Early-Stage Hydration and Product Evolution in Calcium Hydroxide-Activated Lithium Slag

Abstract

This study used calcium hydroxide (CH) to simulate the alkaline environment of cement and to activate lithium slag (LS), aiming to reveal the mechanism of LS in cement. The early-age hydration of LS blended with 10 wt.% CH was monitored via isothermal calorimetry (ICC) at ambient temperature, followed by a comparative analysis of phase assemblage, microstructure, and macroscopic properties under standard and steam curing conditions. The results show that LS exhibits superior early reactivity within the first 9 h, which is attributed to abundant ettringite formation. Two distinct exothermic peaks were identified during LS-CH hydration, corresponding to (i) ettringite formation accompanied by LS dissolution and C–S–H precipitation, and (ii) CaCO₃ crystallization and renewed ettringite formation. The hydrated paste consists of abundant AFt, CaCO₃ polymorphs, unreacted LS particles, and a small amount of C–S–H gel with a low Ca/Si ratio and incorporating Al and S. This unique phase assemblage results in a coarser pore structure and lower specific surface area compared with conventional cement paste. Nevertheless, the system achieves a relatively high 28-day compressive strength, highlighting the promise of LS-CH blends as sustainable cementitious materials.

1. Introduction

Lithium slag (LS) is a byproduct generated from lithium carbonate production via the sulfuric acid method [1], and its accumulation is directly linked to the rapid expansion of the lithium battery industry. In China, the remarkable growth of the new energy vehicle sector has led to a rapid increase in lithium salt production, inevitably resulting in a steady rise in LS emissions, which currently amount to approximately 6 million tons annually [2,3]. This massive accumulation not only occupies farmland but also poses a significant threat to the environment.

Previous studies have shown that the mineral composition of LS includes leached spodumene, quartz, gypsum, and carbonate [4]; and its chemical composition is dominated by SiO₂ (48.6~63.1%), Al₂O₃ (14.0~20.7%), and SO₃ (4.5~9.3%), along with small amounts of CaO (3.6~12.1%) [5,6,7]. Although LS contains high levels of SiO₂ and Al₂O₃, these components are primarily present in the form of leached spodumene rather than in an amorphous phase [8].

The high Al₂O₃ and SiO₂ contents in lithium slag enable its use in preparing zeolite X and ceramic glazed tiles [6,9]. On the other hand, its high gypsum content makes it a high-quality raw material for producing gypsum products. In addition, Ding et al. [10] noted that LS is effective in mitigating expansion induced by the alkali-silica reaction (ASR). However, incorporating LS powder into concrete may lead to micro-expansion due to ettringite formation. Therefore, this micro-expansion should be isolated when evaluating the ASR-inhibiting effect of LS powder. He et al. [11] demonstrated that mine filling materials with excellent unconfined compressive strength can be prepared using cement, lithium slag, and fly ash at a mass ratio of 2:1:1. They further reported that the addition of NaOH can improve the pozzolanic activity of lithium slag in mine backfill binders [12]. Furthermore, LS has been used as a white pigment [13,14] and in the production of lightweight aggregates [15]. Despite its applications in various fields, the utilization rate of lithium slag remains low. Therefore, exploring efficient methods to improve the utilization of LS is of great significance.

In recent years, the use of supplementary cementitious materials (SCMs) in concrete has gained increasing attention owing to their economic benefits, environmental advantages, and technological improvements [16]. LS contains large amounts of SiO₂ and Al₂O₃, which impart pozzolanic activity; thus, it also represents a promising candidate for use as an SCM [17]. However, like other SCMs, LS hinders the early strength development rate of cementitious systems. Zhang [5] reported that a higher LS content leads to greater deterioration in the early strength of concrete. To improve the reactivity of LS, Liu combined alkali activation with thermal treatment, using this method, the 1-day compressive strength of alkali-activated LS was increased from 12.8 MPa to 47.7 MPa [18]. Tan et al. [19,20] found that wet-grinding yields much finer LS particles than dry-grinding, thereby accelerating the formation of hydration products in concrete incorporating LS as an SCM. Previous work by the authors demonstrated that steam curing can significantly improve the demolding strength of LS-blended cement mortar [21,22]. However, the above-mentioned activation methods have issues in terms of environmental friendliness, energy consumption, and cost. Specifically, the strong alkali activation method not only causes severe corrosion to equipment but also increases construction operation risks and material costs. Meanwhile, the drying step after wet-grinding and the steam curing process incur high energy costs.

Although the incorporation of LS can hinder early strength development, it enhances the early crack resistance of concrete by forming more fibrous ettringite [23]. Moreover, when the replacement level is below 20%, LS can improve the long-term (≥28 d) compressive strength of concrete. Li et al. [8,24] reported that incorporating LS can enhance the durability of concrete, including sulfate resistance, elastic modulus, drying shrinkage, and creep behavior. He et al. [25] showed that replacing silica fume with LS can increase the hydration degree of ultra-high-performance concrete. However, at high LS replacement ratios, the high gypsum content in LS may lead to a risk of delayed ettringite formation in concrete. Notably, early-age steam curing has been proven effective in addressing this issue [8].

Not surprisingly, the incorporation of LS reduces the Ca/Si ratio of C–S–H gels. Meanwhile, our study identified the presence of equant grain-shaped C–S–H gels and cubic CaCO₃ in lithium slag-blended cement [22]. Beyond these observations, little is known about the characteristics of hydration products in lithium slag-blended cement. Furthermore, the influence of the high sulfate and carbonate contents in LS on cement hydration remains unclear.

Calcium hydroxide (CH) is one of the main hydration products in cement and the primary reactant for SCMs in concrete. Therefore, to explore the hydration mechanism of SCMs in concrete, CH is often used to simulate the alkaline environment in Portland cement pastes. This approach allows the direct investigation of the reaction between CH and SCMs, providing insight into the hydration process and products of SCMs in concrete. However, unlike common SCMs such as ground granulated blast-furnace slag (GBFS) and fly ash (FA), LS contains carbonates and a higher SO₃ content. The reaction mechanism between lithium slag and CH remains unclear.

Thus, this study first investigates the hydration process of LS blended with 10 wt.% Ca(OH)₂ (LS-CH) using isothermal conduction calorimetry (ICC) and X-ray diffraction (XRD). It then characterizes the hydration products of LS-CH pastes under standard curing conditions using XRD, scanning electron microscopy with energy-dispersive spectroscopy (SEM–EDS), thermogravimetric analysis (TG–DTG), Fourier transform infrared spectroscopy (FTIR), and Brunauer–Emmett–Teller (BET) analysis, while their mechanical properties are evaluated through strength tests. Given the increasing adoption of precast concrete construction, this study further examines the hydration products and mechanical properties of LS-CH pastes under steam curing.

