Effects of curing temperature on sulfate-induced expansion of cement mortars

Effects of curing temperature on sulfate-induced expansion of cement mortars | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Effects of curing temperature on sulfate-induced expansion of cement mortars Qiao Wang, William Wilson, Karen Scrivener This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7347409/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract This study aims to understand the effects of the curing temperature on the phase assemblage and the distribution of aluminum-bearing hydrates (AFm and AFt), and how these affect external sulfate attack. Mortars were prepared with Portland cement (PC) and slag-Portland cement at a water-to-cement (w/c) ratio of 0.5. Specimens were cured at 20°C, 40°C and 60°C for 28 days prior to full immersion in sodium sulfate solutions at 50 g/L. The results showed that curing at higher temperatures shortened the latent period before expansion in both PC and slag systems. High-temperature curing altered both the pore structure and the phases embedded in the C-(A-)S-H matrix, leading to more expansion and degradation. Expansion occurred much later and to a lesser extent in slag-Portland mortars due to lower contents of fine monosulfoaluminate (X-ray amorphous) in the C-(A-)S-H, compared to Portland cement mortars. Cement Chemistry Materials Engineering Sulfate attack Slag-Portland cement Expansion Elevated temperature curing C-(A-)S-H composition Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 1 Introduction Concrete structures are built in a wide variety of environments, which can lead to contact with sulfate ions from seawater or groundwater. External sulfate attack becomes a key durability concern for concrete structures exposed to sulfates and moisture [ 1 ]. In the complex chemical reactions between external sulfates and cement hydration products, the main hydration products involved are monosulfoaluminate (Ms) and portlandite, which can transform into ettringite and/or gypsum. The stability of the ettringite/monosulfoaluminate phases is largely dependent on the curing temperature [ 2 , 3 ] and large concrete masses, such as buried foundations, may experience significant self-heating. This makes the effect of the curing temperature on the sulfate attack of interest. Sulfate attack on cementitious materials has been extensively investigated over the last few decades. However, most studies have been carried out at room temperature with a relatively short exposure time [ 4 , 5 ]. Nevertheless, change of temperature during the sulfate exposure below room temperature (below 25°C down to 1°C) have been investigated in relation with the thaumasite formation [ 6 – 8 ]. Some typical findings are that blended mortars performances are generally inferior at lower temperatures due to the significantly reduced degree of hydration, i.e., fly ash blended mortars. Santhanam et al. [ 9 ] investigated curing temperatures between 5°C and 38°C, which was shown to shorten the induction period before expansion kick off, even if the rate of expansion was not significantly changed. A similar study was conducted in which the temperature of the sodium sulfate solution was raised to 40°C over a period of 300 days. This was found to be beneficial in terms of sulfate resistance, as the physical and mechanical properties of the specimens were improved at 40°C compared to the control at 20°C [ 10 ]. Mangat and Khatib [ 11 ] studied the effect of various relative humidity levels and curing and exposure temperatures of 20°C and 45°C on PC and blended concretes. They revealed that the initial curing at 45°C reduced the sulfate resistance of PC concrete, but had a much smaller impact on blended concretes. However, the carbonated formed during air curing also enhanced sulfate resistance, particularly in blended cement concretes. Increasing the curing temperature is increased above 65–70°C during the initial hydration of the cement paste prevents initial ettringite formation, which can result in delayed ettringite formation (DEF) [ 12 , 13 ]. However, this type of sulfate-related degradation is not the focus of this paper. Instead, it examines the influence of intermediate curing temperatures (between 20°C and 60°C), which may frequently occur in large elements due to self-heating. Considering only the 28-day curing period at elevated temperatures avoids the effects of the solubility of different phases (e.g. ettringite and gypsum) and the differing degrees of pozzolanic reaction of SCMs throughout the sulfate exposure period, which complicates interpretation [ 6 , 14 ]. Compared to cementitious systems cured at ambient temperature, heat curing (at 40°C and 50°C) results in a more heterogeneous distribution of hydration products and a coarser pore structure [ 15 , 16 ]. It has also been suggested that the paste-aggregate bond weakens after heat curing, affecting the transport properties of mortars and concretes [ 1 ]. The coarsened microstructure may increase sulfate ingress rates due to an increased proportion of capillary pores [ 17 , 18 ]. Moreover, the AFm phases (monosulfoaluminate (Ms), monocarboaluminate (Mc) and hemicarboaluminate (Hc)) are more likely to form during high-temperature curing, at the expense of ettringite [ 2 , 16 , 19 ]. The relative amount and distribution of these phases can affect the sulfate attack [ 20 , 21 ]. Curing at higher temperatures also alters the phase composition; for example, it changes the calcium-to-silica ratio (C/S) of the C-(A-)S-H. Previous research has shown that it is only when AFm phases are finely intermixed with C-S-H that expansion occurs upon transformation to ettringite, due to the necessary confinement provided by the surrounding C-S-H matrix [ 22 ]. The curing temperature range (i.e., 40°C and 60°C) between the thaumasite formation and DEF has largely been overlooked and warrants further investigation. This study differs from others in that the temperature effect was isolated to the 28-day curing period prior to sulfate exposure. A robust new tool [ 23 – 25 ] has been developed to help us identify progressive changes in microstructure phases at curing temperatures ranging from 20°C to 40°C to 60°C, and to understand how these phases (i.e., monosulfoaluminate) contribute to the external sulfate attack behaviour. This study compares the sulfate resistance of Portland cement (PC) mortars and slag-blended sulfate-resistant cement mortars, which were cured at 20°C, 40°C and 60°C and then exposed to a sodium sulfate solution for over three years. Initial phase compositions were analyzed using SEM-EDS analyses and the content of aluminum-bearing hydrates was measured using the XRD-Rietveld method. The pore structure of the samples, which were cured at different temperatures, was analyzed using mercury intrusion porosimetry. Samples exposed to sulfates were characterised in terms of sulfate ingress depths prior to expansion/cracking onset, providing insight into sulfate ingress rates influenced by the pore structure. Additionally, the genuine contribution of aluminous hydrates to expansion kinetics post-onset was assessed. 2 Materials and methods 2.1. Types of cement In this study, ordinary Portland cement CEM I 42.5 (PC) and slag-Portland cement were used. The chemical compositions measured by X-ray fluorescence and the main clinker phases according to an X-ray diffraction analysis with Rietveld quantification are given in Table 1 . Using the external-standard XRD–Rietveld method, the slag–Portland cement was found to contain 17% slag and 3.3% C₃A, classifying it as sulfate-resistant compared to ordinary Portland cement (according to the EN 197-1:2011 standard, termed as SR 5, C 3 A content of the clinker ≤ 5%). Table 1 Chemical and phase composition of Portland cement (PC) and slag-Portland cement. Oxides wt. % PC Slag-Portland cement Phase composition wt. % PC Slag-Portland cement SiO 2 20.4 24.4 C 3 S 63.8 49.5 Al 2 O 3 5.1 6.5 C 2 S 13.9 13.1 Fe 2 O 3 1.9 2.7 C 3 A 10.6 3.3 CaO 64.2 55.1 C 4 AF 1.9 6.9 MgO 1.0 3.7 Anhydrite 4 1.3 SO 3 2.9 4.6 Calcite 5.2 0.6 Na 2 O 0.4 0.3 Gypsum 0.6 8.3 K 2 O 0.7 0.9 Slag 17 TiO 2 0.2 0.3 P 2 O 5 0.2 0.2 LOI 1.7 1.3 Figure 1 shows the particle size distribution of PC and slag-Portland cement powders as measured by laser diffraction, which was carried out with a Malvern laser diffractometer. A sample of 0.1 g of cement was dispersed in 50 ml of isopropanol with an ultrasonic probe for 15 minutes. Once the suspension was ready, the solution was transferred with a pipette to the measuring zone. 2.2. Methods Sample preparation Mortar prisms were cast in accordance with the EN 196-1 standard test, with a w/c ratio of 0.5. For each mixture, one part of cement (450 g) and three parts of sand (1350 g) were used. The mortars were then cured in a saturated lime solution at different temperatures (20°C, 40°C and 60°C) for 28 days. Three mortar prisms with embedded studs at both ends were prepared for expansion tests and one prism was prepared for microstructure tests for each system. Full immersion was carried out in a 50 g/L Na 2 SO 4 solution, with the solution being refreshed every month to maintain a constant sulfate concentration. Exposure to the sulfate solution took place at a room temperature of 20 ± 2°C. To investigate the phase assemblages of cement paste samples after curing at different temperatures, samples with a w/c ratio of 0.4 were prepared for XRD analysis. Cement paste was chosen to mitigate the dilution effect of high-intensity quartz in XRD and minimize segregation issues. The cement pastes were mixed for two minutes at a speed of 1600 rpm using a high-shear mixer. After mixing, the cement paste was poured into plastic cylinders (33 mm in diameter, 60 mm in length) and left to harden for one day before demolding. The samples were then transferred to slightly larger plastic cylinders and cured with 10 g of distilled water for 28 days. The curing containers were tightly sealed with lids and subjected to different curing temperatures. Sodium sulfate solution The external sulfate solution was prepared with a pure chemical salt containing over 99% Na 2 SO 4 solid. A concentration of 50 g/L was used in accordance with the ASTM C1012 standard [ 26 ]. Mercury intrusion porosimetry (MIP) The porosity of the mortars was analyzed by mercury intrusion porosimetry (MIP) after 28 days of hydration with Porotec GmbH Pascal 140 and 440 instruments. Slices of 2 mm thickness were cut and then immersed in isopropanol for one week to stop the hydration. During this time, the isopropanol was renewed after 1 hour, 1 day and 3 days. Samples were then dried in the vacuum desiccator for two days. The dried slices were broken into pieces of approximately 1.5 g each, and 3–5 pieces were placed in a glass dilatometer. The intrusion was done with pressure up to 400 MPa and a contact angle of 140˚ was assumed for the analysis. XRD Rietveld method with the external standard method The phase assemblages after 28 days of curing at different temperatures were measured on fresh cement paste (w/c = 0.4) slices. X-ray diffraction (XRD) analyses were carried out with a PANalytical X'Pert Pro instrument (CuKα, λ = 1.54 Angstrom) operating with a Bragg–Brentano geometry over a 2θ range of 5–65°. Quantitative Rietveld analyses were made using the HighScore Plus software and the external standard method, with a phase database adapted from the template provided in [ 27 ]. Titanium dioxide was used as the crystalline external standard (rutile Kronos 2300). Expansion measurements The expansion of each specimen was measured every two weeks, according to the standardized method described in [ 26 ]. Chemical composition of the C-(A-)S-H by SEM-EDS analyses Microstructural test samples were cut perpendicular to their length to obtain a cross-sectional slice approximately 5 mm thick, as shown in Fig. 2. Sampling times were chosen according to expansion; samples were taken both before and after the “take-off” of expansion. Slices were taken at least 10 mm from the longitudinal ends of the sample to avoid end-surface effects. Subsamples were cut from the slices (see Fig. 2d) and impregnated with epoxy resin (see Fig. 2e) before being polished using diamond spray suspensions with particles sizes of 9 