1.Introduction

The modern construction industry faces a global challenge, attempting to reconcile the growing demand for infrastructure with the need for a radical reduction in anthropogenic environmental impact. At the epicenter of this contradiction lies the cement industry – a titanic sector in scale, yet energy- and resource-intensive, responsible for 5–8 % of global CO2 emissions, where the majority of emissions are related to the irreversible process of limestone decarbonation (CaCO3 → CaO + CO2) [1, 2]. In light of the Paris Agreement goals and sustainable development principles, this situation has initiated a widespread search for innovative approaches that go beyond simple optimization of existing processes. The concept of a circular economy comes to the forefront, viewing industrial waste not as an inevitable evil but as potential "secondary resources" or "technogenic raw materials" for related industries, allowing for the simultaneous solution of disposal problems and reduced pressure on natural deposits [3, 4].

One of the most promising and actively studied directions is the partial replacement of primary raw materials (clay, shale, limestone) in the clinker raw mix with wastes containing the necessary oxides (SiO2, Al2O3, Fe2O3, CaO). Successful examples include the use of granulated blast furnace slag, coal fly ash from thermal power plants, nepheline slimes, and metallurgical dusts [5, 6]. However, the effectiveness and safety of such replacement directly depend on the chemical and mineralogical compatibility of the waste with the clinkerization technology, as well as the absence of critical impurities, such as chlorides or excessive alkalis, which can negatively affect the firing process and concrete durability [7, 8].

In the search for such synergistic solutions, the aluminum industry, which generates a range of specific and difficult-to-dispose wastes, is of particular interest. There is already positive experience in using fluorine-carbon-containing gas cleaning sludges as a fluoride mineralizer in cement plants [9]. It has been proven that fluoride ions (F⁻) can lower the temperature of liquid phase formation and reduce the viscosity of the melt in the CaO–SiO2–Al2O3–Fe2O3 system, thereby intensifying the synthesis of the main clinker mineral – alite (C3S) – and reducing specific fuel consumption [10, 11]. This experience was the first step in creating a technological symbiosis between aluminum and cement production.

However, alongside these sludges, another large-tonnage and problematic material remains – spent chamotte lining (SCL) from aluminum electrolyzers. Millions of tons of this material are generated annually worldwide, classified as a Class IV hazardous waste due to significant content of mobile fluorides (up to 15 %) and soluble alkalis (Na2O, K2O) [12]. The lining's service life is 3–7 years, after which it is sent for costly landfill disposal or undergoes inefficient processing, posing a serious environmental and economic problem [13].

The key idea of this research is not merely the disposal of SCL but its deep technological integration into the cement production process after preliminary neutralization of the most hazardous components. We hypothesize that after hydrometallurgical leaching, used for fluoride recovery (e.g., in schemes for obtaining synthetic cryolite), SCL – which is essentially a fired high-alumina clay (main components: SiO2 and Al2O3) – can be considered as a source of an aluminosilicate component. However, its original chemical composition is typically characterized by a relative deficiency of silica and a significant excess of alumina compared to traditional clays, as well as the presence of residual alkalis after leaching [14]. This makes the direct use of SCL impossible without correcting the modular characteristics (silica ratio *n* and alumina ratio *p*) of the raw mix.

To solve this structural problem, we propose the combined joint use of two types of technogenic raw materials: leached SCL and amorphous microsilica (MS). The latter is a finely dispersed (specific surface area 15–25 m2/g) waste from gas cleaning in the production of crystalline silicon or ferrosilicon, consisting of 85–98 % highly reactive amorphous silicon dioxide [15–28]. Thus, the research is aimed not at simple "disposal" of waste but at the targeted design of a new composite raw material component from them. In this composite, MS compensates for the SiO2 deficiency in SCL, while SCL, in turn, serves as a matrix-carrier for the volatile MS, preventing its losses during transportation and dosing. Furthermore, residual traces of fluorine in SCL may continue to perform the function of a mineralizer.

The objective of this work is a comprehensive experimental assessment of the technological feasibility and practical expediency of such a replacement. The following tasks were addressed within the study:

Chemical-mineralogical characterization of the initial wastes (leached SCL and MS) and traditional materials. Calculation and design of the raw mix composition using the "SCL+MS" composite to achieve the target clinker modular characteristics. Laboratory modeling of the firing process with a study of the kinetics of decarbonation and phase formation. Comprehensive analysis of the obtained clinker (phase composition, microstructure) and the Portland cement made from it (standard physico-mechanical properties) in comparison with a control produced by traditional technology.

