A comparison of solvent and gel electrophoretic techniques in the extraction of humic substances from soils

A CompARISON OF SOLVENT AND Gel Electrophoretic Techniques IN THE EXTRACTION OF HUMIC SUBSTANCES FROM SOILS: APPLICATION TO THE Investigation of U-Humic Interactions in A CONIFEROUS FOREST Soil

Margaret C. Graham (Margaret.Graham@ed.ac.uk, Department of Chemistry, University of Edinburgh, Edinburgh, EH9 3JJ, Scotland), Susan I. Vinogradoff, Anthony Abbott (Westlakes Scientific Consulting, Westlakes Technology Park, Moor Row, Cumbria, CA24 3LN, England) and John G. Farmer (Department of Chemistry, University of Edinburgh, Edinburgh, EH9 3JJ, Scotland).

ABSTRACT

Various procedures involving 0.1 M NaOH, 0.045 M Tris-borate or gel electrophoresis were developed to extract humic substances from a coniferous forest soil. The advantage of the electrophoretic extraction technique was that it enabled simultaneous extraction and fractionation of the humic material. The extraction/fractionation methodology was then applied to the investigation of ²³⁸U associations with humic materials with depth down a coniferous forest soil. Variations in the amounts of humic material and associated ²³⁸U extracted by the three methods were found. Importantly, the fractionation by gel electrophoresis revealed changes in humic composition which influenced the distribution of associated ²³⁸U within the humic fraction.

INTRODUCTION

Previous studies have highlighted the importance of organic matter in determining the vertical distribution and associations of heavy metals in forest soils (e.g. Wang and Benoit, 1996). As the active metal-binding component of soil organic matter, humic substances have been implicated in these associations (Graham, 1993a,b, 1995). It has been suggested, however, that the extremes of pH encountered in the traditional humic isolation procedures (pH 13 and pH 1-2) alter the structure of humic compounds by, for example, hydrolysis and auto-oxidation under prolonged exposure to alkaline conditions, and acid-catalysed condensation (Graham, 1995). This has prompted the investigation of milder methods for isolation (Vinogradoff et al, 1998, in press). This study has compared traditional 0.1 M NaOH extraction (without the subsequent acidification step) with 0.045 M Tris-borate solvent extraction and electrophoretic extraction (using 0.045 M Tris-borate as the running buffer). The aim of this work, part of a wider study on actinide associations with humic substances, was to compare the characteristics of the different extracts and to gain an insight into the processes controlling U distribution in forest soils.

METHODS

Humic material from five depths (0-1, 3-4, 5-6, 8-9 and 12-13 cm) of a coniferous forest soil, W Cumbria, England, was extracted (at room temperature) using (i) 0.1 M NaOH (pH 13; 1 h), (ii) 0.045 M Tris-borate (pH 8.5; 1 h) and (iii) horizontal bed gel electrophoresis (1% w/v agarose gel; 10 mA; 3 h; 10 fractions) with 0.045 M Tris-borate (pH 8.5) as the running buffer. The relative concentration of humic material in each of the extracts was determined by UV absorption at 254 nm (presented as an absorption index). This was also compared with the total organic matter content (measured by loss on ignition at 450°C; 24 h) of each soil sample. ‘Pseudo-total’ ²³⁸U concentrations in the soil (30% v/v H₂O₂; 8 M HNO₃ digest; 2 h), as well as ²³⁸U concentrations in the solvent and electrophoretic humic extracts were determined by ICP-MS (VG PQ3).

RESULTS AND DISCUSSION

The coniferous forest soil displayed a typical organic matter concentration profile with values of almost 50% in the near surface soil decreasing rapidly to ~30% by 3-4 cm and then more slowly to ~20% by the bottom of the soil profile (Fig. 1a). A similar trend, presented as relative concentration determined by UV absorption, was also found for the 0.1 M NaOH-extractable humic substances (Fig. 1b). This suggests that there is a good correlation (r²=0.98) between humic substance concentration (as determined by this extraction method) and total organic matter content. Fig. 1c shows data for the milder solvent and the electrophoretic extraction methods. Clearly, the concentration of humic substances extracted by both of these methods is lower than, and decreases more rapidly than, those obtained using 0.1 M NaOH. Further characterisation of the electrophoretic extract was then carried out as this technique enables simultaneous extraction and fractionation. This revealed a change in composition of the electrophoretic humic extract with increasing soil depth (Figs. 1d and 1e). A shift to lower molecular weight humic molecules with increasing soil depth, together with the lower percentage extracted at depth, suggests that a greater proportion of the humic material is in a more intractable (i.e. more humified) form.