µm, 3µm and 1 µm (see Fig. 2f). Energy dispersive spectroscopy (EDS) point analysis was performed on the polished section using a FEI Quanta 200 scanning electron microscope equipped with a Bruker 30 EDS detector. Elemental hyper-imaging was also performed to determine the relative concentration of sulfur from the outer surface to the interior. The accelerating voltage was set to 15.0 kV, with a working distance of 12.5 mm. For the point analysis, the points were carefully selected from the outer C-S-H phase of cement paste in mortar samples. It is important that the operator selects the points as far away as possible from portlandite, large masses of ettringite, and AFm and aggregates. Further details on correctly selecting points from a typical BSE image can be found in [ 22 ]. For each sample, 200 points were selected at a similar distance from the exposed surface. Figure 3 shows an example where the outer C-S-H (intermixed with fine AFt and AFm) can be distinguished by its morphology and the grey level from the different phases. Aggregates, large masses of AFm and AFt, clinker, slag, and portlandite are indicated and should be avoided when selecting the phase of interest. The Al/Si ratio of the inner C-(A-)S-H was obtained for each sample in order to estimate the Al 2 O 3 content embedded inside the microstructure (this was not possible using XRD, since the aluminate hydrates are not fully crystalline). For the hypermaps, the scanned area extended from the sulfate-exposed surface (the starting position) to the inner part (the final position), with sufficient overlap to enable the mappings to be stitched together. An example of a stitched long BSE image is shown in Fig. 4. Each mapping area measured approximately 1720 × 1290 µm (with a nominal magnification of 150) and had a resolution of 1000 × 750 pixels (with a pixel size of 1.72 µm). For each sample, a depth of 10 mm was analysed using 10 hypermaps. The Esprit 1.9 software was used to quantify the hypermaps after calibrating them with standards for each element of interest [ 25 ]. The ten individual maps were stitched together using ImageJ [ 28 ]. The final result is a long BSE map with elemental information over the entire depth. After processing, the maps were quantitatively analyzed using the edxia approach developed by Georget et al. [ 25 ]. In Fig. 4 and the other long maps/profiles presented in this paper, the left side shows the solution-exposed surface and the right side shows the inner part. The degree of reaction of the slag was also measured using SEM-EDS hyperspectral imaging: the slag particles and slag hydrates (both anhydrous and hydrated, the latter being a mixture of hydrotalcite (Ht) and C-(A-)S-H) were segmented using edxia [ 25 ], which enabled the volume fraction of reacted slag to be calculated. This percentage was then used alongside mass balance and other inputs (i.e., XRD) to estimate the total Al 2 O 3 content in crystalline aluminous hydrates and in XRD-amorphous phases, either in the C-(A-)S-H form or as microcrystalline monosulfoaluminate. Mass balance of amorphous and crystalline Al 2 O 3 contents The mass balance was used to determine the amorphous and crystalline Al 2 O 3 contents in hydrates after reaction, based on the procedure summarised below [ 29 ]: 1. All the reacted Fe 2 O 3 was assumed to be bound in the hydrogarnet phase C 3 FS 0.84 H 4.32 [30], enabling the calculation of SiO 2 in this phase; 2. All remaining SiO 2 was assumed to be bound in C-(A-)S-H, and the Al 2 O 3 content in the C-(A-)S-H structure was then estimated using the atomic Al/Si ratio obtained via EDS point analysis; 3. The total Al 2 O 3 content in the hydrates was estimated based on the degree of hydration of the cement and the degree of degree of slag reaction; 4. The X-ray amorphous Al 2 O 3 content was calculated by subtracting the crystalline Al 2 O 3 bound in aluminous hydrates (i.e., ettringite, monosulfoaluminate, hydrotalcite (Ht) and hemi/monocarboaluminate) and in unreacted anhydrous cement and slag from the total initial Al 2 O 3 content; 5. The Al 2 O 3 content in microcrystalline monosulfoaluminate within C-(A-)S-H was obtained by subtracting the Al 2 O 3 in the C-(A-)S-H structure from the total X-ray amorphous Al 2 O 3 content. 3 Results and discussion 3.1 The expansion over time As illustrated in Fig. 5, expansion was measured for up to 3.5 years. Significant differences were observed between samples cured at different temperatures. Generally, expansion 'takes off' earlier in samples cured at higher temperatures, in both PC and slag-Portland cement mortars. However, the latent period before significant expansion occurs differs greatly between the two systems: expansion began after approximately 56 days in PC60°C but only after around 600 days in SlagC60°C. Furthermore, the effect of the high curing temperature was less pronounced for slag-Portland cement than for PC. Nevertheless, PC mortar cured at 60°C exhibited higher ultimate expansion than PC mortar cured at 20°C in the final measurements. Slag-Portland cement mortars cured at 20°C and 40°C did not show any significant expansion at the time of writing. 3.2 The pore structures Figure 6 shows a comparison of the pore structure at two curing temperatures: 20°C and 60°C. For each cement system, higher temperature curing results in higher capillary porosity, a shift in the critical pore entry size towards coarser pores, and higher total porosity. For each temperature, PC systems have coarser pores and higher total porosity than slag-Portland cement systems. This explains why sulfate ingress is faster in PC than in slag-Portland cement, as discussed in the next section. Additionally, faster sulfate ingress is anticipated in mortars cured at 60°C compared to 20°C for each cement system. The finer pores in the slag-Portland cement mortar can partially explain the slow reaction kinetics. Finer pores lead to slower sulfate penetration and lower supersaturation with respect to ettringite [ 31 ], particularly in slag-Portland systems with a limited amount of Al 2 O 3 distributed within the C-S-H, as discussed in Section 3.4. 3.3 The sulfate ingress Figure 7 compares the S/Ca profiles of PC and slag-Portland cement mortars that were cured at different temperatures and then exposed to a sulfate solution for 28 days (before the expansion “take-off” for all systems). The profiles were obtained from SEM mapping and show a general decreasing trend in concentration gradients from the exposed surface to an internal depth of 1 mm. More specifically, the sulfate front in the PC mortar cured at 60°C occurs at a depth of ~ 0.5 mm, whereas in the mortar cured at 20°C, it occurs at a depth of ~ 0.2 mm. This greater sulfate penetration at 60°C can be explained by the coarser pore structure resulting from the increased curing temperature, as demonstrated by the MIP results shown in Fig. 6 . For slag-Portland cement mortars, the effect of porosity on sulfate ingress is less pronounced than for PC systems. The S/Ca value remains below 0.2, the sulfate front is less steep, and the distinction between the affected and sound zones is minimal. Consequently, slag-Portland mortar cured at 60°C exhibits a sulfate front that is very similar to that cured at 20°C. Slag-Portland and PC mortars cured at 60°C and 20°C respectively show a very similar pore size distribution. However, the PC mortar shows a relatively higher S/Ca level, which cannot be explained solely by the effect of pore structure, but also depends on phase assemblages (i.e., AFm phase content). Nevertheless, the faster sulfate ingress in PC mortars cured at 60°C explains the shorter latent period observed in Fig. 5: a larger sulfate-affected zone reduces the ability of the sound bulk mortar to resist expansion forces [ 32 ]. Additionally, a S/Ca ratio of 0.2 appears to be a threshold value for triggering the expansion forces. 3.4 The phase assemblage and the distribution of aluminum The phase assemblage after curing The phase assemblages were characterised in cement pastes that were prepared under the same conditions as the mortars and then left to hydrate for 28 days. As illustrated in Fig. 8, the mineralogical composition was quantified using the XRD-Rietveld method with an external standard. In both systems, increasing the curing temperature resulted in a lower content of anhydrous phases, indicating a higher degree of hydration. For the PC pastes, the dominant aluminous hydrates present at room temperature were ettringite (AFt), hemicarboaluminate (Hc) and monocarboaluminate (Mc). However, at 60°C, these crystalline phases were largely depleted and were replaced by amorphous aluminum-bearing hydrates intermixed with the C-S-H matrix. In the slag-Portland cement pastes, the alumina-bearing phases were primarily composed of ettringite and hydrotalcite (Ht). The degree of slag reaction, as estimated via SEM–EDS hypermaps, was used to assess the residual unreacted slag content. Although a reduction in ettringite formation was observed with elevated curing temperatures in both systems, this effect was less pronounced in the slag-containing pastes. These trends are consistent with earlier thermogravimetric (TGA) data [ 16 ] and thermodynamic modelling studies [ 27 , 33 ]. Monosulfoaluminate was not detected in the XRD patterns, possibly due to its poor crystallinity when embedded within the C-(A-)S-H matrix. This morphology limits its detectability by conventional XRD analysis [ 34 ]. Although the expansion mechanism appeared similar for samples cured at 20°C and 60°C (see Fig. 5), differences in the initial phase assemblages, particularly within the C-S-H, resulted in markedly different expansion kinetics. This suggests that expansion is more dependent on the spatial location and morphology of ettringite formation than its total quantity [ 22 , 23 ]. From a thermodynamic perspective, the chemical composition of the raw materials provides information on the stability of aluminum-bearing hydrates. In the PC system, the mass ratio of C₃A to SO₃ is 3.7, which exceeds the threshold value of 3.3 required for monosulfoaluminate (Ms) formation. In contrast, the slag-Portland cement system exhibits a ratio of 0.7, which is below the required threshold of 1.1 for monosulfoaluminate stabilization and therefore favours the formation of ettringite as the primary aluminous hydrate. Repartition of aluminum between anhydrous phases and hydrates As detailed in the experimental section, the partitioning of aluminum among the different phases was quantified using a mass balance approach. This method incorporated the chemical composition of the raw materials, phase assemblage data obtained from XRD–Rietveld analysis and the degree of slag reaction ascertained via SEM–EDS hypermap segmentation. The degrees of slag hydration were estimated to be 44%, 50% and 58% for curing temperatures of 20°C, 40°C and 60°C, respectively. The total Al₂O₃ content in X-ray amorphous phases was calculated by subtracting the aluminum present in the XRD-detectable crystalline phases from the total alumina content. Two types of X-ray amorphous aluminous phases were considered: (i) alumina in solid solution within the C-(A-)S–H structure, and (ii) amorphous AFm phases physically embedded within the C-S-H matrix. The latter are of particular interest as it is hypothesized that their transformation into ettringite under sulfate exposure generates expansive forces. The amorphous AFm content was obtained by subtracting the Al incorporated in the C-(A-)S-H structure from the total amorphous Al₂O₃ content. The outcomes of this partitioning analysis are presented in Fig. 9 . Following hydration, the Al₂O₃ content becomes distributed among various anhydrous constituents (clinker and slag) and hydration products, including the crystalline phase of ettringite and XRD-detectable AFm phases (e.g., Hc, Mc, and Ht), as well as X-ray amorphous AFm phases which are either intermixed with or in solid solution within the C-(A-)S-H matrix. For both binder systems, a greater proportion of Al₂O₃ from the anhydrous phases is incorporated into the hydrates as the curing temperature increases. In the PC system cured at 60°C, most of the Al₂O₃ is found in X-ray amorphous AFm phases, which are dispersed throughout the C-(A-)S-H. Only a small amount remains as crystalline ettringite. In contrast, although the slag-Portland cement contains a higher total Al₂O₃ content, but much of this is sequestered in hydrotalcite-like (Ht) phases, which are generally considered inert with respect to sulfate-induced transformation. The presence of both Ht and solid-solution alumina in the C-(A-)S-H significantly limits the formation of amorphous AFm phases in the slag system. The reduced content of reactive, amorphous AFm phases may explain the markedly lower expansion observed in slag-Portland cement mortars, as shown in Fig. 5. Aluminum in amorphous phases vs. expansion Figure 10 shows the relationship between expansion behavior and the amount of amorphous Al₂O₃ attributed to AFm phases embedded within the C-(A-)S-H matrix. The onset time of expansion and the subsequent expansion rate both exhibit strong correlations with the quantity of X-ray amorphous AFm. While the absolute values should be interpreted with caution due to the inherent methodological uncertainties inherent in calculating amorphous phase content, consistent relative trends are evident. Systems with higher levels of amorphous AFm phases exhibit significantly shorter latent periods prior to the onset of expansion and a more rapid expansion thereafter (see Fig. 5). These findings provide compelling support for the hypothesis that, when confined within the C-(A-)S-H structure, fine-scale, X-ray amorphous AFm phases are the principal drivers of sulfate-induced expansion [ 22 ]. 