Commercial fast hardening SAC following GB 20472-2006 [41], was used in this study. The raw iron ore tailings (RIOTs) were obtained from a mining enterprise in Huoqiu County of Anhui Province of China, and a small amount of larger particles was removed using a 0.18 mm sieve before the experiment. The chemical compositions of SAC and RIOTs were tested by XRF (see Table 1). According to the chemical composition of RIOTs, the basicity R = CaO/SiO₂ was only 0.08, indicating that RIOTs used here was with a very low pozzolanic reactivity [42], and thus it posed a tremendous technical challenge to apply it to cement-based materials.

The micro-IOTs slurry was prepared in a planetary ball mill with four grinding pots. Each grinding pot included five specification agate balls, with a mass ratio of 10 mm:8 mm:5 mm:3 mm:1 mm = 1:3:6:2:2. The total mass of the agate balls was about 60 g in each grinding pot. A commercially available polycarboxylate superplasticizer (PCE) was used as the dispersant to improve the grinding efficiency [39]. RIOTs (20 g), PCE (0.2 g), and water (20 g) were separately weighed and put into the grinding pot, and the mixture was ground for 2 h with a motor speed of 600 r/m. And then the agate balls were separated, and the micro-IOTs slurry was obtained. Considering that many larger IOTs particles were attached to the bottom of the grinding pot and the surface of the grinding ball in the separation process, the solid-liquid ratio of the micro-IOTs slurry was recalibrated by experiments. 10 g of the slurry was dried, and the weight of solid was 4 g, which meant that the weight ratio of (micro-IOTs + PCE) and water was constant at 4:6 in the slurry.

The particle size distribution of RIOTs, SAC, and micro-IOTs is presented in Fig. 1. It was found that the average diameter of micro-IOTs was 517 nm, indicating that IOTs slurry with micro-scale could be obtained by wet grinding. Besides, it was seen that the particle size of micro-IOTs was far smaller than that of SAC, which was conducive to its filling effect in SAC system.

The XRD patterns of RIOTs and micro-IOTs are presented in Fig. 2. The main crystalline phases were silicon dioxide, grunerite and katophorite, and the diffraction peak intensity of the phases in micro-IOTs was much lower than that in RIOTs. This indicated that the crystalline degree in IOTs was decreased, and the amorphous phase was increased with the wet grinding process.

The influence of wet grinding on IOTs surface characteristics was studied by XPS analysis, and high-resolution scan curves of RIOTs and micro-IOTs for Si2p, Al2p, and O1s were displayed in Fig. 3. The binding energy of silicon was reduced from 102.78 to 102.48 eV, and that of aluminum and oxygen was respectively reduced from 74.33 to 73.78 eV and from 531.98 to 531.68 eV after wet grinding. The decrease in binding energy indicated that atoms’ coupling to electrons was weakened, and chemical bonds were vulnerable to break.

Based on the above results, micro-IOTs have changed physical properties and complex chemical reactions during wet grinding. Firstly, wet grinding significantly decreased the particle size, which is beneficial to play nucleation and filling effects in cement-based materials. Secondly, the mechanical breakage helps decrease silicon dioxide crystallinity in IOTs and improves amorphous phase of physical structure. Moreover, the wet milling decreased the bonding energy, indicating that a chemical reaction occurred during wet grinding. These would contribute to increasing reactivity of micro-IOTs in cement and concrete materials. Similar changes in physical and chemical properties have also been reported in previous research on wet grinding treatment of fly ash [39]. These changes provided the possibility to apply micro-IOTs in improving SAC performance.

The setting time of SAC pastes with different content of micro-IOTs is presented in Fig. 4. It was observed that the initial setting times and final setting times were gradually shortened with increasing micro-IOTs content, demonstrating that micro-IOTs could obviously accelerate the setting and hardening of SAC paste. Since the paste setting depends on the early hydration rate of SAC [44], the shortened setting time showcased that the early hydration of SAC could be accelerated by adding micro-IOTs. The shortening of setting time could be ascribed to the nucleation effect of micro-IOTs particles. It was widely accepted that nanoparticles and microparticles could act as nucleation seeds to accelerate cement’s hydration. After wet grinding, the particles of IOTs were significantly refined, and the micro-IOTs with an average diameter of 517 nm were obtained (Fig. 1). These microparticles could fill the voids of dehydrated SAC particles, offer more nucleation sites for growing hydration products, hasten the early hydration, and help microstructure form skeleton faster, thereby shortening the setting time.

