Printer Friendly

Trace element fractionation processes in resuspended mineral aerosols extracted from Australian continental surface materials.

Introduction

The chemical composition of airborne particulate matter fine enough to be inhaled (< 10 [micro]m, [PM.sub.10]) ultimately depends upon the source materials, combined with its weathering and transport history. Much attention is currently being focused on identifying the relative contributions of anthropogenic [PM.sub.10] sources (traffic, industry, etc.), and to separate these from particles of more 'natural' origin, which derive from marine and erosional continental settings (Parekh et al. 1987; Gordon 1988; Rogge et al. 1993; Lee et al. 1994; Pio et al. 1998; Querol et al. 1998, 2001; Chan et al. 1999; Hosiokangas et al. 1999; Moreno et al. 2005, 2006a, 2006b ; Ariola et al. 2006; Hopke et al. 2006). Difficulties in such source apportionment can arise when different [PM.sub.10] populations mingle and merge into complex mixtures. Thus, continental dusts sourced from rocks and soils may progressively lose their identity in the urban environment as they become repeatedly resuspended by anthropogenic activity (e.g. traffic movement), mixed with other PM (e.g. construction dusts), and act as nucleii for the adsorption and condensation of technogenic pollutants. Such modified continental dusts commonly strongly elevate the mass of atmospheric particulate matter in the built environment, this being relevant to current air quality legislation, which is based on PM mass. In this context it is useful to have a clear idea of what exactly constitutes the mineralogy and geochemistry of continental dusts, how they can be unequivocally identified, and thus, as Zender et al. (2004) have emphasised, the extent to which they are truly 'natural' in origin.

Continental dusts derived from the physical and chemical weathering of lithospheric materials are commonly referred to as 'crustal' and their chemical signature is identified using key elements such as Si, Al K, Ti, Rb, and Sr that are characteristic of geological rock-forming minerals. Given the dominance of felsic silicates in the continental lithosphere, this effectively means those elements associated with quartz, feldspars, micas, and clay minerals, accompanied by elements in the non-silicate minority mineral groups, which are mostly carbonates, sulfates, oxides, hydroxides, and phosphates. This 'crustal' component of [PM.sub.10] is not chemically uniform and can vary considerably before any subsequent mixing with technogenic pollutants during atmospheric transport. In previous publications we have investigated examples of how in some cases a specific geological source of a crustal aerosol sample can be directly identified (Moreno et al. 2003, 2007). However, these 'geological' dusts will vary chemically not only according to their source and weathering history, but also display mineralogical fractionation between their finest grained fraction and the coarser, more arenaceous component. Thus, the inhalable fraction (i.e. [PM.sub.10]) of a geological dust will have a markedly different chemistry to the total suspended dust load lifted from the lithosphere into the atmosphere. The most obvious difference is due to the dominance of quartz in coarser size fractions ('quartz dilution'), and the consequent enrichment of most elements other than silicon in the finer fractions (Glaccum and Prospero 1980; Schutz and Rahn 1982; Hardy and Cornu 2006). However, there are also more subtle effects that produce geochemical variations in natural dusts, particularly those involving the relative susceptibilities of trace elements and their host minerals to surface weathering and transport processes. In this context, this paper investigates the geochemistry of remote-location Australian desert particulate matter as a contribution towards the better understanding of natural fractionation processes influencing [PM.sub.10] and parent materials from which they are derived.

Materials and methods

Sampling procedure

Samples of unconsolidated surface materials were collected under quiet, dry conditions from a range of remote semi-arid locations in outback Australia (Fig. 1), stored in sealed plastic bags, and dried at room temperature for 48h before any analytical treatment. The samples and their locations were as follows:

(i) Red feldspathic silty sand (ORMR) on path north-west of a prominent meander in Ormiston Creek (23[degrees]45'S, 132[degrees] 59'E) and gravels in the meander bed (ORMG), both overlying the granitic bedrock of Ormiston Pound, 100km west of Alice Springs (Northern Territory).

(ii) Red, feldspathic, micaceous, silty sand (SILV) adjacent to pelitic schist exposures offroad near Silverton, 25km north-west of Broken Hill (New South Wales; 31[degrees]52'S, 141[degrees]11'E).

(iii) Red silty sand (WILR; 31[degrees]33'S, 138[degrees]36'E) and fine white sand (WILW, Wilpena Creek bank; 31[degrees]32'S, 138[degrees]33'E) on the outcrop of quartzitic sandstone (Rawnsley Quartzite) offroad in Wilpena Pound, 430km north of Adelaide (South Australia).

(iv) Resuspended red surface particulate material alongside unsealed roads at 3 locations (Fig. 1): minor road adjacent to exposures of Palaeozoic sediments at Gosses' Bluff (Northern Territory, GB; 23[degrees]48'S, 132[degrees] 19'E); Mereenie Loop Road (Northern Territory) between Uluru and Alice Springs (MLR; 23[degrees]53'S, 131[degrees]57'E); road between Pooncarie and Mungo National Park (New South Wales) in the Darling River Basin (PN; 33[degrees]35'S, 142[degrees]29'E).

(v) Resuspended red dusts deposited on the body of a road vehicle immediately after an extended drive on unpaved roads. The first of these was collected at Glen Helen (GH; 23[degrees]41'S, 132[degrees]40'E) after driving the Mereenie Loop Road, and the second was collected after driving the geological trail through the Flinder's Ranges from Wilpena Pound to Parachilna (PC; 31[degrees]08'S, 138[degrees]23'E).

[FIGURE 1 OMITTED]

Analytical methodology

Eight of the 10 samples (PC and OH samples were not plentiful enough) were subjected to resuspension by placing them in a rotating enclosed drum under a unidirectional air flow of 25 L/min for 2 h. The rotating sample, disturbed by regular sliding and cascading during each rotation, resuspends its finer fraction into the airspace of the drum. The resuspended aerosols buoyant enough to stay in resuspension become drawn by the air current out of the drum and pass into a small chamber within which the largest and heaviest particles are deposited. Those particles small and/or light enough to be still capable of being carried by the air current continue their journey through the system, entering a Negretti elutriation filter designed to allow passage to only [PM.sub.10] grade material of average density. Those particles able to travel through this barrier are finally collected on a 47-mm filter. This mechanism, designed to collect readily extractable [PM.sub.10] from unconsolidated sediments, was done several times per sample in order to collect enough [PM.sub.10] to be characterised morphologically, mineralogically and chemically, using nitrate cellulose filters for ICP-AES, ICP-MS, XRD, and SEM analysis. A blank filter was obtained running the drum without any sample inside every time before introducing a new sample in order to determine the blank concentrations and to subtract them from the analysed samples.

