The researchers, led by Li Gong and Xiaoxiao Liu, prepared a family of C50-grade concretes in which desert sand replaced river sand at levels of 0, 20, 40, 60, 80, and 100 percent by mass. Rather than Portland cement, the binder consisted of ground granulated blast-furnace slag and fly ash—industrial by-products activated by a sodium hydroxide and sodium silicate solution. This alkali-activated chemistry sidesteps the carbon-intensive clinker production of ordinary cement while valorizing waste streams that would otherwise be landfilled. The water-to-binder ratio, binder composition, and activator dosage were held constant, making desert sand content the sole experimental variable.

The desert sand itself is a peculiar ingredient. Its particles are smaller, smoother, and more rounded than river sand, with a fineness modulus of just 0.34 compared with 2.78 for the river sand—a sign of extremely fine, poorly graded grains. It also absorbs nearly twice as much water. On paper, these traits look like liabilities, and at high replacement levels they are. But at moderate doses, the fine desert grains slip into the gaps between coarser river sand particles, tightening the packing of the aggregate skeleton in much the same way a well-shuffled mixture of ball bearings and marbles leaves less void space than either alone.

The sweet spot turned out to be 40 percent. The DS40 mix, as the team labeled it, achieved a 28-day compressive strength of 58.77 megapascals and a 91-day strength of 71.25 megapascals—the highest of all six formulations at every age tested. Specimens from this mix also failed more gracefully under compression, developing inclined cracks rather than shattering into fragments, a signature of a dense matrix with strong aggregate-paste bonding and uniform stress transfer. Mixes with 80 or 100 percent desert sand, by contrast, showed irregular failure modes and large-area spalling, evidence that too many fine particles inflate water demand, spawn extra pores, and weaken the interfaces between paste and aggregate.

Workability told the same story. Slump rose steadily from 164 millimeters in the all-river-sand mix to 182 millimeters at 40 percent replacement, then collapsed to just 87 millimeters when desert sand reached 100 percent. The filling effect of the fine grains initially lubricates and homogenizes the fresh mixture, but beyond the optimum the exploding surface area of fine particles soaks up mixing water and starves the paste. Notably, no mix showed bleeding, confirming that the chosen water-to-binder ratio of 0.42 was adequate across the entire replacement range.

The durability trials were the heart of the study. Prismatic specimens were saturated in water and subjected to up to 200 accelerated freeze-thaw cycles, each consisting of 2.5 hours at minus 18 degrees Celsius followed by 1.5 hours at 5 degrees. After 200 cycles, the DS40 mix had lost only 0.23 percent of its mass and retained 94.56 percent of its relative dynamic elastic modulus, the smallest stiffness decline of any formulation. The fully desert-sand mix fared worst, losing 0.38 percent of its mass and nearly 9 percent of its stiffness. Surface damage followed the same nonlinear pattern: the 40 percent blend kept relatively intact faces while the high-desert mixes developed flaky spalling and exposed aggregate.

Peering inside with nuclear magnetic resonance, scanning electron microscopy, and X-ray diffraction, the researchers traced the damage mechanism in detail. Freeze-thaw cycling coarsened the pore network, shifting pores from tiny gel pores toward capillary pores and macropores, and gradually connecting isolated voids into crack networks. Total porosity rose only modestly—from roughly 1.16 to 1.40 percent in the best mix—but the redistribution toward larger, connected pores proved far more consequential than the raw volume change, because it opened highways for moisture migration and concentrated frost-heave stresses at weak interfaces. Statistical analysis reinforced the link: after 200 cycles, porosity correlated strongly with mass loss and inversely with late-age compressive strength, while mass loss and stiffness decay were almost perfectly coupled.

To translate laboratory cycling into real-world longevity, the team fitted a Weibull reliability model to the stiffness-decay data, defining failure as the point where dynamic modulus falls below 60 percent of its initial value. Converting predicted failure cycles to years using an empirical coefficient of 12 and an annual freeze-thaw count of 118 cycles—typical of the alpine cold regions of Northwest China—the DS40 mix yielded an equivalent service life of 113.3 years, with a 95 percent confidence interval stretching from 89.5 to 141.5 years. The all-river-sand control predicted 99.1 years, while the fully desert-sand mix trailed at 75.4 years. The authors are careful to stress that these figures are model-dependent extrapolations beyond the tested 200-cycle range, intended for comparing mixes rather than certifying actual structures.

That caution is well placed. Sensitivity analyses showed the predictions swing substantially with the assumed conversion coefficient and failure threshold: raising the coefficient from 8 to 16 stretched the DS40 estimate from 75.5 to 151.1 years, while tightening the failure criterion shaved its predicted life from 130.6 to 97.2 years. Crucially, however, the ranking of the mixes held steady under every scenario, with the 40 percent blend always on top and the fully desert-sand mix always at the bottom. The framework, in other words, is far more trustworthy as a comparative tool than as a crystal ball.

