Design of New Eco-Cementitious Material Based on Foundry Slag and Lime Sludge

Abstract

Foundry slag has different characteristics from blast furnace slag, such as its high SiO₂ content and low basicity (CaO/SiO₂ < 1), which prevent it from being used as a cementitious component. Lime slurry is a waste product with a high CaO content and can be used to increase the basicity of the mixture. The aim of this study is to obtain new supplementary, eco-cementitious material composed of foundry slag and lime sludge. The compositions were designed with binary basicity (molar ratio of CaO/SiO₂) ranging from 1.0 to 1.4. Clinker was replaced with the proposed material in the range of 6–34 wt% and the performance of the different cement compositions was tested. The results showed that replacing 20 wt% of clinker with the new eco-cementitious material with binary basicity of 1.2 resulted in cement with the same mechanical strength as the reference cement. The new material reacted with free CaO to generate additional calcium silicate hydrate. The initial setting time of the cement containing the new eco-cementitious material was 240 min, acting as hydration reaction retardant. The technical feature of the new eco-cementitious material allows the use of both wastes in cement composition, contributing to the requirements of circular economy.

1. Introduction

Blast furnace slag, a by-product generated in the steel mill sector, is used as a cement material. Granulated blast furnace slag (GBFS) is a type of slag with an amorphous structure. The chemical composition of GBFS includes CaO, Al₂O₃, and SiO₂. According to Markandeya et al. [1], this material is used in cement to decrease the temperature of the hydration reaction and to improve resistance to early cracking. Foundry slag, on the other hand, is not recommended for cementitious components because it inhibits the formation of calcium and aluminum silicates [2,3]. Table 1 shows that foundry slag usually has a higher SiO₂ content and a binary basicity of less than one.

For this reason, several studies have investigated the use of foundry slag as a fine natural aggregate [4,5] and as a substitute for concrete aggregates [6,7]. Studies [8,9] have also investigated the replacement of cement with foundry slag. However, the results indicate a decrease in the mechanical resistance of the cementitious material with the addition of granulated foundry slag. Other studies have discussed the use of foundry slag as fine aggregate for concrete, reporting compressive strength increases of up to 17% when replacing 30% of natural sand [4], and maintaining over 90% of the reference concrete’s strength after one year when replacing 25% of natural aggregate [10].

In this sense, the adjustment of the chemical composition of foundry slag may enable the use of this waste in higher-value applications. One possibility is the use of foundry waste in cementitious materials. However, the Brazilian NBR 16697 standard [11] states that slag must have a (CaO + Al₂O₃ + MgO)/SiO₂ ratio higher than 1 to be used as a cementitious component, which is not met by this material. The use of a material with a higher calcium content could increase the calcium oxide content in the foundry slag composition. Lime sludge is a by-product generated in cellulose production with a higher CaO content and lower contents of other elements (Table 1).

Lime sludge can correct the chemical composition of low-basicity slag, making it more suitable for use as eco-cement. Murali et al. (2024) [12] investigated the replacement of part of the cement with recycled lime sludge combined with calcined clay, showing that a mixture with 15% lime sludge and 30% calcined clay promotes hydration reactions and significant improvements in strength at 28 days. Furthermore, lime slurry acts as a source of CaO, helping in the formation of ettringite and hydrated calcium silicate, contributing to greater density of the cementitious matrix [13].

The aim of this paper is to discuss a method for designing a sustainable product using foundry slag and lime sludge, within the framework described in the new Circular Economy Action Plan of the European Commission and the United Nations’ Sustainable Development Goals.

Figure 1 presents a simplified model through which this study can be implemented in a foundry company. The foundry slag and lime sludge are mixed before reaching the slag pot, eliminating additional costs. However, before implementing this method, the technical viability of this product as a new eco-cementitious material must first be investigated.

Table 1. Chemical compositions of the foundry slag and lime sludge.

