Concrete structures around the world face a quiet chemical threat: the alkali-silica reaction, or ASR, in which alkaline ions from cement attack reactive minerals inside aggregates, producing gels that swell with moisture and crack concrete from within.

The research team, led by Erica Caroline Szydloski with colleagues including Aline Oliveira Mattos, Priscila Ongaratto Trentin, Ronaldo Medeiros-Junior, José Ilo Pereira Filho, Caroline Angulski da Luz and R. Douglas Hooton, tested two supersulfated cement compositions, SSC_75 and SSC_85, containing 75% and 85% slag respectively, alongside anhydrite and 5% Portland cement type V as an alkaline activator. For comparison they used PC_IV, a Portland cement with pozzolan regarded as suitable for ASR control, and PC_V, a pozzolan-free Portland cement used as a reference of unsuitability. All mortars contained the same basaltic fine aggregate from southern Brazil, which prior testing had classified as a potentially reactive grade R2 material because its expansion exceeded the 0.19% limit at 14 days.

The chemistry behind the comparison matters. Supersulfated cement contains higher alumina and lower calcium than ordinary Portland cement, a combination the literature links to ASR mitigation. Alumina is reported to reduce the dissolution of reactive silica by helping form a protective film on aggregate surfaces through the precipitation of zeolites, while calcium raises the system’s pH and promotes silica dissolution, increasing the expansive potential of ASR gels. In the tested cements, PC_IV contained 7.9% alumina and 51.5% calcium, PC_V had only 3.6% alumina but 63.8% calcium, SSC_75 carried 7.9% alumina and 51% calcium, and SSC_85 carried 8.9% alumina and 53% calcium. The supersulfated binders also had far lower equivalent alkali contents, 0.51 to 0.54 in Na2Oeq, than PC_IV at 1.22 and PC_V at 0.99, plus higher sulfate contents of 10 to 20%.

Testing followed the accelerated mortar bar method of NBR 15577-4, the Brazilian standard aligned with ASTM C1260. Mortars mixed at 1:2.25:0.47 cement to sand to water were cast as 25 by 25 by 285 millimeter bars, cured at high humidity, immersed in water at 80 °C for 24 hours, and then held in 1 N sodium hydroxide solution at 80 °C for measurements at 7, 14 and 28 days, with three specimens per cement and age. Beyond the standard window, samples stayed immersed in NaOH but were moved to 23 °C after 42 days, with measurements continuing to 152 days for the SSC mortars, 105 days for PC_IV and 70 days for PC_V. Statistical comparisons used one-way ANOVA with Tukey’s test at 95% reliability.

The length-change results were decisive. SSC_75 expanded just 0.004% at 14 days and 0.031% at 28 days, while SSC_85 actually shrank, by 0.011% at 14 days and 0.005% at 28 days, a contraction the authors attribute to the absence of ASR and to ongoing hydration, particularly C-S-H formation. PC_IV, the reference cement for suitable ASR performance, stayed about 20% below the 0.19% normative limit. PC_V, by contrast, expanded 0.223% at 28 days and kept expanding to 42 days, reaching 0.39%, the highest of the group, so it could not be classified as potentially innocuous. Under the NBR 15577 criteria, both SSC compositions and PC_IV were rated potentially harmless, while PC_V was classified as reactive and potentially deleterious.

Statistical analysis reinforced the pattern. Tukey’s test found no significant difference between the two SSC compositions, nor between SSC_75 and PC_IV, though SSC_85 did differ statistically from PC_IV. PC_V differed significantly from every other cement. Although the two supersulfated cements were statistically indistinguishable, length change decreased as alumina content rose, with SSC_85, the higher-alumina blend, showing the least expansion and results comparable to PC_IV. The authors suggest the low expansions in SSC relate to higher alumina and lower calcium, and note that Ban and colleagues observed that the higher concentration of negatively charged sulfate ions in SSC pore solution can inhibit the diffusion of hydroxide ions from the alkaline NaOH bath through ionic charge balance, which could compensate for SSC_75’s comparatively lower alumina content.

Compressive strength added a second dimension to the comparison. Mortars cured in 80 °C NaOH solution, the aggressive condition of the accelerated ASR test, showed the highest early values, reflecting the impact of elevated curing temperature at 28 days. Crucially, none of the SSC samples lost compressive strength over time in the NaOH solution. The Portland cements behaved differently: PC_IV and PC_V also had higher 28-day strength in NaOH, but at later ages strength increased only in mortars cured in lime-saturated water at 23 °C, while NaOH-cured samples dropped slightly from 28 to 62 days for PC_IV and to 111 days for PC_V. PC_V suffered the steepest decline, a 29% loss in NaOH-cured samples, which the authors link to its intense expansion, nearly 0.4% at 60 days, under the same conditions.

Scanning electron microscopy with energy dispersive spectroscopy, performed after 88 days of NaOH immersion, 42 days at 80 °C and 46 days at 23 °C, revealed the microscopic evidence. In PC_V mortar, pores contained products with characteristic ASR morphology, including crystallized gel in rosette forms and fragmented gel, while PC_IV samples showed pores filled with cracked gel, some partially covered by elongated platelets, and crystallized needle-like products. In both SSC mortars, no ASR gel was observed inside pores; instead the images showed rough pore walls and what appeared to be ettringite needles, a main product of SSC hydration, with some portlandite in SSC_85 attributed to the 5% Portland cement activator. EDS confirmed the chemical divide: PC pores held 32 to 48% SiO2 and 0 to 16% Na2O, typical of ASR, whereas SSC pores contained only 5 to 18% SiO2 and 0 to 5% Na2O, the latter seen only in SSC_75. Products formed in SSC had Ca/Si ratios above 1.5, which are not characterized as ASR products.

The study carries some contextual caveats worth noting. The 28-day readings replaced the standard’s 30-day reading, and PC_V’s was taken at 33 days because restricted laboratory access during the COVID-19 pandemic delayed measurement. Ettringite decomposition in SSC at 80 °C has been reported previously, yet both SSC samples in this investigation appeared to retain a small amount of the phase, which the authors suggest could be compensated by extra ettringite formed during NaOH curing, helping explain the stable strength. The alumina content in the ASR products themselves proved too variable to correlate with a cement’s expansive potential, so the mitigation mechanism cannot be pinned on product chemistry alone. The work also rests on a single basaltic aggregate and accelerated laboratory conditions rather than field structures.

Supersulfated cement is composed of 75 to 90% blast furnace slag, 10 to 20% calcium sulfate and small amounts of alkaline activators, and its low clinker content reduces CO2 emissions. It has already shown efficiency against sulfate attack and chloride ion penetration, but few studies had examined its ASR potential directly.

Subject of Research: Mitigation of the alkali-silica reaction in concrete using supersulfated cement compared with Portland cement

Article Title: Comparison of alkali-silica reaction mitigation between supersulfated cement and Portland cement

Article References: Szydloski, E. C., Mattos, A. O., Trentin, P. O., Medeiros-Junior, R., Pereira Filho, J. I., Angulski da Luz, C., & Hooton, R. D. (2026). Comparison of alkali-silica reaction mitigation between supersulfated cement and Portland cement. Discover Concrete and Cement, 2(1), Article 72. https://doi.org/10.1007/s44416-026-00106-z

Image Credits: AI Generated

DOI: 10.1007/s44416-026-00106-z

Keywords: alkali-silica reaction, supersulfated cement, Portland cement, blast furnace slag, accelerated mortar bar test, concrete durability, alumina, calcium content, compressive strength, SEM-EDS, reactive aggregate, low-clinker cement