Mesoscale Texture of Cement Hydrates
Strength and other mechanical properties of cement and concrete rely upon the formation of calcium–silicate–hydrates (C–S–H) during cement hydration. Departing from traditional models mostly focused on length scales above the micrometer, recent research addressed the molecular structure of C–S–H. However, small-angle neutron scattering, electron-microscopy imaging, and nanoindentation experiments suggest that its mesoscale organization, extending over hundreds of nanometers, may be more important.
Upon dissolution of cement powder in water, calcium–silicate–hydrates (C–S–H) precipitate and assemble into a cohesive gel that fills the pore space in the cement paste over hundreds of nanometers and binds the different components of concrete together. The mechanics and microstructure are key to concrete performance and durability. Calcium–silicate–hydrate (C–S–H) nanoscale gels are the main binding agent in cement and concrete, crucial for the strength and the long-term evolution of the material. Even more than the molecular structure, the C–S–H mesoscale amorphous texture over hundreds of nanometers plays a crucial role for material properties. We use a statistical physics framework for aggregating nanoparticles and numerical simulations to obtain a first, to our knowledge, quantitative model for such a complex material.
To mimic the continuous densification of the C–S–H gel, we combined molecular dynamics (MD) simulations of the interacting nanoscale units with a Grand Canonical Monte Carlo (GCMC) scheme, in which the free energy gain associated to the production of the cement hydrates and the attractive interactions between the nanoscale units drive the aggregation and increase of the density. As a result of the particle aggregation and the progressive densification of the gel, the population of nanopores and mesopores depends on the overall packing fraction: the lower the ? value, the more abundant the mesopores, whereas the nanopores are more abundant as ? increases. The PSDs shown here feature all important characteristics obtained in different experiments ranging from N2 adsorption/desorption to NMR. The statistical analysis of local packing fraction ?local (Materials and Methods) reveals sizeable changes with the total packing fraction. Not only the solid material produced on aggregation and densification is structurally heterogeneous over these length scales, but also, the degree of heterogeneity varies at different stages of the setting process and for different w/c ratios of the cement paste.
The packing fraction distribution obtained in experiments (Materials and Methods and SI Appendix) is compared with the distribution of ?local obtained in the model as shown in Fig. 3C. In these experiments, Portlandite and all other minor phases have been carefully excluded, and the data refer only to the C–S–H phase.
In our results, a complex, extended pore network, as well as a continuum distribution of local densities, compose a physical picture of C–S–H beyond the classic colloidal model based on two distinct local densities.
Mechanical Properties. The stiffness of the interaction between two particles of size ? is defined as the curvature of the generalized Lennard–Jones potential, and at the local equilibrium distance , it equals . Assuming that the cohesive strength between particles is the same as for the calcium–silicate layers within a C–S–H grain, we express the interparticle stiffness as , where is the Young modulus of solid C–S–H 68.4 GPa from atomistic simulations of few nanometers of C–S–H grains (55), and A=?r2 m is the cross-section of the applied force. All of this leads to a unit pressure e=?3 =0.51 GPa. The C–S–H configurations were relaxed to zero stress before computing indentation modulus M and hardness H.
Experiments. Three cement paste specimens were investigated. The first specimen (S1) was synthesized using oil-well cement, class G and quartz hydrated at w/c = 0.43 for over 1 y. Specimens S2 and S3 were made with ordinary Portland cement type I (OPC CEM-I) and hydrated at room temperature with w/c = 0.35. The hydration reaction was stopped at 24 (S2) and 17 (S3) h with the solvent exchange method (59). The S3 sample also contained C–S–H seeds. Force-controlled nanoindentation tests were performed to a maximum indentation depth of 300 nm, with a linear loading and unloading rate of 12 mN/min and a dwell time of 5 s.