The average compressive strengths of SAC pastes comprising different content of micro-IOTs are illustrated in Fig. 5. The content of micro-IOTs significantly impacts the compressive strength of micro-IOTs-SAC binary system. When the replacement ratio is 2%, the compressive strengths at 4 h, 1 d, 3 d, 28 d respectively improved by 22%, 16%, 15%, and 10% compared with those of pure SAC paste. Moreover, when the replacement ratio was 4%, compressive strengths of binary system were respectively higher, about 4%, 4%, 1%, and 2%, than the control group. These results specified that adding low amounts of micro-IOTs to SAC could notably improve SAC’s early strength and was also suitable for developing later strength. This was because micro-IOTs served as ultra-fine grains, filling the voids around the clinker particles and micropores of hydration products, making the microstructure more compact. Furthermore, micro-IOTs provided a great number of nuclear sites to accelerate SAC’s reaction, thus rapidly provide hydration products. These physical effects of micro-IOTs in SAC could increase the bonding force between cement and cement, contributing to the compressive strength growth.

Besides, when the replacement ratio was 8%, the binary system’s compressive strengths at all ages were lower than that of pure SAC paste. Two reasons were speculated: one was that excessive micro-IOTs would reduce the relative content of SAC in the system, thus reducing SAC contribution to the compressive strength. Another possibility was that excessive micro-IOTs were unevenly dispersed, leading to defects in samples and then affecting pastes’ strengths.

From the above analysis, adding low amounts of micro-IOTs indeed reduced the setting time and improved SAC pastes’ compressive strengths at all ages. According to previous reports [26,30], the early strength of SAC is closely associated with hydration degree and hydration products’ microstructure, and later strength maintenance is closely linked to stability of hydrated products and pore structure. A detailed analysis of the mechanism on micro-IOTs, improving the compressive strength of SAC under 2% micro-IOTs dosage, has been given as follows.

The main hydration products of micro-IOTs-SAC system at different curing times were studied by XRD and TG-DTG analyses.

The main components of the SAC are ye’elimite ( C4A3S¯ ) and anhydrite ( CS¯H2 ). According to the previous literature [45], if the proportion of anhydrite is relatively high, the reaction of ye’elimite was shown in Eq. (1). On the contrary, when the anhydrite content was fewer or absent, the reaction of ye’elimite was shown in Eq. (2).

(1) C4A3S¯+2CS¯H2+34H→C6AS3¯H32+2AH3

(2) C4A3S¯+18H→C4AS¯H12(AFm)+2AH3

Fig. 6 shows XRD patterns of SAC and micro-IOTs-SAC at ages of 4 h, 3 d, and 28 d. The hydration phases of SAC and micro-IOTs-SAC were basically the same, highlighting that micro-IOTs could not change hydration products’ phases. The main hydrate was ettringite, and dehydrated products, including ye’elimite, larnite and anhydrite, were also presented. It can be seen that the peak intensity of ye’elimite decreased with the curing age. Conversely, the peak intensity of ettringite was significantly increased. Besides, because several peaks overlapped, identifying other phases was difficult in XRD analysis. However, the dispersion peak at the bottom revealed the possibility of forming poorly crystalline monosulfate (AFm) and aluminum hydroxide (AH₃).

Furthermore, the differences in hydration between SAC and micro-IOTs-SAC could also be observed clearly at the same curing age. The peak intensity of ettringite in micro-IOTs-SAC pastes was higher than that in pure SAC, and the peak intensity of ye’elimite in micro-IOTs-SAC was lower than that in pure SAC, especially for 4 h and 3 d (see Figs. 6b and 6c). It revealed that micro-IOTs could accelerate ettringite formation and hasten hydration of SAC minerals at early ages, which was because micro-IOTs could provide more nucleation sites for ettringite during this period. According to Hargis et al. [45], AH₃ gel was confirmed to be fragile at hydration beginning, and ettringite content was the most critical factor affecting the strength development of SAC. The higher the ettringite content is, the greater the early strength of SAC is [46]. Incorporating micro-IOTs promoted ettringite formation, which contributed to higher compressive strength in SAC with low additions of micro-IOTs at early ages. Besides, Fig. 6c demonstrates that anhydrite in cement is fully hydrated after 28 d, and the diffraction ettringite peak is not higher than 3 d, specifying that the hydration reaction is basically stable after 28 d.

