[en] A scientific literature survey was compiled with the specific objective to find information for smectite mobilization and/or retention in natural clay formations caused by contact with water with low ionic concentrations such as can be expected during and after an ice age. Evidence was sought if smectite particles are lost from the clay to the water and if accessory minerals that remain could form a growing filter slowing down or stopping further loss of smectite. Bentonites are present in geological layers for hundreds of millions of years. There is limited exchange with surrounding layers, eg K transported into the bentonite layer from surrounding shale layers leading to the increased illite % in smectite-illite of the bentonite. Another process is silicification of surrounding layers leading to lowered permeability of surrounding rocks. Geological literature data on historical bentonites do not consider colloid formation in low ionic strength water as relevant mechanism for smectite mobilization. However there are no studied cases where this could be a relevant mechanism (as proposed by colloid release scenario). Soil researchers have studied the mechanism of colloid release in laboratory experiments and have found that there has to be an abrupt change in infiltrating water quality leading to 'osmotic explosion'. Clogging the pores in the lower part of the soil column has followed, leading to dramatic decrease of hydraulic conductivity in vertical profile and increased surface runoff. So, although limited, there are literature evidences of clay colloids release from bentonites/smectites caused by low-ionic circumneutral water. The geological settings to look for natural analogue studies include (1) Bentonite/smectite similar to what is used in repository. (2) Water similar to the composition of glacial meltwater. (3) Scenario similar to what is proposed in the bentonite erosion project. The problem related to the study of historical bentonite profiles is the geological changes that have occurred in bentonites over geological time scales. The bentonite erosion project has to provide the scenarios for approximately 100,000 years, but the geological deposits have been changed and subjected to various conditions, many of those uncertain, during tens and hundreds of millions of years. The field studies of weathering of the historical K-bentonite deposits could provide interesting and new information, but the relevance related to the bentonite erosion project is questionable