3.5 The spatial distribution of phases Segmentation of aluminum-bearing phases Cheng’s study [ 22 ] demonstrated that the spatial distribution of aluminous phases has a critical influence on the development of expansion stresses during the transformation of AFm to ettringite. It was concluded that only ettringite that crystallizes within confined spaces –specifically, in fine pores embedded in the C-(A-)S-H matrix – generates expansive forces capable of inducing damage [ 22 ]. Building on this hypothesis, this section compares the spatial distribution of AFm and ettringite phases in PC and slag-Portland cement systems. Figure 11 shows an example of the segmentation of aluminum-bearing phases in slag–Portland cement paste, as determined using chemical ratio plots and the edxia framework [ 25 ]. Phase identification was verified through correlation with grey-level contrast in BSE micrographs. The presence of hydrotalcite (Ht) is attributed to the magnesium content of the reacted slag grains. A small quantity of monosulfoaluminate, accompanied by minor Mc/Hc, was also detected via SEM–EDS analysis. These phases were found to be finely intermixed within the C-(A-)S-H matrix. Specifically, Ht (highlighted in orange) forms around the periphery of reacted slag grains, while crystalline AFm phases (cyan) appear as discrete, larger domains that are identifiable in the elemental maps. The C-(A-)S-H phase (purple) acts as a continuous binding matrix that incorporates the finely dispersed AFm and Ht phases throughout the microstructure. Distribution of aluminum-bearing phases Figure 12 shows the spatial distribution of the AFm/C-(A-)S-H and Ht/C-(A-)S-H phases in both binder systems after 28 days of hydration at 20°C and 60°C. The aluminum-rich phases were segmented using the edxia framework [ 25 ] and overlaid onto the corresponding BSE micrographs. An increase in AFm content was observed in both the PC and slag–Portland cement systems at 60°C, which is consistent with the XRD–Rietveld quantification (see Fig. 8). Interestingly, the PC system cured at 20°C exhibited a higher AFm content than the slag–Portland system cured at 60°C. In slag–Portland cement hydrated at 20°C, the AFm phase identified by SEM–EDS appeared to be a composite of monosulfoaluminate and C-(A-)S-H, along with a minor fraction of Hc/Mc. The phases were not detected by XRD, due to their low crystallinity or limited abundance. At 60°C, the increase in total aluminum-bearing hydrates in the slag system was attributed to AFm formation, as the Ht/C-(A-)S-H content remained largely unchanged across both curing temperatures. Further analysis using high-magnification SEM imaging (4000×) reveals distinct morphological differences in the distribution of the AFm phase between the two binder systems. In PC pastes, AFm is predominantly localised around clinker grains within the inner C-S-H product. While these phases can be detected by SEM, their encapsulation within the inner C-S-H suggests that they are not mechanically active and therefore do not contribute to expansion. In contrast, AFm in slag–Portland cement pastes appears as large, discrete “pockets” within the matrix. Previous studies have shown that such morphologies are non-expansive, even when these AFm phases subsequently convert to ettringite upon exposure to sulfate [ 22 ]. 4 Conclusions This study highlights the impact of increased curing temperatures on the sulfate resistance of Portland and slag–Portland cement mortars when fully immersed. Based on the experimental findings, the following conclusions can be drawn: Mortars cured at 60°C exhibit significantly higher expansion compared to those cured at 20°C. However, slag–Portland cement systems demonstrate much lower expansion overall, showing reduced sensitivity to curing temperature compared to PC systems. Elevated curing temperatures result in a more porous microstructure characterized by coarser capillary pores, facilitating faster sulfate ingress than in mortars cured at ambient conditions. In PC systems, curing at 40°C and 60°C destabilizes the ettringite and Mc/Hc phases typically present at 20°C, promoting their replacement by X-ray amorphous AFm phases intermixed within the C-S-H. In slag–Portland pastes, ettringite and hydrotalcite remain the dominant alumina-bearing hydrates, with the latter showing little variation with temperature. The X-ray amorphous AFm phases confined within the C-(A-)S-H matrix are identified as the main cause of expansion. Their higher abundance in PC systems cured at 60°C correlates with increased expansion, whereas their limited presence in slag–Portland systems accounts for the substantially lower expansion observed. More broadly, the expansive phase was identified as amorphous Al₂O₃ embedded in the C-(A-)S-H, and its content was found to correlate with both the onset and rate of expansion. The curing temperature influences both the phase assemblage and pore structure, primarily by reducing the stability of ettringite and increasing the formation of amorphous aluminates, thereby altering sulfate resistance. Furthermore, mass balance analysis of Al₂O₃ partitioning provided key insights into the mechanisms underlying the enhanced sulfate resistance of slag–Portland systems, despite their higher total aluminum content. These results suggest that the amorphous Al₂O₃ content within C-(A-)S-H may serve as a potential indicator of sulfate resistance. Future work should focus on establishing quantitative thresholds for this parameter in relation to expansion kinetics and durability performance. References Jan Skalny, Jacques Marchand, Ivan Odler (2002) Sulfate Attack on Concrete. London and New York Chitvoranund N (2021) Stability of hydrate assemblages and properties of cementitious systems with higher alumina content. EPFL Kaufmann J, Winnefeld F, Lothenbach B (2016) Stability of ettringite in CSA cement at elevated temperatures. Advances in Cement Research 28:251–261. https://doi.org/10.1680/jadcr.15.00029 González MA, Irassar EF (1997) Ettringite formation in low C3A Portland cement exposed to sodium sulfate solution. Cement and Concrete Research 27:1061–1071. https://doi.org/10.1016/S0008-8846(97)00093-8 Wang JG (1994) Sulfate attack on hardened cement paste. Cement and Concrete Research 24:735–742. https://doi.org/10.1016/0008-8846(94)90199-6 Hossack AM, Thomas MDA (2015) The effect of temperature on the rate of sulfate attack of Portland cement blended mortars in Na2SO4 solution. Cement and Concrete Research 73:136–142. https://doi.org/10.1016/j.cemconres.2015.02.024 Thaumasite formation in concrete and mortars containing fly ash. Cement and Concrete Composites 25:907–912. https://doi.org/10.1016/S0958-9465(03)00136-7 Schmidt T, Lothenbach B, Romer M, et al (2009) Physical and microstructural aspects of sulfate attack on ordinary and limestone blended Portland cements. Cement and Concrete Research 39:1111–1121. https://doi.org/10.1016/j.cemconres.2009.08.005 Santhanam M, Cohen MD, Olek J (2002) Modeling the effects of solution temperature and concentration during sulfate attack on cement mortars. Cement and Concrete Research 32:585–592. https://doi.org/10.1016/S0008-8846(01)00727-X Aköz F, Türker F, Koral S, Yüzer N (1999) Effects of raised temperature of sulfate solutions on the sulfate resistance of mortars with and without silica fume. Cement and Concrete Research 29:537–544. https://doi.org/10.1016/S0008-8846(98)00251-8 Influence of initial curing on sulphate resistance of blended cement concrete. Cement and Concrete Research 22:1089–1100. https://doi.org/10.1016/0008-8846(92)90039-X Lawrence CD (1995) Mortar expansions due to delayed ettringite formation. Effects of curing period and temperature. Cement and Concrete Research 25:903–914. https://doi.org/10.1016/0008-8846(95)00081-M Taylor HFW, Famy C, Scrivener KL (2001) Delayed ettringite formation. Cement and Concrete Research 31:683–693. https://doi.org/10.1016/S0008-8846(01)00466-5 Perkins RB, Palmer CD (1999) Solubility of ettringite (Ca6[Al(OH)6]2(SO4)3 · 26H2O) at 5–75°C. Geochimica et Cosmochimica Acta 63:1969–1980. https://doi.org/10.1016/S0016-7037(99)00078-2 Haynes H (2002) Sulfate Attack on Concrete: Laboratory vs. Field Experience. CI 24:64–70 Lothenbach B, Winnefeld F, Alder C, et al (2007) Effect of temperature on the pore solution, microstructure and hydration products of Portland cement pastes. Cement and Concrete Research 37:483–491. https://doi.org/10.1016/j.cemconres.2006.11.016 Zhang Z, Zou Y, Yang J, Zhou J (2022) Capillary rise height of sulfate in Portland-limestone cement concrete under physical attack: Experimental and modelling investigation. Cement and Concrete Composites 125:104299. https://doi.org/10.1016/j.cemconcomp.2021.104299 The microstructure of concrete cured at elevated temperatures. Cement and Concrete Research 25:485–490. https://doi.org/10.1016/0008-8846(95)00036-C Thomas JJ, Rothstein D, Jennings HM, Christensen BJ (2003) Effect of hydration temperature on the solubility behavior of Ca-, S-, Al-, and Si-bearing solid phases in Portland cement pastes. Cement and Concrete Research 33:2037–2047. https://doi.org/10.1016/S0008-8846(03)00224-2 Gollop R, Taylor H (1995) Microstructural and microanalytical studies of sulfate attack III. Sulfate-resisting portland cement: Reactions with sodium and magnesium sulfate solutions. https://doi.org/10.1016/0008-8846(95)00151-2 Elahi MMA, Shearer CR, Naser Rashid Reza A, et al (2021) Improving the sulfate attack resistance of concrete by using supplementary cementitious materials (SCMs): A review. Construction and Building Materials 281:122628. https://doi.org/10.1016/j.conbuildmat.2021.122628 Yu C, Sun W, Scrivener K (2013) Mechanism of expansion of mortars immersed in sodium sulfate solutions. Cement and Concrete Research 43:105–111. https://doi.org/10.1016/j.cemconres.2012.10.001 Wang Q, Wilson W, Scrivener K (2023) Unidirectional penetration approach for characterizing sulfate attack mechanisms on cement mortars and pastes. Cement and Concrete Research 169:107166. https://doi.org/10.1016/j.cemconres.2023.107166 Huang Q, Wang Q, Zhu X (2024) Contradict mechanism of long-term magnesium and sodium sulfate attacks of nano silica-modified cement mortars – Experimental and thermodynamic modeling. Cement and Concrete Composites 147:105444. https://doi.org/10.1016/j.cemconcomp.2024.105444 Georget F, Wilson W, Scrivener KL (2021) edxia: Microstructure characterisation from quantified SEM-EDS hypermaps. Cement and Concrete Research 141:106327. https://doi.org/10.1016/j.cemconres.2020.106327 Standard Test Method for Length Change of Hydraulic-Cement Mortars Exposed to a Sulfate Solution. ASTM C1012-18a Wilson W, Gonthier JN, Georget F, Scrivener KL (2022) Insights on chemical and physical chloride binding in blended cement pastes. Cement and Concrete