Exchangeable soil and sediment fractions compared

Soil to plant transfer depends on availability, plant needs and plant excretion. Readily soluble and exchangeable ions represent the minimum available fraction. Traditionally, different extracts have been standardized to monitor just one or two nutrients resp. trace elements. Whereas customers expect a “true value”, leachings with neutral salts lead to slightly different results. Some of the uncertainties are due to the low concentration levels, others due to resorption of reagent blanks at the solid matrix. Thus, the interpretation is of relevance, and not the numeric level.

From the point of view of blanks and ICP-compatibility, ammonium salts or weak organic acids should be preferable extractants.

Data from a ring test of 5 soils and 12 participants show widely overlapping data ranges of extracts obtained with ammonium acetate and ammonium nitrate, but a trend to lower release of Cd, Cu, Pb and Zn into nitrate versus acetate, contrary to Cr and Ni. This is expectable due to complexation capabilities of acetate.

Also from sediments of the River Danube, 1M ammonium acetate released much more Pb, Zn, P, Ca, and Mn than 1M ammonium chloride at the same pH = 7. For Cr, Al, and As, however, this effect was rather reverse. The proportion of acetate versus chloride release did not significantly correlate with loss of ignition, nor with pedogenic of these samples. (The pedogenic oxides were calculated as the sum of Al+Fe+Mn released into oxalate buffer pH 3, and turned into their oxides).

Is it possible, to substitute the fraction exchangeable with LiCl by much cleaner dilute acetic acid? The released amount into LiCl is quite low, and in non-contaminated samples, As, Be, Cd, Cr, Mn, Mo, Pb and V were below detection limits. There is resorption of Cr, Ni, Pb, Zn reagent blanks at the solid, and Li blanks at the ICP for subsequent determinations get reduced. In case of K, Mn, Ba, and B, a good correlation and for S and P a moderate one, had been achieved between concentrations extracted into LiCl and dilute acetic acid, but not for others.

In order to optimum use of ICP-OES as a multi-element instrument, and to cope with decreasing manpower in the labs, the number of extractions should be minimized. Thus, it was tried to simulate the concentration of K and P by CAL extract, B by Baron-extract, and Mg by CaCl₂-extract, by a simple (BCR-like) sequence of 0,16M acetic acid, with subsequent 0,1M oxalate buffer pH 3. This would widen the scope for acid exchangeable trace elements and S also. K, P and B could be very well simulated, and fitted to equal numeric values by partial correlation analysis with other parameters obtained, but for Ca-exchangeable Mg, this was difficult.