Technical Report NTB 17-13

The development of a thermodynamic sorption data base for montmorillonite and the application to bentonite

Over the past two decades measurements on the uptake of radionuclides on montmorillonite in the form of sorption edges and isotherms have been performed. The metals investigated include Co(II), Ni(II), Cd(II), Zn(II), Fe(II), Eu(III), Am(III), Sn(IV), Th(IV), Np(V), Pa(V) and U(VI), and all of these data are presented here. The two site protolysis non electrostatic surface complexation cation exchange (2SPNE SC/CE) sorption model has been applied to successfully model this extensive experimental sorption data set. Previously published results were not always modelled with the same set of thermodynamic data, particularly the hydrolysis constants. In this work the all of the sorption measurements have been consistently (re)modelled with the same selected hydrolysis constant data set and the modelled curves presented.

The conclusions from sorption competition studies are presented and discussed: briefly, chemically similar metals (valence state, hydrolysis behaviour) sorb on the same set of strong sites and exhibit sorption competition, chemically dissimilar metals sorb on different strong site sets and do not compete with one another. This information was used to construct linear free energy relationships (LFERs) between surface complexation constants and the appropriate corresponding aqueous hydrolysis constants separately for divalent transition metals, trivalent lanthanides/actinides and tetravalent actinides. LFERs provide a means for estimating surface site binding constants for metals where the data are either very poor or non-existent and thus allow sorption values to be calculated. How well such a procedure may (or may not) work was illustrated in an exercise where surface complexation constants for specific metals were taken from the appropriate LFERs and used to calculate sorption edges/isotherms which were then compared with the measured sorption values. A general methodology for using LFERs to calculate the sorption of metals for which no sorption data exist is suggested and applied to predict sorption edges/isotherms for Mn(II), Cu(II), Pu(III), U(IV), Np(IV) and Pu(IV).

In a final exercise to test the capabilities of the 2SPNE SC/CE sorption model, and the associated model parameters, to quantitatively describe the uptake of radionuclides in complex MX-80 bentonite/porewater systems, blind model predictions were made of sorption isotherms which were then compared with the measured values. The main assumption was that the sorption on MX-80 bentonite is controlled by the montmorillonite content.

The main conclusion drawn from this work is that the tables of surface complexation constants and cation exchange selectivity coefficients given and used in conjunction with the LFERs and the 2SPNE SC/CE sorption model, provide a powerful means of calculating the sorption of many radionuclides in complex porewater/bentonite systems. Together they constitute a Thermodynamic Sorption Data Base (TSDB) for montmorillonite/bentonite systems.