The broader implications are striking. Deserts hold theoretical sand reserves of roughly six million cubic kilometers, an essentially untapped aggregate bank, and many countries have already restricted river sand mining to protect ecosystems. This study suggests that desert sand need not be an all-or-nothing proposition: blending it at around 40 percent into alkali-activated slag-fly ash concrete simultaneously improves particle packing, boosts strength, and extends frost durability, all while cutting reliance on both virgin cement and river sand. The authors acknowledge that salt freezing, carbonation, drying shrinkage, and sustained loads—none captured by clean-water freeze-thaw testing—still need investigation, and that the gel chemistry of the alkali-activated matrix warrants verification with additional spectroscopic techniques. But as a proof of concept, the work transforms a ubiquitous desert nuisance into a measured, modeled, and surprisingly durable building material.

Subject of Research: Freeze-thaw durability and service-life prediction of alkali-activated concrete made with desert sand

Article Title: Study on freeze–thaw resistance and equivalent service-life prediction of alkali-activated desert sand concrete

Article References: Gong, L., Liu, X., Bu, Y., Yang, T., Hu, H., & Zhang, H. (2026). Study on freeze–thaw resistance and equivalent service-life prediction of alkali-activated desert sand concrete. Case Studies in Construction Materials, 25, Article e06606. https://doi.org/10.1016/j.cscm.2026.e06606

Image Credits: AI Generated

DOI: 10.1016/j.cscm.2026.e06606

Keywords: desert sand concrete, alkali-activated materials, freeze-thaw resistance, Weibull service-life prediction, relative dynamic elastic modulus, pore structure, ground granulated blast-furnace slag, fly ash, sustainable construction, river sand substitution, cold-region durability, Tengger Desert

The core challenge lies in something invisible to the naked eye: the interfacial transition zone, or ITZ, the thin band of cement paste that bonds directly to the surface of each aggregate particle. In ordinary concrete, the ITZ is often the weakest region of the whole material, but coal gangue makes the problem dramatically worse. The rock is riddled with pores, bedding-related defects, pre-existing microcracks, and weak mineral regions, and its exposed surface is dominated by aluminosilicate minerals that react only weakly under normal curing conditions. Although the angular, rough texture of crushed gangue does provide good mechanical interlocking with the surrounding paste, that advantage cannot compensate for the discontinuous, mechanically feeble boundary that forms where aggregate meets binder. When water finds its way into these surface-connected defects, differences in pore structure and stiffness between the aggregate and the paste concentrate stress near the interface and invite cracks to begin exactly where the material can least afford them.

The research team attacked this weakness with three humble modifiers, all readily available in coal-mining regions: sodium silicate, silica fume, and limestone powder. Each works through a different route. Sodium silicate can penetrate and seal open pores and microdefects, while its soluble silicate species participate in reactions near the aggregate surface that improve interfacial continuity. Silica fume, an ultrafine powder of nearly pure amorphous silica, physically fills surface irregularities and interfacial voids and can later react pozzolanically to refine the surrounding cementitious products. Limestone powder contributes mainly through particle packing, heterogeneous nucleation of hydration products, and carbonate-related interactions with aluminate-bearing phases. The researchers applied each modifier at dosages of 2, 3, and 4 percent by aggregate mass, dry-mixing the powder onto washed, surface-dried gangue particles before the concrete was batched.

The concrete itself was deliberately unconventional. Instead of ordinary Portland cement, the team used a coal-based solid-waste binder containing 50 percent benchmark cement, 25 percent calcined coal gangue powder, 10 percent limestone powder, 10 percent fly ash, and 5 percent flue gas desulfurization gypsum, a formulation that turns more mining and power-plant waste into a useful building material. The calcined gangue powder was produced by heating ground raw gangue first to 500 degrees Celsius and then to 800 degrees, activating its aluminosilicate content. All mixtures shared a fixed water-to-binder ratio of 0.40 and an aggregate-to-binder ratio of 1.50, so that the only variable was the type and dosage of the surface coating. Compressive strength screening at 3, 7, and 28 days identified the best dosage for each modifier: 4 percent sodium silicate, 3 percent silica fume, and 4 percent limestone powder.

The most revealing part of the study came from nanoindentation, a technique that presses a microscopic diamond tip into a polished specimen surface to measure local stiffness point by point. The researchers ran indentation paths of 25 points, spaced 10 micrometers apart, across the aggregate, the ITZ, and the paste on either side. In every mixture, the modulus held steady at roughly 50 to 55 gigapascals inside the aggregate, plunged within the interfacial zone, bottomed out near 110 micrometers from the boundary, and then recovered toward the paste. But the depth of that plunge depended strongly on the treatment. The unmodified concrete reached a minimum of about 18 gigapascals; limestone powder raised it to 22, silica fume to 26, and sodium silicate to 33 gigapascals, the latter producing the smoothest mechanical transition across the entire interface.

