Towards a comprehensive model to estimate redox capacity of clay rocks studied as natural barriers for radioactive waste disposal

The Boom Clay (BC) in Belgium and the Opalinus Clay (OPA) in Switzerland are studied as potential host rocks for radioactive waste disposal in the frame of national programmes. In the assessment of the long-term natural barrier evolution, redox capacity is an important physical-chemical parameter to consider as it may affect the speciation and migration behavior of the released radionuclides from the waste. In this respect, iron plays an important role in the electron transfer and thus may influence the speciation and transport of many redox sensitive radionuclides. The clay minerals, commonly present in these rocks, contain iron in the octahedral and tetrahedral sheets of their structure (e.g. illite, smectite, mixed layer illite-smectite (I-S) and chlorite). Despite their relatively high abundance in the sedimentary rocks, the electrochemical activity of iron within clay minerals may vary from as high as 100% in the case of pure smectite, through 41% in the mixed 70/30 layer I-S, 10% in illite and only 2% in chlorite. The electrochemical activities of non-clay iron-bearing minerals, is low in the case of pyrite (from 2 to 12%) and zero in the case of siderite.

In order to estimate the redox capacities of BC and OPA, Fe distribution, redox state and electrochemical activity were determined by a combination of quantitative XRD analysis (QXRD), XRF, phenanthroline, ⁵⁷Fe Mössbauer spectroscopy and mediated electrochemical oxidation (MEO) and mediated electrochemical reduction (MER) measurements. In the bulk OPA clay samples, the Fe²⁺ states dominate over Fe³⁺ states, whereas Fe³⁺ states dominate over Fe²⁺ states in the studied bulk BC samples. In both cases, illite, I-S and smectite are the main Fe³⁺ carriers, while chlorite, pyrite and siderite are the main Fe²⁺ carriers. Fe distribution and the valence state are therefore controlled by the quantitative mineralogical composition in both rocks. The results of the MER and MEO measurements indicate that total electron transfer capacity (ETC) varies in the range between 45 and 64 µmol e^-/g bulk OPA and between 160 and 181 µmol e^-/g bulk BC. In most of the studied samples, the electron accepting capacity (EAC) is higher than the electron donating capacity (EDC). The measured EACs positively correlate with increasing Fe³⁺ contents, smectite content and cation exchange capacities. The correlation between measured EDCs and Fe²⁺ contents is less satisfactory, most likely due to overall low Fe²⁺ electrochemical activities reported for pure pyrite, chlorite and siderite. The quantitative mineralogical composition and available electrochemical data of the pure minerals were used to calculate the theoretical ETCs, EDCs and EACs of the studied OPA and BC samples. Despite relatively large uncertainties in the electrochemical activities in the EDC field, the correlation between theoretical and experimental EACs and ETCs is relatively good in both studied cases (R² = 0.83 and 0.76, respectively for EAC and ETC in the case of OPA and R²=0.74 and 0.84 in the case of BC). The proposed mineral electrochemical model should be tested on larger data sets to verify its general applicability to other clay formations.