Research 156:106747. https://doi.org/10.1016/j.cemconres.2022.106747 Preibisch S, Saalfeld S, Tomancak P (2009) Globally optimal stitching of tiled 3D microscopic image acquisitions. Bioinformatics 25:1463–1465. https://doi.org/10.1093/bioinformatics/btp184 Durdziński PT, Ben Haha M, Zajac M, Scrivener KL (2017) Phase assemblage of composite cements. Cement and Concrete Research 99:172–182. https://doi.org/10.1016/j.cemconres.2017.05.009 Dilnesa BZ, Wieland E, Lothenbach B, et al (2014) Fe-containing phases in hydrated cements. Cement and Concrete Research 58:45–55. https://doi.org/10.1016/j.cemconres.2013.12.012 Scherer GW (2004) Stress from crystallization of salt. Cement and Concrete Research 34:1613–1624. https://doi.org/10.1016/j.cemconres.2003.12.034 Kunther W (2012) Investigation of Sulfate Attack by Experimental and Thermodynamic Means. PhD Thesis, EPFL Lothenbach B, Matschei T, Möschner G, Glasser FP (2008) Thermodynamic modelling of the effect of temperature on the hydration and porosity of Portland cement. Cement and Concrete Research 38:1–18. https://doi.org/10.1016/j.cemconres.2007.08.017 Scrivener K, Snellings R, Lothenbach B (2017) A Practical Guide to Microstructural Analysis of Cementitious Materials. CRC Press, Boca Raton Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7347409","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":498877732,"identity":"c77d53ba-0864-4c99-8aae-fb07f2d27e55","order_by":0,"name":"Qiao Wang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABD0lEQVRIiWNgGAWjYNCCCokEKOsAiGADEYwNeLWcgWlJIFYLYxsDCVr4biQ/e/h1nkUef/sBts+8P+4w8Lcffvbg4w4G2X4cWiRvpJkby26TKJY4k8A8myfhGYPEmTRzw5lnGIxn4rDG4HaCmbTkNonEhhsMzMw5CYfrGw7ksEnztjEkbjiAS0v6N2nJORKJ86FaGOTPv2GT/gvUsh+nlhwzyY8NEokbYFoMbgBtYQTZgssv99+USTMckyg2PJPYzPwn7TCD4Y1nZpK9bRLGM3DYwnfm+DbJHzV1eXLHDx9mnGFzmEHufPIziZ9tNrL9OLwPim5mHjALNSIkcKiHaGH8gVt6FIyCUTAKRgEDAwCMZl/2gE/t4gAAAABJRU5ErkJggg==","orcid":"","institution":"EPFL","correspondingAuthor":true,"prefix":"","firstName":"Qiao","middleName":"","lastName":"Wang","suffix":""},{"id":498877975,"identity":"68b4c0fb-e142-49da-850a-437319a9035b","order_by":1,"name":"William Wilson","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"William","middleName":"","lastName":"Wilson","suffix":""},{"id":498878182,"identity":"5309e0da-c3df-4215-a484-a1d9a14368e7","order_by":2,"name":"Karen Scrivener","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Karen","middleName":"","lastName":"Scrivener","suffix":""}],"badges":[],"createdAt":"2025-08-11 14:37:28","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-7347409/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7347409/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":88878150,"identity":"6d242945-97b8-4e99-ace3-086e1ab6ec3d","added_by":"auto","created_at":"2025-08-12 10:37:39","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":98182,"visible":true,"origin":"","legend":"\u003cp\u003eParticle size distribution of the raw materials used in this study.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7347409/v1/992cc727859d84899aef1527.png"},{"id":88879884,"identity":"155b2d0a-ff0c-4148-be8e-a37f4f93f16b","added_by":"auto","created_at":"2025-08-12 10:53:39","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":1630965,"visible":true,"origin":"","legend":"\u003cp\u003eSEM sample preparation process: (a) mortars exposed in Na\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e solution, (b) position of the slice from the side view, (c) position of the slice from the top view, (d) position of the subsample extracted from the cross section, (e) embedding of SEM sample in epoxy, (f) polished SEM sample.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7347409/v1/2a1ecd0392d56ce2e9e36441.png"},{"id":88879889,"identity":"04a6c529-4e81-47d5-b54c-93605b761c4a","added_by":"auto","created_at":"2025-08-12 10:53:39","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":621634,"visible":true,"origin":"","legend":"\u003cp\u003eExample of an EDS point analysis in the outer C-S-H of a slag-Portland cement mortar sample.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7347409/v1/119c931122582099976b5894.png"},{"id":88879890,"identity":"aca395f3-6ef4-4cb9-9490-a9cea244d80e","added_by":"auto","created_at":"2025-08-12 10:53:39","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":450074,"visible":true,"origin":"","legend":"\u003cp\u003eExample of a stitched long BSE image for PC mortar after sulfate attack with a depth of 11 mm.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7347409/v1/f68bb06e00ea7208cec338c4.png"},{"id":88878158,"identity":"91740c48-c97d-4121-bd45-1ae18c828cc0","added_by":"auto","created_at":"2025-08-12 10:37:39","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":327908,"visible":true,"origin":"","legend":"\u003cp\u003eEvolution of sulfate attack expansion over time at different curing temperatures for PC and slag-Portland cement mortars.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7347409/v1/b2e64059451c148d8eb4883f.png"},{"id":88879078,"identity":"05d39ee4-e2a2-4408-9304-8b1d1f23ac1f","added_by":"auto","created_at":"2025-08-12 10:45:39","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":170881,"visible":true,"origin":"","legend":"\u003cp\u003ePore entry radius and measured pressure of PC and slag-Portland mortars after 28 days of curing at temperatures of 20°C and 60°C.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-7347409/v1/63e68958c2b664d3c4d6df52.png"},{"id":88879082,"identity":"89df6a1e-240a-4406-9036-ff4209977e07","added_by":"auto","created_at":"2025-08-12 10:45:39","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":201738,"visible":true,"origin":"","legend":"\u003cp\u003eSulfate ingress depth after 28 days of sulfate exposure at curing temperatures of 20 °C and 60 °C, for (a) PC mortars (b) the slag-Portland cement mortars.\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-7347409/v1/933c8c3259d13a9094a19e4f.png"},{"id":88878163,"identity":"e260fd23-93c2-4e54-a47e-6c2f8aff9825","added_by":"auto","created_at":"2025-08-12 10:37:39","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":458458,"visible":true,"origin":"","legend":"\u003cp\u003ea) XRD patterns showing the main peaks of ettringite (AFt), hemicarboaluminate (Hc), monocarboaluminate (Mc) and hydrotalcite (Ht) for the investigated systems, and b) Phase assemblage after hydration of 28 days, as obtained with the XRD-Rietveld method.\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-7347409/v1/c54057c235dca4290a3b487b.png"},{"id":88879084,"identity":"cc9de5f6-13c8-4198-ba2f-3cdf63729d9c","added_by":"auto","created_at":"2025-08-12 10:45:39","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":249996,"visible":true,"origin":"","legend":"\u003cp\u003eAl\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e distribution in hydrated phases, for PC and slag-Portland samples after 28 days hydration, where “anhydrous” means unreacted clinker and “slag” means unreacted slag.\u003c/p\u003e","description":"","filename":"9.png","url":"https://assets-eu.researchsquare.com/files/rs-7347409/v1/54bfd0a625666644da78e66c.png"},{"id":88878166,"identity":"a768ecac-0d91-41cc-82fd-6f5b50e2114c","added_by":"auto","created_at":"2025-08-12 10:37:39","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":173712,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelations between the amorphous Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e content in monosulfoaluminate and the “take-off” time or the rate of expansion after “take-off”. The dashed lines were added as eye guides.\u003c/p\u003e","description":"","filename":"10.png","url":"https://assets-eu.researchsquare.com/files/rs-7347409/v1/73718d2fe8be76e3fc007c67.png"},{"id":88879085,"identity":"9e07c471-4d12-46ae-9e8d-d34c85f46413","added_by":"auto","created_at":"2025-08-12 10:45:39","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":675085,"visible":true,"origin":"","legend":"\u003cp\u003eSegmentation of aluminum-bearing phases of slag-Portland cement mortar cured at 20 °C by \u003cem\u003eedxia\u003c/em\u003e: Ms, hydrotalcite, C-(A-)S-H phases.\u003c/p\u003e","description":"","filename":"11.png","url":"https://assets-eu.researchsquare.com/files/rs-7347409/v1/c11c8956ab1869323411996a.png"},{"id":88880484,"identity":"bdf9ef50-c6a1-45aa-a41c-af762493d3fc","added_by":"auto","created_at":"2025-08-12 11:01:40","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":7137000,"visible":true,"origin":"","legend":"\u003cp\u003eBSE micrographs overlain with phase masks for PC mortars (left column) and slag-Portland cement mortars (right column), showing the AFm/C-(A-)S-H mixture in cyan and the Ht/C-(A-)S-H mixture in orange after 28-day curing at 20 °C (a-d) and 60 °C (e-h).\u003c/p\u003e","description":"","filename":"12.png","url":"https://assets-eu.researchsquare.com/files/rs-7347409/v1/ee00dc3a30ae60de74bab0ae.png"},{"id":89065107,"identity":"780ab570-cf24-451c-9c8f-6ebc755038ff","added_by":"auto","created_at":"2025-08-14 10:36:32","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":16931278,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7347409/v1/c6a8de14-d7b8-4067-9ec1-038b6f7b0079.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003e\u003cstrong\u003eEffects of curing temperature on sulfate-induced expansion of cement mortars\u003c/strong\u003e\u003c/p\u003e","fulltext":[{"header":"1 Introduction","content":"\u003cp\u003eConcrete structures are built in a wide variety of environments, which can lead to contact with sulfate ions from seawater or groundwater. External sulfate attack becomes a key durability concern for concrete structures exposed to sulfates and moisture [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. In the complex chemical reactions between external sulfates and cement hydration products, the main hydration products involved are monosulfoaluminate (Ms) and portlandite, which can transform into ettringite and/or gypsum. The stability of the ettringite/monosulfoaluminate phases is largely dependent on the curing temperature [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e] and large concrete masses, such as buried foundations, may experience significant self-heating. This makes the effect of the curing temperature on the sulfate attack of interest.\u003c/p\u003e\u003cp\u003eSulfate attack on cementitious materials has been extensively investigated over the last few decades. However, most studies have been carried out at room temperature with a relatively short exposure time [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Nevertheless, change of temperature during the sulfate exposure below room temperature (below 25\u0026deg;C down to 1\u0026deg;C) have been investigated in relation with the thaumasite formation [\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. Some typical findings are that blended mortars performances are generally inferior at lower temperatures due to the significantly reduced degree of hydration, i.e., fly ash blended mortars. Santhanam et al. [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e] investigated curing temperatures between 5\u0026deg;C and 38\u0026deg;C, which was shown to shorten the induction period before expansion kick off, even if the rate of expansion was not significantly changed. A similar study was conducted in which the temperature of the sodium sulfate solution was raised to 40\u0026deg;C over a period of 300 days. This was found to be beneficial in terms of sulfate resistance, as the physical and mechanical properties of the specimens were improved at 40\u0026deg;C compared to the control at 20\u0026deg;C [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Mangat and Khatib [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e] studied the effect of various relative humidity levels and curing and exposure temperatures of 20\u0026deg;C and 45\u0026deg;C on PC and blended concretes. They revealed that the initial curing at 45\u0026deg;C reduced the sulfate resistance of PC concrete, but had a much smaller impact on blended concretes. However, the carbonated formed during air curing also enhanced sulfate resistance, particularly in blended cement concretes.