Based on the experimental investigation, this study provides a comprehensive understanding of the hydration behavior of the LS-CH system, with detailed descriptions of its microstructural evolution and phase development. These findings clarify the role of lithium slag in cementitious matrices and validate its potential as a sustainable, low-carbon construction material.

2. Results and Discussion

2.1. Results of Material Characterization

As shown in Figure 1, the lithium slag (LS) powder was yellowish-white in appearance. Its chemical composition, analyzed by X-ray fluorescence (XRF), is presented in

Table 1. The chemical compositions of LS, GBFS, FA and Portland cement (PC), wt.%.

Materials	CaO	SiO2	Al2O3	SO3	Fe2O3	MgO	P2O5	Na2O
LS	4.53	62.40	22.1	6.73	1.06	0.49	1.12	0.89
GBFS	36.05	34.67	16.52	2.53	0.29	3.90	4.95	-
FA	6.56	54.71	26.74	1.42	5.00	1.50	1.44	-
PC	64.47	20.87	4.87	2.52	3.59	2.13	-	0.11

. Like ground granulated blast-furnace slag (GBFS) and fly ash (FA), LS is rich in SiO₂ and Al₂O₃, with the main distinction being its higher SO₃ content. SiO₂, Al₂O₃, and SO₃ together account for 89.03 wt.% of LS by mass. Other minor components include CaO, P₂O₅, K₂O, and Fe₂O₃. Notably, the CaO content in LS is relatively low, even lower than that in FA. Accordingly, based on the data in Table 1, the molar Ca/Si ratio of LS is merely 0.08.

LS has a density of approximately 2.50 g/cm³, which is comparable to that of FA. For a LS-water suspension with a water-to-LS mass ratio of 10:1, the pH value reached 7.5 after 4 h of mixing, whereas that of tap water was 7.9. Therefore, LS can be classified as an acidic slag.

The XRD results (Figure 2) indicated that LS was primarily composed of 7.9% quartz (SiO₂), 66.2% leached spodumene (LiAlSi₂O₆), 13.3% gypsum (CaSO₄·2H₂O), and 12.6% amorphous phase, as calculated via the XRD/Rietveld method.

Figure 3 illustrates the morphology of LS particles, which exhibit an angular, distinct crystalline structure. The atomic composition of individual particles, determined by EDS results, is listed in

Table 2. Atomic composition of particles 1~5 in Figure 3, at.%.

Particles	Ca	Si	Al	O	S	Mg
1	0.61	32.21	13.91	52.69	0.58
2	0.47	26.00	11.58	61.96
3		24.38	11.30	63.83	0.49
4	15.43	13.55	7.13	58.73	1.30	3.86
5	9.79	9.89	5.40	66.88	8.05

. Smaller particles (Particles 1, 2, and 3) were predominantly LiAlSi₂O₆, whereas larger particles (Particles 4 and 5) were mainly CaSO₄·2H₂O. This can be attributed to the fact that LiAlSi₂O₆ is readily crushed during grinding [26], while gypsum tends to agglomerate owing to its higher moisture content and viscosity.

Figure 4 presents the particle size distribution of LS powder, measured using a laser particle size analyzer. The particle size of LS powder ranged from 0.6 μm to 296 μm, with an average particle size of 10.45 μm. This value was slightly larger than those of FA (7.43 μm) and GBFS (GBFS, 6.33 μm) used in this study, yet smaller than that of PC (16.70 μm).

Thermogravimetric (TG/DTG) analysis (Figure 5) indicated that, in addition to gypsum, lithium slag contained carbonates, with a loss on ignition of 8.2% below 1300 °C [27,28,29]. The high carbonate content in LS mainly exists in the forms of CaCO₃, Na₂CO₃, and Li₂CO₃, whereas conventional SCMs contain very little carbonate, especially Na₂CO₃ and Li₂CO₃.

The FTIR results (Figure 6) further confirmed the presence of sulfates and carbonates in lithium slag. As reported in our previous study [8,22], these sulfates mainly exist in the forms of gypsum and sodium sulfate.

The 28-day pozzolanic activity of LS was 92%, which was slightly lower than that of GBFS (98%) but higher than that of FA (76%).

2.2. Hydration Heat Evolution of LS-CH Pastes

The cumulative hydration heat and heat flow of LS-CH pastes with a water-to-binder ratio of 0.3 at 20 °C are presented in Figure 7. To clarify the hydration process of LS, the hydration heat results of GBFS-CH and FA-CH (prepared with the same mix proportion as LS-CH) are also included in Figure 7. As observed, the 72 h cumulative hydration heat of GBFS-CH was the highest at 87.2 J/g, followed by LS-CH at 60.30 J/g, while FA-CH exhibited the lowest value at 18.51 J/g. This indicates that the reactivity of LS is significantly higher than that of FA but slightly lower than that of GBFS within 72 h. However, it is noteworthy that the cumulative hydration heat of LS-CH during the first 9 h was the highest, demonstrating that LS possesses the highest hydration reactivity in this early period.

FA is generally regarded as an inert material at very early ages, particularly within the 0–24 h period, and its influence on cement hydration is primarily attributed to the filling effect and the provision of nucleation and precipitation sites for C–S–H gels [30]. Consequently, the hydration exothermic rate of the FA-CH mixture fell below 1 J/(g·h), with only several minor exothermic peaks observed within 30 h, as depicted in Figure 7b. In contrast to FA-CH, both GBFS-CH and LS-CH exhibited relatively high hydration exothermic rates, especially within the first 24 h. Furthermore, the hydration process of the GBFS-CH paste featured a significantly larger and broader exothermic peak, which initiated at approximately 0–1 h and terminated at roughly 48 h. This phenomenon is likely due to the formation of C–S–H gels [31].

However, two distinct exothermic peaks were observed during the hydration heat release of LS-CH pastes. The first exothermic peak was significantly higher than those of both GBFS-CH and FA-CH, whereas the second exothermic peak was relatively small: it was similar to that of FA-CH at approximately 30 h, but much lower than that of GBFS-CH at around 9 h. Notably, the hydration exothermic rate of LS-CH exceeded that of GBFS-CH again at 44 h. Thereafter, the two curves followed a similar trend after 60 h of hydration.

