Size diversity in cement nanoparticles optimizes packing density to give concrete its strength
November 8, 2012
By Denise Brehm
Civil and Environmental Engineering
Concrete may be one of the most familiar building materials on Earth, but its underlying structure remains a bit of a mystery. Materials scientists and concrete engineers still don’t fully understand exactly how the cement paste that works as glue in concrete hardens during the first hours after water and cement powder are mixed.
New technologies are making it possible for researchers in MIT’s Concrete Sustainability Hub to make steady progress toward solving this mystery. First they determined that cement paste is a granular material, where the particles or basic nanoscale units pack together most densely when arranged orderly. A few years later they discovered that the calcium-silicate-hydrate (C-S-H) molecules that make up the basic nanoscale unit of cement have a disorderly geometric arrangement, rather than the orderly crystalline structure scientists had long assumed.
Particle size implications
In new work, they found that the size of C-S-H particles themselves is also somewhat disorderly: The particles form at very diverse sizes and this diversity in the size of the nanoscale units leads to a denser, disorderly packing of the particles, which corresponds to stronger cement paste.
The researchers hope this understanding will allow materials scientists and concrete engineers to alter the C-S-H particles at the molecular level to develop stronger, more durable concrete that will have a reduced environmental footprint. If concrete is stronger, less of it is needed. And if it’s more durable, structures made from it will last longer.
“While previously our models showed that the particles — think of them as homogenous sized oranges — pack together most densely when arranged orderly in a grocer’s pyramid, our new work shows that when C-S-H units form in a variety of sizes, they can pack more densely when in disarray,” says Franz-Josef Ulm, the George Macomber Professor in the Department of Civil and Environmental Engineering (CEE), co-author of a paper published Oct. 12 in Physical Review Letters. “If you imagine a box randomly filled with many types of fruit, you can see that the berries will naturally fill the space between apples and oranges, and the apples and oranges will do the same between larger fruits.”
Ulm’s earlier mathematical models assumed that the nanoscale units were identical in size and shape. Results from those models showed that those units packed naturally at the two densest arrangements possible for homogenous-sized spheres with the highest density occurring when particles arranged in an orderly pattern. However Ulm’s nanoindentation experiments (poking a tiny dot of cement paste with an even smaller needle and measuring the resistance), which relate the hardness of a nanoscale unit to its packing density, indicated that C-S-H particle strength occurs on a spectrum, suggesting that there could be more than two packing densities.
In the new models, parameters for particle size were opened up to range from 3.5 to 35 nanometers. These models’ predictions of C-S-H particle packing densities agree with the spectrum shown by the earlier nanoindentation tests.
A second important insight comes from the model’s assumption that particles interact. When they touch, they stick together, and when they stick together, the cement is stronger. The researchers know this from the atomic-scale modeling work of Roland J.-M. Pellenq, the CEE senior research scientist who determined the basic structure of a C-S-H particle and the behavior of molecules within the particle. Pellenq is a co-author of the latest paper.
“Before we thought of C-S-H particles as grains with no sticking together,” says Enrico Masoero, a CEE postdoctoral associate who is first author of the paper. “Now we assume there are interactions between the particles and the strength of the interaction determines the strength of the material.”
They extrapolated from Pellenq’s work at the nanoscale that the interactions between molecules within a particle would determine the interactions between particles at the meso scale. The meso scale, Masoero says, has been “like this valley of death in between” the nano and macro scales that was difficult to study because some key quantities for the interactions could not be computed from the molecular structure of the particles, before Pellenq’s work. The next step will be to learn how to improve the sticking of the particles and their packing densities by playing with the chemistry of the raw materials.
“We feel our work supports an emerging awareness that fundamental advances in cement research have the potential to transform an entire industry that is critical to global welfare,” says Professor Emeritus Sidney Yip of the Department of Nuclear Science and Engineering and the Department of Materials Science and Engineering. “The challenges involved in the molecular-level understanding of cement hydration are related to problems in the aging and degradation of materials, which are related to environmental sustainability. With the