The P.O42.5R Ordinary Portland cement came from China’s XiNan cement factory. Its specific surface area was 387 m²/kg, the density was 3.15 g/cm³, and the compressive strength was 18.7 MPa at 3 days, 29.8 MPa at 7 days, and 44.5 MPa at 28 days according to the Chinese standard GB175-2007. The chemical composition of cement is displayed in Table 1. A superplasticizer from Laiyang Hongyang construction admixture factory of China was employed to provide a satisfactory workability. Standard sand with a high silica content was used as the reference sand to produce cement mortar; its particle size was kept below 2.35 mm, and it was made by China’s Xia Men ISO Standard Sand Co., Ltd.

Lithium slag (LS) used as a super-fine aggregate to replace standard sand came from China’s Xinjiang lithium salt factory. The specific surface area was 400 m²/kg, and its chemical composition is displayed in Table 1. By comparison, the standard sand is coarser than the lithium slag, which may lead to an increment in water demand (Shi et al., 2020). However, the usage of superplasticizers decreased the need for water. In this study, the standard sand was replaced by an equal weight of lithium slag.

The water–cement ratio was 0.50, the mass ratio of cement and standard sand was 1:3, and the mass ratio of cement:water:standard sand was 450:225:1350. Specimens (40 mm × 40 mm × 160 mm) of each proportional mixture were prepared with standard sand, superplasticizer, Portland cement, water, and lithium slag as a super-fine aggregate, respectively. According to the studies of Shi et al. (2020) and Li et al. (2019), when lithium slag replaced 30% of cement, its performance was basically equivalent to that of pure cement cementitious material because lithium slag could play a micro-aggregate effect. Therefore, 0%, 30%, 50%, 70%, and 100% proportions of lithium slag were considered super-fine aggregates instead of standard sand. The cement paste/mortar was identified as C, and the cement–lithium slag paste/mortar samples with 30%, 50%, 70%, and 100% lithium slag were marked as L1, L2, L3, and L4, respectively. The mix ratios of cement–lithium slag mortar and sample numbers are shown in Table 2. A flow table test was performed on fresh mortar to determine the effect of the lithium slag on workability. Specimens were molded when the flow table test was the same, placed at 20 °C and 95% relative humidity for 24 h and then demolded, and cured at 20 °C ± 1 °C and 95% relative humidity for 7 and 28 days. Then, their hydration product, Ca(OH)₂ content, Fourier spectrum, compressive strength, flexural strength, unit weight, absorption water rate, and stress–strain were investigated.

Cement paste and cement–lithium slag paste (without standard sand) were molded in 10 mL of the centrifugal tube, the cap was tightened, and the samples were sent to the curing room (20 °C ± 1 °C, 95% or higher humidity). After curing for 7 and 28 days, the tube was broken, and the sample was soaked in anhydrous ethanol for 24 h to discontinue hydration and Guaranteed crushing fineness of grinding (sieving 0.1 mm). XRD and thermogravimetric experiments were carried out by SMART APEX Ⅱ-type X-ray polycrystalline crystal diffraction produced by German Brooke AXS Co., Ltd. and a STA449F3-type synchronous thermal analyzer produced by the German Resistance Company, respectively. The finely ground paste was mixed with the dried and grained KBr at a 1:200 ratio, and the Fourier spectrum and the film were tested by WQF-520 type Fourier transform infrared spectroscopy.

Unit weight was tested by the weighing method. The flexural strength and compressive strength were tested using an EPC70-type automatic bending device and a YAW-300B-type computer control electro-hydraulic cement pressure testing machine produced by the Hangzhou Xin High Tech Co. Ltd. according to the “cement mortar strength test method (GB/T 17671-2021)." Stress–strain was tested according to the “Standard for test methods of concrete physical and mechanical properties (GB/T 50081-2019)." Water absorption was tested according to the “Test methods for bulk density, moisture, and water absorption of aerated concrete (GB/T 11969-2008).”

Usually, the main products after cement hydration are calcium silicate hydrate, calcium ferrite hydrate, portlandite, calcium aluminate hydrate, calcium sulphoaluminate hydrate, ettringite, and cement clinker particles that have not been hydrated (Shi et al., 2020). The influence of lithium slag on cement hydration products is shown in Figure 1.

