Accelerated carbonation is induced in pastes and mortars produced from alkali silicate-activated granulated
blast furnace slag (GBFS)–metakaolin (MK) blends, by exposure to CO2-rich gas atmospheres. Uncarbonated
specimens show compressive strengths of up to 63 MPa after 28 days of curing when GBFS is used as the sole
binder, and this decreases by 40–50% upon complete carbonation. The !nal strength of carbonated samples is
largely independent of the extent of metakaolin incorporation up to 20%. Increasing the metakaolin content
of the binder leads to a reduction in mechanical strength, more rapid carbonation, and an increase in
capillary sorptivity. A higher susceptibility to carbonation is identi!ed when activation is carried out with a
lower solution modulus (SiO2/Na2O ratio) in metakaolin-free samples, but this trend is reversed when
metakaolin is added due to the formation of secondary aluminosilicate phases. High-energy synchrotron X-
ray diffractometry of uncarbonated paste samples shows that the main reaction products in alkali-activated
GBFS/MK blends are C–S–H gels, and aluminosilicates with a zeolitic (gismondine) structure. The main
crystalline carbonation products are calcite in all samples and trona only in samples containing no
metakaolin, with carbonation taking place in the C–S–H gels of all samples, and involving the free Na+
present in the pore solution of the metakaolin-free samples. Samples containing metakaolin do not appear to
have the same availability of Na+ for carbonation, indicating that this is more effectively bound in the
presence of a secondary aluminosilicate gel phase. It is clear that claims of exceptional carbonation resistance
in alkali-activated binders are not universally true, but by developing a fuller mechanistic understanding of
this process, it will certainly be possible to improve performance in this area
|
