Understanding cation exchange capacity measurement for bentonite-cement reactions

Bentonite barriers in repositories for high-level radioactive waste disposal are expected to interact with Portland cement, which is commonly used during the underground construction of repositories. These reactions occur at a high pH (>12), leading to the formation of XRD-amorphous calcium silicate hydrate (C-S-H) phase and its crystalline analogue, tobermorite, as well as Ca‑carbonate, Ca- and Mg-silicate phases, including zeolites. The cation exchange capacity (CEC) is a key parameter frequently used to evaluate bentonite's ability to swell and adsorb cationic species released from nuclear waste over its lifespan. In addition to being influenced by the residual smectite content, newly formed phases generated during bentonite-cement reactions may also contribute to the bulk CEC. Moreover, the extremely high pH of the reacted material when in contact with water raises questions about the reliability of CEC measurements under such conditions. In this study, the CEC of both reacted and unreacted bentonite-cement mixtures was tested using hexamminecobalt(III) ([Co(NH 3 ) 6 ] 3+ ) index cation and a spectrophotometric method, under natural pH of the suspension. The extinction of the [Co(NH 3 ) 6 ] 3+ solution was almost constant even at high pH, up to 12.5. Fresh, unwashed bentonite-cement mixtures showed higher-than-predicted CEC values due to the strong contribution of smectite edge charges which develop at high pH. Extended washing of the material via dialysis lowered the pH and reduced the excess Ca 2+ cations content. After dialysis (up to 4 days), the CEC of bentonite-cement mixtures increased and exceeded the CEC predicted from the smectite content alone. In long-term bentonite-cement reactions (3–36 months, at 80 °C), an initial equilibration occurred rapidly, within days, followed by the slow crystallization of tobermorite at the expense of smectite. However, the large effective dimension of the [Co(NH 3 ) 6 ] 3+ cation prevents it from entering the tobermorite structure (as it does to zeolite structures), which results in insignificant contribution of tobermorite to the bulk CEC. All CEC values measured with the [Co(NH 3 ) 6 ] 3+ index cationion reacted bentonite-cement samples were found to be a combination of the exchange by residual smectite and the contribution from new-formed C-S-H phase. The C-S-H phase develops its own surface CEC, which is highly variable and increases as its structural Ca/Si ratio decreases. Soda-soluble silica present in the raw bentonite lowers the Ca/Si ratio in the newly formed C-S-H phase, thereby increasing the CEC of the bulk sample. Also, extensive leaching of Ca 2+ cations in an open system during cement maturation decreases the Ca/Si ratio in the newly formed C-S-H phase, enhancing its CEC contribution. In conclusion, CEC measurements on reacted bentonite-cement mixtures must be interpreted with caution as they likely reflect not only the residual smectite content but also the contributions from newly formed C-S-H phases and other reaction products.