Slag Matrix Research Articles

Bifunctional Pt-loaded steel slag matrix composites for the detection and degradation of tetracycline antibiotics

In this paper, a method for the simultaneous detection and degradation of tetracycline antibiotics (TCs) in water was investigated. Calcium silicate hydrate (CSH), FeO and calcium aluminate hydrate (CAH) were isolated from steel slag as carriers for Pt monomers. The produced Pt-modified modified steel slag (ALANH-Pt) possessed both peroxidase activity and photocatalytic properties. As a result, a sensitive and selective colorimetric sensor was exploited on the basis of ALANH-Pt for tetracycline (TC), oxytetracycline (OTC), and doxycycline (DOX), which exhibited a detection limit (LOD) of 1.696 μM, 0.999 μM and 3.607 μM, respectively. In addition, the degradation rate for TCs could be achieved 82 % within 60 min. The possible mechanisms of detection and degradation are discussed based on ESR spectroscopy, revealing the generation of hydroxyl radicals (OH), superoxide radicals (O2−) and holes (h+). The degradation pathway of TC was inferred by HPLC-MS. The selectivity of the colorimetric sensing platform and the application of the bifunctional ALANH-Pt to real water samples were investigated. This work provides a new idea that allows for the simultaneous detection and degradation of TCs, and offers a new approach to the utilization of steel slag.

Micromechanical and chemical characteristics of interfacial transition zone in alkali-activated slag concrete containing lightweight iron-rich aggregates

This study explored the time-varying alkali-activation effects of the lightweight aggregates (LWA) on the interfacial transition zone (ITZ) of the alkali-activated slag concrete with different curing ages, molar ratios, and alkali concentrations from 28 to 300 d. The improved micromechanical performance of the ITZ at 28 d stemmed from the early-age internal curing effect of the porous LWA. However, the post-28 d internal curing effect diminished, with the alkali-activation dominating the performance of the ITZ from 28 to 300 d. Although alkali-activated cement possessed early strength characteristics, the deconvolved phase results illustrated that alkali activation persisted in the ITZ between the LWA and cement from 28 to 300 d, leading to a high proportion of calcium silica aluminate hydrate. As a marker, Fe from the lightweight iron-rich aggregates mitigated into the alkali-activated cement, demonstrating the alkali-activated reaction between the LWA and alkali-activated slag matrix. Molecular dynamics simulations further characterized that the low content of Fe did not decrease the overall strength of the ITZ. The interfacial interaction energy revealed the alkali-activation between the LWA and cement compensated for the weakening of mechanical properties caused by Fe ions on the ITZ.

First Approach Using Fluidic Force Microscopy (FluidFM®) to Measure Adhesion Forces between Droplets and Flat/Rough Surfaces Immersed in Water

The research program “Engineered Artificial Minerals (EnAM)” addresses the challenge of recycling valuable elements from battery waste streams. These elements, such as lithium (Li), often migrate in the slag phase, in some cases as crystals. EnAM crystals represent concentrated reservoirs of these elements, which can only be effectively recycled if they are extracted from the slag matrix and then separated. Selective wet agglomeration is a separation process based on a three-phase system and is often used in coal and ore processing. The produced agglomerates in this process can be easily separated from the remaining suspension. The precise quantification of the wetting properties and adhesion strength between suspended particles and binding liquid droplets is a scientific challenge. An accurate technique suitable for adhesion force measurements in three-phase systems with micrometer-scale particles is Fluidic Force Microscopy (FluidFM®). An experimental setup with optical control is being developed to measure adhesion forces between droplets and flat/rough surfaces. This will enable precise measurements of adhesion forces between solid EnAM crystals and binding liquid droplets. Based on these measurements, optimal agglomeration conditions can be selected in the future to improve selective wet agglomeration with respect to recycling processes.

