Chlorinated hydrocarbons present remediation challenges, particularly when dissolved plumes are located at depths greater than 25 feet or when plumes are present as saturated zone dense nonaqueous phase liquid (DNAPL) at any depth. The injection of nanoscale bimetallic particles is a promising new technology that can be used to rapidly remediate source areas. The technology also can provide a unique solution in cases where a dissolved phase chlorinated solvent groundwater plume is located at a depth greater than 25 feet. In this case, a permeable reactive barrier can be created by the in situ injection of nanoscale iron particles.
The main objectives of this project were to:
Results generated led to the addition of two subsidiary objectives:
In this project, three types of bench studies were conducted:
Testing focused primarily on four different commercially produced nanoscale irons prepared with widely differing manufacturing processes (PolyMetallix, RNIP, Milled–manufactured through attrition milling by OnMaterials, and Z-Loy). Key results of interest to practitioners follow.
Though acid washing was effective at improving reaction rates, it does not appear to be absolutely necessary before palladization of the nanoscale zero-valent iron (ZVI). The acid wash is thought to remove corrosion products, maximizing the exposure of ZVI surfaces for either TCE reduction or palladization. Acid washing can significantly reduce the amount of palladium hydrogenation catalyst required to achieve a substantial rate increase.
Removing the surface coating of iron oxides with an acid wash will remove material from the ZVI particle resulting in smaller particles. While small particles have a higher specific surface area than large particles, the absolute surface area of a fixed number of particles will decrease through this size reduction. Consistent with the literature, the project team found that acid washing can improve the TCE degradation rate of Milled nanoscale iron. This suggests that there is a trade off in short-term reactivity of the ZVI between the higher reactivity of a freshly exposed ZVI surface over the oxide-coated ZVI surface (which is also reactive) and the lower overall surface area resulting from an acid wash of the ZVI. Factors such as the size of the remaining iron core, reactivity of the ZVI, and activity of the oxide coating may all play important roles in determining whether an acid wash is advantageous over the full lifetime of a ZVI.
Treatment efficiencies for all four irons tested decreased dramatically over several hundred pore volumes of water flow in column tests (conducted over 18 months or less), and all of the irons showed reduced kinetic rate constants in three years or less of storage. Thus, nanoscale irons may not be well suited for barrier applications that require a decade or more of treatment following one injection.
Exposure to waters with high concentrations of carbonate causes rapid decreases in reactivity at least for palladized irons in 150 pore volumes. Even in very low ionic strength deionized water, substantial reactivity was lost in all tested colloids in 100 pore volumes. Exposure to waters containing high concentrations of sulfate also appears to decrease reactivity of nanoscale colloids, but more slowly and less completely than was seen either in deionized or high carbonate waters.
Evidence from theory, column tests, and DNAPL treatment experiments all suggests that the presence of chloride (either natural or from chlorinated volatile organic compound degradation) can provide some at least temporary protection of amorphous irons from passivation. This effect was less often observed with crystalline irons.
All of the tested nanoscale irons generated visible evolution of a gas, believed to be hydrogen, during acid washing, and several did so before acid washing. This suggests that safety precautions are necessary during storing, shipment, and application.
Results for at least one tested iron suggest that substantial (50%) treatment of DNAPL phase chlorinated volatile organic compounds was achieved in less than one month. This occurred without the injection of an oil phase, which others have argued is required for effective treatment of DNAPL with nanoscale ZVI.
Since the initial rate of reaction of a given nanoscale iron is controlled by at least six separate factors, the a priori prediction of reaction rate in support of feasibility or design studies is very difficult to impossible at this time. However, the range of rates achievable is now well documented in several sources. Site-specific treatability testing is required to determine if the treatment rate required to meet the design objectives can be achieved with a given combination of nanoscale iron and aqueous geochemistry. General rules of thumb, however, can be elucidated. Initial rates will generally be most rapid for irons with high surface area, amorphous character, and recent manufacturing date.
The initial rates of reaction of different batches of iron provided by the same manufacturer under the same brand name can differ dramatically. This likely reflects that the manufacturing technology continues to evolve. Therefore, the project team recommends that each individual batch of iron undergo kinetics testing as a quality control step before application in the field. Additives to a mixture containing nanoscale iron can cause unexpected dramatic decreases in performance and thus must be carefully evaluated before implementation.
Design of field-scale systems should carefully consider whether first-order or zero-order kinetics are most likely under a particular set of circumstances. Pseudo first-order kinetics were found to fit the as-received ZVI reaction results slightly better than a zero-order rate law. The majority of ZVI literature has found TCE degradation to be best described by a first-order model. In this model, TCE degrades at a faster rate at higher concentrations than at low concentrations; the rate of degradation is proportional to the TCE concentration. At high concentrations of TCE, above about 6 mg/L, a zero-order model has been reported to best fit TCE degradation with some macro-scale irons. A zero-order model results in a constant rate of TCE degradation regardless of concentration. In these studies of acid washed and/or palladized nanoscale irons, zero-order kinetics were in some cases observed above 2 mg/L TCE.