) to characterize elemental amounts in atmospheric aerosols down to particle sizes of 0.015 um is presented. The procedure was thoroughly evaluated regarding bounce off effects and blank values. Additionally the potential of total reflection X-ray fluorescence–X-ray absorption near edge structure (SR-TXRF-XANES) for speciation of FeII/III down to amounts of 34 pg in aerosols which were collected for 1 h is shown. The aerosols were collected in the city of Hamburg with a low pressure Berner impactor on Si carriers covered with silicone over time periods of 60 and 20 min each. The particles were collected in four and ten size fractions of 10.0–8.0 μm, 8.0–2.0 μm, 2.0– 0.13 μm 0.13–0.015 μm (aerodynamic particle size) and 15–30 nm, 30–60 nm, 60–130 nm, 130–250 nm, 250– 500 nm, 0.5–1 μm, 1–2 μm, 2–4 μm, 4–8 μm, 8–16 μm. Prior to the sampling “bounce off” effects on Silicone and Vaseline coated Si carriers were studied with total reflection X-ray fluorescence. According to the results silicone coated carriers were chosen for the analysis. Additionally, blank levels originating from the sampling device and the calibration procedure were studied. Blank levels of Fe corresponded to 1–10% of Fe in the aerosol samples. Blank levels stemming from the internal standard were found to be negligible. The results from the Synchroton radiation total reflection X-ray fluorescence analysis of the aerosols showed that 20 min of sampling time gave still enough sample material for elemental determination of most elements. For the determination of the oxidation state of Fe in the aerosols different Fe salts were prepared as a reference from suspensions in isopropanol. The results from the Fe K-edge Synchroton radiation total reflection X-ray fluorescence-X-ray absorption near-edge structure analysis of the aerosol samples showed that mainly Fe(III) was present in all particle size fractions. © 2008 Elsevier B.V. All rights reserved.
1. Introduction The importance of aerosols not only for human health but also for the cloud formation and the albedo of the earth have become apparent in recent years [1,2]. The particle size and their chemical load are two significant characteristics to determine their toxic potentials and give information to their origin and their geo potential. Until now characterization of aerosols representative for short time scales (1 h ☆ This paper was presented at the 12th Conference on Total Reflection X-ray Fluorescence Analysis and Related Methods held in Trento (Italy), 18-22 June 2007, and is published in the Special Issue of Spectrochimica Acta Part B, dedicated to that conference. ⁎ Corresponding author. Tel.: +49 40428386099; fax: +49 40428384381. E-mail addresses: ursula.fi[email protected] (U.E.A. Fittschen), [email protected] (F. Meirer), [email protected] (C. Streli), [email protected] (P. Wobrauschek), [email protected] (J. Thiele), [email protected] (G. Falkenberg), [email protected] (G. Pepponi). 0584-8547/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.sab.2008.10.016
and less) remains a challenging analytical aim. Size-resolved determination of major particulate matter components using conventional bulk analytical techniques typically fall short with time resolutions of 6–24 h. Fast bulk on-line methods are so far limited to non-metal components. Single particle on-line or off-line mass spectrometric methods can also detect metals [3]. Since only recently, these methods do allow for quantification on a mass per sampled air volume [4]. Recent studies on aerosol characterization showed that X-ray fluorescence using synchrotron radiation (SR) as excitation source – in total reflection geometry or using thin polymer foils as sample carriers to minimize background radiation – with its high sensitivity is an ideal method for this task [5–7]. The attribute of being non-destructive allows for further analysis of the same sample. Therefore in addition to the elemental amounts other characteristics like the species (chemical state, bonding etc.) can be determined. Only a few techniques are capable to determine the species of metals present in absolute amounts of e.g. 34–920 pg which are the
1490
U.E.A. Fittschen et al. / Spectrochimica Acta Part B 63 (2008) 1489–1495
amounts of Fe found in aerosol samples in this study. Nonetheless speciation is important. The oxidation state of an element gives further important information towards toxicity e.g. Cr(III)/Cr(VI) and As(III)/As(V) or environmental activity e.g. N(III)/N(−III), Fe(II)/Fe(III). As total reflection geometry offers very high sensitivities for elemental determination in X-ray fluorescence analysis, this characteristics can be used to apply this set up for the speciation of minute amounts of metals in X-ray absorption experiments in the fluorescence mode [8]. X-ray absorption near edge structure in total reflection geometry (TXRF-XANES) studies on As in cucumber xylem sap [9] have shown that the oxidation state of As could be determined down to concentrations of 30 ng/mL. In aerosol science it was possible to distinguish between NH+4 and NO−3 in aerosol particulates of different size fractions by means of this method [10]. As elemental determination using SR-TXRF and speciation using TXRF-XANES need essentially the same experimental set up, both analyses can be performed at the same beamline with the same experimental configuration. In this