Alkali–carbonate reaction

The alkali–carbonate reaction is an alteration process first suspected in the 1950s in Canada for the degradation of concrete containing dolomite aggregates.^[1][2]

Alkali from the cement might react with the dolomite crystals present in the aggregate inducing the production of brucite, (MgOH)₂, and calcite (CaCO₃). This mechanism was tentatively proposed by Swenson and Gillott (1964)^[3] and may be written as follows:

${\displaystyle {\ce {CaMg(CO3)2 + 2 NaOH -> Mg(OH)2 + CaCO3 + Na2CO3}}}$

Brucite (Mg(OH)₂), could be responsible for the volumetric expansion after de-dolomitisation of the aggregates, due to absorption of water.

The alkali–carbonate reaction is also catalyzed by the soluble NaOH produced by the reaction of Na
₂CO
₃ with Ca(OH)
₂ (portlandite) present in the hardened cement paste (HCP), therefore perpetuating the reaction indefinitely as observed by Fournier and Bérubé (2000) and Bérubé et al. (2005).^[4][5]

${\displaystyle {\ce {Na2CO3 + Ca(OH)2 -> CaCO3 + 2 NaOH}}}$

The sum of the two above mentioned reactions leading to the ultimate production of brucite and calcium carbonate can be written as follows:

${\displaystyle {\ce {CaMg(CO3)2 + Ca(OH)2 -> Mg(OH)2 + 2 CaCO3}}}$

The alkali-carbonate reaction is much less understood than the alkali-silica reaction. Both reactions share in common the continuous regeneration of the sodium hydroxide (NaOH) after the reaction of soluble sodium carbonate or sodium silicate with calcium hydroxide, Ca(OH)
₂. However, impure dolomitic aggregates also often contain clay impurities, and small amounts of pyrite (FeS
₂) and organic matter. The alkali-carbonate reaction could therefore also simply hide an alkali-silica or an alkali-silicate reaction. Anyway a chemical coupling between ACR and ASR cannot be ruled out.