The concentrations of major and trace elements for the extracted [PM.sub.10] samples were determined by means of inductively coupled plasma-atomic emission spectrometry (ICP-AES) and inductively coupled plasma-mass spectrometry (ICP-MS), respectively. The nitrate cellulose sampled filters were digested using 2.5 mL pure HN[O.sub.3] and 5 mL pure HF for 4 h at 90[degrees]C. Solutions were left to evaporate at 230[degrees]C after adding 2.5 mL pure HC1[O.sub.4] (acid ratios being HN[O.sub.3]:2HF:HClO.sub.4]). The resulting solid was then dissolved adding HN[O.sub.3] and double-distilled water (Mili-Q) obtaining a 5% HN[O.sub.3] solution that can be analysed. Digestion of international reference materials (NIST 1633b ash standard), filter and acid blanks were prepared following the same procedure to confirm the accuracy of the analysis of the acidic digestions. Analytical errors were estimated at <5% for the major elements, and about 10% for the trace elements. In the case of the parent PM samples, these were first milled by an agate mill then dried for 24 h at 60[degrees]C; 0.1 g of each sample was used to determine the concentration of major and trace elements using the same procedure described for the [PM.sub.10] samples.

Results and discussion

Tables 1 and 2 present the ICP results for the parent PM ([PM.sub.PAR], Table 1) and [PM.sub.10] (Table 2), with Post-Archean Australian Shale (PAAS, Nyakairu and Koeberl 2001) for comparison. Table 3 gives [PM.sub.10]/[PM.sub.PAR] ratios for each sample to show how most elements are concentrated into the [PM.sub.10], with the exception of K, Na, Ba, Rb, Sr, Zr, and Hf, which can have [PM.sub.10]/[PM.sub.PAR] ratios <1 due to their concentration in the coarser fraction, a behaviour indicating a mineralogical control involving detrital feldspar and zircon (see below).

Samples in Tables 1-3 are ordered from left to right in order of increasing total trace element content in the [PM.sub.PAR]. Figure 2a-d overviews the geochemical signature of the dust samples, normalising selected elements against chondrite (Fig. 2a) and PAAS (Fig. 2b-d) compositions. Figure 2a, b shows the unfractionated [PM.sub.PAR] samples with the range of fractionated [PM.sub.10] superimposed as a grey background for comparison. Figure 2c shows the [PM.sub.10] fractionates, and Fig. 2d compares the 2 'car dust' resuspended soil samples with the [PM.sub.10] range. The 15 elements selected for Fig. 2 show increasing ionic potential from left to right, with K, Na, Rb, Ba, and Sr representing the large ion lithophile elements (LILE: Z/r < 2), and La-Yb the mostly trivalent lanthanoid (rare earth) elements (which can be grouped with Y which behaves like Yb). The chondrite-normalised diagram (2a) demonstrates similar patterns for all parent PM samples, with a negative slope for REE, a striking depletion in Na and (to a lesser extent) Sr, and especially strong enrichments in Ba, light rare earths (LREE), Zr, Nb, and Th (Fig. 2a). The lowest concentrations overall are found in the Wilpena [PM.sub.PAR] samples, and the highest in those from Ormiston and Silverton, although the latter is somewhat depleted in heavy rare earths (HREE). The shaded [PM.sub.10] range mimics the same general pattern, with the exception of Zr, with enrichment being strongest for the REE and Nb.

[FIGURE 2 OMITTED]

Figure 2b further illustrates the differences between samples. The prominent negative anomaly for Na is less marked and is actually positive with respect to K in 4 of the [PM.sub.PAR] samples. In contrast the difference between Zr content of [PM.sub.PAR] and [PM.sub.10] samples is accentuated. The [PM.sub.10] chemical compositions (Fig. 2c) again all have a broadly similar pattern to each other, although the Ormiston samples show a relative decrease in Ba, St, and Zr, and HREE depletion in the Silverton sample is more pronounced relative to the other [PM.sub.10]. The Wilpena samples are relatively less depleted in trace elements (TE), indicating a strong preference for the finer fraction as already demonstrated by the high [PM.sub.10]/[PM.sub.PAR] ratios in Table 3. Finally, Fig. 2d demonstrates how the 2 'car dust' samples show a similar pattern to the other 8 [PM.sub.PAR] analyses, although with generally high concentrations due to their overall finer grain size (and consequent reduction of quartz dilution effects) so that concentrations normally lie within the field of the [PM.sub.10] samples.

[FIGURE 3 OMITTED]

Controls on fractionation

Several of the key differences between the analysed samples can be understood from Fig. 3, which plots the [PM.sub.10]/[PM.sub.PAR] ratio against 8 TE that are representative of the LILE, REE, and high field strength elements (HSFE), and lists typical mineral hosts. The LILE, represented here by Rb, Ba, and Sr, have a greater mobility and therefore susceptibility to loss by chemical weathering, so that highly feldspathic samples such as those from Ormiston have the lowest ratios. The opposite effect is demonstrated by the Wilpena sediments, which are influenced by contributions from geochemically mature local quartzitic country rocks. Thus, in both WILR and WILW, any residual LILE will mostly be present in clay minerals which are strongly fractionated into the [PM.sub.10], producing a high [PM.sub.10]/[PM.sub.PAR] ratio. The LILE contents of the roadside surface particulate samples, presumably of more mixed origin less influenced by a specifically local geological source, lie between the 2 extremes of Ormiston and Wilpena.

A different behaviour is exhibited by the lanthanoid series elements (one of the 2 REE series), which are less mobile and always present in higher amounts in the [PM.sub.10] fraction than in the parent PM of all samples. Highest absolute REE contents occur in the geochemically immature red dust samples from Ormiston and Silverton. However, as with the LILE, those samples derived from more mature sources (e.g. Wilpena and Gosses) show more marked fraetionation into the [PM.sub.10] due to the presence of relatively fewer coarse REE-bearing particles. A notable exception is provided by the granitic river gravel from Ormiston, in which REE are strongly concentrated in the [PM.sub.10] (note the anomalous leap in La value on Fig. 3). We suggest that the reason for this lies in the physical effect of hydraulic sorting on coarser REE-bearing accessory minerals which have been left behind during fluvial transport of these river gravels. Backscatter SEM study demonstrates the ubiquitous presence of the CeLa-rich phosphatic accessory mineral monazite, and there is always a strong positive correlation between LREE and [P.sub.2][O.sub.5] in any given [PM.sub.PAR] and [PM.sub.10] sample pair (Fig. 4a). Most of the phosphates are <3 [micro]m in size, with larger examples (up to 6 [micro]m) being observed only in the ORMR and SILV samples, which is consistent with their high LREE content. In contrast, no phosphate >2 [micro]m was observed in the granitic fiver gravel; any coarser heavy minerals have been removed.

