Factors affecting the reactivity of granular iron in contact with chlorinated solvents

Abstract

This study investigates, at various scales, the factors that affect the reactivity of granular iron (GI) toward chlorinated solvents and link these scale-specific processes with each other. The Kinetic Iron Model (KIM), provides a separate estimate for both the sorption and reaction processes of contaminant degradation, was used to determine the macro-scale kinetic and sorption parameters in all column experiments associated with this work. The intermediate scale corresponding to pore spaces was investigated by image analysis technique. This technique was used to examine the effects of diluting iron with sand on the availability of reactive iron surface. Two morphological parameters were measured in sections: i) grain perimeter, which reflects surface in contact with solution, and ii) total grain area in section. Morphological analysis showed grain areas exposed in section were highest for 100% iron packings and decreased with increasing sand content. However, the estimated iron grain perimeter length for 85% iron-by weight mixture was found to be the similar to that of 100% iron by weight. This study supported the use of 15% sand (by weight) in iron-sand mixture for the optimum performance of a permeable reactive barrier (PRB). Another pore scale issue examined was the effect of GI packing in column experiments. Among the tested packing variations- vertical packing with long axes preferentially along the flow showed higher reaction rates (2-4 times) compared to packings with long axes preferentially perpendicular to flow (horizontal packing) or randomly arranged (regular packing). The pore-scale differences in grain surface availability to solution through image analysis showed that grain surface availability partially accounted for reactivity differences between columns of different packings. It was suggested that micro-scale changes to the iron surfaces accounted for the remaining differences in reactivity. In order to examine the micro-scale changes that occur on the iron surface due to corrosion and to link these changes with macro-scale KIM parameters, long term column experiments were performed under dynamic flow conditions. Micro-scale grain characteristics were made by recovering single grains from sampling port along the length of columns, and examining them through time using Raman spectroscopy and scanning electron microscopy (SEM)/Energy dispersive spectroscopy (EDS). Trichloroethylene (TCE) reduction kinetics showed considerable changes in both TCE sorption and reaction with time. Similarly, spectroscopic studies also indicated profound changes to the iron grain surfaces. It was found that over time, with exposure to TCE and water, iron tended to loose small number of sorption sites associated with highest reactivities (k) whereas large number of less reactive sorption sites increased in number. Raman spectra collected along the column showed the loss of hematite, and transition of intermediate phases to magnetite. Weakening of Raman signals for surface carbon correspond to declining k values and the non-reactive sorption parameter, suggesting that surface carbon serves as non-reactive sorption sites as well as a reactive one. Further the role of carbon present in GI during reductive dechlorination was assessed by comparing 2 iron types of GI in column experiments: Connelly Iron (GI)(~3% C) and Electrolytic Iron (EI) (¡Ü 0.01% C). Kinetic data suggested a shift in rate constant (k) and sorption parameters for both iron types with time. This work demonstrated the implication of carbon during the retardation (Rapp) of TCE i.e high Rapp for GI and low for EI.