2.Materials

The basis for any modeling in cement technology is the accurate chemical composition of the components. We had four key materials at our disposal, each with its own specifics (Table 1).

Table 1. Chemical composition of raw materials (wt. %).

Component

SiO2

Al2O3

Fe2O3

CaO

MgO

Na2O

K2O

SO3

LOI*

Limestone

1.4

0.6

0.3

47.4

3.1

–

–

0.1

46.5*

Spent Chamotte Lining

59.3

32.0

3.0

0.7

0.6

2.0

0.15

0.3

2.5

Microsilica

92.5

0.8

0.6

0.3

0.5

0.40

0.40

0.1

4.6

Iron-containing Sand

28.7

5.4

57.3

3.3

2.8

–

–

1.5

1.0

*Note: LOI – Loss on Ignition; for limestone, mainly CO₂.

Limestone served as the main source of calcium oxide (CaO). Leached SCL, after pretreatment, lost a significant portion of fluorine but retained the high alumina (Al2O3) content (32 %) characteristic of refractories and a moderate residual Na2O content (~2 %), which required consideration in the alkali balance. Its main drawback for cement raw mix purposes was a pronounced silica (SiO2) deficiency relative to alumina. To compensate for this deficiency, amorphous MS – an ultrafine powder with a content of reactive SiO2 exceeding 92 % – was introduced. Its high specific surface area predetermined both a potential influence on the mixture's reactivity and a possible increase in the water demand of the future cement. The fourth component was an iron-containing sand, the task of which was to adjust the alumina ratio and ensure the formation of the necessary amount of the aluminoferrite phase (C4AF).

2.1.Design of the Raw Mix Composition

The process of designing a raw mix for cement clinker resembles solving a multi-parameter problem. The target values for the lime saturation factor (LSF = 0.90) and silica ratio (SR = 2.3) were set as constants, ensuring the production of clinker with optimal phase composition and hydraulic activity. The main difficulty was working with SCL. Analysis of its composition showed that in its pure form, it cannot serve as a clay analog due to the low SiO2/Al2O3 ratio. To visualize and solve this problem, the classic rule of mixtures – the "Pearson envelope" – was applied. This graphical method clearly showed that to bring the silicate component to the required level (~75 % SiO2), it is necessary to introduce MS in a ratio of 1:0.5 to SCL. Thus, a new, "synthetic" silicate component was created, representing a composite of two technogenic products. Its chemical profile (Table 2), including a reduced content of equivalent alkalis (Na2Oₑq. = 1.3 %), already met the suitability criteria.

Table 2. Chemical composition of the modified silicate additive (SCL+MS) (wt. %).

Compound

SiO2

Al2O3

Fe2O3

CaO

MgO

Na2Oₑq.

Content

75.0

16.0

1.5

0.35

0.27

1.30

Note: Na2Oₑq. = Na2O + 0.658K2O.

Subsequent calculation by solving a system of material balance equations [8] allowed for determining the exact proportions of all components in the raw mix. The final formulation consisted of 81.7 % natural limestone, emphasizing that the proposed technology does not represent a complete replacement of traditional raw materials, but rather a partial, yet significant (14.5 %) replacement of the clayey component. This is a realistic and pragmatic approach for potential industrial implementation.

2.2.Laboratory Modeling of the Firing Process

The transition from calculation formulas to real material began with the preparation of a homogeneous raw mix and pressing it into wet briquettes (Fig. 1).

Figure 1. Briquettes of pressed raw mix.

Firing was carried out in a muffle furnace under a strictly controlled temperature-time regime, which served as a laboratory model of an industrial rotary kiln operation:

from 100 °C to 800 °C – 45 min; from 800 °C to 1100 °C – 50 min; from 1100 °C to 1300 °C – 10 min; from 1300 °C to 1450 °C – 20 min.

Key stages were: slow heating to remove free and chemically bound water, intensive decarbonation in the range of 700–900 °C, formation of intermediate oxide compounds, and finally, the culmination of the process – the synthesis of clinker minerals at 1450 °C. Rapid cooling of the obtained clinker nodules (Fig. 2) was necessary to fix the metastable high-temperature phases, primarily alite (C3S), which is standard practice in cement technology.

Figure. 2. Obtained clinker samples.

For a comprehensive characterization of intermediate and final products, an arsenal of modern physicochemical methods was used.

