The partial replacement of Portland cement by aluminosilicate-rich supplementary cementitious materials (SCMs) represents a valuable approach to lower the CO2 footprint of concrete. However, SCMs alter the composition and structure of the principal binding component, calcium-alumino-silicate-hydrate (C-(A)-S-H), a less-crystalline aluminum-substituted calcium silicate hydrate. The uptake and structural impact of alkalis and aluminum are investigated for synthesized C-(A)-S-H phases with different Ca/Si and Al/Si ratios and alkali (Na+) contents, using solid-state Na-23, Al-27, and Si-29 NMR as principal analytical tools. Si-29{H-1} CP/MAS NMR data propose that the -87 ppm resonance originates from a second type of Q(2) pairing sites most likely in the middle of octameric or longer silicate units. This assignment along with constraints induced by the silicate dreierketten structure results in reliable simulations of the Si-29 NMR spectra. A quantitative 1:1 relation between the Al/Si ratios from Si-29 NMR and the content of fourfold-coordinated Al from Al-27 NMR is observed, which reveals that the Si-29 Q(p)(2)(1Al) resonance is only associated with Al(IV) in C-(A)-S-H. Combined analysis of the Al-27 and Si-29 NMR spectra can account for sixfold-coordinated AI being present in the bridging sites of the silicate chains as proposed very recently. The sodium uptake is found to decrease with increasing Ca/Si ratio. For C-(A)-S-H phases with high Ca/Si ratios (Ca/Si greater than or similar to 1.0), sodium promotes the incorporation of Al(IV) in the silicate chains but not at low Ca/Si ratios (Ca/Si less than or similar to 1.0). The results indicate that the mutual uptake of alkali and AI ions is governed by more than one mechanism and depends on the composition/structure of the C-(A)-S-H phase.