\u003c/p\u003e\u003cp\u003eIncreasing the curing temperature is increased above 65\u0026ndash;70\u0026deg;C during the initial hydration of the cement paste prevents initial ettringite formation, which can result in delayed ettringite formation (DEF) [\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e, \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. However, this type of sulfate-related degradation is not the focus of this paper. Instead, it examines the influence of intermediate curing temperatures (between 20\u0026deg;C and 60\u0026deg;C), which may frequently occur in large elements due to self-heating.\u003c/p\u003e\u003cp\u003eConsidering only the 28-day curing period at elevated temperatures avoids the effects of the solubility of different phases (e.g. ettringite and gypsum) and the differing degrees of pozzolanic reaction of SCMs throughout the sulfate exposure period, which complicates interpretation [\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eCompared to cementitious systems cured at ambient temperature, heat curing (at 40\u0026deg;C and 50\u0026deg;C) results in a more heterogeneous distribution of hydration products and a coarser pore structure [\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. It has also been suggested that the paste-aggregate bond weakens after heat curing, affecting the transport properties of mortars and concretes [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. The coarsened microstructure may increase sulfate ingress rates due to an increased proportion of capillary pores [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e, \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eMoreover, the AFm phases (monosulfoaluminate (Ms), monocarboaluminate (Mc) and hemicarboaluminate (Hc)) are more likely to form during high-temperature curing, at the expense of ettringite [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. The relative amount and distribution of these phases can affect the sulfate attack [\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Curing at higher temperatures also alters the phase composition; for example, it changes the calcium-to-silica ratio (C/S) of the C-(A-)S-H. Previous research has shown that it is only when AFm phases are finely intermixed with C-S-H that expansion occurs upon transformation to ettringite, due to the necessary confinement provided by the surrounding C-S-H matrix [\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e].\u003c/p\u003e\u003cp\u003eThe curing temperature range (i.e., 40\u0026deg;C and 60\u0026deg;C) between the thaumasite formation and DEF has largely been overlooked and warrants further investigation. This study differs from others in that the temperature effect was isolated to the 28-day curing period prior to sulfate exposure. A robust new tool [\u003cspan additionalcitationids=\"CR24\" citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e] has been developed to help us identify progressive changes in microstructure phases at curing temperatures ranging from 20\u0026deg;C to 40\u0026deg;C to 60\u0026deg;C, and to understand how these phases (i.e., monosulfoaluminate) contribute to the external sulfate attack behaviour.\u003c/p\u003e\u003cp\u003eThis study compares the sulfate resistance of Portland cement (PC) mortars and slag-blended sulfate-resistant cement mortars, which were cured at 20\u0026deg;C, 40\u0026deg;C and 60\u0026deg;C and then exposed to a sodium sulfate solution for over three years. Initial phase compositions were analyzed using SEM-EDS analyses and the content of aluminum-bearing hydrates was measured using the XRD-Rietveld method. The pore structure of the samples, which were cured at different temperatures, was analyzed using mercury intrusion porosimetry. Samples exposed to sulfates were characterised in terms of sulfate ingress depths prior to expansion/cracking onset, providing insight into sulfate ingress rates influenced by the pore structure. Additionally, the genuine contribution of aluminous hydrates to expansion kinetics post-onset was assessed.\u003c/p\u003e"},{"header":"2 Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\n \u003ch2\u003e2.1. Types of cement\u003c/h2\u003e\n \u003cp\u003eIn this study, ordinary Portland cement CEM I 42.5 (PC) and slag-Portland cement were used. The chemical compositions measured by X-ray fluorescence and the main clinker phases according to an X-ray diffraction analysis with Rietveld quantification are given in Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e. Using the external-standard XRD\u0026ndash;Rietveld method, the slag\u0026ndash;Portland cement was found to contain 17% slag and 3.3% C₃A, classifying it as sulfate-resistant compared to ordinary Portland cement (according to the EN 197-1:2011 standard, termed as SR 5, C\u003csub\u003e3\u003c/sub\u003eA content of the clinker\u0026thinsp;\u0026le;\u0026thinsp;5%).\u003c/p\u003e\n \u003ctable id=\"Tab1\" border=\"1\" class=\"fr-table-selection-hover\"\u003e\n \u003ccaption language=\"En\"\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eChemical and phase composition of Portland cement (PC) and slag-Portland cement.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eOxides wt. %\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003ePC\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSlag-Portland cement\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003ePhase composition wt. %\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003ePC\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eSlag-Portland cement\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSiO\u003csub\u003e2\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e20.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e24.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eC\u003csub\u003e3\u003c/sub\u003eS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e63.8\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e49.5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAl\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e5.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e6.5\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eC\u003csub\u003e2\u003c/sub\u003eS\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e13.1\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFe\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eC\u003csub\u003e3\u003c/sub\u003eA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e10.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e3.3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCaO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e64.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e55.1\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eC\u003csub\u003e4\u003c/sub\u003eAF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e6.9\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMgO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e3.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eAnhydrite\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e1.3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSO\u003csub\u003e3\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e2.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e4.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eCalcite\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e5.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.6\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNa\u003csub\u003e2\u003c/sub\u003eO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.4\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eGypsum\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.6\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e8.3\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eK\u003csub\u003e2\u003c/sub\u003eO\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.9\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eSlag\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e17\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eTiO\u003csub\u003e2\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eP\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e5\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.2\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eLOI\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.3\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003cp\u003e\u003c/p\u003e\n \u003cp\u003eFigure \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e shows the particle size distribution of PC and slag-Portland cement powders as measured by laser diffraction, which was carried out with a Malvern laser diffractometer. A sample of 0.1 g of cement was dispersed in 50 ml of isopropanol with an ultrasonic probe for 15 minutes. Once the suspension was ready, the solution was transferred with a pipette to the measuring zone.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n \u003ch2\u003e2.2. Methods\u003c/h2\u003e\n \u003cp\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eSample preparation\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eMortar prisms were cast in accordance with the EN 196-1 standard test, with a w/c ratio of 0.5. For each mixture, one part of cement (450 g) and three parts of sand (1350 g) were used. The mortars were then cured in a saturated lime solution at different temperatures (20\u0026deg;C, 40\u0026deg;C and 60\u0026deg;C) for 28 days. Three mortar prisms with embedded studs at both ends were prepared for expansion tests and one prism was prepared for microstructure tests for each system. Full immersion was carried out in a 50 g/L Na\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e solution, with the solution being refreshed every month to maintain a constant sulfate concentration. Exposure to the sulfate solution took place at a room temperature of 20\u0026thinsp;\u0026plusmn;\u0026thinsp;2\u0026deg;C.\u003c/p\u003e\n \u003cp\u003eTo investigate the phase assemblages of cement paste samples after curing at different temperatures, samples with a w/c ratio of 0.4 were prepared for XRD analysis. Cement paste was chosen to mitigate the dilution effect of high-intensity quartz in XRD and minimize segregation issues. The cement pastes were mixed for two minutes at a speed of 1600 rpm using a high-shear mixer. After mixing, the cement paste was poured into plastic cylinders (33 mm in diameter, 60 mm in length) and left to harden for one day before demolding. The samples were then transferred to slightly larger plastic cylinders and cured with 10 g of distilled water for 28 days. The curing containers were tightly sealed with lids and subjected to different curing temperatures.\u003c/p\u003e\n \u003cp\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eSodium sulfate solution\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eThe external sulfate solution was prepared with a pure chemical salt containing over 99% Na\u003csub\u003e2\u003c/sub\u003eSO\u003csub\u003e4\u003c/sub\u003e solid. A concentration of 50 g/L was used in accordance with the ASTM C1012 standard [\u003cspan class=\"CitationRef\"\u003e26\u003c/span\u003e].\u003c/p\u003e\n \u003cp\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eMercury intrusion porosimetry (MIP)\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eThe porosity of the mortars was analyzed by mercury intrusion porosimetry (MIP) after 28 days of hydration with Porotec GmbH Pascal 140 and 440 instruments. Slices of 2 mm thickness were cut and then immersed in isopropanol for one week to stop the hydration. During this time, the isopropanol was renewed after 1 hour, 1 day and 3 days. Samples were then dried in the vacuum desiccator for two days. The dried slices were broken into pieces of approximately 1.5 g each, and 3\u0026ndash;5 pieces were placed in a glass dilatometer. The intrusion was done with pressure up to 400 MPa and a contact angle of 140˚ was assumed for the analysis.