Individual and combined effects of chloride, sulfate, and magnesium ions on hydrated Portland-cement paste

Abstract

Ground water with a high concentration of magnesium ion is known to cause deterioration to portland cement concretes. A proposed mechanism for this deterioration process published previously involves an approximate 1:1 replacement of Ca ions by Mg ions in the crystalline phases of hydrated cement. The current study was undertaken to determine which ions, among magnesium, chloride, and sulfate, cause deterioration; whether their deleterious action is individual or interdependent; and to relate this mechanism of deterioration to the outlook for a 100-yr service life of concretes used in mass placements at the Waste Isolation Pilot Plant. Loss of Ca ion by cement pastes was found to be strongly related to the concentration of Mg ion in simulated groundwater solutions in which the paste samples were aged. This was true of both saltcontaining and conventional cement pastes. No other ion in the solutions exerted a strong effect on Ca loss. Mg ion did not accumulate in the chemically altered cement paste at the same rate that Ca was lost. Ca-ion loss was more than six times the Mg-ion gain. No crystalline Mg-bearing phases were detected in the deteriorated pastes. Ca ion left first from calcium hydroxide in the pastes, depleting all calcium hydroxide by 60 days. Some calcium silicate hydrate remained even after 90 days in the solutions with the highest concentration of Mg ion, while the paste samples deteriorated noticeably. Softening of the samples occurred without complete destruction of calcium silicate hydrate, and with no apparent formation of magnesium silicate hydrate. The results indicated a mechanism that involves dissolution of Ca phases and transport of Ca ions to the surface of the sample, followed by formation of Mg-bearing phases at this reaction surface rather than directly by substitution within the microstructure of hydrated cement. Given that calcium hydroxide and calcium silicate hydrate are the principal strength-giving phases of hydrated cement, this mechanism indicates the likelihood of significant loss of integrity of a concrete exposed to Mg-bearing ground water at the WIPP. The rate of deterioration ultimately will depend on Mg-ion concentration, the microstructure materials of the concrete exposed to that groundwater, and the availability of brine.

Durability of concrete materials in high-magnesium brine

“…As in our previous work (Poole et al, 1994) measurable strength loss occurred well before the detectable appearance of MSH, and deterioration is attributed to loss of calcium from hydrated cement phases, rather than to the formation of MSH.…”

“…In our previous work (Poole et al, 1994), evidence for destruction of CSH was found in loss of more calcium from specimens than could be accounted for by total destruction of calcium hydroxide, and loss of strength occurred without formation of MSH. in the present study, additional evidence for loss of CSH was the appearance of magnesium silicate hydrate (MSH), a non-cementitious phase that forms at the expense of CSH.…”

“…The most notable of these are salts of magnesium and sulfate (Mather, 1966;Oberste-Padtberg, 1985;Bonen and Cohen, 1992). A summary of the most widely proposed mechanism of deterioration of cements in high-Mg brine is given by Poole et al (1994).…”

“…The purpose was to determine _hether the deleterious action of these ions is individual or interdependent. The results indicated a mechanism that involves dissolution of calcium (Ca) phases and transport of Ca ions to the surface of the sample, followed by formation of magnesium (Mg) phases at this reaction surface, rather than directly by substitution within the microstructure of hydrated cement (Poole et al, 1994). The rate of deterioration by this mechanism will depend on Mg-ion concentration, availability of brine, and microstructure and composition of the concrete exposed to that brine.…”

Variability in properties of Salado Mass Concrete

“…This series of concretes called Salado Mass Concrete (SMC) was refined through batching experiments and thermal and mechanical testing (Wakeley et al, 1994a). SMC component materials were also subjected to extensive geochemical tests (Poole et al, 1994;Wakeley et al, 1994b). The geochemical research showed that strength loss of cementitious solids in brines with strong concentrations of magnesium occurred by loss of calcium from cement hydration products, and not by direct replacement of calcium by magnesium, as has been reported elsewhere (Oberste-Padtberg, 1985;Bonen and Cohen, 1992).…”