L1 in Figure 1A shows that after adding 30% lithium slag, the main crystalline hydration products were calcium silicate hydrate, portlandite, unhydrated particles, and a trace of CaCO₃ and Li₂CO₃. Unhydrated particles mainly contained Al₂O₃, SiO₂, and Li₂O·Al₂O₃·SiO₂, and the CaCO₃ due to portlandite was carbonated in the process of grinding. With the increase of lithium slag replacement, the main diffraction peak of unhydrated particles became more acute, which had a low content of portlandite in cement early hydration (such as the curves of Figure 1A). Besides, the activity of lithium slag was relatively lower and the replacement was larger (Wu et al., 2016a; Wu et al., 2016b), so the diffraction peaks of unhydrated particles were acute keen (such as the curves of L4 in Figure 1A), which could be explained by the fact that Al₂O₃ and SiO₂ had not sufficiently participated in the secondary hydration. In addition, it was proposed that ettringite (AFt) was not identified due to the slightly low diffraction peak. Meanwhile, with the increase of curing age from 7 days to 28 days, the diffraction peaks of Al₂O₃, SiO₂, Li₂O·Al₂O₃·SiO₂, and portlandite became significantly lower. Especially in the curves of L2 and L3 in Figure 1B, the diffraction peaks of unhydrated particles and portlandite became significantly lower than the equivalent peaks of L2 and L3 in Figure 1A, indicating that Al₂O₃ and SiO₂ had participated in the secondary hydration, resulting in a significant decrease in portlandite at 18.2° and a significant increase in portlandite at 47.2°. However, after 100% lithium slag replaced the fine aggregate, the diffraction peaks of unhydrated particles were still the most prominent, which once again showed that the activity of lithium slag was low (Shi et al., 2020). Therefore, the diffraction peaks of calcium ferrite hydrate, calcium hydroxide, calcium aluminate hydrate, and calcium sulphoaluminate hydrate were still insignificant.

The hydration product of cement–lithium slag paste can be analyzed qualitatively and quantitatively for calcium silicate hydrate gel, portlandite, and ettringite by thermogravimetric analysis (TG) (Liu et al., 2015; Hu et al., 2021). Figure 2 and Figure 3 show the TG curves of hydration products for pure cement paste and cement–lithium slag paste with 30%–100% lithium slag at 7 days and 28 days. From Figure 2 and Figure 3, it can be seen that the TG curve was divided into four stages, which was similar between pure cement paste and cement-lithium slag paste in 20°C∼900°C. The first stage is the ettringite dehydration stage at 50 °C–200 °C, the second stage is the portlandite dehydration stage at 400 °C–550 °C, the third stage is the calcium carbonate dehydration stage at 550 °C–770 °C, and the fourth stage is the CaCO₃ and Li₂CO₃ dehydration stage at 800 °C–900 °C (Li and Yan, 2010; Chen et al., 2011). Therefore, the content of CH can be calculated according to the dehydration stage at 400 °C–550 °C. The results are shown in Figure 4.

The CH content of pure cement paste was 5.63% at 7 days and 6.12% at 28 days. The CH content of cement–lithium slag paste after adding 30%–100% lithium slag as a super-fine aggregate is shown in Figure 4. It can be seen from Figure 4 that the CH content of the cement–lithium slag paste significantly decreased with the increasing amount of lithium slag. The equations are CH = −0.0148x + 3.933 at 7 days and CH = −0.0152x + 4.007 at 28 days, and the correlation coefficients are R ² = 0.8663 and R ² = 0.8864 at 7 days and 28 days, respectively. We can estimate the content of CH under different replacement amounts of lithium slag using these linear formulas. In addition, the change in the CH content of cement–lithium slag paste with the same replacement ratio at 7 days and 28 days was small, at about 65.0% and 60.6% of pure cement paste, respectively. It can be explained that the activity of lithium slag was also relatively lower, but it could still participate in partial secondary hydration reaction; mainly, Al₂O₃ and SiO₂ in lithium slag reacted with CH to generate calcium aluminate hydrate and calcium silicate hydrate. Therefore, adding 30%–100% lithium slag as a super-fine aggregate in a cement system, which increased the amount of hydrated calcium silicate and hydrated calcium aluminate, also improved the mechanical properties and durability of cement mortar, as will be discussed in the following sections.