To further investigate the underlying causes of the two exothermic peaks formed during the hydration of LS-CH paste, XRD analysis was performed on its hydration products cured at 20 °C for 2 h, 5 h, 40 h, and 50 h in a sealed centrifuge tube. The XRD patterns were collected over a 2θ range of 5° to 55°, with the results presented in Figure 8.

As shown in Figure 8, aside from AFt, no obvious crystalline phases were generated in the LS-CH paste between 2 h and 5 h. This suggests that the first exothermic peak in Figure 7b is primarily attributed to the formation of ettringite, together with the heat released from wetting and dissolution. Notably, the final setting time of the LS-CH paste was determined to be approximately 2 h. This indicates that the C–S–H gels formed by the reaction between LS and CH within 2–5 h further contributed to the hydration heat corresponding to the first exothermic peak.

As illustrated in Figure 7b, a comparison of the hydration heat flow curves of LS-CH and GBFS-CH before 3.9 h shows that, despite GBFS having a higher amorphous phase content than LS, the heat flow of the GBFS-CH system during this period is considerably lower than that of LS-CH. Based on the above analysis, the heat released by GBFS-CH before 3.9 h is mainly attributable to C–S–H formation. Given the lower amorphous content in lithium slag, the thermal contribution from C–S–H formation in the LS-CH system would be correspondingly smaller at the same hydration age. It can therefore be inferred that C–S–H formation is not the primary mechanism responsible for the first exothermic peak observed in LS-CH. In contrast, LS contains a substantially higher SO₃ content. These results suggest that the first exothermic peak in the LS-CH system arises mainly from ettringite formation, with only a limited contribution from C–S–H gels. Accordingly, the enhanced hydration activity of lithium slag within the first 9 h can be attributed primarily to ettringite formation.

Between 30 h and 50 h, CaCO₃ was detected, while the diffraction peak of gypsum decreased further. This suggests that the second exothermic peak can be attributed to the reformation of AFt, along with the formation of CaCO₃. The formation of AFt between 30 h and 50 h is likely related to the renewed dissolution of Al₂O₃, which is similar to the corresponding process during cement hydration [32]. Furthermore, the presence of residual unreacted gypsum in the LS-CH mixture after 50 h of hydration indicates that the second broad exothermic peak in Figure 7b does not correspond to gypsum depletion or AFm formation [33,34].

2.3. Hydration Products of LS-CH Pastes Determined by XRD

2.3.1. Hydration Products of LS-CH Pastes Under Normal Curing Condition

Figure 9 shows the XRD patterns of LS-CH pastes cured for 3 days (N3d) and 28 days (N28d) under normal curing conditions. Figure 10 presents the relative content of each phase in LS-CH pastes, as determined by XRD/Rietveld analysis.

The results show that the diffraction peaks of gypsum (originally from LS) gradually weakened over time, whereas those of quartz remained largely unchanged. This indicates that the quartz in LS is either inert or exhibits extremely low reactivity. Notably, the main diffraction peak of LiAlSi₂O₆ slightly decreased over time, suggesting that, in addition to amorphous phases, LiAlSi₂O₆ can also slowly participate in the hydration reaction. Previous studies have identified leached LiAlSi₂O₆ as a zeolite-like phase [26] with relatively low pozzolanic activity, comparable only to that of FA [35].

At 3 d and 28 d, the primary reactive components in LS consisted of LiAlSi₂O₆ along with amorphous Al₂O₃, SiO₂, and crystalline CaSO₄·2H₂O (gypsum). The major hydration products in LS-CH pastes included C–S–H gels, ettringite (AFt), and CaCO₃ (calcite). Notably, owing to their amorphous nature, C–S–H gels could not be detected by XRD analysis, whereas the crystalline phases (AFt and CaCO₃) were clearly identified. It should be noted that the formation of CaCO₃ arose not only from the reaction between CH and carbonate but also from the carbonation of CH during specimen preparation. Accordingly, the primary phases in hardened LS-CH pastes were unreacted CH, C–S–H gels, AFt, CaCO₃, SiO₂ (quartz), and LiAlSi₂O₆.

Similar to metakaolin [36,37,38], the main chemical reaction in LS-CH pastes involves the reaction between amorphous Al₂O₃, SiO₂, and CH. However, LS contains considerable amounts of SO₃ (primarily present as gypsum and other sulfates) and carbonates (Figure 5). Meanwhile, LiAlSi₂O₆, which is rich in Al₂O₃ and SiO₂, can also participate in hydration reactions, although its reactivity is relatively low. Consequently, the overall hydration process can be described by reactions (1)–(4).

The formation of C–S–H gels:

$$\mathrm{S}\mathrm{i}{\mathrm{O}}_2+\mathrm{C}\mathrm{a}{(\mathrm{O}\mathrm{H})}_2+{\mathrm{H}}_2\mathrm{O}\to \mathrm{C}–\mathrm{S}–\mathrm{H}$$

(1)

The formation of AFt:

$$\mathrm{A}{\mathrm{l}}_2{\mathrm{O}}_3+\mathrm{S}{\mathrm{O}}_3+\mathrm{C}\mathrm{a}{(\mathrm{O}\mathrm{H})}_2+{\mathrm{H}}_2\mathrm{O}\to {\mathrm{C}}_6\mathrm{A}{\overline{\mathrm{S}}}_3{\mathrm{H}}_{32}(\mathrm{A}\mathrm{F}\mathrm{t})$$

(2)

The formation of CaCO₃ [39]:

$${\mathrm{C}{\mathrm{O}}_3}^{2-}+\mathrm{C}{\mathrm{a}}^{2+}\to \mathrm{C}\mathrm{a}\mathrm{C}{\mathrm{O}}_3$$

(3)

$$\mathrm{C}\mathrm{a}{(\mathrm{O}\mathrm{H})}_2+\mathrm{C}{\mathrm{O}}_2\to \mathrm{C}\mathrm{a}\mathrm{C}{\mathrm{O}}_3+{\mathrm{H}}_2\mathrm{O}$$

(4)

The reaction between LS and CH solution proceeds via a mechanism analogous to the pozzolanic reaction of LS in cement systems. The high-pH CH solution actively disrupts Si-O and Al-O bonds, promoting the formation of Al-bearing C–S–H gels via the reaction of SiO₂ and Al₂O₃ (from both amorphous phases and partial dissolution of LiAlSi₂O₆) with CH. The proposed reaction mechanism comprises three key steps:

(1)

CH-activated dissolution of SiO₂ from LS releases SiO₄⁴⁻ species into solution;

(2)

the subsequent reaction of these silicate ions with CH forms C–S–H gels [40,41];

(3)

the concurrent incorporation of aluminum species (released from Al₂O₃ dissolution) into the C–S–H structure yields C–(A)–S–H gels [42].