Fig. 2a shows the vertical distribution of ‘pseudo-total’ ²³⁸U in the coniferous forest soil. The average concentration was 555±39 mg kg^-1, within the range typically found in most soils (Alloway, 1995), and there was a only a slight increase with increasing depth. The vertical trend in the association of ²³⁸U with 0.1M NaOH extractable humic substances (Fig. 2b) appears to reflect the high concentration of organic matter in the near surface soil. Thereafter, where there is a decrease in the amount of extracted humic material (Fig. 1b), there is an increase in the ‘pseudo-total’ ²³⁸U and an increase in humic-bound ²³⁸U (Fig. 2b and Table 1). This is not reflected in the milder solvent extractable ²³⁸U which decreases in a similar manner to the concentration of extracted humic material with increasing soil depth (Figs. 2c and 1c). Surprisingly, although there was good agreement between the amount of humic material extracted by 0.045 M Tris-borate and gel electrophoresis, there is a significant difference in the co-extracted ²³⁸U. Gel electrophoresis extracted two to five times that extracted by the mild solvent, representing about a third to a half of the ²³⁸U extracted by 0.1 M NaOH (Table 1).

Table 1: The percentage of ‘pseudo-total’ ²³⁸U extracted from the coniferous forest soil by 0.1 M NaOH, 0.045 M Tris-borate and electrophoresis.

Soil Depth (cm)	0.1 M NaOH (% of total soil ²³⁸U)	0.045 M Tris-borate (% of total soil ²³⁸U)	Electrophoresis (% of total soil ²³⁸U)
0.5	65	18	30
3.5	49	12	34
6.5	47	7	25
9.5	54	6	31
12.5	58	4	21

On the basis of these findings, the nature of ²³⁸U-humic associations in the electrophoretic extract required some further characterisation. Figs. 2d and 2e show the distribution of ²³⁸U amongst the gel electrophoretic fractions of humic substances for which humic concentration patterns were shown in Figs. 1d and 1e. There is indeed some correlation between the concentration of humic substances and the distribution of ²³⁸U and, moreover, there is an indication of change to association with smaller molecules with increasing depth (F5-10: 43.3% (0.5 cm) vs 57.9% (12.5 cm)). The patterns are complicated but, from Fig. 2e, it also appears that some fractions (e.g. F1, 3 and 4) of humic substances from greater soil depths may have lost a significant proportion of their binding capabilities.

CONCLUSIONS

Although 0.1 M NaOH extracts a greater amount of humic material and of humic-bound ²³⁸U, gel electrophoresis may present an informative alternative method for the extraction and fractionation of soil humic materials. In particular, the alternative approach involves milder chemical conditions, provides an extract representative of a wide humic molecular size range and reveals important changes in composition within the humic fraction. Such changes are important with respect to U associations and potential mobility within forest soils.

REFERENCES

Alloway BJ (1995), In: Heavy Metals in Soils, Glasgow, Blackie and Son Ltd.: p354.

Graham MC, Livens FR, Scott RD (1993a), Proc. 9th Intern. Conf. Heavy Metals Environ., Vol. 2: 243-246.

Graham MC, MacKenzie AB, Allan RL, Cook GT, Scott RD, Pulford ID, Livens FR (1993b), Proc. 9th Intern. Conf. Heavy Metals Environ., Vol. 2: 383-386.

Graham MC, MacKenzie AB, Scott RD, Livens FR (1995), Proc. 10^th Intern. Conf. Heavy Metals Environ., Vol. 1: 73-76.

Vinogradoff SI, Abbott A, Graham MC, Farmer JG (1998), Proc. 6^th Intern. FZK/TNO Conf. Contaminated Soil, Vol. 2: 931-32.

Vinogradoff SI, Graham MC, Farmer JG, Abbott A. Proc. 9^th Intern. Meeting Intern. Humic Substances Soc. (in press).

Wang EX, Benoit G (1996), Environ. Sci. Technol. 30: 2211-2219.

Figure 1: (a) Organic matter content (determined by loss on ignition) in a coniferous forest soil; UV absorption index (based on absorption at 254 nm) showing the relative concentrations of humic substances extracted by (b) 0.1 M NaOH and (c) 0.045 M Tris-borate ( TB) and gel electrophoresis (D EE) and also in electrophoretic fractions of humic substances from (d) 0.5 cm and (e) 12.5 cm samples from a coniferous forest soil; high fraction numbers correspond to high electrophoretic mobility and smaller molecular size (e.g. F1 >200,000 Da; F3 ~200,000 Da, F6 ~14,000 Da).

Figure 2: (a) ‘Pseudo-total’ concentration of ²³⁸U in a coniferous forest soil; concentration of ²³⁸U associated with humic substances extracted by (b) 0.1 M NaOH and (c) 0.045 M Tris-borate ( TB) and gel electrophoresis (D EE) and also in electrophoretic fractions of humic substances from (d) 0.5 cm and (e) 12.5 cm samples from a coniferous forest soil