Mayenite owes much of its chemical personality to its architecture. It crystallizes in a cubic structure with space group I-43d, and its unit cell can be described as a positively charged framework of twelve cage-like units, [Ca24Al28O64]4+, in which two negatively charged free oxygen ions wander randomly among the cages. Because it costs only about one-third as much energy to pull one of these cage oxygens out of mayenite as it does from plain lime, the oxygen behaves almost as if it were free. Water molecules can slip through inter-cage channels into the interior of the structure, meet these quasi-free O2− ions and form hydroxyl groups — a mechanism that explains why C12A7 dissolves nearly instantaneously on contact with water. Even a dose of roughly nine percent can dramatically accelerate the setting and hardening of Portland cement, which is why mayenite sits at the heart of set regulators and spray-applied concrete accelerators.

The classic picture of mayenite hydration, established decades ago, is straightforward but troublesome. Below about 20 °C, C12A7 hydrates into metastable hexagonal phases — platy CAH10 and C2AH8. With time or heat, these convert into the thermodynamically stable cubic hydrogarnet C3AH6, releasing amorphous aluminum hydroxide. The conversion shrinks volume and raises density, enlarging the porosity of the hardened paste and degrading late-age strength. In recent years, researchers have sought to steer this pathway by doping the clinker with foreign ions. Magnesium is an obvious candidate: it enters clinker naturally from limestone raw materials, substitutes for calcium in aluminate lattices during sintering, and during hydration shifts the chemical equilibrium toward hydrotalcite, a layered double hydroxide, and magnesium-modified calcium silicate hydrate. What remained unclear was precisely how temperature modulates this magnesium intervention in mayenite.

To find out, Hu Yaru, Yang Jiahui, Wang Kai, Peng Lingwei, Song Qiang, Cheng Fuan, Chen Yanxin and Pan Zhigang synthesized magnesium-containing C12A7 from analytical-grade calcium carbonate, alumina and basic magnesium carbonate. The magnesium precursor was calcined at 900 °C into MgO powder, all raw materials were sieved to 75 micrometers, and 40 grams of MgO were added per 100 grams of C12A7, achieving a molar magnesium-to-aluminum ratio of one. Pellets were fired to 1450 °C, held for three hours and rapidly air-cooled. X-ray fluorescence showed the resulting clinker contained 43.15 percent MgO, 27.33 percent CaO and 28.86 percent Al2O3. Ground clinker was then hydrated at a water-to-solid ratio of 10:1 in nitrogen-purged sealed bottles, cured at room temperature, 50 °C or 80 °C, and halted at ages from three to twenty-eight days for XRD, FT-IR, TG-DTG and SEM-EDS analysis.

The clinker itself told the first story. Diffraction revealed two phases, C12A7 and periclase (MgO), with no free lime. Backscattered electron imaging showed clean phase separation: black granular MgO grains dispersed in a grey mayenite matrix, with trace C3A inclusions too small for diffraction to detect. Crucially, energy-dispersive spectroscopy showed the mayenite lattice had absorbed a small amount of magnesium — about 0.85 percent of calcium sites were occupied by Mg2+, nudging the Ca/Al ratio slightly below its theoretical value. Line scans across phase boundaries recorded a sharp magnesium rise within about 2.5 micrometers of the interface, evidence of genuine lattice substitution rather than mere mechanical mixing.

At room temperature, hydration followed the textbook route with a twist. After three days, the dominant product was metastable C2AH8, accompanied by traces of hydrotalcite and a diffuse halo of amorphous aluminum hydroxide. Notably, no CAH10 appeared during the first fourteen days, when laboratory temperatures averaged 30 °C — warm enough, it seems, to suppress its formation — but CAH10 precipitated between fifteen and twenty-eight days as the ambient temperature dropped to an average of 25.89 °C. This brackets the stability window of CAH10 in magnesium-bearing systems below 30 °C. Cubic C3AH6 emerged at seven days and grew steadily through the conversion reaction in which three molecules of C2AH8 yield two of C3AH6 plus aluminum hydroxide and water. The metastable hexagonal hydrates, under electron microscopy, appeared as large platy crystals, bent sheets squeezed by neighboring crystals, and multilayer stacks tens of nanometers thick.

Raising the cure to 50 °C changed the cast entirely. From day three, the products were stable C3AH6, amorphous aluminum hydroxide and hydrotalcite, with no CAH10 whatsoever. Thermogravimetric curves lost the low-temperature peaks associated with the metastable phases and gained a new peak near 229 °C, the fingerprint of water leaving hydrotalcite interlayers. By twenty-eight days, hydrotalcite accounted for 34.9 percent of the sample — a striking yield driven by the accelerating dissolution of MgO. Electron microscopy showed polyhedral hydrogarnet grains and spherical periclase particles wrapped in flaky agglomerates whose magnesium-to-aluminum ratio of about two confirmed hydrotalcite; at late ages, hydrotalcite sheets were found embedded within individual C3AH6 crystals, hinting at a direct transformation route. The analysis also revealed ternary calcium–magnesium–aluminum hydrates with an approximate stoichiometry of CaMgAl2(OH)10, an intermediate stage on the way to magnesium dominance.

At 80 °C the transformation became aggressive and visually dramatic. Hydrotalcite reached nearly a quarter of the solid mass within three days, and residual periclase was almost completely consumed by fourteen days. Because C3AH6 does not decompose below 282 °C and no CAH10 was present to feed aluminum, the observed decline of hydrogarnet could only mean one thing: magnesium ions dissolved from MgO were eroding C3AH6 itself, harvesting its aluminum to build hydrotalcite. Electron micrographs captured the aftermath — tiny hydrotalcite flakes nucleating on hydrogarnet surfaces, hollowed-out grains retaining only their outer contours like fossil shells, and finally rose-petaled hydrotalcite aggregates. The driving force is thermodynamic: hydrotalcite carries a smaller solubility product and a lower Gibbs free energy of formation than C3AH6. Broadened diffraction peaks with low-angle shoulders suggested a richer hydrotalcite family at this temperature, including compositions approximating Mg6Al2(OH)18·4.5H2O and Mg2Al(OH)7.

Synthesizing the data, the team proposes a unified pathway. Magnesium-containing mayenite first dissolves to metastable CAH10 and C2AH8, which convert with time and temperature into C3AH6. Meanwhile MgO hydrolyzes slowly, releasing Mg2+ that first reacts with calcium aluminate hydrates to form ternary Ca–Mg–Al intermediates and then fully replaces calcium to yield low-calcium hydrotalcite. Heat accelerates every step: the conversion of metastable phases, the dissolution of MgO, and the erosion of hydrogarnet. For engineers, the message is that temperature is a dial controlling not just the speed but the very destination of mayenite hydration — and that endogenous magnesium, often treated as a clinker nuisance, can be enlisted to steer cement chemistry toward thermodynamically robust, potentially more durable products.

Subject of Research: Hydration characteristics of magnesium-containing mayenite (C12A7) cured at room temperature, 50 °C and 80 °C

Article Title: Hydration characteristics of magnesium-containing C 12 A 7 cured at room temperature, 50 ℃ and 80 ℃

Article References: Hydration characteristics of magnesium-containing C 12 A 7 cured at room temperature, 50 ℃ and 80 ℃. (n.d.). https://doi.org/10.1016/j.cscm.2026.e06495

Image Credits: AI Generated

DOI: 10.1016/j.cscm.2026.e06495

Keywords: mayenite, C12A7, calcium aluminate cement, hydrotalcite, magnesium oxide, hydration, curing temperature, C3AH6, CAH10, C2AH8, XRD, SEM-EDS