A STABLE ISOTOPE APPROACH TO THE STUDY OF THE ASSOCIATION OF HEAVY METALS WITH DIFFERENT SOIL COMPONENTS.

Jeffrey R. Bacon* and Irene J. Hewitt

Macaulay Land Use Research Institute, Craigiebuckler, Aberdeen AB15 8QH, UK

*e-mail: j.bacon@mluri.sari.ac.uk

ABSTRACT

Metals (Cd, Cu, Ni, Pb and Zn) with enriched isotopic composition have been introduced into experimental field plots maintained under grass. The association of the metals with soil components has been studied by thermal ionisation mass spectrometry of metals extracted from soils using the BCR sequential extraction procedure. The amount of the added spikes in each fraction was calculated in order to monitor changes with time for a range of land uses associated with the upland areas of Scotland. An understanding of the processes by which metals bind with the soil components in the field situation will provide information upon which indicators of soil quality can be based.

INTRODUCTION

The heavy metal status of soil is an indicator of soil quality as highlighted in the “UK Strategy for Sustainable Development” (DOE, 1996). A current challenge is to formulate indicators of soil quality which are based on the processes which occur within the soil and which can lead to mobilisation of the metal, uptake by crops or a deleterious effect on microbial activity.

In order to study soil processes involving heavy metals in the field situation, sequential extraction techniques, used to apportion metals to operationally defined soil components, have been applied to soils containing enriched stable isotopes of several metals. The stable isotopes were mixed into homogenised soils from a number of field sites in Scotland in the period 1990-1995. The isotope ratios of the added metals have been measured in each of the operationally defined fractions using thermal ionisation mass spectrometry (TIMS). The isotope analysis allows the proportion of the added spike in each fraction to be calculated and the changes over time to be followed.

METHODS

The procedures used to spike soils, take samples and digest samples for isotopic analysis have been described previously (Bacon et al., 1995).

Sites

Experimental plots, set up in 1990 at the Glensaugh Research Station (NE Scotland, approximately 48 km SW of Aberdeen), contained enriched ²⁰⁶Pb. Additional plots were set up following the same procedures in 1994 and 1995 at the Hartwood Research Station (approximately 10 km to the east of Motherwell, central Scotland), on improved grassland and on a woodland site, and at the Auchincruive site of Scottish Agricultural College (SAC) about 10km east of Ayr in SW Scotland. All of the later plots were spiked with enriched ¹¹⁴Cd, ⁶³Cu, ⁵⁸Ni and ⁶⁴Zn.

Sequential extraction

The first BCR sequential extraction procedure (Quevauviller et al., 1997) was applied to soils collected regularly from the sites. The three–step operationally defined scheme apportions metals to three fractions: acetic acid extractable (F1), reducible (F2) and oxidizable (F3). In addition, the residues after the third extraction (R) were digested by the HF digestion procedure used to determine the total metal concentrations in the whole soil. The sums of the concentrations in the three fractions and the residue were generally within 5% of the total concentration in the whole soil.

Analyses

Digested soil samples and soil extracts were analysed for metal concentrations using inductively coupled plasma atomic emission spectrometry (ICP–AES) or graphite furnace emission atomic absorption spectrometry (GF–AAS). Isotope ratios were measured using a VG354 thermal ionisation mass spectrometer equipped with a 16–sample turret and a five–Faraday–cup simultaneous detection system. Samples were prepared for isotope analysis by anion exchange chromatography and were loaded on to single rhenium filaments using modified silica gel procedures.

RESULTS AND DISCUSSION

Although there were differences in the extractability of lead from the different soils, in particular the acetic acid extractable fraction, the distribution of the added lead spike very rapidly (within 24 h) became that of the native lead in each of the soils and has remained relatively unchanged over several years. (Figure 1)

Figure 1. Distribution of lead between the different fractions (F1, F2, F3 and R) of the BCR sequential extraction procedure. The left hand of each pair is the native lead in the soil, the right hand the added spike. All samples from Glensaugh. A is from improved pasture at time of spiking, B after five years. Samples C and D are from a roadside site and indigenous pasture, respectively, at the time of spiking.

The ²⁰⁶Pb:²⁰⁷Pb ratio, which is a direct measure of the amount of spike in a sample, has increased steadily in all three soil fractions but appears to have levelled off after about five years (Figure 2). The amount of spike in the three extractable fractions has therefore decreased with time. It is to be noted that the isotope ratios consistently follow the same order for the three fractions with the acetic acid extractable fraction having the lowest ratio.

Figure 2. Change with time of ²⁰⁶Pb:²⁰⁷Pb ratio in soil fractions extracted from an upland site at Glensaugh, Scotland

In contrast to lead, the added zinc spike did not assume the distribution of the native zinc (Figure 3) even over a year after spiking. The added spike was predominantly extractable by acetic acid whereas most of the native zinc was not extracted and remained in the residual phase. Virtually none of the zinc in the woodland soil (40% organic matter) but a significant proportion in the grassland soils (4–6% organic matter) was associated with the oxidizable fraction. If zinc is associated with the organic matter in organic soils, it is extracted in a different way to that associated with the organic component in more mineral soils. The amount extractable in the oxidizable fraction appears to be decreasing for the latter soils with a corresponding increase remaining in the residual phase. This is confirmed (Figure 4) by a decrease in the proportion of zinc in the oxidizable fraction that is derived from the spike.

Figure 3. Distribution of zinc between the different fractions (F1, F2, F3 and R) of the BCR sequential extraction procedure. The left hand of each pair is the native zinc in the soil, the right hand the added spike. Samples a and b are from Auchincruive 1 day and 4 weeks, respectively, after spiking; c, d and e are from improved grassland at Hartwood 1, 7 and 61 weeks, respectively, after spiking; f and g are from woodland at Hartwood 1day and 5 weeks, respectively, after spiking.

Figure 4. Change with time of the percentage of the oxidizable fraction of zinc that is derived from the added spike. Improved grassland site at Hartwood.

Whereas the sensitivity of isotope analysis allowed the concentrations of metals to be increased by only 2-3% in the spiking process, thereby introducing little perturbation, the low concentration of native Cd in soils has resulted in a relatively large increase in concentration upon addition of spike. Most cadmium was extracted predominantly in the first step (acetic acid) with consequent limitations on the data that could be acquired. Preliminary data (Figure 5) indicate a possible slow decrease in the proportion of the cadmium extracted in this first step that is derived from the spike.

Figure 5. Change, over the first five weeks after spiking, of the percentage of the acetic acid extractable fraction of cadmium that is derived from the added spike. Replicate plots at Auchincruive (A) and woodland site at Hartwood (W).

REFERENCES

Bacon JR, Berrow ML, Shand CA (1995), Chem. Geol. 124:125-134.

Department of the Environment (1996). Indicators of Sustainable Development for the United Kingdom. London, HMSO.

Quevauviller P, Rauret G, Lopez-Sanchez, JF, Rubio R, Ure A, Muntau, H (1997), Sci. Total Environ. 205: 223–234