[FIGURE 4 OMITTED]

Of the 4 HFSE presented on the right side of Fig. 3, 3 of them (Y, Th, and Nb), are also preferentially fractionated into the [PM.sub.10]. The notable exception to this general rule of TE enrichment in the [PM.sub.10] is provided by Zr, which shows a drop in [PM.sub.10]/[PM.sub.PAR] ratio, falling to <1 in the TE-enriched samples SILV and ORMR due to its preference for the coarser size fraction. Thorium and Nb show a similar behaviour to the LREE, being most abundant in the red dust samples from Ormiston and Silverton, but depleted in the Ormiston river gravel. The presence of Th-bearing phosphates identified under SEM at both Ormiston and Silverton indicates that these are important host minerals for this element, and (as discussed below) most of the Nb is deduced to be present in accessory oxide minerals. Yttrium is most abundant in the Ormiston samples, and, given that this locality was the only one to yield xenotime under SEM, is presumably concentrated like Th and the REE into phosphatic accessory minerals. Yttrium behaves anomalously in the Silverton sample, emulating the HREE (Fig. 3; Table 3) with a very low [PM.sub.10]/[PM.sub.PAR] ratio. This unusual lack of concentration of Y and HREE in the finer fraction, combined with a lack of the normally strongly positive correlation of Y, Ti, and [P.sub.2][O.sub.5] from [PM.sub.PAR] to [PM.sub.10] (Fig. 4b), indicates the absence of fine-grained phosphatic and titaniferous accessory minerals hosting these elements. Instead, the data are consistent with the controlling presence of a Y + HREE-bearing host mineral in the coarser size fraction, and given the metamorphic geological context, a likely phase responsible for the Y + HREE anomaly is metamorphic garnet. Further details on specific TE behaviour are discussed below.

Rubidium

In fresh rocks Rb is concentrated in K-bearing silicates such as micas and alkali feldspar, thus producing high concentrations in the samples from Ormiston (K-feldspar) and Silverton (muscovite). However, during surface weathering ion exchange and differential adsorption mechanisms tend to favour the retention of Rb over K in clay minerals such as illite (Heier and Billings 1970). This is borne out by lower K/Rb ratios in our [PM.sub.10] samples when compared to the parent PM, and [PM.sub.10.sup.K/Rb]/[PM.sub.PAR.sup.K/Rb] < 1 (see data in Table 3), which demonstrate preferential concentration of Rb into the fine fraction. The only exception is the sample from Silverton which is very rich in muscovite, this being a mica with a high K/Rb (higher than, for example, K-feldspar). Thus, the 2 samples from Ormiston both have [PM.sub.10.sup.K/Rb]/[PM.sub.PAR.sup.K/Rb] of 0.7, whereas the equivalent ratio in Silverton is 1.5.

Barium and strontium

Both of these elements substitute for K and Ca in rock-forming minerals, and so are again concentrated in the feldspathic Ormiston and micaceous (and more Ca-rich) Silverton samples. Again as with Rb, these 2 elements are strongly adsorbed onto clay minerals, producing [PM.sub.10]/[PM.sub.PAR] ratios > 1 in those samples lacking coarse unweathered feldspars and/or micas. Thus, the geochemically mature samples from Wilpena have the lowest Ba and Sr concentrations in their parent PM (few remaining feldspars and micas), but the highest [PM.sub.10.sup.Ba]/[PM.sub.PAR.sup.Ba] and (especially) [PM.sub.10]Sr/[PM.sub.PAR]sr ratios due to the presence of these elements on fine clay particles. An anomalously low [PM.sub.10.sup.Ba]/[PM.sub.PAR.sup.Ba] of < 1 in the Mereenie loop road sample is tentatively attributed to contamination by coarse traffic particles (e.g. from brakes; Pacyna 1986; Birmili et al. 2006), this location having the highest traffic volume of all our localities.

[FIGURE 5 OMITTED]

Lanthanoids

The REE in clastic sediments are concentrated in detrital accessory minerals, mainly phosphates (notably monazite and apatite for LREE and xenotime for HREE), minor silicates (notably allanite, titanate, and zoisite for LREE, garnet and zircon for HREE), and adsorbed to clays (especially HREE). Figure 5 demonstrates not only that these elements are always preferentially fractionated into the [PM.sub.10], but that there is usually a divergence in behaviour with increasing atomic number. The Ce/La ratio (Fig. 5a) is similar for both [PM.sub.PAR] and [PM.sub.10], showing the ~2:1 ratio typical of the common host mineral monazite and many rocks and dusts (e.g. Sugimae 1980; Yang et al. 2007). However, with increasing atomic number the REE show a separation between the [PM.sub.PAR] and [PM.sub.10], with the heavier elements demonstrating a distinct relative enrichment (Fig. 5b-d), excepting the HREE-depleted SILV sample. Thus, whereas the LREE distribution in [PM.sub.PAR] and [PM.sub.10] can be satisfactorily explained by the controlling influence of a host mineral such as monazite in both (Ce/La = 2 : 1), that of the HREE is more complex. In all samples excepting the HREE-depleted SILV, there is at least one other mineral host that is concentrating the heavier REE in the [PM.sub.10]. Given the general rarity of obvious strongly HREE-bearing accessory minerals (such as xenotime, identified under SEM only in ORMR), and the demonstrable relative loss of zircon in the [PM.sub.10] compared to the coarser [PM.sub.pAR], we deduce that this fractionation of HREE into the finest size fractions is most likely due to their preferential adsorption on to clay minerals. Illite and montmorillonite in particular are known to favour HREE over LREE (Maza-Rodriguez et al. 1992; Uysal and Golding 2003).

Zirconium and hafnium

These 2 elements are geochemically very similar, so that most crustal rocks retain near-chondritic Zr/Hfratios of ~35-40 (Ahrens and Erlank 1969; Hoskin and Schaltegger 2003). However, as zircon is by far the dominant Zr-Hf host mineral and preferentially incorporates Zr, the ratio declines in magrnatically evolved felsic igneous rocks such as many granites because zircon has been fractionated out (Linnen and Keppler 2002; Lowery Claiborne et al. 2006). Our dust samples similarly show a subtle tendency towards relative Zr loss, with an average Zr/Hf slightly lower than chondrite (32.5), ranging from 36 down to 24. We attribute this tendency to zircon loss resulting from a combination of physical winnowing during desert resuspension and geochemical derivation of our continental dusts ultimately from felsic igneous rocks and siliciclastic sediments. Zirconium concentrations in our samples are low (240-40ppm) when compared with the average content of 375 ppm Zr in loess recorded by McLennan and Murray (1999).

[FIGURE 6 OMITTED]

With regard to fractionation of Zr and Hf between [PM.sub.pAR] and [PM.sub.10], our dusts exhibit 3 distinct behaviours. The geochemically immature red dust samples from Ormiston (ORMR) and Silverton have much higher concentrations in the parent materials (>200 ppm Zr, >6 ppm Hf) than in the [PM.sub.10] (around 100ppm Zr, 3.5 ppm HI), indicating abundant coarse zircon (Fig. 6). In contrast, the more geochemically mature surface dust samples (WILR, WILW, PN, MLR, GB) all show much lower concentrations in the [PM.sub.pAR] (40-120ppm Zr, <3.5ppm Hf) but higher amounts in the [PM.sub.10] (96-151ppm Zr, 3-4ppm Hf). Thus, while there is clearly less fresh, coarse zircon, as would be expected from samples with a long weathering history, the finest grained component has not lost its Zr and Hf content. As Fig. 6 shows, this resistant reservoir of fine-grained zircons has the effect of creating a relatively restricted range for Zr and Hf in all [PM.sub.10] samples, whatever their origin. The third type of Zr behaviour is shown by the fiver gravel sample from Ormiston (ORMG), which has depleted levels of Zr and no marked size fractionation, a pattern which we attribute to loss of coarse zircons by hydraulic sorting of heavy minerals. Finally, the highest Zr and Hf concentrations of all our samples are shown by GH (297 : 8.5 ppm) and PC (239:6.6ppm), these being the 2 'car dust' samples already subjected to size fractionation by traffic resuspension, and therefore finer grained than the surface particulate materials but coarser than the [PM.sub.10]. This suggests a preference for grains >10 [micro]m, and is consistent with previous observations that detrital zircons most commonly measure 10-50 [micro]m (Eltayeb et al. 2001; Schutz and Rahn 1982). Road resuspension of continental surface PM thus seems to favour the concentration of natural zircon.