Simultaneously, in the presence of gypsum and other sulfates, dissolved Al₂O₃ reacts with CH to initially form AFm phases, which subsequently convert to ettringite (AFt). Notably, carbonate ions may partially or fully replace sulfate in the ettringite structure through anion substitution [40]. Furthermore, carbonate species derived from LS can react directly with CH in aqueous conditions to precipitate CaCO₃ (calcite). Additionally, CaCO₃ may form indirectly through the carbonation of CH upon exposure to atmospheric CO₂.

Unlike GBFS, LS contains both amorphous and crystalline phases. The crystalline phases in LS (e.g., gypsum and carbonates) mainly contribute to its early hydration activity within 9 h, as evidenced by the significantly higher hydration heat of the LS-CH system relative to GBFS-CH in Figure 7. In contrast, the amorphous phase and spodumene in LS are primarily responsible for its later-stage reactivity. In the late hydration stage, spodumene and the amorphous phase in LS can react continuously with CH, releasing hydration heat steadily. Nevertheless, the performance is limited by several factors: the reactivity of spodumene is only comparable to that of FA; LS does not undergo the high-temperature water quenching process used for GBFS, leading to lower activity of its amorphous phase; and the amorphous phase content in LS is considerably lower than in GBFS. Owing to the combined effect of these factors, the hydration heat release of the LS-CH system after 9 h is lower than that of GBFS-CH.

2.3.2. Hydration Products of LS-CH Pastes Under Steam Curing Condition

As steam curing is widely used in precast concrete production, this study also investigated the hydration products of LS-CH paste under steam-cured conditions. Figure 11 shows the XRD patterns of LS-CH pastes cured at 80 °C for 7 h (S7h) and 7 days (S7d) under steam curing. Analysis indicated that the main hydration products were C–S–H gels, ettringite (AFt), monosulfoaluminate (AFm), bassanite (CaSO₄·0.5H₂O), and calcite (CaCO₃), together with unreacted LiAlSi₂O₆ and quartz. A comparison of Figure 10 and Figure 11 demonstrated that steam curing significantly accelerates the hydration rate of LiAlSi₂O₆ compared with standard curing conditions, which is consistent with previous results [8].

Taylor et al. [43] demonstrated that excess gypsum can stabilize ettringite (AFt) even under elevated temperature, which explains its persistence after 7 h and 7 d of steam curing in this study. However, given the well-documented thermal instability of AFt at 80 °C [44], we observed its gradual conversion to monosulfoaluminate (AFm) and bassanite (CaSO₄·0.5H₂O) after the initial 7 h of curing, a finding consistent with previous reports [45]. Concurrently, elevated temperatures also accelerated the decomposition of gypsum into bassanite [27].

With prolonged steam curing (7 d), two distinct phenomena were observed:

(1)

further decomposition of AFt, as indicated by the progressive weakening of its characteristic XRD peaks (Figure 11);

(2)

the reappearance of gypsum diffraction signals.

This latter observation suggests the existence of a dynamic equilibrium involving sulfate release from continuous AFt decomposition and subsequent rehydration of bassanite back to gypsum under the steam curing conditions.

2.4. Morphology of Hydration Products in LS-CH Pastes Under SEM

2.4.1. LS-CH Pastes After 28 Days of Standard Curing

Given the high SO₃ content in LS, SEM-EDS was used to conduct an in-depth investigation into the morphology and composition of hydration products in LS-CH pastes. The representative SEM-EDS results are presented below. Since no significant differences were observed in the hydration products of LS-CH pastes cured for 3 and 28 days, Figure 12 only presents the morphology and EDS results of hydration products in the 28-day LS-CH paste.

As shown in Figure 12, the hardened LS-CH paste consisted of six types of hydration products along with unreacted LS particles. These six hydration products were needle-like AFt with a length of 1~10 μm (Figure 12a,d), spherical CaCO₃ with a particle size of approximately 1 μm (Figure 12a), porous reticular C–S–H (Figure 12b,d), cubic CaCO₃ with a particle size of about 1 μm (Figure 12c), ellipsoidal CaCO₃ with a particle size of approximately 2 μm (Figure 12d), and hexagonal prismatic gypsum with a particle size of approximately 1 μm (Figure 12d).

Ettringite, which contains 32 bound water molecules, has a large molar volume that contributes to an increased solid volume and reduced porosity in hardened pastes [46]. However, when formed in excess, the generation of loosely packed, polydisperse ettringite needles (Figure 12a) can reduce the packing density and hinder the densification of the LS-CH matrix. Additionally, the high carbon content in the EDS analysis of ettringite (Figure 12a) suggests that carbonates from the lithium slag were involved in its formation.

It is well established that CaCO₃ crystals typically exhibit three distinct morphologies: cubic calcite, spherical vaterite, and needle-like aragonite. Among these, calcite is the most thermodynamically stable under ambient conditions, whereas vaterite and aragonite tend to convert to calcite over time [47]. Furthermore, the morphology of CaCO₃ is highly dependent on precipitation conditions, including solution pH, temperature, the presence of foreign ions (e.g., Mg²⁺) and organic additives, as well as the degree of supersaturation [47,48].

In this study, the low alkalinity of the LS-CH hardened paste, carbonates present in LS, elevated curing temperature, and the use of acetone to halt hydration are believed to contribute to the formation of multiple CaCO₃ morphologies. Notably, CO₃²⁻ and Ca²⁺ initially react to form metastable spherical or ellipsoidal vaterite (Figure 12d), which subsequently undergoes gradual transformation into calcite. This indicated that ellipsoidal vaterite acts as an intermediate phase during calcite formation under the given conditions.

Additionally, the presence of unreacted gypsum (Figure 12d) after 28 days of standard curing suggests incomplete sulfate consumption in the LS-CH paste. Consequently, special attention should be paid to the risk of delayed ettringite formation (DEF) when lithium slag is used as an SCM in concrete applications.