study elements in atmospheric aerosols were determined with TXRF and the oxidation state of Fe was studied with XANES using the same SR-TXRF set up. Iron is the most abundant transition metal in atmospheric aerosols. It is transported from the great landmasses into the ocean. This transport is of utmost importance because Fe as a micro nutrient is essential for phytoplankton, the basis of marine life. As plankton is responsible for the Dimethyl sulphide (DMS) production in the oceans and DMS is known to drive cloud formation over the oceans [11] it is also relevant to geo climate. The Fe oxides from the earth crust contain mainly Fe(III). Many studies on redox cycles in the atmosphere indicate that the oxidation state of Fe differs by photo reduction in the atmosphere. Several studies on the oxidation state of Fe in rain water reported that Fe(II) was present in the samples in concentrations from up to 70% [12]. Additionally it has been proposed that it is reduced by sulphur or organic compounds [13]. As the uptake of Fe(II) is easy for marine organisms in contrast to the uptake of Fe(III) a change in the oxidation state is important. As it is proven that Fe undergoes reactions in the atmosphere [14], it is likely that the oxidation state changes also between the particle size fractions. Several studies deal with speciation of Fe in aerosol particulates [15–19]. Major concern for all speciation procedures is the alteration of the species during sampling and analysis. Oxidation of Fe(II) to Fe(III) is the most likely alteration process during sampling of the aerosols. Long storage on the filters in a continuous air stream can cause changes in the oxidation state. Although the use of online speciation techniques, like DOAS and aerosol mass spectrometry can prevent for unreasonable speciation results not all elemental species can be addressed using this methods by now. Online aerosol mass spectrometry has been used to determine oxidation state of Cr and S distinguishing between the mass spectra of their oxides [20,21]. Fe forms numerous different oxides some easily converted into each other [22] and rearrangements in the mass spectrometer itself can't be excluded [23]. Therefore most studies on speciation of Fe oxidation state uses methods like: wet extraction and photometric methods [15,24] or electrochemical methods [25], Moessbauer spectroscopy, which needs very high sample amounts (mg range) [26] or XANES where less sample amounts are needed [15–19]. The direct analysing techniques described above are not capable for analysing the small amounts analyzed in this study using SR-TXRFXANES. In consequence the sampling time used in those studies is minimum 24 h. Reduction of the sampling time to 1 h using SR-TXRFXANES therefore is a great advantage. Hoffmann et al. [15] and Zhang [17] collected aerosols over a very long time period of 5 months and 24 h respectively. In the study performed by Hoffman et al. [15] aerosols were collected on filters from a ventilation system without size fractionation. Electron micro probe analysis (EMPA) with energy dispersive detection was applied in addition to the X-ray absorption studies. The results give information about the chemical composition of single particles but
do not provide information about the bulk aerosol or the oxidation state of Fe. In the study from Hirabayashi et al. [16] no information towards sampling time are given and the statistic for the XANES analyses was very poor. All the same, the three studies found predominantly Fe(III) in the aerosol particulates and as well did A. Ohta et al. and B. J. Majestic et al. [18,19] using sampling time of 24 h. Liquid extraction followed by photometric or electrochemical analysis addresses only the species of a small fraction of the aerosol 0– 20% and inconsistencies occurring when results obtained from different methods are compared indicate that redox reaction influence the results from the wet analysis [19]. Beside Fe as micro nutrient the results for Pb will be discussed more detailed according to its ecotoxicity and human toxic potential as described in Ref. [27]. In the EU a limit value for Pb of 2 μg/m3 (24 h mean) was defined. Since the 1970s, abatement measures for Pb as additive in fuel, led to reduced levels in the atmospheric environment [28]. Nonetheless Pb is again object of public discussions according to its concentrations in electronic units and toys. As fine dust (PM10, Particulate matter b10 μm, aerodynamic particle size) und ultra fine dust (PM 2.5) have been related to severe health damage [29] this study focuses on elemental determination according to the particle size with special interest in small aerosol particles. Additionally a good size resolution was desired to enable the detection of Fe(II) which was expected to be enriched in specific particle size fractions according to its generation. A 12-stage round nozzle low pressure Berner impactor which enables sampling of particles down to 15 nm in aerodynamic particle size was used to collect aerosol particulates over sampling times up to 1 h. Other Berner type impactors have been used for aerosol analysis with TXRF [30–32], but none of these achieved comparable size resolution and a low particle cut off performance. The sampling time for the aerosols was kept as short as possible to avoid oxidation of Fe during sampling and to enable the detection of changes that may occur due to day and night time and different meteorological condition in the future. To achieve very low detection limits a direct analysis of the aerosols without digestion prior to the SR-TXRF analysis as described by Fittschen et al. [6] was applied. 