Niobium

Although Nb in fresh rock can be present within various rock-forming silicates such as biotite, titanate, and zircon, in siliciclastic sediments it is most commonly hosted within detrital Ti-bearing oxides, especially rutile, which is highly resistant to weathering (Zack et al. 2004). Niobium is therefore closely associated with Ti, as well as with Ta (Nb and Ta have the same ionic radius), and our samples show a predictably strong correlation between all 3 of these elements. Our data demonstrate strong fractionation of Nb into the [PM.sub.10] component so that zircon (which shows the opposite) can be eliminated as an important host mineral for this element in the finer grained dusts, further implicating a dominating control by Ti-oxides. Furthermore, most of the analyses on Table 1 show Nb/Ti[O.sub.2] ratios lying in the 9-27 range characteristic of rutile (Plank and Langmuir 1998; Barth et al. 2000), the only exceptions being the geochemically immature samples from Ormiston and Silverton which are relatively enriched in Nb. These latter samples are those most likely to preserve relatively unweathered Nb-bearing primary phases such as biotite or ilmenite (Schroeder et al. 2002), or even (in the case of the granite-derived Ormiston sediments) fresh grains of accessory Nb-minerals such as pyrochlore or columbitetantalite. The Silverton sample is also exceptional in displaying an abnormally low Nb [PM.sub.10]/[PM.sub.pAR] ratio, which we suggest is due to the presence of relatively fresh rutile derived from the local high grade metamorphic rocks. Rutile, which characteristically crystallises in high-grade metamorphic rocks, is extremely compatible for Nb compared to most metamorphic phases and therefore concentrates some 90% of the available Nb and Ti (Zack et al. 2004).

Conclusions

Given the data discussed above, we can recognise that the main controlling influences on the trace element chemistry of the [PM.sub.10] component of continental dusts are both physical and chemical. Physical fractionation will be produced by any of the following mechanisms:

1. The preference of a given trace element in chemically unweathered rocks for primary rock-forming silicates (such as fresh feldspar) and relatively large accessory minerals (such as fresh zircon). Thus, geochemically immature alkali feldspathic parent PM (e.g. those derived from granites) will contain more Na, K, Rb, Ba, as well as coarse fresh zircons that will harbor most of the Zr and Hf in a sample. [PM.sub.10]/[PM.sub.PAR] ratios of such elements in such materials will be low (e.g. ORMR, SILV).

2. Hydraulic sorting of particles will favour the loss of coarse higher density minerals so that a PM transport history that includes movement in water (e.g. by rivers or desert sheet flooding) may become depleted in heavy accessory phases such as monazite and xenotime and the trace elements they contain (e.g. REE and Y in ORMG).

3. Physical attrition of hard, durable particles during a long history of surface transport will reduce them to finer sizes. Thus, zircons, which when fresh tend to be >10 [micro]m, may eventually become fine enough to enter the [PM.sub.10] size fraction, elevating the [PM.sub.10]/[PM.sub.PAR] ratio in old, polycyclic continental dusts such as those resuspended from WILR and GB. As we illustrate in Fig. 6, the Zr and Hf concentration range for [PM.sub.10] derived from natural continental surface PM is telescoped by the combined opposing effect of the loss of coarse zircons in geochemically immature dusts, and the gain of fine, physically worn-down zircons in the more mature dusts. Although zircon is recognised as a likely host mineral for HSFE in sediments (e.g. Nyakairu and Koeberl 2001), we suggest that its role as such in [PM.sub.10] is limited by the negative anomaly characteristic of the finer size fraction (Fig. 2c).

Superimposed on these physical effects will be those of chemical fractionation, most of which will ultimately depend on the ionic potential of the individual trace element and its consequent mobility in the surface weathering environment. The main chemical controls on trace element fractionation into [PM.sub.10] will therefore be:

1. Vulnerability to dissolution by weathering, this being most characteristic of the low ionic potential elements, the most extreme example being Na (e.g. Fig. 2c), which has no low-solubility salts so that once it is in solution it tends to remain in the dissolved form and be lost. Different minerals hold onto cations with different degrees of tenacity; thus, K in muscovite is much less easily released than in biotite (White et al. 2002), incongruent weathering of alkali feldspar dissolves Na more rapidly than K (Lee and Parsons 1995), K-feldspar is more resistant than plagioclase (Nesbitt et al. 1997; White et al. 2001), and so on. The tendency towards relative loss of LILE from our [PM.sub.10] fractionates, in comparison to PAAS, is clear from Fig. 2c.

2. Ability to become adsorbed to clays after chemical breakdown of its thermodynamically unstable original mineral host. Thus, Rb, Cs, Sr, and Ba, all released by the weathering of rock-forming minerals (notably feldspar and mica), are strongly concentrated in the [PM.sub.10] of our most geochemically mature samples due to their affinity for clay minerals (WILR, WILW; Table 3). Similarly, REE can be mobilised in acidic soils but then adsorb on clays and hydroxides and so are not lost to the system (Nesbitt 1979)

3. Propensity to be hosted in small, resistant accessory minerals such as zircon (Zr, Hf), rutile (Ti, Nb, Ta), and monazite (La, Ce). The trace elements most involved here are all of high ionic potential and so have the combined advantage of resistance to both chemical and physical attack. Although clay adsorption of these elements will play a role, much of the HFSE content of natural continental desert [PM.sub.10] is contained within tiny nuggets of phosphates and Ti-oxides, and these have the potential to be extremely long-lived and recyclable (e.g. Poitrasson et al. 2002). We can see the result of this in the high [PM.sub.10]/[PM.sub.PAR] ratios for trace elements such as REE, Nd, and Th in our geochemically mature samples WILR, WILW, and GB (Table 3). It is these resistant elements with high ionic potential that are potentially most useful in defining the geochemical signature of continental dusts.

Acknowledgments

Thanks to Jim West (New South Wales Department of Primary Industries) for help supplying the 1:25 000 Silverton geological map (Geological Survey of New South Wales).

