Le Chatelier’s concrete conundrum

Concrete, that cold, hard, formless mass, unloved yet ubiquitous, is on the rebound—not only as the centennial-old backbone material of our societies’ need for housing and infrastructure but as a generational challenge for scientists and industry alike to meet our societies’ climate policy goals. The latter is due to the fact that the annual worldwide cement production of four Giga-tons currently accounts for ca. 8% of global CO₂ emissions (1, 2). Although there is an ever-increasing abundance of new concrete propositions emerging to decarbonize cement manufacturing and compositions, there is an urgent need to return to the very foundations of cement science and concrete engineering to enhance the longevity of buildings and infrastructure to be built with both classical and new cement binders. Paramount among the factors affecting concretes’ longevity are volume changes that occur when cement and water react to form—at room temperature—the cement paste which glues sand and aggregates together. Therefore, it is not surprising that volume changes in such a chemically reactive medium have been the focus of scientific inquiry in step with the industrialization of Portland cement concretes; first in Europe in the late 19th century by Le Chatelier (3), and later in the 20th century in America by Powers (4, 5), once concrete had become the material of capital expansion, from grain silos on the Great Lakes, to Ford’s River Rouge plant in Detroit, to Hoover Dam, each a shrine to American achievement (6). Despite more than a century of research, a question that has never been fully resolved is whether the hydration of cement involves expansion or contraction or both? In PNAS, Rosa et al. (7) address this question head on by combining volume changes measured on cement paste flakes of 500 $\mathrm{\mu }$m thickness with poromechanic analysis to solve Le Chatelier’s concrete conundrum.

Le Chatelier’s conundrum is best summarized in his own words that 1) “the hydration of all hydraulic binders is accompanied both by an increase in the apparent volume and a decrease in the absolute volume;” and that 2) “our current state of knowledge falls short to explain the increase in apparent volume as a simple and more general phenomenon” (8). The absolute volume change, often referred to as Le Chatelier contraction or chemical shrinkage, refers to the observation that an open system composed of a hydrating cement sample in a water bath sees its free water volume reduced due to the difference in the average density between the reactants and the products of the hydration reactions (Fig. 1A). In contrast, Le Chatelier’s apparent volume change anticipates modern continuum-based deformation theory and particle-to-particle potential-of-mean-force theory: defined as the change in total volume contained between two surfaces [“la totalité du volume compris entre les surfaces limitant la masse” (3)], it is attributed to the relative position of the different solid parts [“[...] exclusivement determiné par les positions relatives des différentes parties solides” (3)]. The complexity of the two phenomena arises from their simultaneous and interconnected occurrence. For instance, advanced spectroscopy measurements, such as Raman spectroscopy, provide us today with continuous time-space resolved positions of reactants and products (11) (Fig. 1B), and thus a means to link the heterogeneity of the chemical reactions to Le Chatelier’s absolute volume change. On the other hand, the relative volume change can only be assessed experimentally for specific thermal, hydraulic, and deformation or stress boundary conditions [see, for instance, (12, 13)] or indirectly by means of molecular and mesoscale simulations of atom-to-atom (14, 15) and particle-to-particle interactions considering the open and out-of-equilibrium nature of the hydrating system (16, 17). The renewed interest in Le Chatelier’s conundrum regarding apparent volume changes, particularly within the statistical physics community, seems to stem from measurements of interaction forces reported in 2005 between a micrometric flat C-S-H surface and a C-S-H nanocrystal at the top of an Atomic Force Microscope tip immersed in lime solutions of different concentrations (9). While Le Chatelier maintained that the hydration of lime always results in expansion (3), this 2005 study revealed a shift in force from repulsion (linked to expansion) to attraction (linked to contraction) as lime concentration increased (Fig. 1C). These experimental results not only emphasized the dynamic behavior of hydrating matter but also opened the door to studying how forces between particles change with their relative positions and chemical environment, offering valuable insights into the microscopic behavior of cement-based materials (Fig. 1D) (10). That is, amid the convergence of multiscale experiments and simulations, the new experimental results reported in PNAS (7) help resolve a century-old conundrum by offering a benchmark measurement and interpretation of Le Chatelier’s apparent volume increase during cement hydration. This is achieved by means of a laser microscopy-topography technique on samples of a submillimeter size that allows continuous resaturation of pores, canceling a variety of other sources of deformation such as self-desiccation (12, 18). Yet, the story may still not be complete: As the out-of-equilibrium system transitions from hydration to aging, the apparent volume increase driven by repulsive forces is expected to gradually shift toward contraction. In fact, as evidenced by AFM force measurements (Fig. 1C) and predicted by particle simulations of highly packed cement pastes (Fig. 1D), attractive forces between particles define the cohesive nature of the C-S-H gel as it approaches thermodynamic equilibrium, contributing to the dimensional stability and longevity of concrete.

Fig. 1.

Le Chatelier’s Conundrum: (A) Solid volume fractions in function of the hydration degree as predicted by the Powers model (5) Notable is the Le Chatelier contraction (or chemical shrinkage) during hydration. (B) Raman spectroscopy map of a cement paste after 72 h of hydration (Courtesy of A. Masic), showing the main solid phases accessible with the vibration technique (C-S-H $=$ Calcium-Silica-Hydrates; CH $=$ Portlandite). (C) AFM-measured repulsion and attraction forces for different lime concentration (adapted from ref. 9, Copyright 2005 American Chemical Society). (D) Mesostructure of a cement paste with packing densities ($\eta$) obtained from particle-to-particle force simulations (adapted from ref. 10).

This convergence of experiments and simulations has never been more important than now in the context of infrastructure development in the United States. As the country reinvests heavily in its built infrastructure, leveraging science-enabled engineering with industrial developments becomes essential for designing, testing, and improving systems that are safe, efficient, and resilient. To cite a few applications, it is expected that the findings of Rosa et al. (7) will be critical for mitigating the risk of leakage of gas wells often caused by early-age volume changes (19, 20); for increasing the longevity of pavements, bridges, and power plants, specifically those built with high cement binder content; and —as benchmark—for the cement and concrete industry—in the urgent transition to sustainable binder phases (21). For this transition, the words of Henri Louis Chatelier (1850–1936) have lost none of their relevance: “In the absence of such collaboration [between sciences and industry], science, deprived of all effective control, is lost in vain imaginations; and industry, deprived of a precise direction, becomes immobilized in empirical trial and error with no way out” (22).

It is expected that the findings of Rosa et al. will be critical for mitigating the risk of leakage of gas wells often caused by early-age volume changes; for increasing the longevity of pavements, bridges and power plants, specifically those built with high cement binder content; and -as benchmark-for the cement and concrete industry- in the urgent transition to sustainable binder phases.