2. Experimental 2.1. Chemicals Standard solutions containing 1 g/L of cobalt (Merck, Darmstadt, Germany), subboiled HNO3 prepared from HNO3 (65%) from Merck and bi-distilled water were used in this work. As reference for the Fe SR-TXRF-XANES experiments following substances were used: Fe2O3, Fe3O4, Fe(II)oxalate, Fe(II)sulphide, Fe(II)chloride, Fe(II)sulphate, Fe (III)sulphate, Fe(II)ammonium sulphate and Fe(III)ammonium sulphate (Merck). These reference compounds were suspended in isopropanol (Merck) and pipetted on the reflectors. 2.2. Instruments For aerosol sampling a modified 12-stage low-pressure Berner impactor covering the size range from 16 to 0.015 μm (aerodynamic particle diameter, Haucke KG, Gmunden, Austria) was used at a flow rate of 1.5 m3/h. The impactor flow was controlled several times using a calibrated flowmeter (Gilibrator, Gilian) [33]. The ring-shaped steel impaction plates (with outer and inner diameter 70 and 30 mm respectively) were modified such that Si-wafer supports could be inserted on the plate. The supports were embedded in cavities in such a way that the level of the impaction side of the plate remained unchanged. The cut off were not recalibrated. The cut offs given here rely on those given by the manufacturer and may be changed slightly by the modification. Nonetheless, Particle sizes determined by SEM
U.E.A. Fittschen et al. / Spectrochimica Acta Part B 63 (2008) 1489–1495
analysis, which are not presented here, indicate that mean particle sizes collected on the different stages remained mainly unchanged (estimated deviation 20%). Due to the construction of this type of impactor, the particulate matter is deposited in small spots on ringshaped targets being placed on stagnation plates. The impactor was cleaned after and before each sampling with pure water in an ultra sonic bath for 1 h. Silicon wafers were chosen as material for the impaction plates, because they show very good performance in TXRF in comparison with quartz and glassy carbon carriers as a result of their low surface roughness [34]. The silicon carriers were cut in pieces of 24.7 × 15.0 mm from a silicon wafer (diameter 200 mm) with the aid of a diamond saw. To make sure that the aerosols were collected exclusively in the middle of the silicone carriers, the other parts were protected by cellulose nitrate filters from Sartorius (pore diameter: 3 μm, filter diameter: 40 mm) that were removed after the sampling. To study “bounce off” effects during sampling on the silicon reflectors these were covered with medical Vaseline or with a silicone layer. Therefore every reflector was treated with 10 μL of 10% silicone solution in isopropanol from SERVA (Heidelberg, Germany). For calibration one droplet containing 160 pg of cobalt was spotted with an HP DeskJet 500C onto the aerosol as described in Ref. [6]. 2.3. Measurements Particles were collected with a Berner impactor on silicon carriers on a university building in 10 m height in Hamburg twice for 1 h and once for 20 min. The particles on the silicon carriers from stages containing particles with aerodynamic particle diameter from 16– 8 μm, 8–2 μm, 2 –0.130 μm and 0.130–0.015 μm were analysed with both Fe K-edge TXRF-XANES and SR-TXRF. In a second experiment particles were collected as described above but in ten size fractions with aerodynamic particle diameter from 16–8 μm, 8–4 μm, 4–2 μm, 2–1 μm, 1–0.5 μm, 0.5–0.25 μm, 0.25–0.130 μm, 0.130–0.060 μm, 0.060–0.030 μm, 0.030–0.015 μm. The samples were kept under Argon atmosphere and analysed by Fe K-edge TXRF-XANES. The TXRF measurements for the study off “bounce off” effects and blank levels were performed on an EXTRAII TXRF device (Seifert, Ahrensburg). The SR-TXRF measurements were performed at the setup for TXRF comprising a vacuum chamber at HASYLAB bending magnet beamline L [7], recently equipped with a sample changer capable to carry eight samples. A peltier-cooled silicon drift detector with an active area of 50 mm2 (VORTEX, SII NanoTechnology, Northridge, USA) was used. Aerosol samples were excited with synchrotron-radiation monochromized to 18.5 keV with the aid of a multilayer monochromator and an integration time of 100 s was used. The sample was mounted vertical in front of the side-looking detector and the beam dimensions were set to 200 μm × 1400 μm (height × width). Vertical line scans with 200 μm step width have been performed in order to localize the collected aerosol and the applied picodroplet containing the Co standard. To assure an accurate quantification the single spectra of each line scan have been summed to make sure that the total amounts of aerosol and standard were measured respectively. The Kα-lines were used for the determination of all elements investigated, except for Pb, for which L-lines were used. The amounts
1491
Fig. 1. Lateral distribution of elements on a “clean” Si-wafer that was imprinted with one drop of a Co standard solution obtained by means of a SR-TXRF line scan (50 um beam heights, 50 um step width). The constant amounts of Fe and Cu indicate that these blank values are not from the printed solution.