Component	Foundry Slag			Lime Sludge
	Cardoso et al. [4]	Ceccato et al. [5]	Devi et al. [14]	Singh et al. [8]	Maheswaran et al. [9]	Suthar, Aggarwa [15]
CaO	2.58	21.78	7.0	85.50	40.74	57.11
SiO2	71.6	49.2	41.92	11.13	5.78	2.3
MgO	2.16	11.0	1.24	1.18	-	1.22
MnO	7.04	2.81	6.61	-	-	-
Al2O3	11.0	10.68	20.55	0.24	0.18	0.21
Na2O	0.62	-	-	0.99	-	0.84
K2O	0.5	-	0.53	0.20	2.03	0.16
P2O5	-	0.02	-	0.31	0.19	0.41
ZrO2	0.27	-	-	-	-	-
Fe2O3	3.39	-	18.92	0.18	0.15	0.28
FeO	-	2.97	-	-	-	-
Cl	0.11	-	-	-	-	-
S	-	0.64	-	-	-	-
V2O5	-	0.67	-	-	-	-
TiO2	-	0.67	1.49	-	-	-
Cr2O3	-	0.05	0.53	-	-	-
CuO	-	-	-	0.02	-	-
SO3	-	-	-	0.17	0.28	0.64
Loss on ignition	-	-	-	-	50.04	36.67
CaO/SiO2	0.03	0.44	0.17	7.68	7.04	24.83

Table 1. Chemical compositions of the foundry slag and lime sludge.

Component	Foundry Slag			Lime Sludge
	Cardoso et al. [4]	Ceccato et al. [5]	Devi et al. [14]	Singh et al. [8]	Maheswaran et al. [9]	Suthar, Aggarwa [15]
CaO	2.58	21.78	7.0	85.50	40.74	57.11
SiO2	71.6	49.2	41.92	11.13	5.78	2.3
MgO	2.16	11.0	1.24	1.18	-	1.22
MnO	7.04	2.81	6.61	-	-	-
Al2O3	11.0	10.68	20.55	0.24	0.18	0.21
Na2O	0.62	-	-	0.99	-	0.84
K2O	0.5	-	0.53	0.20	2.03	0.16
P2O5	-	0.02	-	0.31	0.19	0.41
ZrO2	0.27	-	-	-	-	-
Fe2O3	3.39	-	18.92	0.18	0.15	0.28
FeO	-	2.97	-	-	-	-
Cl	0.11	-	-	-	-	-
S	-	0.64	-	-	-	-
V2O5	-	0.67	-	-	-	-
TiO2	-	0.67	1.49	-	-	-
Cr2O3	-	0.05	0.53	-	-	-
CuO	-	-	-	0.02	-	-
SO3	-	-	-	0.17	0.28	0.64
Loss on ignition	-	-	-	-	50.04	36.67
CaO/SiO2	0.03	0.44	0.17	7.68	7.04	24.83

2. Materials and Methods

2.1. Raw Materials and Characterization

The eco-cementitious material was produced from lime sludge and foundry slag using the following procedure. The lime sludge was first dried at 110 °C for 48 h. The sample was then quartered to produce cementitious material, which was further characterized. Foundry slag was provided as an agglomerate, which was then milled to a particle size of less than 1 mm in a ball mill for further characterization and to conduct melting tests to produce the eco-cementitious material.

The chemical compositions of lime sludge and foundry slag were determined by inductively coupled plasma optical emission spectrometry (ICP-OES). The particle size distribution of the lime sludge was determined using a CILAS 1064 laser diffraction equipment. Water was used as the liquid and sodium polyacrylate as a dispersant. Thermogravimetric tests were conducted in an STA 449F3 Jupiter thermobalance over a range of 30–1000 °C, under a nitrogen flow of 60 mL/min and a heating rate of 10 °C/min. A Shimadzu X-ray diffractometer equipped with Cu Kα (λ = 1.5418 Å) was used to determine the crystalline phases. A step scan mode was employed, with a scan range of 3–80°, scan speed of 2°/min, voltage of 25.0 kV, and current of 25.0 mA.

CPV-ARI (composed of 90–100 wt% clinker and 0–10 wt% carbonate material) was used as the reference cement. CP II-E 32, a commercial Portland cement composed of clinker and granulated blast furnace slag in the range of 6–34 wt%, was also used as a reference cement. This type of cement may also contain up to 10 wt% of carbonate material. The chemical compositions of the cement types were determined using AXIOS MAX Panalytical X-ray fluorescence equipment.