Thermogravimetric Analysis (TGA/DTA) acted as a "real-time observer" of the process, allowing for the recording of mass loss during decarbonation and thermal effects accompanying phase transformations. X-ray Diffraction Analysis (XRD) allowed not only qualitative identification of crystalline phases in the clinker but also their quantitative assessment using the Rietveld method. Scanning Electron Microscopy (SEM) provided a unique opportunity to see the material's microstructure: the shape, size, and mutual arrangement of crystals, which directly correlates with the strength properties of the future cement. Standard physico-mechanical tests of cement paste and mortar became the final and most important criterion for success, translating scientific data into the language of practical construction standards (GOST).

3.Results and Discussion

Comparative analysis of thermograms of the experimental and control mixtures revealed important differences (Fig. 3).

Figure 3. Comparative thermograms (TGA/DTA) of the experimental and control raw mixes.

The main endothermic peak, corresponding to the decomposition of calcium carbonate (CaCO3 → CaO + CO2), in the mixture containing SCL and MS, was shifted to a lower temperature region by 20–30 °C. This can be interpreted as a result of two factors. Firstly, finely dispersed MS, possessing an enormous specific surface area, creates an extensive contact zone with limestone particles, facilitating diffusion processes. Secondly, residual fluoride ions, preserved in SCL after leaching, may act as a catalytic mineralizing additive. It is known that even small amounts of fluorine contribute to the destruction of the crystal lattices of initial oxides and reduce the viscosity of the forming melt, thereby intensifying the reactions of solid-state and liquid-phase synthesis of calcium silicates [9]. This effect is also confirmed by the shift of the onset of the exothermic effect associated with alite (C3S) formation towards lower temperatures. Thus, the first important conclusion is that the introduction of the technogenic composite not only does not impair but potentially improves the kinetics of the firing process, promising direct energy benefits on an industrial scale.

The obtained diffractograms, presented in Fig. 4, clearly demonstrate that the main qualitative composition of the clinkers is identical.

Figure 4. X-ray diffractograms of the experimental and control clinker.

In both cases, peaks characteristic of alite (C3S), belite (C2S), tricalcium aluminate (C3A), and tetracalcium aluminoferrite (C4AF) are clearly identified. However, quantitative calculation (Table 3) revealed nuances.

Table 3. Phase composition of clinkers (wt. %), determined by the Rietveld method.

Sample

C3S (Alite)

C2S (Belite)

C3A

C4AF

Glass phase / Other

Experimental

52 ± 2

25 ± 2

8 ± 1

12 ± 1

3

Control

56 ± 2

22 ± 2

7 ± 1

13 ± 1

2

In the experimental clinker, the proportion of alite – the most valuable and fast-hardening phase – was 4 % lower (52 % vs. 56 %), while the proportion of belite, which gains strength more slowly but more stably, was correspondingly higher.

The explanation for this phenomenon can be found when examining the microstructure (Fig. 5).

a

b

Figure 5. SEM micrographs (BSE mode) and EDS mapping of clinker samples. (a) Control clinker: prismatic alite (C₃S, light grey, 20–35 µm) with rounded belite (C₂S, darker grey) and interstitial phase (bright, Fe-rich). (b) Experimental clinker with SCL+MS additive: alite crystals (15–40 µm) show slightly irregular morphology; belite content visually higher.

SEM micrographs show that alite crystals in the experimental sample have a slightly less regular shape and a larger size distribution (15–40 µm vs. 20–35 µm in the control). This may be related to changes in the rheological properties of the melt. Highly dispersed MS, actively interacting with the melt, may locally increase its viscosity. This creates kinetic barriers for the growth of large, ideal alite crystals and complicates the final stage of its formation from belite and free lime. Interestingly, energy-dispersive X-ray analysis (EDX) confirmed the hypothesis about alkali distribution: sodium primarily concentrated in the inter-crystalline glassy phase rather than entering the lattice of the main silicate minerals. This is a favorable scenario, as such immobilization reduces the risk of harmful alkali-silica reaction in concrete [6].

The compressive strength was determined at 28 days of age on a hydraulic press in accordance with GOST 310.4. Standard specimens made from the experimental and control cements were tested (Fig. 6).

a

b

Figure 6. Process of testing cement specimens for compressive strength: (a) specimen installed in the grips of the testing machine; (b) general view of the hydraulic press during testing.

The key confirmation of the new cement's quality is the compliance of its properties with standard requirements. The data presented in Table 4 and Fig. 7 allow for a definitive conclusion: the cement based on the experimental clinker fully meets the requirements of PC 300 grade.

Table 4. Properties of Portland cements based on experimental and control clinker.