were calculated from fundamental parameters [35].The same set up was utilized to carry out K-edge XANES measurements for Fe in the fluorescence mode. The exciting energy was tuned with a Si(111) double crystal monochromator from 6950 eV to 7600 eV in varying steps (10 eV far above the edge to 0.5 eV at the absorption edge) across the iron K-edge at 7111 eV. The fluorescence spectra were recorded using the silicon drift detector described above. The acquisition time for the spectra of a scan was set from 2 to 10 s depending on the sample mass to get a reasonable signal to noise ratio. For each specimen at least three repetitive scans were performed to improve the measurement statistic. To check the energy calibration the absorption by an elemental Fe foil was recorded in transmission mode simultaneously for each scan. The first inflection point (i.e. the first maximum of the derivative spectrum) of the Fe metal foil scan was assumed to be 7111 eV (Fe–K edge). The energy scale of each XANES scan (standards and samples) was corrected with respect to the Fe–K edge. 3. Results and discussion Prior to the aerosol analysis blank values originating from the internal standard and originating from the sampling device were studied. Additionally the loss of aerosol particles according to the “bounce off” effect was investigated.
Table 1 Elemental amount of Fe in aerosol particulates collected for 1 h under class ISO 5 clean bench with an low pressure berner impactor on Si-wafer plates coated with silicone Fraction size
Fe pg/m3
Fe pg
16–8 μm 8–2 μm 2–0.13 μm 0.13–0.015 μm
415 283 410 n. d.
5±1 17 ± 4 14 ± 3 n.d.
Elemental amounts of lead were below the detection limits. n. d.: not detected.
Fig. 2. Low pressure Berner impactor with a modified impaction plate carrying four Siwafer plates which were directly used as sample carrier in SR-TXRF analysis.
1492
U.E.A. Fittschen et al. / Spectrochimica Acta Part B 63 (2008) 1489–1495
Table 2 Elemental amounts of S, Fe and Zn in aerosol particulates (2–0.060 μm aerodynamic diameter) collected for 2 h on Si-wafer plates coated with Vaseline and coated with silicone
S Fe Zn
Silicone coated ng
Vaseline coated ng
5±1 0.10 ± 0.01 0.037 ± 0.009
1.1 ± 0.51 0.053 ± 0.009 0.033 ± 0.008
3.1. Blank values Some studies report severe blank values of Fe, Cr, Ni and Mn stemming from a slightly different impactor itself [31,32]. To estimate contamination arising from this source aerosol were collected in a clean bench (class ISO 5) for 1 h. The results are given in Table 1. Chromium, Mn, Pb and Ni could not be detected. Iron was present in absolute amounts from “not detected” to 17 pg depending on the particle size fraction. That corresponds to 1–10% of the concentrations of Fe found in atmospheric aerosol samples collected under similar conditions. For calibration one droplet of a cobalt standard solution (1 g/L) containing 160 pg cobalt was spotted as reference onto the aerosol particulates with an ink-jet printer [6,36]. Blank values stemming from this procedure were determined prior by spotting one droplet of cobalt solution (1 g/L) with a printer on a “clean” reflector followed by a line scan with SR-TXRF (50um beam heights, 50um step width). The results show that the blank values are of the same order of magnitude as the blank values for non-spotted areas on the reflectors themselves (Fig. 1). 3.2. Bounce off “Bounce off” effects were monitored by collecting aerosols on Siwafer plates with different surface coatings. For this experiment an impaction plate was constructed which can carry four Si-wafer plates
simultaneously (Fig. 2). For the aerosol sampling less modified impaction plates having a cavity to carry one Si-wafer were used. The coating of the surface with Vaseline is said to prevent “bounce off” to a high extent [32]. A few other authors report that silicon coating was used to prevent “bounce off” effects [37,38] but didn't examine its performance. Because the Vaseline was sometimes contaminated with Ca or Fe and the preparation of thin and homogeneous layers of Vaseline was not always satisfactory silicone layers were tested to prevent “bounce off” effects. As silicone solutions are commonly used to generate smooth hydrophobic surfaces in TXRF, the preparation is a routine and contamination known to be negligible. The coatings form thin elastic layers and seemed therefore suitable to prevent for “bounce off” effects during aerosol impaction. The loss of aerosols depending on the surface coating was studied using Si-wafer plates covered with Vaseline as well as coatings with silicone. The results are given in Table 2. For S and Fe the amounts collected on Silicone coated surface are higher than on Vaseline. This can be explained by the high background in the spectra taken from samples collected on Vaseline. Amounts of Zn were more or less the same. The same set up was used to determine the error of the results when different sample spots from the same particle size fraction are analyzed and found to be between 5 and 18%. For this experiment only Si-wafer plates covered with silicone were used. 3.3. Aerosol analysis During the first campaign atmospheric aerosols were collected for 20 and 60 min in the city centre of Hamburg. Fe K-edge TXRF-XANES on Fe were performed prior to the TXRF analysis, because the addition of the reference element necessary for quantification in SR-TXRF could change the Fe species. The results from the SR-TXRF analysis of the collected aerosols are presented in Fig. 3. A spectrum of the aerosols collected on stage 4, size fraction 2–0.130 μm, is shown in Fig. 4. It can be seen that 20 min of sampling time gave still enough sample material for elemental
Fig. 3. Amount of elements determined in atmospheric aerosol from three different samplings: 60 min day time (a) 20 min day time (b) and 60 min night time (c) in four size fractions 10.0–8.0 μm, 8.0–2.0 μm, 2.0–0.13 μm and 0.13–0.015 μm.
U.E.A. Fittschen et al. / Spectrochimica Acta Part B 63 (2008) 1489–1495
1493
Fig. 4. SR-TXRF spectrum obtained from aerosol particulates size fraction 2–0.130 μm at 18.5 keV excitation energy and a life time of 100 s.
determination of most elements for the size fraction 16–8 μm, 8–2 μm and 2–0.130 μm, but not enough for the determination of elements in the smallest size fraction. For the first three stages the TXRF results show significant differences between the elemental compositions, as could be expected according to the studies of Ref. [5]. The results from the first and the last 60 min sampling vary in the composition which indicates a variation of aerosol composition in the atmosphere between the two sample periods. As the quantification is based on single measurements overall uncertainty of the determination was estimated from the errors of the single methods and regarding the representativeness of the taken sample to be 20%. 3.4. Iron and lead in aerosols The concentration of Fe and Pb in the particles varied depending on the particle size from 3–24 ng/m3 and 0.045–4 ng/m3, respectively (Table 3). The highest amounts were found in the 8 to 2 μm particle size fraction for Fe and in the 2 to 0.130 μm fraction for Pb. The lowest amounts for both elements were found in the smallest particles. Summation of Pb concentrations over all stages gives bulk elemental concentration of 5 ng/m3 for sampling period “a” and 2 ng/m3 for the sampling period “c”. These concentrations are in the same order of magnitude as the concentrations (1.6–90 ng/m3) determined during a large-scale campaign on post-abatement Pb-levels in several urban and rural sites in Germany. In this campaign the aerosols were collected over a time period of 24 h [39]. Absolute amounts from 34–920 pg of Fe and 3–260 pg of Pb per sample were found. Detection limits for Fe were 100–400 fg (100 s life time). To estimate detectable levels of Fe in the aerosol particulates,
blank values stemming from the impactor have to be taken into account. Those were below the detection limits for the smallest particles but 5–17 pg for the other size fractions (Table 1). Therefore detectable levels of Fe in the aerosol correspond to ca. 0.8–3.2 pg/m3 (hourly sampling) for the smallest size fraction and to 283–415 pg/m3 (hourly sampling) for the other size fractions. Detection limits for Pb were found to vary from 200–500 fg depending on the individual sample. These values correspond to detectable levels of approx. 1.5– 8 pg/m3 (hourly sampling). The detection limits found here would allow for an hourly monitoring of Pb in the four mentioned size fractions. If a better fractionation of the particles is necessary the sampling time would have to be elongated approximately to 2 or 3 h. To learn more about the redox characteristics of Fe present in atmospheric particulate matter the oxidation state of Fe was analysed by Fe K-edge TXRF-XANES with respect to the particle size. As several studies reported, Fe is photo chemically reduced in the atmosphere it seemed possible that Fe(II) was enriched in specific particle size fractions according to its generation.
Table 3 Elemental amounts of lead in atmospheric aerosols in four size fractions: 16–8 μm, 8– 2 μm, 2–0.130 μm and 0.130–0.015 μm from three sampling for a (60 min), b (20 min) and c (60 min) in the city of Hamburg 16–8 μm Iron a Iron b Iron c Lead a Lead b Lead c
8–2 μm
2–0.13 μm
0.13–0.015 μm
pg/m3
pg
pg/m3
pg
pg/m3
pg
pg/m3
pg
7000 12,000 4000 80 n. d. 50
800 400 500 11 n. d. 6
24,000 19,000 22,000 300 n. d. 200
800 200 700 10 n. d. 7
15,000 5000 3000 4300 200 1600
900 70 200 300 3 100
n. d. n. d. 3000 n. d. n. d. 200
n. d. n. d. 30 n. d. n. d. 3
n. d.: not detected.
Fig. 5. Fe K-edge XANES spectra from different Fe species: Fe-metal, Fe(II)sulphide, Fe(II) oxalate, Fe(II)sulphate, Fe(II)chloride, Fe(II+III)oxide, AmmoniumFe(II)sulfate, AmmoniumFe(III)sulfate and Fe(III)oxide.
1494
U.E.A. Fittschen et al. / Spectrochimica Acta Part B 63 (2008) 1489–1495
verification of these results is necessary. In future studies LA-ICPTOFMS will be used to determine Pb in aerosols for comparison. The TXRF-XANES of Fe in the aerosol samples gave the result that Fe was present in the oxidation state of three and very likely in form of Fe(III)oxide in all collected aerosols. This corresponds to other studies on the oxidation state of Fe in aerosols particulates [15–19] though in rain and cloud samples high amounts of Fe(II) are reported [12]. As oxidation of Fe(II) by air flow should be drastically reduced due to the short sampling times (1 h) the fact, that the aerosols analyzed here were composed of Fe(III) may be due to the location, especially the height, where the aerosols were collected. Nonetheless, it can't be excluded that Fe(II) present in the aerosols is oxidized during sampling. Acknowledgements
Fig. 6. Fe-XANES spectra from different aerosol particle size fractions: particles size from 15–30 nm, 30–60 nm, 60–130 nm, 130–250 nm, 250–500n, 0.5–1 μm, 1–2 μm, 2– 4 μm, 4–8 μm, 8–16 μm, in addition to the aerosols one Fe-XANES curve from Fe(II) sulfate and one from Fe(III)oxide are shown.
Different Fe salts were analyzed as reference (Fig. 5). The references were prepared on silicon reflectors from low concentrated suspension of different Fe salts in isopropanol to avoid oxidation which more easily occurs in aqueous solutions. The curves were energy calibrated using the absorption edge of metallic Fe as reference. The curves were fitted using IFFEFIT Athena software [40– 42]. The shape of the XANES-curves matches very well with those that can be found in literature [43]. This shows that the procedure used here is suitable to prepare reference samples from the different Fe species for TXRF-XANES analysis. According to the XANES analysis of the aerosol collected during the first campaign, Fe was present in the oxidation state (III) in all four particle size fractions. Nonetheless it could not be excluded that small amounts of Fe(II) were present but could not be detected because of the dominance of the Fe(III) species. Therefore in a second campaign aerosol particles were collected from ten size fractions. The samples were kept under Ar atmosphere to prevent oxidation and determined by Fe K-edge TXRF-XANES (Fig. 6). It can be seen from the curves, that all fraction contain Fe predominantly in the oxidation state of three. Because the amounts of Fe on the collection plates were very low around 500–900 pg all repetitive energy scans of the same sample were merged to improve statistics. Repetitive scans of the same sample showed no changes of the XAFS confirming that no oxidation of the samples took place during the measurements (which were all performed in vacuum). 4. Conclusions Using SR-TXRF elemental amounts in atmospheric aerosols were determined down to particle sizes of 15 nm aerodynamic particle diameter and sampling times of 20 min. The blank values for Fe stemming from the impactor were 10–100 times lower then the amounts found in the aerosols. Nonetheless this subject has to be studied more detailed in the future. To address the sources of post-abatement levels of Pb in West Europe, information about the variation of Pb in the atmosphere on short time scales and information on Pb concentration according to the particle sizes are necessary. The detection limits found here will allow for a large-scale study of Pb in the atmosphere and offer the possibility of further speciation e.g. by Pb L-edge XANES. The Pb level found here 2–5 ng/m3 in particles from 10–0.015 μm are similar to those found in other studies [39]. Nonetheless a
The authors would like to thank Prof. Gerhard Lammel from the Max Plank Institute of Meteorology for the allowance to use his impactor. This work was supported by the European Community — Research Infrastructure Action under the FP6 “Structuring the European Research Area” Programme — Integrating Activity on Synchrotron and Free Electron Laser Science (I-03064EC) and the Austrian Science Foundation: FWF (Fonds zur Förderung der wissenschaftlichen Forschung), Projekt: P18299. This study was done within the frame of CRP (G4.20.02./1271) “Unification of nuclear spectrometries: integrated techniques as a new tool for material research.” References [1] K. Katsouyanni, G. Touloumi, E. Samoli, A. Gryparis, A. Le Tertre, Y. Monopolis, G. Rossi, D. Zmirou, F. Ballester, A. Boumghar, H.R. Anderson, B. Wojtyniak, A. Paldy, R. Braunstein, J. Pekkanen, C. Schindler, J. Schwartz, Confounding and effect modification in the short-term effects of ambient particles on total mortality: results from 29 European cities within the APHEA2 project, Epidemiology 12 (2001) 521–531. [2] J.E. Penner, M. Andreae, H. Annegarn, L. Barrie, J. Feichter, D. Hegg, A. Jayaraman, R. Teaitch, D. Murphy, J. Nganga, G. Pitari, Aerosols, their direct and indirect effects, in: Climate Change 2001 — the scientific basis, Contribution of Working Group I to the Third Assessment Report of the Intergovernmental Panel on Climate Change (J.T. Houghton et al., eds.), Cambridge University Press, Cambridge, UK, pp. 289–348. [3] S.H. Pastor, J.O. Allen, L.S. Hughes, P. Bhave, G.R. Cass, K.A. Prather, Ambient single particle analysis in Riverside, California by aerosol time-of-flight mass spectrometry during the SCOS97-NARSTO, Atmos. Environ. 37 (2003) 239–258. [4] J.L. Jimenez, J.T. Jayne, Q. Shi, C.E. Kolb, D.R. Worsnop, I. Yourshaw, J.H. Seinfeld, R.C. Flagan, X. Zhang, K.A. Smith, J.W. Morris, P. Davidovits, Ambient aerosol sampling using the Aerodyne Aerosol Mass Spectrometer, J. Geophys. Res. 108 (2003) 8425, doi:10.1029/2001JD001213. [5] N. Bukowiecki, M. Hill, R. Gehrig, C.N. Zeicky, P. Lienemann, F. Hedegüs, G. Falkenberg, E. Weingartner, U. Baltensperger, Trace metals in ambient air: hourly size-segregated mass concentrations determined by synchrotron-XRF, Environ. Sci. Technol. 39 (2005) 5754–5762. [6] U.E.A. Fittschen, S. Hauschild, M.A. Amberger, G. Lammel, C. Streli, S. Förster, P. Wobrauschek, C. Jokubonis, G. Pepponi, G. Falkenberg, J.A.C. Broekaert, A new technique for the deposition of standard solutions in total reflection X-ray fluorescence spectrometry (TXRF) using pico-droplets generated by inkjet printers and its applicability for aerosol analysis with SR-TXRF, Spectrochim. Acta Part B 61 (2006) 1098–1104. [7] C. Streli, G. Pepponi, P. Wobrauschek, C. Jokubonis, G. Falkenberg, G. Zaray, A new SR-TXRF vacuum chamber for ultra trace analysis at HASYLAB, beamline L, X-Ray Spectrom 34 (2005) 451–455. [8] G. Falkenberg, G. Pepponi, C. Streli, P. Wobrauschek, Comparison of conventional and total reflection excitaion geometry for fluorscence X-ray absorption spectroscopy on droplet samples, Spectrochim. Acta Part B 58 (2003) 2239–2244. [9] F. Meirer, G. Pepponi, C. Streli, P. Wobrauschek, V.G. Mihucz, G. Zaray, V. Czech, J.A.C. Broekaert, U.E.A. Fittschen, G. Falkenberg, Arsenic Speciation in cucumber (Cucumis sativus L.) xylem sap by K-edge TXRF-XANES, X-ray Spectrom. 36 (2007) 408–412. [10] S. Török, J. Osan, B. Beckhoff, G. Ulm, Ultratrace speciation of nitrogen compounds in aerosols collected on silicon wafer surfaces by means of TXRF-NEXAFS, Powder Diffr. 19 (1) (2004) 81–86. [11] L. Spokes, T.D. Jickells, B. Lim, Solubilisation of aerosol trace metals by cloud processing: a laboratory study, Geochim. Cosmochim. Acta 58 (1994) 3281–3287. [12] G. Zhuang, Z. Yi, G.T. Wallace, Iron(II) in rainwater, snow, and surface seawater from coastal environment, Mar. Chem. 50 (1995) 41–50. [13] Y. Ertel, S.O. Pehkonen, M.R. Hoffmann, Redox chemistry of Iron in fog and stratus clouds, J. Geophys. Res. 98 (1993) 18423–18434. [14] M.O. Andrea, P.J. Crutzen, Atmospheric aerosols: biogeochemical sources and role in atmospheric chemistry, Science 276 (1997) 1052–1056.
U.E.A. Fittschen et al. / Spectrochimica Acta Part B 63 (2008) 1489–1495 [15] P. Hoffmann, A.N. Dedik, J. Ensling, S. Weinbruch, S. Weber, T. Sinner, P. Gütlich, H.M. Ortner, Speciation of iron in atmospheric aerosol samples, J. Aerosol Sci. 27 (1996) 325–337. [16] M. Hirabayashi, M. Matsuo, Characterization of iron in airborn particulate matter by X-ray absorbtion fine structure technique, Anal. Sci. 17 (2001) i1581–i1584. [17] G. Zhang, Atmospheric particulate matter studied by Mössbauer spectroscopy and XAFS, Hyperfine Interact. 151/152 (2003) 299–306. [18] A. Ohta, H. Tsuno, H. Kagi, Y. Kanai, M. Nomura, R. Zhang, S. Tershima, N. Imai, Chemical composition and XANES speciation of Fe, Mn and Zn from aerosols collected in China and Japan during dust events, Geochem. J. 40 (2006) 363–376. [19] B.J. Majestic, J.J. Schauer, M.M. Shafer, Application of synchrotron radiation for measurement of iron red-ox speciation in atmospherically processed aerosols, Atmos. Chem. Phys. Discuss. 7 (2007) 1357–1390. [20] K.R. Neubauer, M.V. Johnston, A.S. Wexler, Chromium speciation in aerosols by rapid single-particle mass spectrometry, Int. J. Mass Specrom. Ion Process. 151 (1995) 77–87. [21] K.R. Neubauer, S.T. Sum, M.V. Johnston, Sulfur speciation in individual aerosol particles, J. Geophys. Res. 101 (1996) 18701–18707. [22] A. Navrotzky, L. Mazeina, J. Majzlan, Size-driven structural and thermodynamic complexity in Iron oxides, Science 319 (2008) 1635–1638. [23] D.M. Murphy, D.S. Thomson, M.J. Mahoney, In situ measurement of organics, meteoritic material, mercury and other elements in aerosols at 5 to 19 kilometers, Science 282 (1998) 1664–1669. [24] R.J. Kieber, K. Williams, J.D. Willey, S. Skrabal, G.B. Avery, Iron speciation in coastal rainwater: concentration and deposition to seawater, Mar. Chem. 73 (2001) 83–95. [25] E. Viollier, P.W. Inglett, K. Hunter, A.N. Roychoudhury, P. Van Capellen, The ferrozine method revisited: Fe(II)/Fe(III) determination in natural waters, Appl. Geochem. 15 (2000) 785–790. [26] H.M. Ortner, Characterization of individual aerosol particles with special reference to speciation techniques in Handbook of elemental speciation: techniques and methodology, Whiley-VCH Weinheim (2003), 505–526. R. Cornelis, H. Crews, J. Caruso, K.Heumann. [27] T.C. Hutchinson, K.M. Meema (eds), lead, mercury, cadmium in the environment, SCOPE vol. 31, New York, Wiley. [28] R.M. Harrison, W.T. Sturges, The measurement and interpretation of Br/Pb ratios in airborn particles, Atmos. environ. 17 (1983) 311–318. [29] U. Lahl, W. Steven, Pneumology 59 (2005) 704–714.
1495
[30] B. Schneider, The determination of atmospheric trace metal concentrations by collection of aerosol particles on sample holders for total-reflection X-ray fluorescence, Spectrochim. Acta Part B 44 (1989) 519–524. [31] M. Schmeling, D. Klockow, Sample collection and preparation for analysis of airborne particulate matter by total reflection X-ray fluorescence spectrometry, Anal. Chim. Acta 346 (1997) 121–126. [32] R. Klockenkämper, H. Bayer, A.v. Bohlen, M. Schmeling, D. Klockow, Collection of airborne particulate matter for a subsequent analysis by Total Reflection X-ray Fluorescence, Anal. Sci. 11 (1995) 495–498. [33] G. Lammel, D.G. Baumgardner, U.E.A. Fittschen, B. Peschel, Evolution of anthropogenic aerosols in the coastal town of Salina Cruz, Mexico: part III sizesegregated elemental composition analyzed by total-reflection X-ray fluorescence spectrometry, Int. J. Environ. Anal. Chem. 87 (2007) 659–672. [34] F. Esaka, K. Watanabe, T. Onodera, T. Taguchi, M. Magara, S. Usuda, The use of Si carriers for aerosol particle collection and subsequent elemental analysis by totalreflection X-ray fluorescence spectrometry, Spectrochim. Acta Part B 58 (2003) 2145–2155. [35] D.K.G. de Boer, Fundamental parameters for X-ray fluorescence analysis, Spectrochim. Acta Part B 44 (1989) 1171–1190. [36] U.E.A. Fittschen, N.H. Bings, S. Hauschild, S. Foerster, A.F. Kiera, E. Karavani, A. Froemsdorf, J. Thiele, G. Falkenberg, Picoliter droplet dried residues as standards for direct analysis techniques, Anal. Chem. 80 (2008) 1967–1977. [37] J. Injuk, R. Van Grieken, Optimisation of total-reflection X-ray fluorescence for aerosol analysis, Spectrochim. Acta Part B 50 (1995) 1787–1803. [38] N. Streit, E. Weingartner, C. Zellweger, M. Schwikowski, H.W. Gaeggeler, U. Baltensperger, Characterization of size-fractionated aerosol from the Jungfraujoch (3580 m asl)using total reflection X-ray fluorescence (TXRF), Int. Environ. Anal. Chem. 76 (1) (2000) 1–16. [39] G. Lammel, A. Röhrl, H. Schreiber, Atmospheric lead and bromine in Germany, Environ. Sci. Pollut. Res. 9 (2002) 397–404. [40] http://cars9.uchicago.edu/ifeffit/, The IFEFFIT homepage. 2007. [41] M. Newville, IFEFFIT: interactive XAFS analysis and FEFF fitting, J. Synchrotron Radiat. 8 (2001) 322–324. [42] M. Newville, EXAFS analysis using FEFF and IFEFFIT, J. Synchrotron Radiat. 8 (2001) 96–100. [43] http://cars9.uchicago.edu/xafs/xas_fun/xas_fundamentals.pdf.