Sands with particle sizes in the ranges of 1.2–2.4 mm, 0.6–1.2 mm, 0.3–0.6 mm, and 0.15–0.3 mm were used to produce the proof bodies for mechanical strength tests.

2.2. Manufacture of the Eco-Cementitious Material

Three mixtures were produced using foundry slag and lime sludge in the proportions required to obtain the eco-cementitious material with binary basicity (CaO/SiO₂) values of 1.0, 1.2, and 1.4. The binary basicity values were calculated from the mass balance of the chemical compositions of the input materials. The minimum binary basicity value was stipulated based on the Brazilian NBR 16697 standard [11], which specifies the minimum binary basicity value for blast furnace slag used in cement composition. The medium binary basicity value (1.2) was calculated as the average composition of blast furnace slag [11,16]. The effect of excess binary basicity on the composition of the eco-cementitious material was also investigated. This corresponded to the geometric point with a binary basicity of 1.4.

The raw materials were placed in a plastic bag and homogenized manually for 5 min. Then, the homogeneous mixture was transferred to an Al₂O₃ crucible and placed in an electric furnace. The mixture was heated in air from 30 °C to 1300 °C at a rate of 20 °C/min. The final temperature was maintained for 30 min to complete melting. The melted mixture was poured into water to promote rapid cooling and to obtain an amorphous structure. Part of the melted mixture was kept in an Al₂O₃ crucible and returned to the furnace for slow cooling to room temperature. This step was conducted to investigate the potential of the eco-cementitious material to form silicates.

2.3. Characterization of the Eco-Cementitious Material

The eco-cementitious materials were characterized by DTA/TG analysis using STA 499-F3 Jupiter, Netzsch equipment (Selb, Germany) under a flow of 50 mL/min of argon, at a temperature range of 30–1500 °C, with a heating rate of 10 °C/min. X-ray diffraction analysis was carried out under the conditions mentioned in Section 2.1. Thermodynamic simulations were performed using Thermo-Calc software (version 2020b) to investigate the phases of the eco-cementitious material, using the TCOX10 database. The chemical composition (

Table 2. Chemical composition of the eco-cementitious material calculated via mass balance from the lime sludge and foundry slag compositions.

Components	Basicity 1.0	Basicity 1.2	Basicity 1.4
Al2O3	14.0	13.1	12.4
CaO	29.4	33.4	36.6
Fe2O3	5.0	4.7	4.5
MnO	19.2	18.0	17.0
MgO	1.1	1.1	1.1
K2O	0.5	0.4	0.4
Na2O	1.5	1.7	1.9
SiO2	29.2	27.5	26.1

) of the eco-cementitious material used as input data in the simulation was determined by mass balance.

2.4. Cement Preparation

The procedures and performance classifications adopted in this study follow the Brazilian standard NBR 16697 [11], which is functionally equivalent to ASTM C150 [17] and EN 197-1 [18] in terms of cement composition, strength classes, and general performance requirements. Five cement compositions (

Table 3. Experimental programs of the cement produced with the eco-cementitious material.

Types of Cement	Binary Basicity	Eco-Cementitious Material (wt.%)
B1.0-S6	1.0	6
B1.0-S34	1.0	34
B1.4-S6	1.4	6
B1.4-S34	1.4	34
B1.2-S20 (1)	1.2	20
B1.2-S20 (2)	1.2	20

) were prepared to investigate the applicability of the eco-cementitious material produced in this work as a supplementary cementitious material. A 2² + 2 experimental program was used to obtain these compositions. The variables in this program were the required binary basicity (between 1 and 1.4) and the percentage of eco-cementitious material addition (6–34 wt%, according to the Brazilian NBR 16697 standard [11]). The experimental program was designed with two degrees of freedom.

The symbols B and S represent the binary basicity and the content of eco-cementitious material (wt%), respectively. The subscript numbers indicate the value of the corresponding variable in the mixture composition. For example, the B_1.0-S₆ cement contains 6 wt% of eco-cementitious material with a binary basicity of 1.0. B_1.2-S₂₀ (1) and B_1.2-S₂₀ (2) were the central points of the experimental program.