Figure 5 shows the spectral characteristic of cement–lithium slag paste under different replacement amounts. It can be seen from Figure 5 that the characteristic peaks of cement–lithium slag paste were mainly 3467 cm⁻¹, 3438 cm⁻¹, 2923 cm⁻¹, 2348 cm⁻¹, 1638 cm⁻¹, 1429 cm⁻¹, 1111 cm⁻¹, 1000 cm⁻¹, 768 cm⁻¹, 696 cm⁻¹, and 462 cm⁻¹ at 7 days and 28 days, respectively. With increasing lithium slag replacement, the characteristic 7-day peak in 3467 cm⁻¹ became smaller and even disappeared when the replacement rate reached 100%. This behavior is the reverse of the behavior of the characteristic peak in 3438 cm⁻¹; namely, its intensity peaked when the replacement rate reached 100%, revealing that with increasing lithium slag replacement, the content of hydration product CSH increased, but the content of CH decreased. This illustrates that lithium slag can participate in the secondary hydration reaction; it consumes CH and forms a certain amount of CSH. The characteristic peak in 2923 cm⁻¹ increased with the increase of lithium slag replacement; it was the largest when the replacement rate reached 100%. This illustrates that the O–H content increases, which is mainly composed of CSH gel in the paste and free water adsorption of lithium slag; therefore, the O–H content is considerably larger when the lithium slag replacement rate is 100%. The change law of wave number 2348 cm⁻¹ was similar to that at 2923 cm⁻¹, but the peak intensity at 28 days was bigger than that of the others at 7 days. The size of this peak is mainly caused by Li₂CO₃ in lithium slag and carbonized CH in paste, which leads to the peak intensity being larger at 2348 cm⁻¹. The change law of wave numbers 1638 cm⁻¹ and 2348 cm⁻¹ was similar to that of 2348 cm⁻¹ and 2923 cm^-1, and the variation was also small as the curing age increased from 7 days to 28 days. The change is mainly caused by CaSO₂·2H₂O and AFt in paste, which can effectively improve the properties of the paste. Simultaneously, the change law of wave numbers 696 cm⁻¹ and 462 cm⁻¹ at 7 days and 28 days was similar to that of 2348 cm⁻¹, 2923 cm⁻¹, and 1638 cm⁻¹ with increasing lithium slag replacement.

The unit weight of cement mortar with lithium slag as a super-fine aggregate is presented in Figure 6.

From Figure 6, it can be seen that the unit weight of cement mortar with lithium slag at the age of 7 days and 28 days showed a downward trend with increasing amounts of replacement, and the change of their peak unit weight depended on the content of lithium slag.

As a control group, the unit weight of cement mortar was 2065.04 kg/m³ at 7 days and 2017.68 kg/m³ at 28 days. After adding 30%, 50%, 70%, and 100% lithium slag as a super-fine aggregate, 7-day unit weights varied between 1983.91 and 1560.12 kg/m³, while 28-day weights changed between 2017.34 and 1582.54 kg/m³, a decrease of 3.95%–24.45% at 7 days, and 2.62%–23.61% at 28 days, compared with the control group. Therefore, as expected, adding lithium slag as a super-fine aggregate resulted in lower unit weight at 7 and 28 days than pure cement mortar. This reason is that the specific gravity of lithium slag is 2.48 g/cm³, which is less than that of standard sand. Thus, the more lithium slag is added, the smaller the unit weight of mortar is. (Snellings et al. (2022) classify as lightweight, half-structural, and isolation concrete according to their unit weights, which are 0–2000, 2000–2,800, and above 2,800 kg/m³, respectively. Lightweight concrete was achieved in this study when the replacement amount was 50%, 70%, and 100%. This reduction significantly reduced the dead load of the concrete structure (Xian and Shao, 2021). In addition, the relationship between the unit weight of mortar and lithium slag content is also presented in Figure 6. There is a high linear correlation between the unit weight of mortar and lithium slag content, at R ² = 0.9753 at 7 days and R ² = 0.9307 at 28 days. Therefore, y = −5.3029x + 2100.3 and y = −5.224x + 2127.5 could be used to predict the unit weight of mortar with lithium slag at 7 days and 28 days, respectively. As clearly shown from these findings, a high content of lithium slag will result in reduced unit weight.