2.4.2. LS-CH Pastes Steam-Cured for 7 h and 7 Days

Figure 13 and Figure 14 present the microstructural characteristics of hydration products in LS-CH pastes after steam curing for 7 h and 7 days, respectively. The main hydration phases formed under steam curing include: (1) loose reticular C–S–H gels (Figure 13a,c and Figure 14a); (2) fibrous AFt with a length of 5 μm and a diameter of 0.5 μm (Figure 13b); (3) hexagonal plate-like AFm with a size close to 1 μm (Figure 13b and Figure 14b); (4) ellipsoidal CaCO₃ with a particle size of 0.5~1 μm (Figure 13a and Figure 14b); and (5) short prismatic gypsum crystals with a particle size close to 1 μm (Figure 13b). Notably, the appearance of partially eroded LiAlSi₂O₆ after heat treatment indicates its involvement in the hydration reaction, especially at elevated temperatures. This observation is in good agreement with the XRD results.

Consistent with the high SiO₂, Al₂O₃, and SO₃ contents in lithium slag, the C–S–H gels in the LS-CH paste exhibited lower Ca/Si ratios (0.28~0.59) and higher Al/Si ratios (0.37~0.49) (Figure 12b, Figure 13a and Figure 14a), along with significant sulfate adsorption (Figure 12b, Figure 13a and Figure 14a). Previous studies have established that sulfate primarily adsorbs onto Si-O sites as CaSO₃ complexes without being incorporated into the C–S–H gel structure [46,49], this adsorption behavior may subsequently promote the formation of AFt and AFm phases.

Interestingly, the C–S–H gels maintained a relatively loose, highly porous structure (Figure 12b, Figure 13a and Figure 14a) regardless of curing conditions. This observation aligns with the existing literature: Bérodier et al. [49] reported that elevated sulfate content reduces the bulk volume of C–S–H gels while increasing capillary porosity, whereas Adu-Amankwah et al. [50] observed that excess sulfate lowers gel water content and increases porosity. Although our results support these findings, the underlying mechanisms warrant further investigation.

2.5. TG/DTG Analysis Results of Hydration Products in LS-CH Pastes

Figure 15 presents the TG/DTG results of LS-CH pastes under different curing conditions. The characteristic mass loss events within specific temperature ranges, along with their corresponding chemical reactions, are systematically summarized in

Table 3. Reactions occur within specific temperature ranges [27,28,29,51,52].

Temperature Ranges	Reactions
60–150 °C	The loss of bound water from the decomposition of gypsum, ettringite, and C–S–H
150–400 °C	The decomposition of monosulfates, aluminate hydrates, hydrogarnet series phases, and calcium aluminosilicate hydrates, e.g., gehlenite hydrate, stratlingite, etc.
400–550 °C	Dehydroxylation of CH
550–1000 °C	Decarbonation of carbonates

. In addition,

Table 4. The mass loss of LS-CH pastes determined by TG analysis within specified temperature ranges, along with the CH content (wt.%).

Sample	60–150 °C	150–400 °C	400–550 °C	550–1000 °C	Total Mass Loss	CH
S7h	4.93	1.75	1.36	2.12	10.16	5.59
N3d	4.56	1.83	1.75	2.32	10.46	7.19
S7d	4.98	2.30	1.33	2.32	10.93	5.47
N28d	5.38	2.35	1.68	2.31	11.72	6.90

provides a quantitative analysis of total weight loss percentages within defined temperature intervals for all investigated LS-CH paste samples.

The mass loss observed between room temperature and 400 °C primarily corresponds to the decomposition of newly formed hydration products in LS-CH paste, thus serving as a reliable indicator of hydration degree. Based on this analysis, the relative hydration degrees under different curing conditions followed the order: N28d > S7d > S7h > N3d. Although steam curing at early ages promotes hydration, prolonged steam curing (7 days) appeared to inhibit further hydration development. This phenomenon may be attributed to the gradual decomposition of ettringite under extended thermal treatment, as well as the limited CH content in the LS-CH mixture.

In addition, trace carbonate phases were detected in all four LS-CH paste samples. These carbonates may originate from: (1) reactions between inherent carbonates in LS and CH, (2) carbonation occurring during sample preparation or storage, or (3) residual unreacted carbonates in the original LS material.

Besides the non-evaporable water, the mass loss of the LS-CH paste between room temperature and 1000 °C also reflects the decomposition of carbonates, unreacted gypsum, and residual CH. Nevertheless, the total mass loss remained considerably lower than that of pure cement paste [8], indicating that the hydration degree of LS-CH pastes is relatively low compared with conventional cement systems.

2.6. Hydration Products of LS-CH Pastes Determined by FTIR

To better understand the distinct hydration characteristics of LS-CH paste compared with conventional cement systems, Fourier transform infrared spectroscopy (FTIR) was employed to analyze the chemical composition of hydration products. Figure 16 shows the comparative FTIR spectra of: (1) LS-CH paste and (2) Portland cement paste (w/c = 0.3) after 28 days of hydration under standard curing conditions.

As shown in Figure 16, the distinct band at 3640 cm⁻¹, assigned to the stretching vibration modes of H-OH groups, corresponded to calcium hydroxide in both the hardened cement paste and hardened LS-CH pastes. Meanwhile, the band at 875 cm⁻¹ is associated with the out-of-plane bending (ν₂) modes of CO₃²⁻ and the antisymmetric stretching modes of AlO₄ groups [53]. This result further confirms that, similar to its role in hardened cement paste, the carbonate phase is also an important hydration product in hardened LS-CH paste, which is consistent with SEM observations.

The main differences between the hydration products of LS-CH pastes and PC paste appeared in the characteristic bands within the ranges of 1000–970 cm⁻¹ and 550–520 cm⁻¹ [54]. This can be attributed to two factors: firstly, the presence of unreacted spodumene and quartz; secondly, in LS-CH pastes, the band at 970 cm⁻¹, which is characteristic of the Si-O stretching vibrations in C–S–H gels, exhibited a significant shift toward higher wavenumbers compared to that in PC paste (from 970 cm⁻¹ to 1000 cm⁻¹). This phenomenon further indicates an increased content of Al incorporated into C–S–H gels, as well as the gradual polymerization of silicate chains in C–S–H gels accompanied by a decreasing Ca/Si ratio [55]. These findings are consistent with the aforementioned SEM-EDS analysis results.

2.7. Hydration Products of LS-CH Pastes Determined by BET Analysis

To further analyze the differences in properties between the hydration products of LS-CH pastes and those of 28-day PC paste, the multipoint Brunauer–Emmett–Teller (BET) analysis was performed to characterize the hydration products of LS-CH paste, with the results presented in

Table 5. BET results of LS-CH pastes.

