GEMAS: prediction of solid-solution partitioning coefficients (Kd) for cationic metals in soils using mid-infrared...

Area of Science:

• Environmental Science
• Geochemistry
• Analytical Chemistry

Background:

• Soil metal cation distribution coefficients (Kd) are crucial for assessing environmental risk.
• Predictive models for Kd can enhance soil monitoring and management strategies.

Purpose of the Study:

• To develop and apply partial least squares regression (PLSR) models for predicting metal cation distribution coefficients (Kd) in European soils.
• To assess the mobility and availability of selected metal cations using spectroscopic data.

Main Methods:

• Mid-infrared (MIR) diffuse reflectance Fourier-transformed (DRIFT) spectroscopy was used to analyze soil samples.
• PLSR models were developed using MIR-DRIFT spectra and soil pH to predict Kd values for various metal cations.
• Models were validated and applied to a large dataset of European soils (GEMAS program).

Main Results:

• Optimized PLSR models accurately predicted log Kd+1 for Co, Mn, Ni, Pb, and Zn (R(2) ≥ 0.83).
• Predictions for Ag, Cu, and Sn were less accurate (R(2) < 0.50).
• The models were successfully applied to predict Kd values in over 4000 unknown soils.

Conclusions:

• PLSR models using MIR-DRIFT spectra and soil pH provide a rapid and cost-effective tool for estimating metal cation mobility in European soils.
• The models can aid in assessing the potential availability of metallic cations in agricultural soils.
• Further research is needed to improve model accuracy for certain metals and expand applicability to diverse global soil types.