Although the compressive strengths of cement mortar and concrete are considered in concrete structural design, their flexural strengths are also important in applications such as dams and highways (Youm et al., 2014). The flexural strength change of cement mortar with lithium slag as a super-fine aggregate is presented in Figure 7. It can be seen from Figure 7 that the flexural strength development of cement mortar presented an increasing trend and then decreased with the increment of lithium slag content. The change of the peak flexural strength depended on the content of lithium slag.

As a control group, the flexural strength of cement mortar was 5.82 MPa at 7 days and 6.46 MPa at 28 days. The peak flexural strength value that appeared in the replacement amount was 30%, which was 11.11 MPa at 7 days and 14.70 MPa at 28 days, an increase of 90.89% and 127.55% compared with the control group. As the replacement amount increased, the flexural strength of the mortar decreased. For example, when the replacement amount of lithium slag was 100%, the flexural strength decreased by 29.21% at 7 days and increased by 37.77% at 28 days compared with the control group. Longer curing time can result in increasing flexural strength. On the one hand, it is clear from Figure 7 that the volcanic ash activity of the lithium slag had an important influence on flexural strength, which led to an increase of 90.89% at 7 days and 127.55% at 28 days. On the other hand, adding 30%, 50%, 70%, and 100% lithium slag as a super-fine aggregate could improve the interfacial transition zone of mortar. In addition, the hydration rate of cement decreased as the amount of lithium slag increased. The flexural strength of mortar with lithium slag as a super-fine aggregate decreased with a higher replacement ratio.

Compressive strength, a vital element of concrete structural design, is the most important property of cement mortar and concrete. Compressive strength results of cement mortar with lithium slag as a super-fine aggregate are presented in Figure 8. From Figure 8, it can be seen that the compressive strength of cement mortar with lithium slag at the age of 7 days and 28 days showed a downward trend after rising first, and the change of the peak compressive strength depended on the content of lithium slag.

The compressive strength of the control group cement mortar was 32.4 MPa at 7 days and 45.3 MPa at 28 days. After adding 30%–100% lithium slag as a super-fine aggregate, the compressive strength varied between 57.91 MPa and 21.70 MPa at 7 days and between 82.72 MPa and 68.81 MPa at 28 days, a decrease of 78.70%∼−33.02% at 7 days and 82.56–51.89% at 28 days, compared with the control group. Like the trend of flexural strength with content, the peak value of compressive strength appeared in the 30% replacement amount, where it was an increase of 78.70% at 7 days and 82.56% at 28 days compared with the control group. The compressive strength decreased by 33.02% for 7 days, while it increased by 51.88% for 28 days when the replacement rate of lithium slag was 100%. As clearly shown here, the compressive strength of mortar with 100% lithium slag is higher than that of pure cement mortar. The reason is that lithium slag is finer than cement and better fills gel pores, improving the compactness of mortar. The hydration of cement can be increased by adding lithium slag, and cementitious hydrated calcium silicate can be formed. In addition, the connection of cement particle hydration product was dispersed due to the lower activity of lithium slag; the higher the replacement rate was, the more significant the dispersion effect. The thermo-gravimetric and infrared spectra results also showed the same finding; the higher the content of non-binding substances such as lithium slag and lithium carbonate was, the more limited the formation of hydrated calcium silicate was, which affected the mechanical properties of the mortar. Overall, the mechanical property of mortar reached the highest value when the lithium slag content was 30%. The relationship between compressive strength and flexural strength is presented in Figure 9. There was a high linear correlation between compressive strength and flexural strength, with R ² = 0.9661 at 7 days and R ² = 0.8232 at 28 days. Compressive strength is directly related to flexural strength. The strength of mortar with 100% lithium slag as a super-fine aggregate instead of standard sand was higher than that of pure cement mortar, which means a higher usage ratio of lithium slag as a super-fine aggregate instead of standard sand is possible for the production of lightweight concrete.