Sample	Average Pore Width (nm)	Cumulative Pore Volume (cm3/g)	N2-Specific Surface Area (m2/g)
LS	19.96	0.0285	4.400
S7h	43.61	0.1631	11.94
N3d	38.01	0.1638	11.49
S7d	40.49	0.1580	12.39
N28d	25.88	0.1663	17.96
28d PC	19.09	0.1530	29.98

. BET analysis is well suited for characterizing the gel pores in calcium silicate hydrate (C–S–H) and partial transitional pores in hardened LS-CH paste [56]. Accordingly, BET results can provide insights into the structural characteristics of C–S–H to some extent. Specifically, pore structure reflects the compactness of C–S–H, whereas specific surface area acts as an indicator of its cementitious potential.

As shown in Table 5, LS-CH pastes exhibited a coarser pore structure than PC paste, characterized by larger average pore diameters (25.88–43.61 nm vs. 19.09 nm), lower specific surface areas (11.49–17.96 m²/g vs. 29.98 m²/g), and higher cumulative pore volumes (0.1580–0.1638 cm³/g vs. 0.1530 cm³/g).

The phase composition of conventional cement hydration products typically consists of 50–60% C–S–H, 20–25% calcium hydroxide, and 5–10% ettringite [57,58]. In contrast, the LS-CH system displayed a markedly distinct phase assemblage: C–S–H content below 24.88% (Figure 10), calcium hydroxide ranging from 5.47% to 7.19% (Table 4), and ettringite between 10.60% and 13.34% (Figure 10). This unique phase composition, differing from that of traditional hardened cement pastes, is the primary cause of the coarser pore structure and lower specific surface area observed in the hardened LS-CH paste. These characteristics are mainly attributed to the low C–S–H content, high ettringite content, and the large proportion of unreacted lithium slag.

In addition, the reticular C–S–H structure formed in the LS-CH paste in this study was relatively loose, with its gel pore size significantly larger than that of C–S–H gel in hardened cement paste [59], reaching several tens of nanometers (Figure 12b, Figure 13a and Figure 14a). Meanwhile, a large number of unreacted lithium slag particles were present in the LS-CH paste. Owing to the inherently porous structure of lithium slag (with an average pore size of 19.96 nm, as shown in Table 5), the incorporation of these unreacted particles further increased the porosity of the paste. Furthermore, it should be noted that although a moderate amount of ettringite can refine the microstructure via pore-filling effects, excessive ettringite formation may instead increase porosity [60].

Extending the curing age promotes the formation of C–S–H (Table 4). Consequently, under both standard and steam curing regimes, prolonged curing (N28d vs. N3d; S7d vs. S7h) yields a finer pore structure, characterized by smaller average pore diameters and larger specific surface areas. This refinement is attributed to the inherently high specific surface area and dense nature of C–S–H gel. In contrast, early-age steam-cured samples (S7h) exhibited larger average pore diameters and specific surface areas than their standard-cured counterparts (N3d). This result arises from a dual effect: while steam curing accelerates early-age C–S–H formation, it also induces rapid hydration that can lead to an uneven distribution of products, thereby introducing more and larger pores. Consistent with this mechanism, S7d samples displayed a larger pore diameter than N28d samples. Nevertheless, owing to the lower total volume of C–S–H formed in S7d samples, their pore volume and specific surface area remained lower than those of N28d.

2.8. Mechanical Properties

Figure 17 presents the mechanical strength development of LS-CH pastes under various curing regimes. Under standard curing conditions, the 3-day compressive and flexural strengths reached 16.6 MPa and 2.7 MPa, respectively, and increased to 32.9 MPa and 4.2 MPa after 28 days of hydration. As discussed earlier, this significant strength development is primarily attributed to the substantial formation of AFt and C–(A)–S–H gels.

Steam-cured specimens exhibited enhanced early-age strength, with 7 h compressive and flexural strengths reaching 23.0 MPa and 4.1 MPa, respectively. These values exceeded those of 3-day normally cured samples. This significant improvement clearly demonstrates the beneficial effect of elevated temperature in activating the hydration potential of lithium slag. However, extending the steam curing duration to 7 days resulted in only moderate strength gains (28.3 MPa compressive and 5.6 MPa flexural), representing increases of just 23.0% and 36.6%, respectively, relative to the 7 h steam-cured specimens.

The compressive strengths of the four specimens—N3d, S7h, S7d, and N28d—increased sequentially. This can be attributed to the gradual rise in the content of newly formed hydration products (Table 4) and the corresponding increase in N₂ specific surface area (Table 5), which serves as an indicator of C–S–H formation.

Notably, the compressive strength of the 7-day steam-cured specimen (S7d) was lower than that of the 28-day standard-cured specimen (N28d). This behavior is not only associated with the lower quantity of hydration products generated, but also with microstructural defects induced by prolonged high-temperature exposure—especially microcracks (Figure 14a) and a coarsened pore structure (Table 5)—which collectively impaired strength development. Interestingly, however, S7d exhibited a higher flexural strength. This may be attributed to its lower CH content (Table 4). Wang et al. [61] previously reported that a lower content of CH, along with a smaller CH particle size, can improve the interfacial transition zone and thereby help enhance the flexural strength of concrete.

In addition, after 28 days of normal curing, the flexural and compressive strengths of LS-CH are approximately one-third of those of P·I 42.5 cement under the same conditions [62]; whereas under steam curing at 80 °C for 7 days, the compressive strength and flexural strength of LS-CH reach 3.49 times and 1.70 times those of ferronickel slag powder activated by 10% CH, respectively, under identical curing conditions [63]. These favorable mechanical properties not only confirm the high pozzolanic reactivity of lithium slag but also demonstrate the potential of LS-CH paste as a viable cementitious material.

2.9. Discussion

2.9.1. Differences Between LS and Traditional SCMs

Distinct from conventional SCMs, lithium slag is characterized by elevated SO₃ and carbonate contents. These components exist in specific chemical forms: SO₃ is present mainly as calcium sulfate (both hydrated and anhydrous CaSO₄) and sodium sulfate (Na₂SO₄), while carbonates occur as CaCO₃, Na₂CO₃, and Li₂CO₃. SO₃ plays a crucial role in controlling setting behavior through a dual mechanism: (1) at low concentrations, it retards hydration by moderating C₃A reactivity, thereby prolonging setting time; (2) at high concentrations, it accelerates hydration through rapid ettringite formation, which may induce flash setting [46,49].