The different types of cement were then prepared in two steps: (1) the eco-cementitious materials were first milled to obtain particle sizes smaller than 0.075 µm, and (2) the required mass (calculated from the mass balance) of the milled eco-cementitious material was mixed with pure commercial cement (CP V-ARI).

2.5. Statistical Analysis

Statistical analysis of the data was performed using Statistica^® software (version 10.1 trial), with a significance level of α = 0.05. Analysis of variance (ANOVA) was applied to evaluate the effects of the addition of cementitious material and basicity on the compressive strength of concrete at 3, 7, and 28 days.

The experimental design followed a 2² factorial plan with two central points, totaling six conditions evaluated. The interactions between the factors were disregarded in the statistical analysis because they were not significant. The value of the standardized effects and Pareto index were used to interpret the results, as well as the response surface graph.

2.6. Test of Mechanical Strength

Mechanical tests were performed to determine the maximum mechanical compressive strength of the samples containing the types of cement produced in this study. Specimens were also made with two commercial types of cement (CP V-ARI and CP-II-E 32) to compare the results. All test specimens were produced in accordance with standard NBR 7215 [19]. A 1:3:0.5 cement–sand–water ratio was kept. The fine aggregate contained an equal proportion of standardized sand with particle sizes of 1.2–2.4 mm, 0.6–1.2 mm, 0.3–0.6 mm, and 0.15–0.3 mm.

Specimens 50 mm in diameter and 100 mm in height were fabricated and kept in a room at 23 ± 2 °C for 24 h. The proof bodies were submerged in a tank containing water and lime until the mechanical tests were conducted at 3, 7, and 28 days of curing. The mechanical strength was measured in triplicate, using a PC200CS EMIC Universal testing machine (São José dos Pinhais, Brazil) with a load speed of 0.25 MPa/s.

2.7. Superficial Area

The Blaine method determined the superficial area through air permeability using the cement [20].

2.8. Initial Setting Time

The initial setting time was tested according to the Brazilian NBR 16607 standard [21]. This test determined the time range during which water was added to the cement until the corresponding Vicat needle penetrated the paste to a depth of 4 ± 1 mm. In other words, this test determines the moment at which the chemical reaction of cement hydration begins.

2.9. Hydration Heat Test

A homemade apparatus (Figure 2) was developed to determine the types of cement hydration heat. This apparatus consisted of an insulating box filled with glass wool, a reactor made of polyvinyl chloride, a type-K thermocouple, a data acquisition system (Novus Datalogger), and a computer.

The test consisted of mixing 450 g of cement with the same water content as described in the initial setting time test. The cement and water were mixed for 10 min and placed into the reactor, which was then closed. A thermocouple was inserted into the sample, and the insulating box was sealed. The hydration heat test was performed on the B_1.2-S₂₀ (1) cement (as it presented the highest mechanical strength value), as well as on the CP II-E 32 and CP V-ARI cements for comparison.

3. Results and Discussion

3.1. Wastes Characterization

As shown in

Table 4. Chemical composition of the lime sludge, foundry slag, and commercial types of cement.

Chemical Composition	Lime Sludge (wt%)	Foundry Slag (wt%)	Cement CPV—ARI (wt%)	Cement CPII—E (wt%)
Al2O3	0.18	19.95	6.46	5.49
CaO	55.53	1.96	52.21	55.03
Fe2O3	0.18	7.04	2.97	2.2
MnO	0.03	27.51	0.08	0.2
MgO	0.34	1.36	5.78	2.41
K2O	0.01	0.61	1.12	0.93
Na2O	2.58	0.34	0.27	0.32
SiO2	1.0	41.23	22.84	21.83
TiO2	-	-	0.35	0.27
P2O5	-	-	-	0.15
Loss on Ignition	40.07	-	5.15	8.72

, lime sludge is predominantly composed of CaO (55.53 wt%). The X-ray diffraction (XRD) results (Figure 3a) demonstrated that calcium was present as calcium carbonate (card 5-0586). Peaks of silicon oxide (card 46-1045) were also identified. Thermogravimetric analysis (Figure 4) indicated that the lime sludge underwent a mass loss of 40.07 wt% between 600 and 800 °C, which was attributed to the decomposition of calcium carbonate [22].