The stress–strain curve, a relationship model for concrete structure non-linear analysis and ultimate bearing capacity analysis, is an important basis for studying the bearing capacity and deformation of a concrete structure (Guo et al., 2022). Figure 10 shows the stress–strain curve of mortar at different replacement percentages and curing ages. It changed at different replacements and curing ages, following a consistent pattern. After adding 30%, 50%, 70%, and 100% lithium slag as a super-fine aggregate, the peak stress of mortar at 7 days was 2.04, 1.52, 1.14, and 0.95 times that of the control group, while the peak stresses of mortar at 28 days were 1.39, 1.73, 1.23, and 1.21 times that of the control group. Therefore, it was feasible to prepare lightweight concrete using lithium slag instead of standard sand, which could increase the use of solid waste in mortar and concrete. The displacement of mortar with lithium slag was larger than that of cement mortar at 7 days and 28 days. All were within 0.2 mm at 7 days, while the displacement could reach 0.4 mm at 28 days. Based on the literature method of calculating external load work (Jian et al., 2016), it was found that the external load work of mortar decreased from 66.91 MPa to 37.78 MPa when the replacement percentage of lithium slag increased from 30% to 100%. However, the external workload of mortar with increasing percentages of lithium slag was nevertheless 2.99, 2.60, 1.84, and 1.69 times higher than that of pure cement mortar at 7 days, while it was 1.68, 2.44, 1.42, and 1.33 times higher than pure cement mortar at 28 days.

The relationship between external load work and the replacement rate of lithium slag was established to analyze the mechanical properties of mortar. The results are shown in Figure 11, which shows that the peak stress ratio of mortar at 7 days and 28 days decreased linearly as the replacement increased from 30% to 100%. The relationships were R ² = 0.9543 at 7 days and R ² = 0.5639 at 28 days, respectively. The change laws of external load work were similar to peak stress; their relationships are R ² = 0.9465 at 7 days and R ² = 0.5439 at 28 days. It is also difficult to find that the mechanical property of mortar was better than that of pure cement mortar, except that the replacement rate was 100% at 7 days. In addition, the external load work ratio of mortar with 30%–100% lithium slag was above 1.0 at 7 days and 28 days, which further illustrated the advantage of mortar with lithium slag over that of pure cement mortar.

The water absorption rate is one way to measure cement mortar and cement concrete permeability, which is an important parameter to evaluate the durability of mortar and concrete. After adding 30%–100% lithium slag, the water absorption rate of mortar at 7 days and 28 days is shown in Figure 12, Table 3 and Table 4.

Figure 12A and Table 3 show that the water absorption rate of mortar increased as the percentage of lithium slag increased and reached a maximum at 483 h; after this, its change was less than that of pure cement mortar. When 100% lithium slag was added as a super-fine aggregate, the water absorption rate of mortar was 3.55 times that of pure cement mortar within 2 h. However, the change was within 12.6%, except when the replacement of lithium slag was 100%. The water absorption rate change of mortar with 0%–100% lithium slag was fast, which was from 1.37%, 0.39%, 0.39%, 0.36%, and 0.43% to 12.30%, 3.24%, 1.47%, 1.71%, and 4.93% within 2 h–216 h, while it was very slow after 216 h. Particularly, after 390 h, the water absorption rate fell by 0.5% every 50 h. The water absorption time was prolonged to 480 h, and the water absorption rate of the mortar increased gradually. It was 48.51%, 60.00%, 148.09%, and 416.89% that of pure cement mortar. A similar pattern is shown in Figure 12B and Table 4. Therefore, the water absorption rate was lower with lithium slag replacement and was less than 50% than that of pure cement mortar.

Based on the aforementioned experimental study, the replacement rate of lithium slag as a super-fine aggregate is recommended to be not more than 50%, which is 675 g in this experiment. When using lithium slag as an admixture instead of cement, the content should not be more than 30%; in this test, that amount is 135 g (Shi et al., 2020). Therefore, five times as much lithium slag can be used as a super-fine aggregate than as an admixture instead of cement, and using lithium slag can not only greatly affect energy conservation and emission reduction but also improve the performance of mortar and concrete.