The material at the heart of the study is (TiZrVCrAl)N, a nitride of a high-entropy alloy containing five metallic elements: titanium, zirconium, vanadium, chromium, and aluminum. High-entropy alloys, first articulated as a class in landmark reviews by Brian Cantor and by Daniel Miracle and Oleg Senkov, abandon the traditional idea of one dominant base metal. Instead, five or more elements are mixed in near-equal proportions, producing extreme configurational entropy that stabilizes unexpected crystalline phases. When such alloys are reacted with nitrogen, the resulting nitride films inherit this chemical complexity, and with it a combination of high hardness, thermal stability, and wear resistance that conventional binary nitrides such as TiN struggle to match.

High-entropy nitride coatings of the TiZrVCrAl family have already shown promise in demanding applications. Earlier work published in the Journal of Vacuum Science and Technology A reported that a TiZrVCrAl coating on aerospace bearings reduced the friction coefficient in oil-liquid mixed media, while magnetron-sputtered TiVZrCrAl nitride films have achieved super-hard behavior. The new study adds a critical piece to this picture by examining how these coatings form on the workhorse titanium alloy Ti-6Al-4V, the material that carries much of the structural load in modern aircraft, and by asking whether the way that alloy is processed changes the coating that grows on it.

To explore this question, the team prepared two versions of the Ti-6Al-4V substrate. One had a conventional coarse-grained structure. The other was converted to an ultrafine-grained state using severe plastic deformation, a family of processing techniques, including equal-channel angular pressing, that refines grains to sub-micrometer dimensions by imposing enormous strain without changing the shape of the part. Ultrafine-grained titanium alloys are prized for their enhanced strength, fatigue resistance, and creep performance, and previous studies by the same research groups had already shown that such substrates can improve the scratch resistance and adhesive strength of protective coatings. The new work investigates what happens at the finest structural level when a high-entropy nitride is deposited on each substrate type.

The coatings themselves were produced by vacuum arc physical vapor deposition, or arc-PVD, a technique in which an electric arc vaporizes a metallic cathode, ionizes the vapor, and accelerates the ions toward a negatively biased substrate. Arc-PVD is prized industrially because it produces dense, strongly adherent films at reasonable deposition rates. The researchers deposited two architectures on both substrate types: a monolayer coating of (TiZrVCrAl)N and a multilayer coating in which (TiZrVCrAl)N layers alternate with layers of conventional TiN. Multilayer designs are widely used to toughen hard coatings, because interfaces between layers can block dislocation motion and deflect cracks.

The first major result concerns the internal structure of the monolayer coating. Using X-ray diffraction and transmission electron microscopy, the team found that the monolayer (TiZrVCrAl)N film is not structurally uniform through its thickness. Instead, it exhibits a pronounced phase gradient. At the interface where the coating meets the titanium alloy, the film is largely amorphous, lacking long-range crystalline order. Moving toward the outer surface, the structure progressively crystallizes into a face-centered cubic lattice, the same rock-salt-type structure adopted by classical transition-metal nitrides. This gradient makes physical sense: the earliest atoms to arrive on the substrate condense onto a chemically different surface with imperfect lattice matching, so the first atomic layers cannot easily find a crystalline template. As the film thickens, however, newly arriving atoms nucleate on the already-formed crystalline material beneath them, and the face-centered cubic phase establishes itself and grows toward the surface.

The multilayer (TiZrVCrAl)N/TiN coating told a different story. Rather than a single smooth gradient, it displayed pronounced structural and chemical periodicity, with alternating regions of varying degrees of crystallinity repeating through the film thickness. This periodicity mirrors the alternating layer architecture imposed by the deposition sequence, confirming that the multilayer design is faithfully reproduced at the nanometer scale and that each layer transition resets, at least in part, the structural state of the growing film. Such periodic interfaces are precisely what multilayer coating designers intend, and their clear observation in a five-element high-entropy system is an encouraging sign that sophisticated architectures can be engineered in these chemically complex nitrides.

The most striking discovery, however, is the influence of the substrate itself. When the coating was deposited on the ultrafine-grained Ti-6Al-4V rather than the coarse-grained alloy, the coherent scattering region in the coating, the size of the domains over which the crystal lattice remains perfectly aligned, dropped from 14.0 plus or minus 0.3 nanometers to 8.2 plus or minus 0.6 nanometers. In other words, the crystalline domains in the film became roughly forty percent smaller simply because the metal underneath had a finer grain structure. The effect extended to grain morphology as well: near the substrate-coating interface, the ultrafine-grained substrate produced more equiaxed grains, meaning grains with roughly equal dimensions in all directions rather than elongated columnar shapes, and their average size was about twenty percent smaller than the corresponding grains grown on the coarse-grained substrate.