The foundry slag comprised 41.23 wt% of SiO₂, 27.51 wt% of MnO, and 19.95 wt% of Al₂O₃. Smaller amounts of Fe₂O₃, CaO, MgO, K₂O, and Na₂O were also detected. The XRD spectrum (Figure 3b) did not exhibit crystalline peaks, indicating that the foundry slag possessed an amorphous structure. The results of the thermogravimetric analysis (Figure 4) did not indicate any mass loss.

3.2. Characterization of the Eco-Cementitious Material

The XRD spectra shown in Figure 5 demonstrate that the eco-cementitious materials were predominantly amorphous after rapid cooling in water, although a low-intensity peak of magnetite was detected (card 19-0629). The amorphous phase is predominant in blast furnace slag [16,23], which is specified for use in Portland cement [22,24].

However, the XRD spectra shown in Figure 6 demonstrate that the eco-cementitious materials synthesized under slow cooling are composed of calcium silicate (CaSi₂—card 16-0135) and calcium aluminum-silicon (Ca₃Al₂Si₂—card 01-083-1278), which are present in cementitious compounds [19,20]. Peaks corresponding to aluminum manganese silicon (Al₈₅Mn₁₄Si—card 00-040-1111) and iron (Fe—card 00-006-0696) were also observed.

The thermodynamic analysis (Figure 7) demonstrated that calcium silicate (Ca₃O₃·Si₂O₄) was the main phase in the eco-cementitious materials. This phase reacts with water to form the calcium silicate hydrate (CSH) phase (CaO_x·SiO₂·H₂O_γ), which alters the physical and mechanical properties of the cement [22,25,26]. The simulation also indicated the formation of manganese aluminate (MnO·Al₂O₃) and manganese silicate (Mn₂O₂·SiO₂), which supports the results presented in Figure 6. Additionally, the simulation showed that eco-cementitious materials with a binary basicity of 1.2, as shown in Figure 7b, yielded the highest calcium silicate content (Ca₃O₃·Si₂O₄).

The DTA analysis of the eco-cementitious materials upon rapid cooling (Figure 8) showed exothermic peaks at 883 °C, 885 °C, and 887 °C due to crystallization processes [27], and endothermic peaks at 1075 °C, 1093 °C, and 1095 °C corresponding to the melting temperature. Increasing the binary basicity up to 1.2 resulted in a higher exothermic peak, indicating that the formation of silicates might be favored. However, increasing the binary basicity to 1.4 reduced the exothermic peak signal, suggesting that silicate formation in the slag was less likely.

3.3. Cement Characterization

Table 5. Surface area of the cement produced in the research.

Cementitious Materials	Surface Área (m2/kg)
B1.0-S6	1438.56
B1.0-S34	1337.56
B1.4-S6	1519.79
B1.4-S34	1342.54
B1.2-S20 (1)	1400.94
B1.2-S20 (2)	1396.17
CP II	1013.93
CP V	1222.84

displays the surface area values of the types of cement containing the eco-cementitious material and the commercial materials (CP V–ARI and CP II-E). The addition of eco-cementitious material increased the surface area of the cement types, which in turn enhanced the mechanical strength [28].

Figure 9 shows that the CP V–ARI cement reached the highest mechanical strength values at 3, 7, and 28 curing days (26.71 MPa, 25.23 MPa, and 30.08 MPa, respectively). The B_1.2-S₂₀ (1) cement attained mechanical strengths of 19.23 MPa at 3 days, 19.74 MPa at 7 days, and 26.82 MPa at 28 days, indicating that these types of cement met the minimum strength required by the NBR 16697 standard [11] for a CP II-E 25 MPa cement (a cement with 35 MPa at 28 days containing blast furnace slag as a supplementary material). The B_1.2-S₂₀ (1) cement showed higher mechanical strength than the commercial CP II-E cement, especially at 3 curing days. This demonstrates that the eco-cementitious material contributes to an earlier and faster hydration reaction compared to the commercial CP II-E cement, which may have resulted in the production of additional CSH.

