Cold Water Extraction as a method to determine the free alkali content of cementitious binders

The alkali content (Na₂O and K₂O) of cement is of major importance when evaluating the risk of developing Alkali-Silica Reaction (ASR) in concrete. During hydration, alkali metals (Na and K) can either be bound to unreacted particles, bound to hydrates, or free in the pore solution. The latter category is the most critical one with respect to ASR, but its quantification remains more complex than the total content.

Blended cements containing e.g. fly ash or calcined clay are known to have less free alkalis than ordinary Portland cement, despite a larger total amount. Thus, using the total content to classify cements or calculate the alkali loading of a concrete mix may result in irrelevant figures regarding ASR. A procedure enabling to quantify free alkalis, which can be performed in most cement laboratories, would therefore be a useful tool.

This paper introduces the possibility to use Cold Water Extraction (CWE) as a method to determine the free alkali content of a cementitious binder. CWE is a technique aiming at extracting the pore solution of hardened samples, which can be performed with standard laboratory equipment. Three different cement types were investigated: a CEM I, a CEM II/B-M (35 % clinker replacement) and a prototype CEM II/C-M (50 % clinker replacement), all manufactured with the same clinker. The more traditional Pore Water Extraction method was also performed to benchmark CWE results. The free alkali content was determined after performing extraction on paste samples cured for 28 days at 20°C. Both methods indicated the same trends, namely lower relative and absolute free contents for blended cements compared to CEM I. The results were also in agreement with expansions obtained with the Danish accelerated mortar bar test (TI-B 51), which is currently prescribed in the Danish standards when total alkalis requirements cannot be met.