Why should the substrate grain size matter so much? The interface between a coating and its substrate is a zone of intense atomic activity during deposition. Nucleation of the first crystalline layers depends on the topography, chemistry, and defect structure of the surface, all of which differ between coarse-grained and ultrafine-grained titanium. An ultrafine-grained surface presents a far higher density of grain boundaries, which can act as preferential nucleation sites and disrupt the lateral growth of columnar grains. The result is a finer, more equiaxed initial grain structure in the coating, and that initial structure propagates its influence through the film as growth continues, reflected in the reduced coherent scattering domain size measured by diffraction. The study thus demonstrates that substrate microstructure is not a passive backdrop but an active participant in coating formation.

The practical implications are considerable. The authors connect these microstructural differences to the mechanical properties of the coatings, discussing how composition, architecture, and now substrate grain structure jointly determine performance. Smaller crystalline domains and more equiaxed grains near the interface are generally associated with higher hardness and better load-bearing capacity, following the well-established Hall-Petch-type reasoning that grain boundaries impede dislocation motion. Combined with earlier findings that ultrafine-grained titanium substrates enhance coating adhesion and scratch resistance, the new results support a coherent design strategy: process the titanium alloy into an ultrafine-grained state first, then deposit the high-entropy nitride coating, and obtain a component whose surface layer is finer, more defect-controlled, and potentially longer-lasting than anything achievable by coating alone. For aerospace components such as bearings, shafts, and fasteners, where Ti-6Al-4V must resist wear, corrosion, and fatigue simultaneously, that strategy could translate directly into longer service intervals and lighter, more efficient machines. The work also fits into a broader vision, championed by proponents of severe plastic deformation, of so-called superfunctional materials in which bulk nanostructuring and surface engineering are combined deliberately rather than by accident. As high-entropy coatings mature from laboratory curiosities into industrial workhorses, studies like this one make clear that the full performance of these remarkable films can only be unlocked by treating the substrate and the coating as a single engineered system, designed together from the grain boundaries of the metal to the outermost atomic layers of the film.

Subject of Research: Microstructure of high-entropy (TiZrVCrAl)N nitride coatings deposited on coarse-grained and ultrafine-grained Ti-6Al-4V titanium alloy

Article Title: Microstructure of high-entropy (TiZrVCrAl)N coatings deposited on Ti–6Al–4V alloy with different grain structures

Article References: Valiev, R. R., Ovchinnikov, S. V., Savina, Y. N., Belyakov, A. N., Modina, I. M., Nikolaev, A. A., Nazarov, A. Y., Ramazanov, K. N., Caron, A., & Valiev, R. Z. (2026). Microstructure of high-entropy (TiZrVCrAl)N coatings deposited on Ti–6Al–4V alloy with different grain structures. Journal of Materials Science. https://doi.org/10.1007/s10853-026-13813-6

Image Credits: AI Generated

DOI: 10.1007/s10853-026-13813-6

Keywords: high-entropy alloys, thin films, arc-PVD, Ti-6Al-4V, ultrafine-grained titanium, severe plastic deformation, nitride coatings, transmission electron microscopy, X-ray diffraction, coating microstructure, aerospace materials, nanocrystalline materials

The difficulty with C–S–H begins with its structure. Unlike the tidy crystals that fill mineralogy textbooks, C–S–H is X-ray amorphous, meaning its atomic arrangement is too disordered to resolve with conventional diffraction. Its calcium-to-silicon ratio varies widely, its silicate chains are broken at unpredictable points, and its water content shifts with the surrounding environment. This heterogeneity has frustrated experimentalists for decades and has equally hampered computationalists. Density functional theory, the gold standard of atomistic simulation, delivers benchmark accuracy but only for systems of a few hundred atoms over picoseconds. Classical molecular dynamics can reach larger scales, but its reliability rests entirely on the quality of empirical force fields, and the force fields developed for cement each carry serious compromises.

ClayFF, for example, covers a broad range of elements cheaply but was parameterized for clay minerals and struggles to reproduce the mechanics of C–S–H. The CSH-FF improves on crystalline tobermorite but cannot capture reactive chemistry. ReaxFF introduces bond breaking, useful for hydration reactions, yet suffers from poor mechanical accuracy and punishing computational demands. The newer ERICA FF extends elemental coverage and agrees better with quantum calculations, but it remains anchored to fixed functional forms. Machine learning potentials offer a way out of this dilemma: trained on quantum-mechanical data, they reproduce near-DFT accuracy while running orders of magnitude faster than ab initio molecular dynamics. Previous efforts had built such potentials for tobermorite or for narrow slices of C–S–H compositional space, but none had confronted the full chemical and structural wildness of the real material.

The heart of the new work is not the neural network itself but the dataset behind it. Because no public database of experimentally derived C–S–H atomic structures exists, the researchers constructed one from scratch using the pyCSH structure-generation program. They began by generating roughly 500,000 candidate supercells spanning calcium-to-silicon ratios from 0.83 to 2.5, then screened them on four key structural parameters, including silanol density, calcium hydroxylation, and mean silicate chain length, to eliminate redundancy while preserving diversity. The result was 1,926 distinct models covering 688 separate compositions. These were equilibrated with the ERICA force field, refined at the density functional theory level, and then augmented in two stages: deliberate structural perturbations of up to twenty percent generated 9,450 high-energy configurations, and configurations harvested from one hundred molecular dynamics trajectories added another 39,057. In total, more than 50,000 DFT-labeled configurations now constitute what the authors describe as the largest and most comprehensive openly accessible C–S–H dataset to date, a first step toward what they call a cement genome.