Manuscript received 18 August 2007, accepted 15 January 2008

References

Ahrens LH, Erlank AJ (1969) Hafnium. In 'Handbook of geochemistry'. Sections B-O, II/5. (Ed. KH Wedepohl) (Springer-Verlag: New York)

Ariola V, D'Alessandro A, Lucarelli F, Marcazzan G, Mazzei F et al. (2006) Elemental characterization of [PM.sub.10], PM2.5 and PM1 in the town of Genoa (Italy). Chemosphere 62, 226-232. doi: 10.1016/j. chemosphere .2005.05.004

Barth M, McDonough W, Rudnick R (2000) Tracking the budget of Nb and Ta in the continental crust. Chemical Geology 165, 197-213. doi: 10.1016/S0009-2541(99)00173-4

Birmili W, Allen A, Bary F, Harrison R (2006) Trace metal concentrations and water solubility in size-fractionated atmospheric particles and influence of road traffic. Environmental Science & Technology 40, 1144-1153. doi: 10.1021/es0486925

Chan YC, Simpson RW, Mctainsh GH, Vowles PD, Cohen DD, Bailey GM (1999) Source apportionment on PM2.5 and [PM.sub.10] aerosols in aerosols on Brisbane (Australia) by receptor modelling. Atmospheric Environment 33, 3251-3268. doi: 10.1016/S1352-2310(99)00090-4

Eltayeb M, Injuk J, Maenhaut W, Van Grieken R (2001) Elemental composition of mineral aerosol generated from Sudan Sahara sand. Journal of Atmospheric Chemistry 40, 247-273. doi: 10.1023/ A:1012272208129

Glaccum R, Prospero J (1980) Saharan aerosols over the tropical North Atlantic-Mineralogy. Marine Geology 37, 295-321. doi: 10.1016/0025-3227(80)90107-3

Gordon GE (1988) Receptor models. Environmental Science & Technology 22, 1132-1142. doi: 10.1021/es00175a002

Hardy M, Cornu S (2006) Location of natural trace elements in silty soils using particle-size fractionation. Geoderma 133, 295-308. doi: 10.1016/ j.geoderma.2005.07.015

Heier KS, Billings GK (1970) Rubidium. In 'Handbook of geochemistry'. (Ed. KH Wedepohl) pp. 37-C-1-37-N-1. (Springer-Verlag: Berlin and Heidelberg)

Hopke PK, Ito K, Mar T, Christensen WF, Eatough DJ et al. (2006) PM source apportionment and health effects: 1. Intercomparison of source apportionment results. Journal of Exposure Science and Environmental Epidemiology 16, 275-286. doi: 10.1038/sj.jea.7500458

Hosiokangas J, Ruuskanen J, Pekkanen J (1999) Effects of soil dust episodes and mixed fuel sources on source apportionment of [PM.sub.10] particles in Kuopio, Finland. Atmospheric Environment 33, 3821-3830. doi: 10.1016/S1352-2310(98)00400-2

Hoskin P, Schaltegger U (2003) The composition of zircon and igneous and metamorphic petrogenesis. In 'Zircon'. Mineralogical Society of America Revue. (Eds JM Hanchar, PWO Hoskin). Mineralogy and Geochemistry 53, 27-62.

Lee DS, Garland JA, Fox A (1994) Atmospheric concentrations of trace elements in urban areas of the United Kingdom. Atmospheric Environment 28, 2691-2713. doi: 10.1016/1352-2310 (94)90442-1

Lee MR, Parsons I (1995) Microtextural controls of weathering of perthitic alkali feldspars. Geochimica et Cosmochimica Acta 59, 4465-4492. doi: 10.1016/0016-7037(95)00255-X

Linnen R, Keppler H (2002) Melt composition control of Zr/Hf fractionation in magmatic processes. Geochimica et Cosmochimica Acta 66, 3293-3301. doi: 10.1016/S0016-7037(02)00924-9

Lowery Claiborne L, Miller CF, Walker BA, Wooden JL, Mazdab FK, Bea F (2006) Tracking magmatic processes through Zr/Hf ratios in rocks and Hf and Ti zoning in zircons: an example from the Spirit Mountain batholith, Nevada. Mineralogical Magazine 70, 517-543. doi: 10.1180/0026461067050348

Maza-Rodriguez J, Olivera-Pastor P, Bruque S, Jimenez-Lopez A (1992) Exchange selectivity of lanthanide ions in montmorillonite. Clay Minerology 27, 81-89. doi: 10.1180/claymin. 1992.027.1.08

McLennan SM, Murray RW (1999) Geochemistry of sediments. In 'Encyclopedia of geochemistry'. (Eds CP Marshall, RW Fairbridge) pp. 282-292. (Kluwer Academic Publishers: Dordrecht, The Netherlands)

Moreno T, Alastuey A, Querol X, Font O, Gibbons W (2007) Identification of metallic pathfinder elements in airborne particulate matter derived from fossil fuels at Puertollano, Spain. International Journal of Coal Geology 71, 122-128. doi: 10.1016/j.coal.2006.08.001

Moreno T, Gibbons W, Jones T, Richards R (2003) The geology of ambient aerosols: characterising urban and rural/coastal silicate [PM.sub.10] & [PM.sub.2.5-0.1] using high volume cascade collection and scanning electron microscopy. Atmospheric Environment 37, 4265-4276. doi: 10.1016/S1352-2310(03)00534-X

Moreno T, Querol X, Alastuey A, Garcia do Santos S, Fernandez Patier R, Artinano B, Gibbons W (2006a) PM source apportionment and trace metallic aerosol affinities during atmospheric pollution episodes: a case study from Puertollano, Spain. Journal of Environmental Monitoring 8, 1060-1068. doi: 10.1039/b608321h

Moreno T, Querol X, Alastuey A, Viana M, Gibbons W (2005) Exotic dust incursions into central Spain: implications for legislative controls on atmospheric particulates. Atmospheric Environment 39, 6109-6120. doi: 10.1016/j.atmosenv.2005.06.038

Moreno T, Querol X, Castillo S, Alastuey A, Cuevas E, Herrmann L, Mounkaila M, Elvira J, Gibbons W (2006b) Geochemical variations in mineral aerosols from the Sahara-Sahel Dust Corridor. Chemosphere 65, 261-270. doi: 10.1016/j.chemosphere.2006.02.052

Nesbitt H (1979) Mobility and fractionation of REE during weathering of granodiorite. Nature 279, 206-210. doi: 10.1038/279206a0

Nesbitt H, Fedo C, Young G (1997) Quartz and feldspar stability, steady and non steady state weathering and petrogenesis of siliciclastic sands and muds. Journal of Geology 105, 173-191.

Nyakairu G, Koeberl C (2001) Mineralogical and chemical composition and distribution of rare earth elements in clay-rich sediments from central Uganda. Geochemical Journal 35, 13-28.

Pacyna JM (1986) Emission factors of atmospheric elements. In 'Toxic metals in the atmosphere'. (Eds JO Nriagu, CI Davidson) (Wiley: New York)

Parekh P, Ghaudri B, Siddiqi Z, Husain L (1987) The use of chemical and statistical methods to identify sources of selected elements in ambient air aerosol in Karachi, Pakistan. Atmospheric Environment 21, 1267-1274.