Since some amorphous and poor crystalline products are difficult to be distinguished using XRD due to poor crystallinity, the hydration products were further analyzed by TG. TG-DTG analysis of SAC and micro-IOTs-SAC system at 4 h, 3 d, and 28 d is depicted in Fig. 7. At the same age, the overall mass loss of micro-IOTs-SAC system was more than that of SAC. For instance, the overall mass loss at ages of 4 h, 3 d, and 28 d was 22.36 wt%, 28.32 wt%, and 29.39 wt% in micro-IOTs-SAC system compared with 20.62 wt%, 27.56 wt%, and 29.05 wt% in pure SAC, respectively. The overall mass loss could directly prove the accelerating effect of micro-IOTs on SAC hydration. It indicated that adding micro-IOTs enhanced ye’elimite dissolution and induced precipitating more hydration products.

Moreover, three weight loss peaks could be observed obviously from the DTG curve. According to previous literature [47–49], the mass loss peak at 100–150°C was ascribed to ettringite decomposition, and that at 250–280°C was mainly because of AH₃ dehydration. The peak at 600–800°C was assigned to calcium carbonate decomposition produced by carbonation during sample preparation or curing. Besides, the weak peak around 160–200°C could be ascribed to AFm [48].

The calculated weight loss of ettringite at 100–150°C was presented in Fig. 8a. Adding micro-IOTs was found to increase the weight loss by 23% and 2% at 4 h and 3 d, compared with pure SAC paste. The micro-IOTs promoted ettringite formation at early age, consistent with XRD results. This could conclude that pastes containing micro-IOTs exhibited higher compressive strength at the early stage.

Furthermore, the weight loss of AH₃ at 250–280°C was calculated and illustrated in Fig. 8b. At ages of 4 h, 3 d, and 28 d, micro-IOTs increased the content of AH₃ by 11%, 5%, and 12% compared to the blank. AH₃ was found to have a small contribution to the early strength of SAC but had a more significant impact on later strength. According to previous research [50], the more the content of AH₃, the less the content of ettringite, the higher the compressive strength of SAC at 28 d. Fig. 8a showcases that although ettringite’s content in 28 d of the sample containing micro-IOTs was similar to that of pure SAC sample, higher content of AH₃ could be filled in ettringite pores to ensure later SAC strength. Therefore, the higher content of AH₃ in micro-IOTs-SAC was one of the reasons for strength development at 28 d.

The pore structures of hardened paste samples were revealed using MIP. The results of porosity, median pore size, and cumulative pore volumes are shown in Table 3, and the pore size distribution results are presented in Fig. 9.

At the age of 4 h, cumulative pore volumes of paste were found to decrease by 13% (from 0.2426 to 0.2312 mL/g), and porosity decreased from 35.73% to 35.09% with adding 2.0% micro-IOTs. Moreover, the median diameter (volume) of reference group was 82.6 nm, while that of SAC mixed with 2% micro-IOTs was 61.8 nm. These findings highlighted that micro-IOTs could reduce the total porosity and refine the pore structure of SAC at an early age. Moreover, according to previous reports [26], the pores in cement-based materials are generally divided into four types: macropores (diameter ≥1000 nm), capillary pores (diameters, 100–1000 nm), transitional pores (diameter, 10–100 nm), and gel pores (diameters ≤ 10 nm). It can be observed from Fig. 9a that micro-IOTs significantly decreased the number of macropores and increased the number of transitional pores. It is assumed that the larger pores represent the spaces in the unhydrated cement particles, and the smaller pores represent the spaces between the hydration products [51]. Since the uniformly distributed micro-IOTs could serve as a nucleation site for the hydration product of SAC, it was beneficial to accelerating the early hydration of SAC, thus producing dense microstructure in micro-IOTs-SAC system [52]. These results further explained the reason why micro-IOTs could improve SAC’s early strength.