The training data were validated with structural fingerprinting. Using the smooth overlap of atomic positions method and principal component analysis, the team mapped the configurational landscape of their dataset and confirmed that it spreads across a genuinely broad space rather than clustering around a few crystal-like basins. The optimization sub-dataset sits near the origin, providing stable energy minima, while the perturbed and trajectory-sampled configurations fan outward, covering the non-equilibrium territory a simulation must navigate. This deliberate coverage of metastable and high-energy states is precisely what separates a robust potential from one that fails the moment atoms stray from perfect crystals, and it explains much of what follows.

The accuracy numbers are striking. CSH-MLP reproduces DFT atomic energies with root mean squared errors of 1.18 and 1.50 meV per atom on training and test sets respectively, and forces with errors of roughly 0.1 eV per angstrom, placing it toward the low end of the accepted range for machine learning potentials. During production simulations, the model maintained consistent accuracy across compositions spanning Ca/Si ratios from 0.91 to 2.06, and it stably equilibrated large-scale models it had never seen, including an 8,456-atom bulk structure and a 10,109-atom gel containing a water-filled nanopore, both built with methods entirely different from the training pipeline. When benchmarked on lattice constants, CSH-MLP deviated from DFT by an average of only 0.14 percent for the C–S–H benchmark model, and it even predicted tobermorite structures, which were absent from training, with errors mostly below one percent.

The comparisons with rival methods were revealing. The general-purpose foundation model MACE-MP-0, despite covering more than eighty elements, performed worse than CSH-MLP on both C–S–H and tobermorite, running at only 0.8 nanoseconds per day on a single GPU versus 21.8 for CSH-MLP. The earlier csh_cp potential, trained largely on crystalline configurations, produced spurious oxygen clustering and outright simulation failure on the gel system, while ClayFF showed unphysical water diffusion and ReaxFF showed artifacts of its own. CSH-MLP, by contrast, reproduced experimental pair distribution functions measured by two independent research groups, matched ab initio molecular dynamics on individual atomic pair correlations, and predicted a water diffusion coefficient in confined nanopores squarely within the range reported by prior studies.

Perhaps the most scientifically satisfying result concerns a long-standing contradiction. High-pressure X-ray diffraction experiments suggest that C–S–H stiffens monotonically as the calcium-to-silicon ratio rises, while decades of simulation work predict that higher calcium content breaks silicate chains and softens the material. CSH-MLP resolved the tension with a non-monotonic prediction: the bulk modulus climbs from 45.57 gigapascals at Ca/Si of 1.3 to a peak of 53.47 gigapascals at 1.5, then falls to 48.29 gigapascals at 1.9. Below the peak, added calcium densifies the interlayer and strengthens cohesion faster than chain scission weakens it; above it, progressive chain fragmentation and interlayer disorder take over. The low-calcium value agrees almost exactly with the experimental figure of 45.6 gigapascals, and the two regimes reconcile the densification story told by X-ray experiments with the defect story told by atomistic simulation as complementary halves of one physical picture.

Speed is the final piece of the puzzle. At the conservative 0.2 femtosecond timestep required by ReaxFF, CSH-MLP runs 3.5 to 6 times faster, and because it remains stable at timesteps up to 1.0 femtosecond, its effective advantage reaches 16.8 to 29.9 times. That means simulations of more than 10,000 atoms over nanosecond timescales at near-quantum accuracy, something previously infeasible for cementitious systems. The dataset and model are openly available on Figshare and the AIS Square platform, and the authors envision fine-tuning the framework toward carbonation, sulfate attack, alkali uptake, and interfaces. If the cement genome project succeeds, the humble glue holding the built environment together may finally yield its secrets, and with them, a credible path toward concrete that builds the future without overheating the planet.

Subject of Research: A machine learning interatomic potential for calcium silicate hydrate, the binding phase of cement, trained on a diverse DFT-labeled dataset

Article Title: Machine Learning Potential for Calcium Silicate Hydrates With Broad Compositional and Structural Diversity

Article References: Li, Y., Chen, C., Pellenq, R. J.-M., Li, Z., & Liu, J. (2026). Machine Learning Potential for Calcium Silicate Hydrates With Broad Compositional and Structural Diversity. Advanced Science, Article e77942. https://doi.org/10.1002/advs.77942

Image Credits: AI Generated

DOI: 10.1002/advs.77942

Keywords: calcium silicate hydrate, machine learning potential, cement, molecular dynamics, density functional theory, concrete, nanocrystalline materials, force fields, mechanical properties, carbon emissions, tobermorite, materials genome