Pio C, Ramos M, Duarte A (1998) Atmospheric aerosol and soiling of external surfaces in an urban environment. Atmospheric Environment 32, 1979-1989. doi: 10.1016/S1352-2310(97)00507-4

Plank T, Langmuir C (1998) The chemical composition of subducting sediment and its consequences for the crust and mantle. Chemical Geology 145, 325-394. doi: 10.1016/$0009-2541(97)00150-2

Poitrasson F, Hanchar JM, Schaltegger U (2002) The current state and future of accessory mineral research. Chemical Geology 191, 3-24. doi: 10.1016/S0009-2541(02)00146-8

Querol X, Alastuey A, Puicercus JA, Mantilla E, Miro JV, Lopez-Soler A, Plana F, Artinano B (1998) Seasonal evolution of suspended particles around a large coal-fired power station: particulate levels and sources. Atmospheric Environment 32, 1963-1978. doi: 10.1016/S1352-2310 (97)00504-9

Querol X, Alastuey A, Rodriguez S, Plana F, Ruiz C, Cots N, Massague G, Puig O (2001) PM 10 and PM2.5 source apportionment in the Barcelona metropolitan area, Catalonia, Spain. Atmospheric Environment 35, 6407-6419. doi: 10.1016/S1352-2310(01)00361-2

Rogge WF, Hildemann LM, Mazurek MA, Cass G, Simoneit B (1993) Sources of fine organic aerosol. 3. Road dust, tire debris and organometallic brake lining dust: road as sources and sinks. Environmental Science & Technology 27, 1892-1904. doi: 10.1021/ es00046a019

Schroeder PA, Le Golvan JJ, Roden MF (2002) Weathering of ilmenite from granite and chlorite schist in the Georgia Piedmont. The American Mineralogist 87, 1616-1625.

Schutz L, Rahn K (1982) Trace elements in erodible soils. Atmospheric Environment 16, 171-176. doi: 10.1016/0004-6981(82) 90324-9

Sugimae A (1980) Atmospheric concentrations and sources of rare earth elements in the Osaka area, Japan. Atmospheric Environment 14, 1171-1175. doi: 10.1016/0004-6981(80)90181-X

Uysal IT, Golding S (2003) Rare earth element fractionation in authigenic illite-smectite from Late Permian clastic rocks, Bowen Basin, Australia: implications for physico-chemical environments of fluids during illitization. Chemical Geology 193, 167-179. doi: 10.1016/S0009-2541 (02)00324-8

White A, Blum AE, Schulz MS, Huntington TG, Peters NE, Stonestrom DA (2002) Chemical weathering of the Panola granite: solute and regolith elemental fluxes and the dissolution rate of biotite. In 'Water--rock interaction, ore deposits, and environmental geochemisty: a tribute to David A. Crerar'. Special Publication 7. (Eds R Hellmann, SA Wood) pp. 37-59. (The Geochemical Society: St Louis, MO)

White A, Bullen TD, Schulz MS, Blum AE, Huntington TG, Peters NE (2001) Differential rates of feldspar weathering in granitic regoliths. Geochimica et Cosmochimica Acta 65, 847-869. doi: 10.1016/S0016-7037(00)00577-9

Yang X, Liu Y, Li C, Song Y, Zhu H, Jin X (2007) Rare earth elements of aeolian deposits in Northern China and their implications for determining the provenance of dust storms in Beijing. Geomorphology 87, 365-377. doi: 10.1016/j.geomorph.2006.10.004

Zack T, von Eynatten H, Kronz A (2004) Rutile geochemistry and its potential use in quantitative provenance studies. Sedimentary Geology 171, 37-58. doi: 10.1016/j.sedgeo.2004.05.009

Zender CS, Miller RL, Tegen I (2004) Quantifying mineral dust mass budgets: terminology, constraints, and current estimates. EOS Transactions American Geophysical Union 85, 509-512. doi: 10.1029/2004EO480002

Teresa Moreno (A'C), Fulvio Amato (A), Xavier Querol (A) Andres Alastuey (A), and Wes Gibbons (B)

(A) Institute of Earth Sciences 'Jaume Almera', CSIC, Lluis Sole i Sabaris s/n, Barcelona 08028, Spain.

(B) AP 23075, Barcelona 08080, Spain.

(C) Corresponding author. Email: tmoreno@ija.csic.es
Table 1. ICP-AES (major element, wt% oxides) and ICP-MS (trace
element, ppm) analyses of unfractionated parent PM ([PM.sub.PAR])

Samples ordered by ascending total trace element content from left to
right. CIA, Chemical index of alteration; WILR, Wilpena Pound red;
WILW, Wilpena Creek white; GB, Gosses' Bluff; PN, Pooncarie; MLR,
Mereenie Loop Road; ORMG, Ormiston Pound; ORMR, Ormiston Creek;
SILV, Silverton; GH, Glen Helen; PC, Parachilna; DL, Detection limit

 WILR WILW GB PN MLR

[Al.sub.2][O.sub.3] 1.91 1.69 2.26 3.54 5.03
CaO 0.05 0.05 0.07 5.93 0.23
[K.sub.2]O 0.24 0.38 1.21 2.71 1.90
[Na.sub.2]O 0.05 0.10 0.05 0.24 0.11
MgO 0.10 0.12 0.09 0.29 0.30
[Fe.sub.2][O.sub.3] 1.39 0.42 0.93 1.23 2.78
MnO <DL <DL 0.01 0.02 0.19
[P.sub.2][O.sub.5] 0.02 0.01 0.03 0.03 0.03
Ti[O.sub.2] 0.10 0.17 0.10 0.32 0.26
S[O.sub.3] 0.01 0.02 0.01 0.02 0.01

CIA 78.03 66.27 47.77 38.46 56.19

Li 8 8 4 12 11
Sc 1 1 1 2 3
V 25 12 12 25 40
Cr 11 6 7 12 22
Co 1 1 1 3 16
Ni 3 2 3 6 10
Cu 3 4 3 9 25
Zn 6 8 6 17 18
Ga 3 2 3 5 6
As 2 1 1 2 4
Se <DL <DL <DL <DL 1
Rb 11 14 30 28 60
Sr 10 9 21 60 37
Y 4 4 5 7 8
Zr 41 76 63 80 118
Nb 2 3 2 5 6
Mo <DL <DL <DL <DL 1
Cd <DL <DL <DL <DL <DL
Sn 0 1 1 1 2
Sb <DL <DL <DL <DL <DL
Cs 1 1 1 1 2
Ba 57 67 231 232 718
La 5 7 10 8 17
Ce 9 14 18 22 49
Pr 1 2 2 2 3
Nd 5 8 9 10 14
Sm 1 1 1 2 2
Eu <DL <DL <DL <DL 1
Gd 1 1 1 1 2
Tb <DL <DL <DL <DL <DL
Dy 1 1 1 2 2
Ho <DL <DL <DL <DL <DL
Er <DL 1 1 1 1
Tm <DL <DL <DL <DL <DL
Yb <DL 1 1 1 1
Hf 2 2 2 2 3
Ta 1 3 2 4 7
Pb 5 6 7 9 14
Th 2 4 4 4 9
U <DL 1 1 1 2

