Despite a millennial history and the ubiquitous presence of cement in everyday life, the molecular processes underlying its hydration behavior, like the formation of calcium–silicate–hydrate (C–S–H), the binding phase of concrete, are mostly unexplored. Using time-resolved potentiometry and turbidimetry combined with dynamic light scattering, small-angle X-ray scattering, and cryo-TEM, we demonstrate C–S–H formation to proceed via a complex two-step pathway. In the first step, amorphous and dispersed spheroids are formed, whose composition is depleted in calcium compared to C–S–H and charge compensated with sodium. In the second step, these amorphous spheroids crystallize to tobermorite-type C–S–H. The crystallization is accompanied by a sodium/calcium cation exchange and aggregation. Understanding the formation of C–S–H via amorphous liquid precursors may allow for a better understanding of the topography of the nucleation in cement paste and thus the percolation of hydration products leading to the mechanical setting as well as the retarding effect of known chemical species like aluminum ions and polycarboxylate ethers.

Amorphous calcium carbonate (ACC) is an important precursor in the biomineralization of crystalline CaCO 3 . The lifetime of transient ACC in nature is regulated by an organic matrix, to use it as an intermediate storage buffer or as a permanent structural element. The relevance of ACC in material science is related to our understanding of CaCO 3 crystallization pathways. ACC can be obtained by liquid−liquid phase separation, and it is typically stabilized with the help of macromolecules. We have prepared ACC by milling calcite in a planetary ball mill. The ball-milled amorphous calcium carbonate (BM-ACC) was stabilized with small amounts of Na 2 CO 3 . The addition of foreign ions in form of Na 2 CO 3 is crucial to achieve complete amorphization. Their incorporation generates defects that hinder recrystallization kinetically. In contrast to wet-chemically prepared ACC, the solvent-free approach makes BM-ACC an anhydrous modification. The amorphization process was monitored by quantitative Fourier transform infrared (FTIR) spectroscopy and solid-state 23 Na magic angle spinning nuclear magnetic resonance ( 23 Na MAS NMR) spectroscopy, which is highly sensitive to changes in the symmetry of the local sodium environment. The structure of BM-ACC was probed by vibrational spectroscopy (FTIR, Raman) and solid-state MAS NMR ( 23 Na, 13 C) spectroscopy. A structural model revealing the partly unsaturated coordination sphere for the Ca 2+ ions was derived from the analysis of total scattering data with high-energy synchrotron radiation. Our findings aid in the understanding of mechanochemical amorphization of calcium carbonate and emphasize the effect of impurities on the stabilization of the amorphous phase, which allowed the synthesis of a so far unknown defect variant of ACC with new properties. This may also represent a general approach to obtain new amorphous phases in a variety of different systems.

Crystallization via metastable phases plays an important role in chemical manufacturing, biomineralization, and protein crystallization, but the kinetic pathways leading from metastable phases to the stable crystalline modifications are not well understood. In particular, the fast crystallization of amorphous intermediates makes a detailed characterization challenging. To circumvent this problem, we devised a system that allows trapping and stabilizing the amorphous intermediates of representative carbonates (calcium, strontium, barium, manganese, and cadmium). The long-term stabilization of these transient species enabled a detailed investigation of their composition, structure, and morphology. Total scattering experiments with high-energy synchrotron radiation revealed a short-range order of several angstroms in all amorphous intermediates. From the synchrotron data, a structural model of amorphous calcium carbonate was derived that indicates a lower coordination number of calcium compared to the crystalline polymorphs. Our study shows that a multistep crystallization pathway via amorphous intermediates is open to many carbonates. We could isolate and characterize these transient species, thereby providing new insights into their crystallization mechanism.