Mineralogical investigations of the interaction between iron corrosion products and bentonite from the NF-PRO experiments (Phase 2)

Milodowski, A.E.; Cave, M.R.; Kemp, S.J.; Taylor, H.; Green, K.; Williams, C.L.; Shaw, R.A.; Gowing, C.J.B.; Eatherington, N.D.. 2007 Mineralogical investigations of the interaction between iron corrosion products and bentonite from the NF-PRO experiments (Phase 2). Nottingham, UK, British Geological Survey, 76pp. (CR/07/225N) (Unpublished)

Before downloading, please read NORA policies.
[img] Text

Download (8Mb)
Official URL:


This report describes the findings of a second programme of work (Phase 2) undertaken by the British Geological Survey (BGS) on behalf of Svensk Kärnbränslehantering AB (SKB), to characterise the mineralogical alteration of samples of compacted bentonite from experiments that SKB have co-funded in a study by Serco Assurance (Culham Laboratory, Oxfordshire, United Kingdom) to investigate the interaction of iron and bentonite, within the EU Framework 6 NFPRO Project (Smart et al., 2006). Reacted bentonite residues from four NF-PRO Experiments – NFC1, NFC4, NFC7 and NFC13 were examined by BGS using; X-ray diffraction analysis (XRD); petrographical analysis with backscattered scanning electron microscopy (BSEM) with energy-dispersive X-ray microanalysis (EDXA) techniques, cation exchange capacity (CEC) and exchangeable cation analysis; and sequential chemical extraction. In addition, background chemical analysis of altered and background bentonite were also obtained by X-ray fluorescence spectrometry (XRFS). Bentonite immediately adjacent to corroding steel wires was found to have interacted with Fe released from the corroding metal. This resulted in the formation of narrow haloes of altered bentonite around the corroding steel wires, in which the clay matrix was significantly enriched in Fe. Similar observations were observed in bentonite around corroded iron coupons (observed in experiments NFC4 and NFC7 only), although the alteration zones were not as well developed in comparison to those around corroded steel wires. Detailed petrographical observation found no evidence for the formation discrete iron oxide or iron oxyhydroxide phases within the clay matrix but appeared to show that the clay particles themselves had become enriched in Fe. However, data from sequential chemical extraction suggests that a significant proportion (26 to 68 %) of the iron in the altered bentonite is present as amorphous iron oxide or crystalline iron oxides (15 to 33 % of the total iron). Some of the crystalline iron is present as primary magnetite and ilmenite present from the original MX-80 bentonite but part of this will also probably be secondary magnetite formed as a corrosion product of the steel. Nevertheless, sequential chemical extraction analyses also suggest that a large proportion of the iron (11-38 %) may be present within the silicate/clay mineral lattice. The implication of this would be that there has been significant conversion of the original montmorillonite to an Fe-rich clay mineral within these alteration haloes. Although XRD does not detect very much change in clay mineralogy, and suggests that the smectite in the altered bentonite is dioctahedral, it is likely that the subsampling for XRD analysis was on too coarse a scale to be able to resolve the alteration within these very narrow reaction zones around the corroded wires. The alteration observed around the corroded steel wires in experiments NFC4, NFC7 and NFC13 is more complex than that in NFC1 or earlier experiments studied in Phase 1 (Milodowski et al., 2007) or previously by Smart et al. (2006). The reacted bentonite from these experiments exhibited the formation of a Mg-Fe-rich clay mineral or aluminosilicate alteration product. This was formed within the Fe-enriched alteration halo but appears to have formed relatively early and was subsequently partially overprinted or replaced by more Fe-rich aluminosilicate. EDXA microchemical mapping did suggest some slight Mg enhancement in the reacted bentonite from NFC1 but no discrete Mg-rich phase was detected. Whilst Mg may potentially have been derived from the “Allard” reference water used in experiment NFC4, in the case of NFC7 and NFC13 it could only have been derived from the breakdown of the bentonite itself since the porefluid only contained NaCl in these two experiments. XRD observations indicated a slight increase in d002/d003 peak ratio, which could possibly be accounted for by a small amount of substitution of Fe into the octahedral layers of the smectite. This is not supported by exchangeable cation analyses, which show very little exchangeable Fe to be present within the altered bentonite. The cation exchange capacity (CEC) and exchangeable cation chemistry of the bentonite show very little difference in properties between reacted and background bentonite. However, it is also possible that the subsampling for exchangeable cation analysis was also on too coarse a scale to be able to resolve such changes within the fine alteration haloes. Fe released from the corroding steel was also observed to displace Ca2+ from the interlayer cation sites in the montmorillonite component. This was manifested by the marked concentration of Ca at the interface with the corroding metal and along the leading edges of ‘fronts’ of Fe diffusing into the bentonite matrix. The displaced Ca was seen to have reprecipitated as aragonite. The petrographical observations show that the bentonite within the alteration zone, that has reacted with and is enriched by Fe, has a tendency to show significantly reduced shrinkage on sample drying than the unaltered bentonite. Conversely, this would suggest that the reacted and altered clay will also have less ability to swell on hydration with water. This behaviour might be consistent with the partial conversion of the montmorillonite to an iron rich dioctahedral smectite such as nontronite. If this is the case, then this may have important implications for the longterm behaviour of bentonite seals around radioactive waste canisters made of iron or steel.

Item Type: Publication - Report (UNSPECIFIED)
Programmes: BGS Programmes > Chemical and Biological Hazards
Funders/Sponsors: NERC, Svensk Karnbranslehantering AB
Additional Information. Not used in RCUK Gateway to Research.: This item has been internally reviewed but not externally peer-reviewed
Additional Keywords: Bentonite, Iron, Oxides
NORA Subject Terms: Earth Sciences
Related URLs:
Date made live: 18 Jun 2009 10:10

Actions (login required)

View Item View Item

Document Downloads

More statistics for this item...