Similarly, the presence of carbonates in LS also exerts a dual effect on the LS-CH system. On the one hand, carbonates react with Ca²⁺ in the system to form CaCO₃, lowering the Ca²⁺ concentration. This in turn promotes further dissolution of CH and releases more OH⁻, increasing the alkalinity of the system and accelerating the dissolution of the amorphous phase in LS [64]. Meanwhile, carbonate ions can partially replace sulfate ions during ettringite formation, accelerating the setting of the system [40]. On the other hand, the reduced Ca²⁺ concentration retards the nucleation of C–S–H, thereby prolonging the setting time [65].

Although LS is rich in SiO₂ and Al₂O₃ (Table 1), these components mainly exist as crystalline LiAlSi₂O₆ rather than in reactive amorphous phases. The low reactivity of LiAlSi₂O₆ explains the weaker pozzolanic activity of LS at later ages compared with GBFS, which contains a higher proportion of amorphous aluminosilicate phases. However, high-temperature treatment can significantly improve the reactivity of spodumene (Figure 10 and Figure 13c and Table 4). Therefore, lithium slag is highly suitable for applications in components cured at high temperatures [8,22], or it can be effectively utilized after a high-temperature pretreatment process [66].

2.9.2. Differences Between LS-CH and Traditional Cement Products

Additionally, owing to the high SO₃ and carbonate contents in lithium slag, ettringite and CaCO₃ constitute a considerable proportion of the hydration products in hardened LS-CH pastes (Figure 10 and Table 4). Their contents are not only markedly higher than those in conventional cement-based materials but also in some cases even exceed that of C–S–H gel in LS-CH pastes. This distinctive chemical composition exerts several critical influences on the properties of LS-CH paste:

(1)

Reduced fluidity and enhanced early strength: The rapid formation of ettringite accelerates paste hardening and improves early-age strength (Figure 17). Meanwhile, it consumes considerable free water, thereby reducing the fluidity of the fresh paste (the fluidity of the LS-CH mortar was 120 mm).

(2)

Pore structure: The LS-CH paste displays a coarser pore structure and a smaller specific surface area (Table 5).

(3)

Thermal stability of ettringite: Prolonged steam curing provides only limited benefits to the LS-CH paste. Even when the curing duration is increased from 7 h to 7 days, the increases in hydration products and mechanical properties remain negligible, as shown in Table 4 and Figure 15. This limited improvement is mainly controlled by two factors: the thermal instability of ettringite and thermal damage caused by steam curing (Figure 14a), both of which restrict further property development. Therefore, from the perspective of balancing energy costs and strength gain, prolonged high-temperature steam curing should be applied cautiously to LS-CH specimens. It is also worth noting that this limited strength enhancement may be associated with the low CH content in the LS-CH system. Hence, future research could further explore the performance evolution of LS-CH with different CH dosages under various steam curing conditions.

(4)

Delayed ettringite formation: The high SO₃ content raises potential concerns regarding delayed ettringite formation (DEF), which requires further investigation. However, existing literature indicates that in low-alkalinity lime-based systems, carbonation tendencies may suppress the stability of ettringite [67].

2.9.3. Thermal Damage Caused by Steam Curing

As shown in Figure 14a and Table 5, microcracks and pore coarsening appeared in the LS-CH paste under steam curing conditions, indicating that steam curing caused thermal damage to LS-CH.

Thermal damage caused by steam curing to concrete typically includes three types: unrecoverable expansion deformation, surface layer defects, and embrittlement effects, all of which are detrimental to its durability [68]. This can be attributed to three main factors: (1) The rapid and uneven formation of hydration products leads to coarsening and deterioration of the pore structure [69]. (2) Thermal expansion mismatch induces microcracking. At the macroscopic level, the temperature gradient arising from differential heat dissipation between the surface and interior creates thermal stress gradients, resulting in surface microcracks. At the mesoscopic level, the mismatch in thermal expansion coefficients between aggregate and cement paste imposes additional stresses on the interfacial transition zone (ITZ) during heating and cooling. Furthermore, a higher steam-curing temperature leads to a more uneven microhardness distribution in the ITZ and a greater number of interfacial defects [70]. (3) When concrete is heated, the internal free water expands and partially vaporizes, generating internal steam pressure [71].

3. Conclusions

To explore the hydration mechanism of lithium slag (LS) in cementitious systems, this study adopts a systematic approach by simulating the alkaline environment of cement using calcium hydroxide (CH). The investigation first compares the hydration processes of the LS-CH system (containing 10 wt.% CH) with those of parallel GBFS-CH and FA-CH systems. Subsequently, the research comprehensively examines the hydration products and mechanical properties of LS-CH paste under both standard curing (3 days and 28 days) and steam curing (80 °C for 7 h and 7 days) conditions. The principal findings can be summarized as follows:

(1)

Under CH activation, the reactivity of LS within 72 h is slightly lower than that of GBFS but significantly higher than that of FA. Notably, during the initial 9 h of hydration, the reactivity of LS is superior to that of GBFS, owing to the abundant formation of ettringite in the LS-CH system.

(2)

Within 72 h, the hydration heat flow curve of LS-CH exhibits two distinct exothermic peaks. The first peak (0–10 h) corresponds to three simultaneous processes: (I) ettringite formation, (II) dissolution of LS particles in the CH solution, and (III) rapid precipitation of C–S–H gel. In contrast, the secondary peak (30–50 h) arises mainly from calcite crystallization and renewed ettringite formation.

(3)

Under standard curing conditions, lithium slag reacts with Ca(OH)₂ solution to form several typical hydration products: needle-like AFt crystals, spherical/cubic/spindle-shaped CaCO₃ polymorphs, and loose reticular C–S–H gels. When subjected to 80 °C steam curing, the product assemblage is extended to include AFm phases and calcium sulfate hydrates (both dihydrate and hemihydrate forms). Notably, the persistence of calcium sulfate in the products implies potential risks of delayed ettringite formation (DEF), which deserves careful consideration when using lithium slag as a supplementary cementitious material in concrete.

(4)

SEM-EDS and FTIR analyses indicate that the C–S–H gel, the main binding phase in LS-CH, has a unique chemical composition characterized by a low Ca/Si ratio and significant incorporation of aluminum and sulfur. Compared with conventional hardened cement paste, the LS-CH paste shows a larger average pore diameter (25.88–43.61 nm vs. 19.09 nm) and higher pore volume (0.1580–0.1638 cm³/g vs. 0.1530 cm³/g), but a lower specific surface area (11.49–17.96 m²/g vs. 29.98 m²/g).