\u003c/p\u003e\n \u003cp\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eXRD Rietveld method with the external standard method\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eThe phase assemblages after 28 days of curing at different temperatures were measured on fresh cement paste (w/c\u0026thinsp;=\u0026thinsp;0.4) slices. X-ray diffraction (XRD) analyses were carried out with a PANalytical X\u0026apos;Pert Pro instrument (CuK\u0026alpha;, \u0026lambda;\u0026thinsp;=\u0026thinsp;1.54 Angstrom) operating with a Bragg\u0026ndash;Brentano geometry over a 2\u0026theta; range of 5\u0026ndash;65\u0026deg;. Quantitative Rietveld analyses were made using the HighScore Plus software and the external standard method, with a phase database adapted from the template provided in [\u003cspan class=\"CitationRef\"\u003e27\u003c/span\u003e]. Titanium dioxide was used as the crystalline external standard (rutile Kronos 2300).\u003c/p\u003e\n \u003cp\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eExpansion measurements\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eThe expansion of each specimen was measured every two weeks, according to the standardized method described in [\u003cspan class=\"CitationRef\"\u003e26\u003c/span\u003e].\u003c/p\u003e\n \u003cp\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eChemical composition of the C-(A-)S-H by SEM-EDS analyses\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eMicrostructural test samples were cut perpendicular to their length to obtain a cross-sectional slice approximately 5 mm thick, as shown in Fig.\u0026nbsp;2. Sampling times were chosen according to expansion; samples were taken both before and after the \u0026ldquo;take-off\u0026rdquo; of expansion. Slices were taken at least 10 mm from the longitudinal ends of the sample to avoid end-surface effects. Subsamples were cut from the slices (see Fig.\u0026nbsp;2d) and impregnated with epoxy resin (see Fig.\u0026nbsp;2e) before being polished using diamond spray suspensions with particles sizes of 9 \u0026micro;m, 3\u0026micro;m and 1 \u0026micro;m (see Fig.\u0026nbsp;2f).\u003c/p\u003e\n \u003cp\u003eEnergy dispersive spectroscopy (EDS) point analysis was performed on the polished section using a FEI Quanta 200 scanning electron microscope equipped with a Bruker 30 EDS detector. Elemental hyper-imaging was also performed to determine the relative concentration of sulfur from the outer surface to the interior.\u003c/p\u003e\n \u003cp\u003eThe accelerating voltage was set to 15.0 kV, with a working distance of 12.5 mm. For the point analysis, the points were carefully selected from the outer C-S-H phase of cement paste in mortar samples. It is important that the operator selects the points as far away as possible from portlandite, large masses of ettringite, and AFm and aggregates. Further details on correctly selecting points from a typical BSE image can be found in [\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e]. For each sample, 200 points were selected at a similar distance from the exposed surface. Figure \u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e shows an example where the outer C-S-H (intermixed with fine AFt and AFm) can be distinguished by its morphology and the grey level from the different phases. Aggregates, large masses of AFm and AFt, clinker, slag, and portlandite are indicated and should be avoided when selecting the phase of interest. The Al/Si ratio of the inner C-(A-)S-H was obtained for each sample in order to estimate the Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e content embedded inside the microstructure (this was not possible using XRD, since the aluminate hydrates are not fully crystalline).\u003c/p\u003e\n \u003cp\u003eFor the hypermaps, the scanned area extended from the sulfate-exposed surface (the starting position) to the inner part (the final position), with sufficient overlap to enable the mappings to be stitched together.\u003c/p\u003e\n \u003cp\u003eAn example of a stitched long BSE image is shown in Fig. 4. Each mapping area measured approximately 1720 \u0026times; 1290 \u0026micro;m (with a nominal magnification of 150) and had a resolution of 1000 \u0026times; 750 pixels (with a pixel size of 1.72 \u0026micro;m). For each sample, a depth of 10 mm was analysed using 10 hypermaps. The Esprit 1.9 software was used to quantify the hypermaps after calibrating them with standards for each element of interest [\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e]. The ten individual maps were stitched together using ImageJ [\u003cspan class=\"CitationRef\"\u003e28\u003c/span\u003e]. The final result is a long BSE map with elemental information over the entire depth. After processing, the maps were quantitatively analyzed using the \u003cem\u003eedxia\u003c/em\u003e approach developed by Georget et al. [\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e]. In Fig. 4 and the other long maps/profiles presented in this paper, the left side shows the solution-exposed surface and the right side shows the inner part.\u003c/p\u003e\n \u003cp\u003eThe degree of reaction of the slag was also measured using SEM-EDS hyperspectral imaging: the slag particles and slag hydrates (both anhydrous and hydrated, the latter being a mixture of hydrotalcite (Ht) and C-(A-)S-H) were segmented using \u003cem\u003eedxia\u003c/em\u003e [\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e], which enabled the volume fraction of reacted slag to be calculated. This percentage was then used alongside mass balance and other inputs (i.e., XRD) to estimate the total Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e content in crystalline aluminous hydrates and in XRD-amorphous phases, either in the C-(A-)S-H form or as microcrystalline monosulfoaluminate.\u003c/p\u003e\n \u003cp\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eMass balance of amorphous and crystalline Al\u003c/span\u003e\u003csub\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e2\u003c/span\u003e\u003c/sub\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eO\u003c/span\u003e\u003csub\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003e3\u003c/span\u003e\u003c/sub\u003e \u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003econtents\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eThe mass balance was used to determine the amorphous and crystalline Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e contents in hydrates after reaction, based on the procedure summarised below [\u003cspan class=\"CitationRef\"\u003e29\u003c/span\u003e]:\u003c/p\u003e\n \u003cp\u003e1. All the reacted Fe\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e was assumed to be bound in the hydrogarnet phase C\u003csub\u003e3\u003c/sub\u003eFS\u003csub\u003e0.84\u003c/sub\u003eH\u003csub\u003e4.32\u003c/sub\u003e [30], enabling the calculation of SiO\u003csub\u003e2\u003c/sub\u003e in this phase;\u003c/p\u003e\n \u003cp\u003e2. All remaining SiO\u003csub\u003e2\u003c/sub\u003e was assumed to be bound in C-(A-)S-H, and the Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e content in the C-(A-)S-H structure was then estimated using the atomic Al/Si ratio obtained via EDS point analysis;\u003c/p\u003e\n \u003cp\u003e3. The total Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e content in the hydrates was estimated based on the degree of hydration of the cement and the degree of degree of slag reaction; \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e4. The X-ray amorphous Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e content was calculated by subtracting the crystalline Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e bound in aluminous hydrates (i.e., ettringite, monosulfoaluminate, hydrotalcite (Ht) and hemi/monocarboaluminate) and in unreacted anhydrous cement and slag from the total initial Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e content;\u003cp\u003e\u003c/p\u003e\n\u003c/div\u003e\u003cp\u003e5. The Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e content in microcrystalline monosulfoaluminate within C-(A-)S-H was obtained by subtracting the Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e in the C-(A-)S-H structure \u0026nbsp;from the total X-ray amorphous Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e content. \u0026nbsp; \u0026nbsp;\u0026nbsp;\u003c/p\u003e"},{"header":"3 Results and discussion","content":"\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\n \u003ch2\u003e3.1 The expansion over time\u003c/h2\u003e\n \u003cp\u003eAs illustrated in Fig. 5, expansion was measured for up to 3.5 years. Significant differences were observed between samples cured at different temperatures. Generally, expansion \u0026apos;takes off\u0026apos; earlier in samples cured at higher temperatures, in both PC and slag-Portland cement mortars. However, the latent period before significant expansion occurs differs greatly between the two systems: expansion began after approximately 56 days in PC60\u0026deg;C but only after around 600 days in SlagC60\u0026deg;C. Furthermore, the effect of the high curing temperature was less pronounced for slag-Portland cement than for PC. Nevertheless, PC mortar cured at 60\u0026deg;C exhibited higher ultimate expansion than PC mortar cured at 20\u0026deg;C in the final measurements. Slag-Portland cement mortars cured at 20\u0026deg;C and 40\u0026deg;C did not show any significant expansion at the time of writing.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\n \u003ch2\u003e3.2 The pore structures\u003c/h2\u003e\n \u003cp\u003eFigure\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e shows a comparison of the pore structure at two curing temperatures: 20\u0026deg;C and 60\u0026deg;C. For each cement system, higher temperature curing results in higher capillary porosity, a shift in the critical pore entry size towards coarser pores, and higher total porosity. For each temperature, PC systems have coarser pores and higher total porosity than slag-Portland cement systems. This explains why sulfate ingress is faster in PC than in slag-Portland cement, as discussed in the next section. Additionally, faster sulfate ingress is anticipated in mortars cured at 60\u0026deg;C compared to 20\u0026deg;C for each cement system. The finer pores in the slag-Portland cement mortar can partially explain the slow reaction kinetics. Finer pores lead to slower sulfate penetration and lower supersaturation with respect to ettringite [\u003cspan class=\"CitationRef\"\u003e31\u003c/span\u003e], particularly in slag-Portland systems with a limited amount of Al\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3\u003c/sub\u003e distributed within the C-S-H, as discussed in Section 3.4.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\n \u003ch2\u003e3.3 The sulfate ingress\u003c/h2\u003e\n \u003cp\u003eFigure 7 compares the S/Ca profiles of PC and slag-Portland cement mortars that were cured at different temperatures and then exposed to a sulfate solution for 28 days (before the expansion \u0026ldquo;take-off\u0026rdquo; for all systems). The profiles were obtained from SEM mapping and show a general decreasing trend in concentration gradients from the exposed surface to an internal depth of 1 mm. More specifically, the sulfate front in the PC mortar cured at 60\u0026deg;C occurs at a depth of ~\u0026thinsp;0.5 mm, whereas in the mortar cured at 20\u0026deg;C, it occurs at a depth of ~\u0026thinsp;0.2 mm. This greater sulfate penetration at 60\u0026deg;C can be explained by the coarser pore structure resulting from the increased curing temperature, as demonstrated by the MIP results shown in Fig. \u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003e.\u003c/p\u003e\n \u003cp\u003eFor slag-Portland cement mortars, the effect of porosity on sulfate ingress is less pronounced than for PC systems. The S/Ca value remains below 0.2, the sulfate front is less steep, and the distinction between the affected and sound zones is minimal. Consequently, slag-Portland mortar cured at 60\u0026deg;C exhibits a sulfate front that is very similar to that cured at 20\u0026deg;C.\u003c/p\u003e\n \u003cp\u003eSlag-Portland and PC mortars cured at 60\u0026deg;C and 20\u0026deg;C respectively show a very similar pore size distribution. However, the PC mortar shows a relatively higher S/Ca level, which cannot be explained solely by the effect of pore structure, but also depends on phase assemblages (i.e., AFm phase content). Nevertheless, the faster sulfate ingress in PC mortars cured at 60\u0026deg;C explains the shorter latent period observed in Fig.