Portland cements (PCs) and cement blends are multiphase materials of different
fineness, and quantitatively analysing their hydration pathways is very chal-
lenging. The dissolution (hydration) of the initial crystalline and amorphous
phases must be determined, as well as the formation of labile (such as ettringite),
reactive (such as portlandite) and amorphous (such as calcium silicate hydrate
gel) components. The microstructural changes with hydration time must also be
mapped out. To address this robustly and accurately, an innovative approach is
being developed based on in situ measurements of pastes without any sample
conditioning. Data are sequentially acquired by Mo K�1 laboratory X-ray
powder diffraction (LXRPD) and microtomography (mCT), where the same
volume is scanned with time to reduce variability. Wide capillaries (2 mm in
diameter) are key to avoid artefacts, e.g. self-desiccation, and to have excellent
particle averaging. This methodology is tested in three cement paste samples: (i)
a commercial PC 52.5 R, (ii) a blend of 80 wt% of this PC and 20 wt% quartz, to
simulate an addition of supplementary cementitious materials, and (iii) a blend
of 80 wt% PC and 20 wt% limestone, to simulate a limestone Portland cement.
LXRPD data are acquired at 3 h and 1, 3, 7 and 28 days, and mCT data are
collected at 12 h and 1, 3, 7 and 28 days. Later age data can also be easily
acquired. In this methodology, the amounts of the crystalline phases are directly
obtained from Rietveld analysis and the amorphous phase contents are obtained
from mass-balance calculations. From the mCT study, and within the attained
spatial resolution, three components (porosity, hydrated products and unhy-
drated cement particles) are determined. The analyses quantitatively demon-
strate the filler effect of quartz and limestone in the hydration of alite and the
calcium aluminate phases. Further hydration details are discussed.
