Sulphosilicate clinker is considered as a low-carbon cementitious material due to its high sulfur/silicon contents, low sintering temperature and low carbon emissions. Despite its promising performance and environmental advantages, the feasibility and mechanism of incorporating supplementary cementitious materials such as slag powder into sulphosilicate cement remain unclear. This study was to investigate the effects of slag powder and quicklime into sulphosilicate cement on its hydration behavior and mechanical performance.

The sulphosilicate clinker primarily comprised Belite (C₂S), Yeelimite (C₄A₃$) and Ternesite (C₅S₂$), with minor amounts of calcium sulfate and ferrite phases. Four types of cement were prepared, i.e., the reference (i.e., sample A-1), 30% slag powder (i.e., sample A-2), 30% slag powder with 1.5% quicklime (i.e., sample A-3) and 30% slag powder with 3% quicklime (i.e., sample A-4). The cement paste was analyzed for pH value, setting time, strength and hydration production at different ages of 3, 7 d and 28 d. The hydration products, microstructure evolution and thermodynamic behavior were determined by X-ray diffraction (XRD), thermogravimetric and differential thermal analysis (TG-DTG), scanning electron microscopy (SEM) and isothermal calorimetry, respectively.

The results indicate that the addition of 30% slag powder (sample A-2) extends the initial setting time by 12 min and improves workability (9.5% increase in flow), but significantly reduces early compressive strength due to the dilution effect thus diminishing active mineral content and delaying hydration product formation. The sample A-2 exhibits a 40% and 38% reduction in compressive strength at 3 d and 28 d, respectively, compared to the reference. The incorporation of quicklime significantly changes pH value in early hydration stages and promotes a pozzolanic reactivity of the slag. Compared to the sample A-2, the sample A-3 achieves a 35% increase in 28 d compressive strength (i.e., 46.3 MPa) meets the 42.5-grade requirement. However, the sample A-4 exhibits less compact microstructure and diminishes strength development, compared to the sample A-3, indicating that excessive quicklime can hinder mineral reactivity and hydration kinetics.

The analysis of isothermal calorimetry shows distinct hydration profiles among the cement systems. The samples A-1 and A-2 display two distinct peaks (initial wetting and C₄A₃$ hydration), while the samples A-3 and A-4 show only one sharp early peak due to the exothermic quicklime hydration, with later stages subdued due to an inhibited mineral reactivity. The XRD patterns reveal that the ettringite content in the sample A-3 increases from 32% at 3 d to 51% at 28 d, which is substantially higher than that of the sample A-2 (i.e., from 26% to 42%). In contrast, the sample A-4 plateaues 33% ettringite at 28 d. The TG-DTG analysis and bound water quantification reinforce these findings, showing that the sample A-3 has the maximum cumulative increase in hydration products over time among the blended systems. The SEM images confirm that the early formation of ettringite is suppressed, and the resulting ettringite transforms from dense columnar crystals in the sample A-1 to finer needle-like structures in the sample A-2, indicating a weaker skeletal framework and less compact microstructure. The sample A-3 develops a denser network of interlocked hydration products, while the sample A-4 shows persistent voids and unreacted particles, illustrating a delicate balance required for optimal alkaline activation.

This study demonstrated that the combination of slag powder and quicklime content could synergistically improve the mechanical performance and sustainability of sulphosilicate cement. At a 30% replacement level, slag powder significantly extended the initial setting time. However, the dilution effect led to a reduction in active mineral content, causing the amount of ettringite at 3 d and 28 d to decrease by 23% and 17%, respectively. This corresponded to a decline in compressive strength by 40% at 3 d and 38% at 28 d. The microstructural analysis revealed that the addition of slag powder could cause the morphology of ettringite to transform from columnar to needle-like/rod-like (with an aspect ratio > 10), and a higher residue of unhydrated slag particles led to a more porous microstructure. An appropriate amount of quicklime enhanced the alkalinity and properties of sulphosilicate cement. At a dosage of 1.5%, quicklime promoted slag hydration and increased the strength of the samples. The ettringite content in these samples was increased by 6% at 3 d and 10% at 28 d, while the chemically bound water content was increased by 6.1% and 4.5%, respectively. Consequently, the 28-d compressive strength was increased by 35%. In addition, distinct etching marks appeared on the slag surfaces along with the formation of a small amount of needle-like ettringite in 7 d samples with 1.5% quicklime, enhancing the density of the composite system. Conversely, a 3% quicklime dosage significantly inhibited the hydration of C₄A₃$ and C₅S₂$. The resulting excessive Ca(OH)₂ delayed the second hydration heat release peak to 40 h and reduced the total cumulative heat to 116.76 J/g.

The combined use of slag powder and quicklime profoundly could affect sulphosilicate cement performance. Incorporating 1.5% quicklime, 30% slag powder and an appropriate amount of anhydrite into the sulphosilicate clinker could allow for the preparation of cement that meeted the 42.5 strength grade. However, when slag was used alone, its activity was difficult to be activated due to the insufficient alkalinity of the cement liquid phase. Thus, it was not recommended as a standalone mineral admixture for this type of cement. Quicklime could significantly improve alkalinity, the performance of the cement was highly sensitive to its dosage. The optimal quicklime content could be determined based on the specific mineral composition of the clinker or the final cement blend.