\u0026nbsp;5: a larger sulfate-affected zone reduces the ability of the sound bulk mortar to resist expansion forces [\u003cspan class=\"CitationRef\"\u003e32\u003c/span\u003e]. Additionally, a S/Ca ratio of 0.2 appears to be a threshold value for triggering the expansion forces.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\n \u003ch2\u003e3.4 The phase assemblage and the distribution of aluminum\u003c/h2\u003e\n \u003cp\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eThe phase assemblage after curing\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eThe phase assemblages were characterised in cement pastes that were prepared under the same conditions as the mortars and then left to hydrate for 28 days. As illustrated in Fig.\u0026nbsp;8, the mineralogical composition was quantified using the XRD-Rietveld method with an external standard. In both systems, increasing the curing temperature resulted in a lower content of anhydrous phases, indicating a higher degree of hydration. For the PC pastes, the dominant aluminous hydrates present at room temperature were ettringite (AFt), hemicarboaluminate (Hc) and monocarboaluminate (Mc). However, at 60\u0026deg;C, these crystalline phases were largely depleted and were replaced by amorphous aluminum-bearing hydrates intermixed with the C-S-H matrix. In the slag-Portland cement pastes, the alumina-bearing phases were primarily composed of ettringite and hydrotalcite (Ht). The degree of slag reaction, as estimated via SEM\u0026ndash;EDS hypermaps, was used to assess the residual unreacted slag content. Although a reduction in ettringite formation was observed with elevated curing temperatures in both systems, this effect was less pronounced in the slag-containing pastes. These trends are consistent with earlier thermogravimetric (TGA) data [\u003cspan class=\"CitationRef\"\u003e16\u003c/span\u003e] and thermodynamic modelling studies [\u003cspan class=\"CitationRef\"\u003e27\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e33\u003c/span\u003e].\u003c/p\u003e\n \u003cp\u003eMonosulfoaluminate was not detected in the XRD patterns, possibly due to its poor crystallinity when embedded within the C-(A-)S-H matrix. This morphology limits its detectability by conventional XRD analysis [\u003cspan class=\"CitationRef\"\u003e34\u003c/span\u003e]. Although the expansion mechanism appeared similar for samples cured at 20\u0026deg;C and 60\u0026deg;C (see Fig.\u0026nbsp;5), differences in the initial phase assemblages, particularly within the C-S-H, resulted in markedly different expansion kinetics. This suggests that expansion is more dependent on the spatial location and morphology of ettringite formation than its total quantity [\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e, \u003cspan class=\"CitationRef\"\u003e23\u003c/span\u003e].\u003c/p\u003e\n \u003cp\u003eFrom a thermodynamic perspective, the chemical composition of the raw materials provides information on the stability of aluminum-bearing hydrates. In the PC system, the mass ratio of C₃A to SO₃ is 3.7, which exceeds the threshold value of 3.3 required for monosulfoaluminate (Ms) formation. In contrast, the slag-Portland cement system exhibits a ratio of 0.7, which is below the required threshold of 1.1 for monosulfoaluminate stabilization and therefore favours the formation of ettringite as the primary aluminous hydrate.\u003c/p\u003e\n \u003cp\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eRepartition of aluminum between anhydrous phases and hydrates\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eAs detailed in the experimental section, the partitioning of aluminum among the different phases was quantified using a mass balance approach. This method incorporated the chemical composition of the raw materials, phase assemblage data obtained from XRD\u0026ndash;Rietveld analysis and the degree of slag reaction ascertained via SEM\u0026ndash;EDS hypermap segmentation. The degrees of slag hydration were estimated to be 44%, 50% and 58% for curing temperatures of 20\u0026deg;C, 40\u0026deg;C and 60\u0026deg;C, respectively. The total Al₂O₃ content in X-ray amorphous phases was calculated by subtracting the aluminum present in the XRD-detectable crystalline phases from the total alumina content.\u003c/p\u003e\n \u003cp\u003eTwo types of X-ray amorphous aluminous phases were considered: (i) alumina in solid solution within the C-(A-)S\u0026ndash;H structure, and (ii) amorphous AFm phases physically embedded within the C-S-H matrix. The latter are of particular interest as it is hypothesized that their transformation into ettringite under sulfate exposure generates expansive forces. The amorphous AFm content was obtained by subtracting the Al incorporated in the C-(A-)S-H structure from the total amorphous Al₂O₃ content. The outcomes of this partitioning analysis are presented in Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e9\u003c/span\u003e.\u003c/p\u003e\n \u003cp\u003eFollowing hydration, the Al₂O₃ content becomes distributed among various anhydrous constituents (clinker and slag) and hydration products, including the crystalline phase of ettringite and XRD-detectable AFm phases (e.g., Hc, Mc, and Ht), as well as X-ray amorphous AFm phases which are either intermixed with or in solid solution within the C-(A-)S-H matrix. For both binder systems, a greater proportion of Al₂O₃ from the anhydrous phases is incorporated into the hydrates as the curing temperature increases.\u003c/p\u003e\n \u003cp\u003eIn the PC system cured at 60\u0026deg;C, most of the Al₂O₃ is found in X-ray amorphous AFm phases, which are dispersed throughout the C-(A-)S-H. Only a small amount remains as crystalline ettringite. In contrast, although the slag-Portland cement contains a higher total Al₂O₃ content, but much of this is sequestered in hydrotalcite-like (Ht) phases, which are generally considered inert with respect to sulfate-induced transformation. The presence of both Ht and solid-solution alumina in the C-(A-)S-H significantly limits the formation of amorphous AFm phases in the slag system. The reduced content of reactive, amorphous AFm phases may explain the markedly lower expansion observed in slag-Portland cement mortars, as shown in Fig. 5.\u003c/p\u003e\n \u003cp\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eAluminum in amorphous phases vs. expansion\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eFigure\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e10\u003c/span\u003e shows the relationship between expansion behavior and the amount of amorphous Al₂O₃ attributed to AFm phases embedded within the C-(A-)S-H matrix. The onset time of expansion and the subsequent expansion rate both exhibit strong correlations with the quantity of X-ray amorphous AFm. While the absolute values should be interpreted with caution due to the inherent methodological uncertainties inherent in calculating amorphous phase content, consistent relative trends are evident.\u003c/p\u003e\n \u003cp\u003eSystems with higher levels of amorphous AFm phases exhibit significantly shorter latent periods prior to the onset of expansion and a more rapid expansion thereafter (see Fig.\u0026nbsp;5). These findings provide compelling support for the hypothesis that, when confined within the C-(A-)S-H structure, fine-scale, X-ray amorphous AFm phases are the principal drivers of sulfate-induced expansion [\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e].\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e\n \u003ch2\u003e3.5 The spatial distribution of phases\u003c/h2\u003e\n \u003cp\u003e\u003cspan type=\"Underline\" class=\"Underline\" name=\"Emphasis\"\u003eSegmentation of aluminum-bearing phases\u003c/span\u003e\u003c/p\u003e\n \u003cp\u003eCheng\u0026rsquo;s study [\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e] demonstrated that the spatial distribution of aluminous phases has a critical influence on the development of expansion stresses during the transformation of AFm to ettringite. It was concluded that only ettringite that crystallizes within confined spaces \u0026ndash;specifically, in fine pores embedded in the C-(A-)S-H matrix \u0026ndash; generates expansive forces capable of inducing damage [\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e]. Building on this hypothesis, this section compares the spatial distribution of AFm and ettringite phases in PC and slag-Portland cement systems.\u003c/p\u003e\n \u003cp\u003eFigure\u0026nbsp;11 shows an example of the segmentation of aluminum-bearing phases in slag\u0026ndash;Portland cement paste, as determined using chemical ratio plots and the edxia framework [\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e]. Phase identification was verified through correlation with grey-level contrast in BSE micrographs. The presence of hydrotalcite (Ht) is attributed to the magnesium content of the reacted slag grains. A small quantity of monosulfoaluminate, accompanied by minor Mc/Hc, was also detected via SEM\u0026ndash;EDS analysis.\u003c/p\u003e\n \u003cp\u003eThese phases were found to be finely intermixed within the C-(A-)S-H matrix. Specifically, Ht (highlighted in orange) forms around the periphery of reacted slag grains, while crystalline AFm phases (cyan) appear as discrete, larger domains that are identifiable in the elemental maps. The C-(A-)S-H phase (purple) acts as a continuous binding matrix that incorporates the finely dispersed AFm and Ht phases throughout the microstructure.\u003c/p\u003e\n \u003cdiv\u003e\n \u003cdiv align=\"left\" class=\"colspec\"\u003e\u003cbr\u003e\u003cu\u003eDistribution of aluminum-bearing phases\u003c/u\u003e\u003c/div\u003e\n \u003c/div\u003e\n \u003cp\u003eFigure\u0026nbsp;12 shows the spatial distribution of the AFm/C-(A-)S-H and Ht/C-(A-)S-H phases in both binder systems after 28 days of hydration at 20\u0026deg;C and 60\u0026deg;C. The aluminum-rich phases were segmented using the edxia framework [\u003cspan class=\"CitationRef\"\u003e25\u003c/span\u003e] and overlaid onto the corresponding BSE micrographs. An increase in AFm content was observed in both the PC and slag\u0026ndash;Portland cement systems at 60\u0026deg;C, which is consistent with the XRD\u0026ndash;Rietveld quantification (see Fig.\u0026nbsp;8).\u003c/p\u003e\n \u003cp\u003eInterestingly, the PC system cured at 20\u0026deg;C exhibited a higher AFm content than the slag\u0026ndash;Portland system cured at 60\u0026deg;C. In slag\u0026ndash;Portland cement hydrated at 20\u0026deg;C, the AFm phase identified by SEM\u0026ndash;EDS appeared to be a composite of monosulfoaluminate and C-(A-)S-H, along with a minor fraction of Hc/Mc. The phases were not detected by XRD, due to their low crystallinity or limited abundance. At 60\u0026deg;C, the increase in total aluminum-bearing hydrates in the slag system was attributed to AFm formation, as the Ht/C-(A-)S-H content remained largely unchanged across both curing temperatures.\u003c/p\u003e\n \u003cdiv\u003eFurther analysis using high-magnification SEM imaging (4000\u0026times;) reveals distinct morphological differences in the distribution of the AFm phase between the two binder systems. In PC pastes, AFm is predominantly localised around clinker grains within the inner C-S-H product. While these phases can be detected by SEM, their encapsulation within the inner C-S-H suggests that they are not mechanically active and therefore do not contribute to expansion. In contrast, AFm in slag\u0026ndash;Portland cement pastes appears as large, discrete \u0026ldquo;pockets\u0026rdquo; within the matrix. Previous studies have shown that such morphologies are non-expansive, even when these AFm phases subsequently convert to ettringite upon exposure to sulfate [\u003cspan class=\"CitationRef\"\u003e22\u003c/span\u003e].\u003c/div\u003e\n\u003c/div\u003e"},{"header":"4 Conclusions","content":"\u003cp\u003eThis study highlights the impact of increased curing temperatures on the sulfate resistance of Portland and slag\u0026ndash;Portland cement mortars when fully immersed. Based on the experimental findings, the following conclusions can be drawn:\u003c/p\u003e\u003cp\u003e\u003cul\u003e\u003cli\u003e\u003cp\u003eMortars cured at 60\u0026deg;C exhibit significantly higher expansion compared to those cured at 20\u0026deg;C. However, slag\u0026ndash;Portland cement systems demonstrate much lower expansion overall, showing reduced sensitivity to curing temperature compared to PC systems.