Characterization of the spatial distribution of iron- and manganese oxides in the Sterksel Formation at the Maalbeek quarry, The Netherlands
MetadataShow full item record
Iron and manganese oxides are common components in subsurface environments. They influence the groundwater flow, sediment chemistry and contaminant transport. The oxides can form coatings on the surface of sand particles and thereby provide a reactive interface between mineral grains and groundwater. Concentrations of iron and manganese oxides can be very high in sediments. Because of their adsorption capacity, naturally occurring iron- and manganese-bearing phases are known to strongly influence the transfer of inorganic species and ionizable organic compounds. This project has focused on the spatial distribution and concentration of iron oxides and total iron in the Sterksel Formation. This formation was deposited by the Rhine during the colder intervals in the middle Pleistocene. The studied site is the Maalbeek quarry in Limburg. At this site, the formation consists mostly of coarse sand and pebbles, and has a highly heterogeneous distribution of iron and manganese. This study is aimed to characterize the spatial distribution of iron and manganese within the sediment at this quarry. Three different methods have been used to quantify the concentration of iron and other components. The Citrate- Bicarbonate Dithionite method was used to extract the reactive iron from the sediment. Ingestion with aqua regia was used to extract acid soluble iron, manganese and other metals from the sediment. Measurements with XRF were done to get a view on the total composition of the sediment. The outer layer of coating of the grains was mainly measured. In general, enrichment in oxides is in the lower parts of the formation. Manganese oxides are very locally enriched in thin layers and are in a range of 5.42 ppm to 589 ppm manganese. Reactive iron concentrations were measured in the range from 139 to 3000 ppm, and total iron concentrations range from 515 to 515 to 71000 ppm. It is argued that the actual reactive iron concentration should in many cases be higher than measured here. Possible reasons for the low concentrations are shortage of reagents, incomplete dissolution due to crystallinity of oxides and degradation of dithionite. The iron measured by XRF is in the range of 1335 to 12406 ppm. This is a higher value than the aqua regia measured iron because mainly the coating was measured and not the entire sand grains. An overestimation might occur here because the lighter elements are underestimated relative to the heavier elements. The total iron distribution has a lognormal distribution and the reactive and XRF-measured iron has a bimodal distribution. There is a weak correlation between iron and manganese, a reason could be that manganese has a very local enrichment compared to iron. There is a strong iron-titanium correlation, suggesting presence of ilmenite in the sediment. Iron-silicon measured by XRF has a negative correlation; this could be because the iron oxides form coatings around the grains. The distinguish between reactive and non-reactive iron was estimated with theFe2O3/Al2O3 ratio and by comparing the iron measured by different methods. This results in an approximate ‘background’ value of iron oxide of 0.26 times aluminium oxide concentration. The variograms that were created result in a horizontal correlation range of maximum 5 meters, and the vertical correlation range was maximum 1meter. Because the iron enrichment has a layered structure, it was expected that the vertical correlation range would be lower than the horizontal range. The sampling grid is not dense enough to get a representative view on the spatial distribution, because the nugget of the variograms was very high. It is recommended to make the sampling locations much smaller, with minimum sampling distances of 5 cm in the vertical direction and 10 cm in the horizontal direction. The results of the smaller sampling locations could be extrapolated to the whole area.