(5)

Mechanical tests demonstrate that LS-CH exhibits promising potential as a sustainable alternative to traditional cementitious materials.

4. Materials and Methods

4.1. Materials

The lithium slag (LS) used in this study was supplied by Jiangsu Rongda New Material Co., Ltd. (Nantong, China).

For comparison, Type S95 ground granulated blast-furnace slag (GBFS) conforming to Chinese National Standard GB/T 18046-2008 [72], Type I fly ash (FA) compliant with Chinese National Standard GB/T 1596-2017 [73], and P·II 52.5 Portland cement (PC) meeting Chinese National Standard GB 175-2007 [74] were employed. It should be clarified that GBFS, FA, and PC were mainly used as reference groups to investigate the hydration process of LS and the characteristics of its hydration products. GBFS and FA were only used in hydration heat tests, whereas PC was solely used to cast specimens with a water-to-cement ratio of 0.3 for FTIR and BET analyses of the 28-day hydration products.

To eliminate the interference of other impurity ions, deionized water and analytically pure calcium hydroxide (Ca(OH)₂, purity > 95 wt.%) were used to prepare lithium slag-Ca(OH)₂ pastes.

Standard sand with a fineness modulus of 2.75, conforming to GB/T 17671-2021 [75], was used to prepare LS-CH mortar and perform fluidity test.

4.2. Sample Preparation and Test Methods

4.2.1. Sample Preparation

To investigate the mechanism of LS in cement, this study employs CH to simulate the alkaline environment of cement. The hydration process of LS is activated by CH, and the types and characteristics of the hydration products are analyzed to directly reflect the influence of LS on the cement hydration process. Since a CH dosage of 10% has been widely used to activate SCMs [63,76] and to facilitate comparison with existing literature, this study adopts the same CH dosage at the same water-to-binder ratio to activate lithium slag. The detailed mix proportions of the LS–Ca(OH)₂ mixture (abbreviated as LS–CH) are presented in

Table 6. Mix proportion (wt.%), setting time and fluidity of LS-CH pastes.

LS	CH	Water	Water/(LS + CH) Ratio	Setting Time (h: min)		Fluidity (mm)
				Initial	Final
90	10	30	0.3	1:06	2:25	120

. Meanwhile, in accordance with the GB/T 17671-2021 standard [75], three LS–CH prismatic specimens measuring 40 mm × 40 mm × 160 mm were prepared for mechanical property testing.

As shown in Table 6, the setting time of LS-CH, measured in accordance with GB/T 1346-2024 [77], is comparable to that of ordinary Portland cement. The fluidity of the LS-CH mortar was only 120 mm, which can be attributed to the low solubility of calcium hydroxide and the high water demand of lithium slag [78].

For comparison, a paste of PC with a water-to-cement ratio (w/c) of 0.3 was also prepared and hydrated for 28 days.

To investigate the characteristics of hydration products in LS-CH under different curing conditions, this study employed both normal curing (NC) and steam curing (SC) regimes, as detailed in previous studies [63,79] and illustrated in Figure 18. Specifically, the normal curing regime included two durations: curing at 20 ± 2 °C with a relative humidity (RH) greater than 95% for 3 days and for 28 days, respectively. Meanwhile, the steam curing regime involved constant-temperature curing at 80 °C for 7 h and for 7 days. For clarity, the specimens are designated as follows: N3d and N28d denote LS-CH samples under normal curing for 3 and 28 days, respectively, whereas S7h and S7d represent those subjected to steam curing for 7 h and 7 days.

Compressive strength and flexural strength tests were conducted in accordance with the Chinese standard GB/T 17671-2021 [75].

4.2.2. Test Methods

The setting time of the LS-CH paste was determined using a Vicat apparatus (GB/T 1346-2024 [77]), and its fluidity was tested by the jump table method (GB/T 2419-2024 [80]).

The 28-day pozzolanic activity of SCM was evaluated by the compressive strength ratio of 40 mm × 40 mm × 160 mm prismatic cement mortar specimens. These specimens were prepared with a 30% SCM replacement level and subjected to 28 days of standard curing, with the results compared to those of pure cement specimens.

After the strength tests at the designated ages, hydration of the specimens was stopped by immersion in anhydrous ethanol. The paste samples were then ground into dried powder or cut into approximately 5 mm blocks for characterization by XRD, SEM-EDS, TG-DTG, FTIR, and BET.

X-ray diffraction (XRD) was employed to quantitatively analyze the mineralogical compositions of the LS and LS-CH systems using a Bruker D8-Discover X-ray diffractometer (Karlsruhe, Germany) combined with the Rietveld full-pattern fitting method [8].

The microstructure, phase morphology, and elemental composition of LS and LS-CH were examined using a Sirion field-emission scanning electron microscope (SEM) equipped with an energy-dispersive X-ray spectrometer (EDS) (FEI, Hillsboro, OR, USA).

Thermogravimetric analysis (TG/DTG) and Fourier transform infrared (FTIR) spectroscopy were performed to investigate the phase evolution and structural characteristics of the LS and LS-CH samples, using a STA 449 F3 Jupiter^® (Selb, Germany) and a Deaupos Scientific FTIR spectrometer (Berlin, Germany), respectively.

The pore structure parameters and specific surface area of the powdered LS and LS-CH paste were determined from nitrogen adsorption–desorption isotherms using the BET method with a fully automated gas sorption analyzer (ASIQM0001-5, Quantachrome Instruments, Boynton Beach, FL, USA) [81].

Isothermal conduction calorimetry (ICC) test: to clarify the hydration behavior of lithium slag in a CH solution, ICC tests were performed at 20 °C using a TAM Air Calorimeter (Thermometric, Stockholm, Sweden) on pastes with a water-binder ratio of 0.3. To minimize the interference of CH dissolution on LS hydration, CH was pre-dissolved in water in a sealed plastic bottle for 24 h prior to testing. Owing to the small sample mass required for the ICC test, the LS-CH paste was manually mixed at room temperature using a glass rod. For comparison, two commonly used SCMs, namely GBFS and FA, were selected as reference materials. The hydration processes of GBFS-CH and FA-CH pastes were also tested under identical conditions.

Furthermore, to ensure the reliability and reproducibility of the experimental results, all tests were performed in duplicate or triplicate.