Characteristic

Experimental Cement

Control Cement

Norm for PC 300 (GOST 31108)

Specific surface, m2/kg

355

350

Not less than 300

Normal consistency, %

28.50

25.75

–

Initial setting time, min

220

200

Not less than 45

Final setting time, min

430

470

Not later than 600

Compressive strength, MPa:

3 days

12.5 ± 0.8

14.1 ± 0.9

≥ 12.0

7 days

21.3 ± 1.2

22.8 ± 1.3

≥ 19.0

28 days

29.6 ± 1.5

29.8 ± 1.5

≥ 28.0

56 days

34.2 ± 1.6

33.1 ± 1.5

≥ 32.0

Some differences from the control sample are characteristic and explainable. The increased normal consistency (28.5 %) is a direct consequence of the presence of ultrafine MS, which requires more water for wetting. This is a technological nuance that must be considered when selecting water-reducing admixtures in concrete mix design.

The kinetics of strength gain (Fig. 7) demonstrate a classic pattern for clinker with a slightly increased belite content.

Figure 7. Variation of compressive strength with curing time (3, 7, 28, and 56 days).

In the first three days, the strength of the experimental cement slightly lags, which correlates with the slightly lower amount of fast-reacting alite. However, by 7 days, the gap narrows, and by 28 days, the strengths become statistically indistinguishable. This vividly illustrates the compensatory role of belite: although its hydration proceeds more slowly, it continues for a long time, ensuring confident strength growth in the medium- and long-term perspective. Thus, a minor shift in the clinker's phase composition did not lead to a deterioration of its key strength characteristic.

Comparison with previous studies. The observed reduction in decarbonation temperature (20–30 °C) due to the combined effect of fluoride residues and amorphous silica is consistent with the findings of 9. K. Vance et al. [9], who reported that fluoride ions lower the liquid phase formation temperature by 15–40 °C in cement raw mixes. Similarly, Katey O’Quinn et al. [11] demonstrated that fluoride-containing mineralizers enhance alite formation at lower temperatures. However, unlike previous studies that used pure fluoride salts, the present work achieved a comparable effect using residual fluorides in a waste material (SCL), which is more economically and environmentally attractive. The 28-day compressive strength of 29.6 MPa obtained in this study is slightly lower than the 32–35 MPa reported by Abdul,  W. et al. [27] for SCL-fly ash blends, but this difference is attributed to the higher alkali content in our SCL (2.0 % Na2O vs. 0.8 % in their study). Vuthaluru [19] also noted that alkalis above 1.5 % Na2Oₑq. can suppress alite formation, which agrees with our observed 4 % reduction in C3S content compared to the control.

4.Conclusion

The conducted study, consistently from the calculation model to the testing of the finished product, proved the fundamental possibility of deep technological integration of wastes from aluminum and silicon production into the Portland cement manufacturing process. We did not merely add an inert filler but purposefully designed a new composite component (SCL+MS), which in quantities up to 14.5 % of the raw mix mass can fully replace traditional clay.

The main conclusions of the work are of both scientific and practical application value:

The combined use of leached SCL and MS allows for the mutual compensation of their shortcomings: MS compensates for the silica deficiency in SCL, while SCL serves as a matrix-carrier for finely dispersed MS. Residual fluorides in SCL and the high reactivity of MS have a positive effect on the firing kinetics, lowering the temperature of key reactions, which opens the way to fuel savings. Despite a slight modification of the phase composition (shift in the C3S/C2S balance) and the clinker's microstructure, the obtained Portland cement satisfies all regulatory requirements for PC 300 grade, demonstrating full competitiveness. For successful implementation, it is necessary to control the alkali level in the incoming SCL (Na2Oₑq. ≤ 2.0 %) and its proportion in the raw mix (8–12 %), as well as to account for the increased water demand of the cement when designing concrete mixes.

Prospects for further research logically follow from the obtained results. The next necessary step should be pilot industrial trials on an operating rotary kiln, which will allow assessing the real impact on energy consumption, process stability, and emissions of volatile compounds. In parallel, an in-depth study of the durability characteristics of concrete based on the new cement is required, especially its resistance to freeze-thaw cycles and corrosive aggression. Future work should also include a comprehensive evaluation of concrete durability, including freeze-thaw resistance, sulfate attack, and alkali-silica reaction, as the presence of residual alkalis may pose long-term risks.

In a broader context, this work is a clear example of implementing circular economy principles in heavy industry. It offers a specific, technically justified solution for two environmental problems simultaneously, transforming the costs of hazardous waste disposal into investments in the production of a demanded construction material.