Statistical analysis of the 28-day compressive strength results indicated that only the addition of cementitious material had a statistically significant influence on the final strength (p = 0.04), while basicity, in the range of 1.0 to 1.4, had no significant effect (p = 0.44). The interaction between the factors was also not significant (p = 0.46). The Pareto index was −3.37, indicating that mechanical strength decreased with increasing material addition.

The response surface (Figure 10) shows a tendency for strength to decrease with increasing addition content, regardless of the level of basicity. The highest mechanical strength (29.49 MPa) was recorded under conditions of low addition (6%) and low basicity (1.0), suggesting that excessive addition of the material compromises the integrity of the cementitious matrix. The midpoint, with 20% addition, obtained a mechanical result only 9% lower than the best result, which is a tolerable addition level for this work.

For the 3-day tests, although the factors did not reach predetermined statistical significance (p < 0.05), the addition factor had a value of p = 0.07, indicating, with 93% reliability, that increasing the addition content also contributes to reducing the initial strength. For the 7-day tests, the values of p = 0.63 (addition) and p = 0.23 (basicity) indicate the absence of statistically relevant effects of the factors analyzed at this stage of curing.

Cement B1.2-S20 (1) presented the highest value of the initial setting time (240 min) required by the Brazilian NBR 16697 standard [11]. The cement CPV-ARI presented an initial setting time of 150 min, while the cement CPII-E exhibited an initial setting time of 210 min.

The chemical composition of the B1.2-S20 (1) cement (

Table 6. Chemical composition of the cement B_1.2-S₂₀ (1) obtained from mass balance.

Components	B1.2-S20 (1) (wt%)
Al2O3	7.79
CaO	48.45
Fe2O3	3.32
MnO	3.66
MgO	4.84
K2O	0.98
Na2O	0.56
SiO2	23.77

) showed a CaO content of 48.45 wt%, which was lower than that of CP V–ARI (52.21 wt%) and CP II–E (55.03 wt%). Increasing the CaO content also increased the exothermic peaks from the cement hydration reaction [29], indicating a decrease in the activation energy of the hydration process and, therefore, resulting in an increased initial setting time. Moreover, the commercial cements CP V–ARI and CP II–E exhibited lower surface areas (1222.84 m²/kg and 1013.93 m²/kg, respectively) compared to B_1.2-S₂₀ (1) cement (1400.94 m²/kg). A higher specific surface area is expected to enhance the hydration rate, as finer particles provide more surface for water interaction, accelerating the reaction kinetics [30].

Figure 11 shows the temperature reached during the hydration reaction for the B_1.2-S₂₀ (1), CP V–ARI, and CP II–E cements. The CP V–ARI cement yielded the highest hydration temperature with a shorter reaction time (approximately 12.5 h). The B_1.2-S₂₀ (1) cement reached its maximum hydration temperature at 13.9 h. The commercial CP II–E cement presented a lower hydration temperature, occurring at approximately 20 h. According to Neville [31], the hydration reaction takes place over approximately 10 h, during which CSH is formed. According to Scrivener et al. [32], the temperature release is due to the exothermic reaction that forms CSH phases, which contributes to the increase in mechanical strength of the cement. The CP V–ARI cement contained a higher percentage of clinker than B_1.2-S₂₀ (1) and CP II–E, which implies higher hydration heat and mechanical strength [33].

4. Conclusions

This study demonstrated the technical feasibility of employing lime sludge (55.53 wt% CaO, predominantly as calcium carbonate) in combination with foundry slag (41.23 wt% SiO₂, amorphous structure) for the production of a new eco-cementitious material. The chemical adjustment between these two wastes enabled binary basicity values higher than 1, a requirement established by NBR 16697 for application in cementitious systems.

The partial replacement of 20 wt% clinker with this material resulted in compressive strength of 25.0 MPa at 28 curing days, a performance comparable to the reference cements CP II-E and CP V-ARI. Additional calcium silicate hydrate was formed through the reaction with free CaO, contributing to strength development. The initial setting time was extended to 30–90 min, a beneficial feature for applications requiring long transportation times or reduced use of chemical retarders.

Therefore, this research confirms the possibility of valorizing underutilized industrial waste streams, reducing clinker dependency, mitigating CO₂ emissions, and aligning the process with the European Commission’s Circular Economy Action Plan and the United Nations’ Sustainable Development Goals.