Scanning electron microscopy gave those numbers a visible structure. The unmodified interface showed coarse residual particles, loose agglomerates, open pores, and discontinuous patches. Sodium silicate treatment, by contrast, produced abundant flocculent, fibrous, and fine granular products distributed around the aggregate surface, yielding a comparatively continuous and compact boundary. Silica fume yielded a more uniform fine-grained morphology, though some spherical residual particles remained visible. Limestone powder produced a heterogeneous picture, with compact product-rich regions coexisting alongside loose, porous zones and needle-like, plate-like reaction products morphologically consistent with carbonate-bearing AFm-type phases such as monocarboaluminate and hemicarboaluminate, though the authors caution that SEM alone cannot confirm phase identity.

Then came the durability trials, framed in an unusual but physically precise vocabulary. Rather than the conventional labels of wet-dry and freeze-thaw cycling, the researchers speak of liquid-vapor and liquid-solid phase cycling, emphasizing what the pore water itself is doing. In the liquid-vapor test, specimens spent 8 hours immersed in water, 8 hours drying naturally, and 8 hours in a 75-degree-Celsius oven, repeating the absorption-evaporation cycle up to 30 times. In the liquid-solid test, a rapid freeze-thaw apparatus swung the specimen core temperature between minus 35 and plus 20 degrees Celsius every 5 hours, forcing pore water to freeze, redistribute, and melt again and again. The results diverged sharply. After 30 liquid-vapor cycles, silica fume-modified concrete lost only 9.3 percent of its compressive strength, dropping from 27.9 to 25.3 megapascals, while the unmodified control lost 18.8 percent. Freezing proved far more brutal: after 15 liquid-solid cycles, the unmodified and limestone-modified specimens had lost roughly 70 and 67 percent of their strength, and by 30 cycles both had collapsed beyond the point of valid testing.

The survivors told the most important story. Sodium silicate-modified concrete, which had not been the best performer in the wet-dry regime, proved the champion of freeze-thaw resistance, retaining the highest residual strength, 9.7 megapascals after 30 cycles, and enough structural integrity to keep testing. Post-cycling nanoindentation after 15 cycles confirmed that the damage was concentrated precisely where the team expected: the aggregate region barely changed, the paste weakened moderately, and the ITZ deteriorated most severely. Silica fume’s interface lost only 3.8 percent of its minimum modulus under liquid-vapor cycling, while sodium silicate’s interface retained both the highest absolute modulus and the smallest relative loss, 24.8 percent, under liquid-solid cycling. The lesson is that different coatings build different kinds of interfacial quality, and the best choice depends on the environment the concrete will face.

The mechanistic explanation ties the whole picture together. Surface modification does not eliminate water-induced deterioration; it delays its initiation, localization, and propagation within the ITZ. Silica fume’s fine particles plug the small voids where moisture redistribution would otherwise nucleate microcracks, making it ideal for hot, drying climates with repeated wetting. Sodium silicate, supplying soluble silicate species into an alkaline local environment, appears to foster a more continuous, product-rich interfacial fabric that can distribute the internal pressures generated by ice formation and delay debonding and crack coalescence, making it the better choice for cold regions. Limestone powder, with its more heterogeneous product distribution, offered the least protection against either regime. The authors note that water-migration pathways were inferred from combined macroscopic, microscopic, and nanoindentation evidence rather than directly measured, and that local alkalinity changes and individual reaction products were not quantified.

What makes this work resonate beyond the laboratory is its double dividend: it converts an environmental liability into infrastructure material while simultaneously making that material tougher than anyone expected. Because the binder itself is built from calcined coal gangue, fly ash, and desulfurization gypsum, an entire concrete could one day be assembled almost entirely from industrial byproducts, with nothing more exotic than sodium silicate, silica fume, or ground limestone standing between a weak interface and a durable one. For coal regions searching for a future after mining, and for a construction sector under pressure to cut its carbon and resource footprint, the humble act of coating a waste rock particle before it enters a mixer may prove to be one of the most consequential steps in sustainable materials engineering.

Subject of Research: Surface modification of coal gangue aggregate to strengthen the interfacial transition zone in coal-based solid-waste concrete

Article Title: Surface modified coal gangue aggregate in coal-based solid-waste concrete: ITZ strengthening and resistance to water phase-transition cycling

Article References: Wang, Y., Yang, Y., Yu, C., Li, R., Chang, J., & Zhang, T. (2026). Surface modified coal gangue aggregate in coal-based solid-waste concrete: ITZ strengthening and resistance to water phase-transition cycling. Case Studies in Construction Materials, 25, Article e06572. https://doi.org/10.1016/j.cscm.2026.e06572

Image Credits: AI Generated

DOI: 10.1016/j.cscm.2026.e06572

Keywords: coal gangue, concrete, interfacial transition zone, surface modification, sodium silicate, silica fume, limestone powder, freeze-thaw resistance, solid waste recycling, nanoindentation, sustainable construction, coal-based binder