 ORMG ORMR SILV GH PC

[Al.sub.2][O.sub.3] 9.59 12.32 11.93 10.08 10.26
CaO 0.79 0.85 0.36 4.14 0.82
[K.sub.2]O 4.66 4.11 2.78 2.48 0.92
[Na.sub.2]O 1.66 1.95 1.07 0.31 1.14
MgO 0.19 0.25 0.86 1.34 2.50
[Fe.sub.2][O.sub.3] 0.81 1.92 4.68 4.44 5.09
MnO 0.01 0.04 0.04 0.05 0.07
[P.sub.2][O.sub.5] 0.02 0.03 0.07 0.07 0.13
Ti[O.sub.2] 0.09 0.34 0.64 0.76 0.94
S[O.sub.3] 0.01 0.02 0.02 0.08 0.11

CIA 46.66 54.79 64.30 59.31 78.06

Li 6 15 11 26 34
Sc 2 4 11 9 11
V 13 27 81 88 125
Cr 5 11 42 38 51
Co 1 3 11 8 12
Ni 2 4 15 12 21
Cu 2 9 21 19 32
Zn 8 20 48 42 63
Ga 11 15 19 15 15
As 1 2 3 6 9
Se <DL 1 1 2 2
Rb 203 205 171 120 122
Sr 125 131 84 86 157
Y 5 16 13 30 21
Zr 56 232 217 297 239
Nb 4 15 21 17 16
Mo <DL <DL 2 1 1
Cd <DL <DL <DL <DL <DL
Sn 1 2 2 4 3
Sb <DL <DL <DL <DL 1
Cs 2 5 3 5 6
Ba 811 769 619 603 687
La 7 43 50 54 39
Ce 18 90 105 107 77
Pr 2 7 9 10 7
Nd 7 30 40 44 33
Sm 1 4 6 6 5
Eu 1 1 1 1 1
Gd 1 4 5 6 4
Tb <DL 1 1 1 1
Dy 1 4 4 6 5
Ho <DL 1 1 1 1
Er 1 2 2 3 3
Tm <DL <DL <DL 1 <DL
Yb 1 3 2 3 3
Hf 2 7 6 8 7
Ta 7 23 23 21 15
Pb 28 31 23 28 20
Th 5 34 28 31 13
U 1 4 7 4 3

Table 2. ICP-AES (major element, wt% oxides) and ICP-MS (trace
element, ppm) analyses of [PM.sub.10] component fractionated from
[PM.sub.PAR]

Samples ordered as in Table 1. CIA, Chemical index of alteration;
PAAS, Post-Archean Australian Shale (Nyakairu and Koeberl 2001);
WILR, Wilpena Pound red; WILW, Wilpena Creek white; GB, Gosses'
Bluff; PN, Pooncarie; MLR, Mereenie Loop Road; ORMG, Ormiston
Pound; ORMR, Ormiston Creek; SILV, Silverton; GH, Glen Helen; PC,
Parachilna, DL, Detection limit

 WILR WILW GB PN MLR

[Al.sub.2][O.sub.3] 19.82 13.32 12.80 15.96 18.53
CaO 0.45 0.28 0.47 5.21 0.60
[K.sub.2]O 1.40 2.22 4.07 2.46 2.91
[Na.sub.2]O 0.18 0.34 0.10 0.31 0.12
MgO 0.89 0.78 0.72 2.00 1.55
[Fe.sub.2][O.sub.3] 8.86 2.52 5.78 6.25 7.29
MnO 0.06 0.01 0.05 0.12 0.20
[P.sub.2][O.sub.5] 0.16 0.10 0.16 0.15 0.14
Ti[O.sub.2] 0.86 0.94 0.57 0.67 0.85
S[O.sub.3] 0.08 0.11 0.03 0.09 0.11

CIA 90.7 82.4 73.4 66.7 83.6

Li 40 26 19 27 37
Sc 15 <DL 3 4 15
V 183 88 92 106 151
Cr 67 47 55 55 69
Co 14 4 15 14 29
Ni 27 11 15 28 34
Cu 30 22 17 18 72
Zn 54 71 49 86 65
Ga 28 18 18 21 27
As 11 10 9 8 14
Se 2 3 9 7 3
Rb 110 89 119 72 149
Sr 96 123 78 236 100
Y 24 23 30 22 28
Zr 148 112 133 96 151
Nb 15 14 14 10 22
Mo 2 <DL <DL <DL 1
Cd 1 1 2 <DL 1
Sn 4 3 3 2 5
Sb 1 2 <DL <DL 1
Cs 7 5 5 5 8
Ba 283 370 561 345 601
La 32 74 39 25 50
Ce 82 106 69 54 99
Pr 9 18 9 7 10
Nd 38 75 39 29 44
Sm 6 11 6 5 7
Eu 2 2 <DL <DL 2
Gd 5 7 5 4 6
Tb 1 <DL <DL <DL 1
Dy 6 6 6 5 7
Ho 1 <DL <DL <DL 1
Er 3 3 3 2 4
Tm 1 <DL <DL <DL 1
Yb 4 3 3 3 4
Lu 0 <DL <DL <DL 1
Hf 5 4 4 3 4
Ta 14 8 5 4 17
Pb 25 43 10 4 22
Th 16 32 14 8 18
U 2 2 <DL <DL 3

 ORMG ORMR SILV PAAS

[Al.sub.2][O.sub.3] 17.72 21.59 20.88 18.90
CaO 1.24 0.82 1.86 1.30
[K.sub.2]O 2.66 2.38 3.66 3.70
[Na.sub.2]O 0.70 0.73 0.51 1.20
MgO 1.52 1.01 1.81 2.20
[Fe.sub.2][O.sub.3] 6.02 6.83 7.90 7.22
MnO 0.10 0.13 0.07 0.11
[P.sub.2][O.sub.5] 0.17 0.16 0.22 0.16
Ti[O.sub.2] 0.57 0.66 0.75 1.00
S[O.sub.3] 0.15 0.09 0.09 --

CIA 79.4 84.6 77.6 75.30

Li 43 55 27 --
Sc <DL 5 15 16
V 76 87 142 150
Cr 29 30 79 110
Co 7 16 19 23
Ni <DL 14 39 55
Cu <DL 24 46 50
Zn 73 93 101 85
Ga 25 37 33 --
As 5 7 6 --
Se 6 5 4 --
Rb 167 176 155 160
Sr 107 103 376 200
Y 37 45 17 27
Zr 62 88 117 210
Nb 25 54 28 19
Mo <DL 1 3 --
Cd <DL 1 1 --
Sn 3 <DL 2 --
Sb <DL <DL 1 --
Cs 10 10 6 15
Ba 400 419 826 650
La 64 93 91 38
Ce 107 187 182 80
Pr 15 22 27 --
Nd 61 86 113 32
Sm 9 12 13 6
Eu <DL 2 2 1
Gd 8 9 10 5
Tb <DL 1 1 1
Dy 9 11 7 --
Ho <DL 2 1 --
Er 5 6 2 --
Tm <DL <DL 0 --
Yb 5 7 2 3
Lu <DL <DL 0 --
Hf <DL 4 3 5
Ta 13 34 23 --
Pb 27 36 52 --
Th 38 58 40 15
U 7 9 9 3