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eElevated curing temperatures result in a more porous microstructure characterized by coarser capillary pores, facilitating faster sulfate ingress than in mortars cured at ambient conditions.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eIn PC systems, curing at 40\u0026deg;C and 60\u0026deg;C destabilizes the ettringite and Mc/Hc phases typically present at 20\u0026deg;C, promoting their replacement by X-ray amorphous AFm phases intermixed within the C-S-H. In slag\u0026ndash;Portland pastes, ettringite and hydrotalcite remain the dominant alumina-bearing hydrates, with the latter showing little variation with temperature.\u003c/p\u003e\u003c/li\u003e\u003cli\u003e\u003cp\u003eThe X-ray amorphous AFm phases confined within the C-(A-)S-H matrix are identified as the main cause of expansion. Their higher abundance in PC systems cured at 60\u0026deg;C correlates with increased expansion, whereas their limited presence in slag\u0026ndash;Portland systems accounts for the substantially lower expansion observed.\u003c/p\u003e\u003c/li\u003e\u003c/ul\u003e\u003c/p\u003e\u003cp\u003eMore broadly, the expansive phase was identified as amorphous Al₂O₃ embedded in the C-(A-)S-H, and its content was found to correlate with both the onset and rate of expansion. The curing temperature influences both the phase assemblage and pore structure, primarily by reducing the stability of ettringite and increasing the formation of amorphous aluminates, thereby altering sulfate resistance. Furthermore, mass balance analysis of Al₂O₃ partitioning provided key insights into the mechanisms underlying the enhanced sulfate resistance of slag\u0026ndash;Portland systems, despite their higher total aluminum content.\u003c/p\u003e\u003cp\u003eThese results suggest that the amorphous Al₂O₃ content within C-(A-)S-H may serve as a potential indicator of sulfate resistance. Future work should focus on establishing quantitative thresholds for this parameter in relation to expansion kinetics and durability performance.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eJan Skalny, Jacques Marchand, Ivan Odler (2002) Sulfate Attack on Concrete. London and New York\u003c/li\u003e\n\u003cli\u003eChitvoranund N (2021) Stability of hydrate assemblages and properties of cementitious systems with higher alumina content. EPFL\u003c/li\u003e\n\u003cli\u003eKaufmann J, Winnefeld F, Lothenbach B (2016) Stability of ettringite in CSA cement at elevated temperatures. Advances in Cement Research 28:251\u0026ndash;261. https://doi.org/10.1680/jadcr.15.00029\u003c/li\u003e\n\u003cli\u003eGonz\u0026aacute;lez MA, Irassar EF (1997) Ettringite formation in low C3A Portland cement exposed to sodium sulfate solution. Cement and Concrete Research 27:1061\u0026ndash;1071. https://doi.org/10.1016/S0008-8846(97)00093-8\u003c/li\u003e\n\u003cli\u003eWang JG (1994) Sulfate attack on hardened cement paste. Cement and Concrete Research 24:735\u0026ndash;742. https://doi.org/10.1016/0008-8846(94)90199-6\u003c/li\u003e\n\u003cli\u003eHossack AM, Thomas MDA (2015) The effect of temperature on the rate of sulfate attack of Portland cement blended mortars in Na2SO4 solution. Cement and Concrete Research 73:136\u0026ndash;142. https://doi.org/10.1016/j.cemconres.2015.02.024\u003c/li\u003e\n\u003cli\u003eThaumasite formation in concrete and mortars containing fly ash. Cement and Concrete Composites 25:907\u0026ndash;912. https://doi.org/10.1016/S0958-9465(03)00136-7\u003c/li\u003e\n\u003cli\u003eSchmidt T, Lothenbach B, Romer M, et al (2009) Physical and microstructural aspects of sulfate attack on ordinary and limestone blended Portland cements. Cement and Concrete Research 39:1111\u0026ndash;1121. https://doi.org/10.1016/j.cemconres.2009.08.005\u003c/li\u003e\n\u003cli\u003eSanthanam M, Cohen MD, Olek J (2002) Modeling the effects of solution temperature and concentration during sulfate attack on cement mortars. Cement and Concrete Research 32:585\u0026ndash;592. https://doi.org/10.1016/S0008-8846(01)00727-X\u003c/li\u003e\n\u003cli\u003eAk\u0026ouml;z F, T\u0026uuml;rker F, Koral S, Y\u0026uuml;zer N (1999) Effects of raised temperature of sulfate solutions on the sulfate resistance of mortars with and without silica fume. Cement and Concrete Research 29:537\u0026ndash;544. https://doi.org/10.1016/S0008-8846(98)00251-8\u003c/li\u003e\n\u003cli\u003eInfluence of initial curing on sulphate resistance of blended cement concrete. Cement and Concrete Research 22:1089\u0026ndash;1100. https://doi.org/10.1016/0008-8846(92)90039-X\u003c/li\u003e\n\u003cli\u003eLawrence CD (1995) Mortar expansions due to delayed ettringite formation. Effects of curing period and temperature. Cement and Concrete Research 25:903\u0026ndash;914. https://doi.org/10.1016/0008-8846(95)00081-M\u003c/li\u003e\n\u003cli\u003eTaylor HFW, Famy C, Scrivener KL (2001) Delayed ettringite formation. Cement and Concrete Research 31:683\u0026ndash;693. https://doi.org/10.1016/S0008-8846(01)00466-5\u003c/li\u003e\n\u003cli\u003ePerkins RB, Palmer CD (1999) Solubility of ettringite (Ca6[Al(OH)6]2(SO4)3 \u0026middot; 26H2O) at 5\u0026ndash;75\u0026deg;C. Geochimica et Cosmochimica Acta 63:1969\u0026ndash;1980. https://doi.org/10.1016/S0016-7037(99)00078-2\u003c/li\u003e\n\u003cli\u003eHaynes H (2002) Sulfate Attack on Concrete: Laboratory vs. Field Experience. CI 24:64\u0026ndash;70\u003c/li\u003e\n\u003cli\u003eLothenbach B, Winnefeld F, Alder C, et al (2007) Effect of temperature on the pore solution, microstructure and hydration products of Portland cement pastes. Cement and Concrete Research 37:483\u0026ndash;491. https://doi.org/10.1016/j.cemconres.2006.11.016\u003c/li\u003e\n\u003cli\u003eZhang Z, Zou Y, Yang J, Zhou J (2022) Capillary rise height of sulfate in Portland-limestone cement concrete under physical attack: Experimental and modelling investigation. Cement and Concrete Composites 125:104299. https://doi.org/10.1016/j.cemconcomp.2021.104299\u003c/li\u003e\n\u003cli\u003eThe microstructure of concrete cured at elevated temperatures. Cement and Concrete Research 25:485\u0026ndash;490. https://doi.org/10.1016/0008-8846(95)00036-C\u003c/li\u003e\n\u003cli\u003eThomas JJ, Rothstein D, Jennings HM, Christensen BJ (2003) Effect of hydration temperature on the solubility behavior of Ca-, S-, Al-, and Si-bearing solid phases in Portland cement pastes. Cement and Concrete Research 33:2037\u0026ndash;2047. https://doi.org/10.1016/S0008-8846(03)00224-2\u003c/li\u003e\n\u003cli\u003eGollop R, Taylor H (1995) Microstructural and microanalytical studies of sulfate attack III. Sulfate-resisting portland cement: Reactions with sodium and magnesium sulfate solutions. https://doi.org/10.1016/0008-8846(95)00151-2\u003c/li\u003e\n\u003cli\u003eElahi MMA, Shearer CR, Naser Rashid Reza A, et al (2021) Improving the sulfate attack resistance of concrete by using supplementary cementitious materials (SCMs): A review. Construction and Building Materials 281:122628. https://doi.org/10.1016/j.conbuildmat.2021.122628\u003c/li\u003e\n\u003cli\u003eYu C, Sun W, Scrivener K (2013) Mechanism of expansion of mortars immersed in sodium sulfate solutions. Cement and Concrete Research 43:105\u0026ndash;111. https://doi.org/10.1016/j.cemconres.2012.10.001\u003c/li\u003e\n\u003cli\u003eWang Q, Wilson W, Scrivener K (2023) Unidirectional penetration approach for characterizing sulfate attack mechanisms on cement mortars and pastes. Cement and Concrete Research 169:107166. https://doi.org/10.1016/j.cemconres.2023.107166\u003c/li\u003e\n\u003cli\u003eHuang Q, Wang Q, Zhu X (2024) Contradict mechanism of long-term magnesium and sodium sulfate attacks of nano silica-modified cement mortars \u0026ndash; Experimental and thermodynamic modeling. Cement and Concrete Composites 147:105444. https://doi.org/10.1016/j.cemconcomp.2024.105444\u003c/li\u003e\n\u003cli\u003eGeorget F, Wilson W, Scrivener KL (2021) edxia: Microstructure characterisation from quantified SEM-EDS hypermaps. Cement and Concrete Research 141:106327. https://doi.org/10.1016/j.cemconres.2020.106327\u003c/li\u003e\n\u003cli\u003eStandard Test Method for Length Change of Hydraulic-Cement Mortars Exposed to a Sulfate Solution. ASTM C1012-18a\u003c/li\u003e\n\u003cli\u003eWilson W, Gonthier JN, Georget F, Scrivener KL (2022) Insights on chemical and physical chloride binding in blended cement pastes. Cement and Concrete Research 156:106747. https://doi.org/10.1016/j.cemconres.2022.106747\u003c/li\u003e\n\u003cli\u003ePreibisch S, Saalfeld S, Tomancak P (2009) Globally optimal stitching of tiled 3D microscopic image acquisitions. Bioinformatics 25:1463\u0026ndash;1465. https://doi.org/10.1093/bioinformatics/btp184\u003c/li\u003e\n\u003cli\u003eDurdziński PT, Ben Haha M, Zajac M, Scrivener KL (2017) Phase assemblage of composite cements. Cement and Concrete Research 99:172\u0026ndash;182. https://doi.org/10.1016/j.cemconres.2017.05.009\u003c/li\u003e\n\u003cli\u003eDilnesa BZ, Wieland E, Lothenbach B, et al (2014) Fe-containing phases in hydrated cements. Cement and Concrete Research 58:45\u0026ndash;55. https://doi.org/10.1016/j.cemconres.2013.12.012\u003c/li\u003e\n\u003cli\u003eScherer GW (2004) Stress from crystallization of salt. Cement and Concrete Research 34:1613\u0026ndash;1624. https://doi.org/10.1016/j.cemconres.2003.12.034\u003c/li\u003e\n\u003cli\u003eKunther W (2012) Investigation of Sulfate Attack by Experimental and Thermodynamic Means. PhD Thesis, EPFL\u003c/li\u003e\n\u003cli\u003eLothenbach B, Matschei T, M\u0026ouml;schner G, Glasser FP (2008) Thermodynamic modelling of the effect of temperature on the hydration and porosity of Portland cement. Cement and Concrete Research 38:1\u0026ndash;18. https://doi.org/10.1016/j.cemconres.2007.08.017\u003c/li\u003e\n\u003cli\u003eScrivener K, Snellings R, Lothenbach B (2017) A Practical Guide to Microstructural Analysis of Cementitious Materials. CRC Press, Boca Raton\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[{"identity":"e9fb9da4-4838-482a-af15-03ae006556c8","identifier":"10.13039/501100004543","name":"China Scholarship Council","awardNumber":" 201806050061","order_by":0}],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"École Polytechnique Fédérale de Lausanne","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Sulfate attack, Slag-Portland cement, Expansion, Elevated temperature curing, C-(A-)S-H composition","lastPublishedDoi":"10.21203/rs.3.rs-7347409/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7347409/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThis study aims to understand the effects of the curing temperature on the phase assemblage and the distribution of aluminum-bearing hydrates (AFm and AFt), and how these affect external sulfate attack. Mortars were prepared with Portland cement (PC) and slag-Portland cement at a water-to-cement (w/c) ratio of 0.5. Specimens were cured at 20\u0026deg;C, 40\u0026deg;C and 60\u0026deg;C for 28 days prior to full immersion in sodium sulfate solutions at 50 g/L. The results showed that curing at higher temperatures shortened the latent period before expansion in both PC and slag systems. High-temperature curing altered both the pore structure and the phases embedded in the C-(A-)S-H matrix, leading to more expansion and degradation. Expansion occurred much later and to a lesser extent in slag-Portland mortars due to lower contents of fine monosulfoaluminate (X-ray amorphous) in the C-(A-)S-H, compared to Portland cement mortars.\u003c/p\u003e","manuscriptTitle":"Effects of curing temperature on sulfate-induced expansion of cement mortars","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-12 10:37:34","doi":"10.21203/rs.3.rs-7347409/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"35ae79f5-f05b-43ae-8453-fe8fd34cc1eb","owner":[],"postedDate":"August 12th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":52987759,"name":"Cement Chemistry"},{"id":52987760,"name":"Materials Engineering"}],"tags":[],"updatedAt":"2025-08-12T10:37:35+00:00","versionOfRecord":[],"versionCreatedAt":"2025-08-12 10:37:34","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-7347409","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7347409","identity":"rs-7347409","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}