Assessment of Long-Term Performance and Chromate Reduction Mechanisms in a Field Scale Permeable Reactive Barrier BETTINA FLURY, † JAKOB FROMMER, ‡ URS EGGENBERGER, † URS M ¨ ADER, †, * MAARTEN NACHTEGAAL, § AND RUBEN KRETZSCHMAR ‡ Rock-Water Interaction Group, Institute of Geological Sciences, University of Bern, Switzerland, Institute of Biogeochemistry and Pollutant Dynamics, Department of Environmental Sciences, ETH Zu ¨rich, Switzerland, and Paul Scherrer Institute, General Energy Research Department, Laboratory for Energy and Materials Cycles, Villigen, Switzerland Received December 11, 2008. Revised manuscript received May 13, 2009. Accepted May 14, 2009. An innovative full-scale implementation of a permeable reactive barrier, consisting of a double-row of cylinders filled with zerovalent iron shavings, for chromate remediation was monitored over four years. Solid samples were analyzed to elucidate (i) the relevant corrosion mechanisms and products, (ii) the pathways of chromate reduction and immobilization, and (iii) the long-term performance of the barrier situated in a hydrological and geochemical complex groundwater regime. Sampling and analysis of groundwater and reactive material revealed an oxidative iron corrosion zone evolving in the inflow and a zone of anaerobic iron corrosion in the center and outflow of the barrier. Chromate reduction was mainly confined to the inflow region. The formation and thickness of corrosion rinds depended on sampling time, position, and depth, as well as on the size, shape, and graphite content. In the inflow, the corrosion rinds mostly consisted of goethite and ferrihydrite. X-ray absorption fine structure spectroscopy revealed two distinct Cr III species, most likely resulting from homogeneous and heterogeneous redox reaction pathways, respectively. The longevity and long-term effectiveness of the PRB appears to be primarily limited by reduced corrosion rates of the ZVI-shavings because of the thick layers of Fe-hydroxides. Introduction Zerovalent iron (ZVI) is a highly reactive material that can be used for the remediation of contaminated groundwater. It is often applied in a permeable reactive barrier (PRB) system, in which contaminated subsurface water is forced to pass through a zone containing granular iron. Pilot and full-scale installations have demonstrated the success of ZVI- PRBs in removing organic and inorganic contaminants (1-5). The corrosion of ZVI results in the formation of redox-reactive interfaces and solution species, leading to reductive degra- dation of organic pollutants or immobilization of trace elements (e.g., Cr VI ,U VI ) by reduction. Despite its initial success some important challenges remain, including as- sessing the long-term performance (6), interferences with the hydrology (1, 3), and the need for more material- and cost-efficient strategies (7-10). Predicting the long-term performance of PRBs and developing alternative approaches require an improved understanding of the geochemical processes involved in the contaminant treatment and of the influence of the ground- water chemistry on the PRB material. Little is known on whether processes identified in column studies apply to field scale installations (3). A common application of ZVI-PRBs is the removal of mobile and toxic hexavalent chromium (Cr VI , chromate) from groundwater by reduction to less soluble and less toxic Cr III . Different pathways have been proposed for the reduction of chromate by ZVI, including a direct electron transfer from the metal surface (11), homogeneous solution phase reactions with Fe II resulting from the metal corrosion, and heteroge- neous reactions at the interface of neo-formed Fe II -containing phases (4). One way to explore reaction pathways in complex field systems is to first asses the speciation of the reduced Cr III and its association with other elements. This information is used to infer the relevant molecular mechanisms respon- sible for the formation of the Cr III species observed (4). The required solid-state speciation data can be assessed by microfocused X-ray absorption fine-structure (XAFS) spec- troscopy in a unique way at the micrometer scale. In this study, we investigated reactive ZVI material sampled one to four years after installation of a PRB system designed to remediate a Cr VI contaminated aquifer in Switzerland. The innovative design minimizes the amount of reactive material needed and partly relies on a dispersive Fe II -plume extending the reductive capacity beyond the barrier itself. Furthermore, the installation allows a collection of solid samples and a monitoring of the groundwater chemistry at different locations within the barrier. In an accompanying study, the functionality of this installation has been shown (12). The objectives of the present study were to (i) assess the state of the reactive material in this innovative PRB system with respect to the known problems of surface passivation and pore-space reduction and (ii) to examine the molecular- level mechanisms of the occurring corrosion and redox processes on the micrometer scale. Information from the decameter scale (groundwater monitoring) to the molecular scale (XAFS) was integrated to cope with the complexity of this PRB system. Materials and Methods Study Site. The remediation site is located in Willisau, Switzerland, where copper-chromate solutions have been used in an industrial wood preservation process. From 1958 to 1987, hexavalent chromium seeped into the subsoil and accumulated in the unsaturated zone (13). The groundwater table considerably varies between 12-16 m depth (transition zone), and the aquifer thickness is 25 m. Groundwater velocity was established to be 5-6 m/day, which is rather high compared to most PRB settings (12). Contaminants ac- cumulated in the unsaturated zone are leached into the groundwater and transported in flow direction of the groundwater with little vertical dispersion over several decameters. The groundwater chemistry at the site is dominated by calcium and bicarbonate with an electric * Corresponding author address: Rock Water Interaction Group, Institute of Geological Sciences, University of Bern, Baltzerstrasse 3, 3012 Bern, Switzerland; phone: ++41 (0)31 631 45 63; fax: ++41 (0)31 631 48 43; e-mail:urs.maeder@geo.unibe.ch. † University of Bern. ‡ ETH Zu ¨ rich. § Paul Scherrer Institute. Environ. Sci. Technol. 2009, 43, 6786–6792 6786 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 17, 2009 10.1021/es803526g CCC: $40.75  2009 American Chemical Society Published on Web 07/24/2009