Table 3. [PM.sub.10]/[PM.sub.PAR] ratios (element concentrations
of fractionated particulate matter <10 [micro]m divided by
concentrations in the unfractionated parent rock)

Note ratios <1 indicate a preference for the coarser fraction.
TE-enriched samples (right side of Table) have more coarse rock
forming silicates such as alkali feldspar, mica, or zircon and show
correspondingly low ratios for elements such as Na, K, Rb, Sr, Ba,
Zr, and Hf. Conversely, the highest ratios are characteristic of the
TE-poor, geochemically recycled samples (left side of Table). WILR,
Wilpena Pound red; WILW, Wilpena Creek white; GB, Gosses' Bluff; PN,
Pooncarie; MLR, Mereenie Loop Road; ORMG, Ormiston Pound; ORMR,
Ormiston Creek; SILV, Silverton; GH, Glen Helen; PC, Parachilna

 WILR WILW GB PN

[Al.sub.2][O.sub.3] 10.4 7.9 5.7 4.5
CaO 9.8 5.3 6.9 0.9
[K.sub.2]O 5.8 5.8 3.4 0.9
[Na.sub.2]O 3.5 3.6 1.9 1.3
MgO 9.0 6.3 8.3 6.8
[Fe.sub.2][O.sub.3] 6.4 6.0 6.2 5.1
MnO 11.4 6.5 8.5 6.7
[P.sub.2][O.sub.5] 10.3 9.3 6.3 4.8
Ti[O.sub.2] 8.7 5.5 5.7 2.1

Li 5.2 3.2 5.2 2.2
Sc 16.9 -- 4.4 1.6
V 7.4 7.5 7.7 4.3
Cr 6.2 7.8 8.2 4.4
Co 11.5 6.8 11.3 5.4
Ni 10.5 6.9 6.0 5.0
Cu 8.9 5.9 5.5 2.1
Zn 8.4 8.9 8.2 5.2
Ga 10.3 8.2 7.0 4.6
As 6.9 12.0 7.8 3.9
Se 13.5 246.2 114.8 41.9
Rb 10.0 6.4 3.9 2.6
Sr 9.4 13.1 3.8 3.9
Y 6.7 5.5 6.5 3.3
Zr 3.6 1.5 2.1 1.2
Nb 8.0 4.8 6.1 2.0
Mo 7.1 -- -- --
Cd 6.0 11.4 16.3 --
Sn 8.1 5.1 4.5 1.7
Sb 34.1 98.6 -- --
Cs 9.3 7.3 6.4 3.5
Ba 4.9 5.6 2.4 1.5
La 6.9 10.2 4.0 3.0
Ce 9.3 7.7 3.9 2.5
Pr 8.2 10.9 4.7 3.1
Nd 8.0 9.9 4.5 2.9
Sm 8.6 9.9 5.4 3.2
Eu 9.3 10.7 -- --
Gd 6.7 7.2 5.0 3.0
Tb 5.8 -- -- --
Dy 7.0 6.7 5.9 3.1
Ho 6.9 -- -- --
Er 7.3 6.1 5.5 3.0
Tm -- -- -- --
Yb 7.1 5.2 5.5 2.8
Lu -- -- -- --
Hf 3.1 2.1 2.2 1.3
Ta 11.4 2.9 2.6 1.1
Pb 5.3 7.8 1.4 0.4
Th 6.5 8.5 3.9 2.0
U 4.4 3.4 -- --

 MLR ORMG ORMR SILV

[Al.sub.2][O.sub.3] 3.7 1.8 1.8 1.7
CaO 2.6 1.6 1.0 5.2
[K.sub.2]O 1.5 0.6 0.6 1.3
[Na.sub.2]O 1.1 0.4 0.4 0.5
MgO 5.2 7.9 4.0 2.1
[Fe.sub.2][O.sub.3] 2.6 7.4 3.6 1.7
MnO 1.0 8.5 3.4 1.9
[P.sub.2][O.sub.5] 4.1 8.0 4.8 3.3
Ti[O.sub.2] 3.3 6.3 2.0 1.2

Li 3.3 7.2 3.6 2.3
Sc 4.7 -- 1.2 1.4
V 3.8 5.8 3.3 1.8
Cr 3.1 5.5 2.8 1.9
Co 1.9 5.3 5.4 1.7
Ni 3.4 -- 3.8 2.5
Cu 2.9 -- 2.9 2.3
Zn 3.7 9.2 4.6 2.1
Ga 4.7 2.4 2.5 1.8
As 3.4 5.3 4.0 2.0
Se 3.5 26.8 9.7 3.2
Rb 2.5 0.8 0.9 0.9
Sr 2.7 0.9 0.8 4.5
Y 3.3 7.5 2.8 1.3
Zr 1.3 1.1 0.4 0.5
Nb 3.6 6.1 3.6 1.3
Mo 2.2 -- 3.8 1.5
Cd 3.8 -- 4.3 1.9
Sn 2.9 2.3 -- 1.1
Sb 4.5 -- -- 2.5
Cs 4.3 4.0 1.9 1.6
Ba 0.8 0.5 0.5 1.3
La 3.0 8.7 2.2 1.8
Ce 2.0 6.0 2.1 1.7
Pr 3.3 9.4 3.0 2.9
Nd 3.1 8.7 2.9 2.8
Sm 3.4 8.6 3.0 2.4
Eu 3.0 -- 2.6 1.9
Gd 2.8 7.9 2.5 2.0
Tb 2.6 2.4 1.7
Dy 3.3 7.9 3.0 1.8
Ho 3.0 2.8 1.6
Er 3.2 7.6 2.8 1.6
Tm 3.1 -- -- 1.4
Yb 3.0 7.3 2.5 1.4
Lu 3.1 -- -- 1.3
Hf 1.3 -- 0.5 0.6
Ta 2.6 1.8 1.5 1.0
Pb 1.6 1.0 1.2 2.2
Th 2.1 7.7 1.7 1.4
U 1.5 4.8 2.2 1.2
COPYRIGHT 2008 CSIRO Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2008 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Moreno, Teresa; Amato, Fulvio; Querol, Xavier; Alastuey, Andres; Gibbons, Wes
Publication:Australian Journal of Soil Research
Article Type:Report
Geographic Code:8AUST
Date:Mar 1, 2008
Words:8940
Previous Article:Comparison of solid-state [sup.13]C NMR spectra of soil organic matter from an experimental burning site acquired at two field strengths.
Next Article:Microbial biomass carbon and some soil properties as influenced by long-term sodic-water irrigation, gypsum, and organic amendments.
Topics:


Related Articles
X-ray scans for trace elements.
A new product line of yeast extracts. (Ingredients).
Fractionation of trace elements in some non-agricultural Australian soils.
If rocks could talk ... The ion microprobe extracts hidden clues about our planet's history and evolution.
Igneous rock associations 8. Arc magmatism II: geochemical and isotopic characteristics.
Chemical fractionation of fluorine in soils with a long-term phosphate fertiliser history.
Greens Pak To Go.
Evidence for biocycling from Ba/Ca, Sr/Ca, and [sup.87]Sr/[sup.86]Sr in soils (Red Brown Earths) from South Australia.

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters