In situ measurements of heavy metal distributions in soil columns during miscible displacement experiments using portable X‐ray fluorescence

Miscible displacement experiments traditionally rely on measured effluent concentrations of a given chemical with time to characterize its transport through soils. Time-dependent distributions of heavy metals within a soil column are difficult to obtain and are typically ignored. We developed a methodology for measuring time-dependent distributions of zinc (Zn) and nickel (Ni) within soil using a Kapton film (KF) column and portable X-ray fluorescence (pXRF) device. Matrix-matched calibrations were developed for each matrix/solute combination based on linear regressions between known and pXRF-measured Zn and Ni concentrations within the KF column. We assessed the accuracy of pXRF measurements using mass balance calculations; our matrix-matched calibration significantly reduced cumulative mass balance errors for three of the four experimental datasets. We compared pXRF-measured Zn or Ni concentrations at various column lengths to those predicted by a single-site, nonlinear kinetic model. The model accurately predicted the timing of peak Zn or Ni concentrations at each column length in reference sand and Wolfpen soil; however, overall model performance was element- and matrix-dependent. Our results demonstrate that use of a KF column and pXRF device enables the acquisition of time-dependent distributions of Zn or Ni distributions in soil columns. We expect that this approach will be appropriate for other heavy metals, particularly those with higher energy fluorescent X-rays (atomic number > 30) where less X-ray attenuation by the KF column and soil matrix is expected. Accurate descriptions of heavy metal distributions with column length and time will provide an additional metric to validate reactive transport models.