The results at 28 days seemed to be different as few changes were found in cumulative pore volumes of pastes without micro-IOTs or with 2% micro-IOTs, and porosity increased from 30.59% to 31.19% with adding 2.0% micro-IOTs. However, adding micro-IOTs reduced the median diameter from 87.8 nm in blank to 64.8 nm. Adding micro-IOTs shifted the differential intrusion curves from larger pores to smaller pores. It was seen from Fig. 9b that micro-IOTs significantly refined both capillary pores and decreased the number of macropores. These results demonstrated that adding micro-IOTs was beneficial to pore structure refinement of SAC pastes in the later stage, which was related to the filling effect of micro-IOTs and the accelerated hydration of SAC via nucleation effect. It was reported that MIP could only detect interconnected pores [53]. Capillary pores and macropores were easily blocked by micro IOT and hydration products (including additional AH₃), resulting in interconnection pore decrease and pore structure improvement. This is an important reason why the incorporation of micro-IOT caused the increase in later strength.

Fig. 10 displays SEM images of hydrated samples at different ages. As demonstrated in Fig. 10a, after 4 h of hydration, loose micro morphology and needle-like ettringite with a very low draw ratio were observed from the pure SAC sample. However, ettringite morphology in the sample with 2% micro-IOTs was mainly ribbon-like and columnar, and only a few were needle-like, which was remarkably different from ettringite morphology in the pure SAC. According to a previous report [33], similar ribbon-like ettringite was also observed in the SAC paste with graphene oxide (GO), and formation and growth of ribbon-like ettringite further led to improved early strength of GO-SAC. As depicted in Fig. 10b, the ribbon-like crystals and needle crystals tended to crisscross and interweave from different directions, which helped develop early strength.

After 3 d of hydration, the sample’s microstructure containing 2% micro-IOTs displayed a more dense structure and fewer voids than the pure SAC sample. Moreover, ettringite was distributed randomly in the aluminum hydroxide matrix that covered it to a great extent. Besides, some needle-like ettringite still could be observed from the pure SAC sample in Fig. 10c, while almost all of the ettringite in the samples with micro-IOTs were columnar in Fig. 10d. It was reported that columnar ettringite was more conducive to SAC strength’s stability in a later stage than needle-like ettringite [50]. At the age of 28 d, the pure SAC sample presented a loose structure with several pores, while the microstructure of pastes containing 2% micro-IOTs manifested a more compact structure, consistent with pore structures. These also provided viewable evidence that the growth of compressive strengths was closely correlated with higher hydration degree of SAC and more stable ettringite at ages of 3 d and 28 d.

Since ettringite is the main hydration product of SAC, its content and morphology have an important influence on SAC performance. It was confirmed that the early strength of SAC mainly depended on the mechanical occlusion between the ettringite crystal clusters, and the late strength was closely related to ettringite stability [45]. The addition of micro-IOTs changed the hydration environment and formed a new morphology hydrate (ribbon-like ettringite), see Fig. 10b. Meanwhile, the ettringite intertwined to form a network structure, leading to improved early strength of SAC. Moreover, more columnar ettringite and higher degree of hydration due to micro-IOTs ensured the stable growth of late strength of SAC.

In conclusion, the use of micro-IOTs to improve SAC performance is a method of high value-added recycling and utilization of IOTs, which is worthy of popularization and application. Based on the above results, micro-IOTs, which can enhance SAC’s mechanical properties and improve the pore structure of SAC, can be used as micron mineral admixture in the SAC system to expanding the application fields of SAC. Although preparing micro-IOTs will inevitably generate some energy consumption, the total cost is much cheaper than common nanomaterials due to completely free raw materials. In addition, the micro-IOTs can be obtained through environmentally friendly wet milling technology [39]. Compared to traditional dry-milling, wet grinding shows higher efficiency and can avoid dust generation. Since most of IOTs are directly discharged into the river or exposed to rainwater, wet grinding can avoid the necessary pre-drying procedure in the traditional dry grinding [40]. More importantly, micro-IOTs are in a slurry state, which is easier to disperse evenly in SAC and conducive to the formation of a denser microstructure of SAC. Hence, using micro-IOTs as a mineral admixture in the SAC system is expected to be developed.