Laser-induced reduction of iron ore: Trends of the chemical distribution across the domains

The reduction of iron ore powder in a laser-induced thermal cycle using several reducing agents was studied. The laser-assisted reduction process resulted in the formation of iron-rich domains, irregularly embedded in a slag matrix, and transitional phases. The appearance of these various chemical phases was categorized and geometrically evaluated with respect to representative dimensions using scanning electron microscopy. The statistical trends of the morphology are presented, in context with trends of the chemical composition across the domains, to gain a better understanding of the mechanisms behind the reduction process. Iron domains were predominately observed in the vicinity of the Si-rich zones, indicating the occurrence of the reduction reaction as a result of the Fe2O3-Si interaction. Furthermore, different appearances of the Fe-rich domains and other phases in the close proximity to iron were analyzed and discussed based on diffusion and coalescence phenomena. The obtained results show that the reduction occurs, but the process is still uncontrolled and only partially understood. Further analysis and experiments are, therefore, needed to investigate the prospects of the proposed method.

Viscosity experiments of slag-spinel suspensions: The effect of volume fraction, particle size, and shear rate

Heterogeneous slag viscosity measurements have a wide variety of parameters, such as particle size, shape, solid volume fraction, and shear rate, which affect the final measured viscosity. Often, some of these parameters are neglected or predicted via models, since it is experimentally difficult to determine them during high-temperature slag viscosity measurements. In this work, a viscosity apparatus was used, which allows quenching of the sample after measurement. This way, all relevant parameters could be experimentally determined. The viscosities of three datasets were studied with various spinel sizes: small (13 μm), medium (34 μm), and large particles (76 μm). Within each dataset, the composition of the liquid slag matrix was kept constant to uniquely measure the effect of solids. Shear thinning, i.e., a decreasing viscosity with increasing shear rate, was observed for all samples, even at a low vol. % of 1.8. Moreover, stronger shear thinning was observed at a higher vol. % and for smaller spinel particles. On the basis of these results, the phenomenon of shear thinning was discussed and mainly attributed to the particle–particle orientation in the suspension. The obtained viscosity values were used to optimize a Krieger–Dougherty equation, which describes the viscosity increase caused by the presence of a certain vol. % of spinel particles, with a certain size and at a specific shear rate.

Improved Settling Mechanisms of an Industrial Copper Smelting Slag by CaO Modification

A significant source of copper losses from pyrometallurgical copper extraction is attributed to dissolved and entrained copper in discarded slag. Entrained copper can be recovered via pyrometallurgical slag cleaning in a settling furnace, where the slag viscosity and copper droplet size distribution (CDSD) are critical parameters. Reduced copper losses suggest improved raw material efficiency, and the slag becomes a more environmentally safe byproduct. In this study, iron silicate slags from a smelting furnace are industrially CaO modified in a fuming furnace to contain a CaO content between 8 and 18 wt.%. The viscosity of slags with and without CaO modifications is evaluated in the temperature range from 1423 to 1723 K. The influences of the CaO modifications on the CDSD, slag matrix copper content, and total copper content were determined. The results show that the slag viscosity decreases with increasing CaO concentrations in the slag. In addition, the copper content decreases in the slag phase, and the CDSD shifts to contain relatively larger droplets. The effect of CaO slag modification reveals a linear relationship between the overall copper recovery and viscosity, which increased from 63 to 88% when the viscosity in the respective batch was 0.51 and 0.25 Pa·s.Graphical

O3/UV-type POAs integrated with catalytic material based on scrap iron and mineral clay to degrade 2,4 and 2,6-dinitrotoluene in Red Water

Nitroaromatic compounds have significant economic and industrial relevance in various inputs such as dyes, inks, agrochemicals, and explosives. In the manufacture of the explosive 2,4,6 trinitrotoluene (TNT), there is the formation of contaminated wastewater that is difficult to treat because it contains, among other compounds, high concentrations of 2,4 and 2,6-trinitrotoluene (DNT). In this work, the application of advanced oxidation processes of the O3/UV type was studied in sequential integration with catalytic material based on metallic iron slag (SZVI) and mineral matrix to treat an industrial effluent from the explosives industry contaminated with compounds nitroaromatics. Two types of mineral matrix composition were studied, kaolinite (Kau) and pumice powder (Pum), having been observed in the conditions of best efficiency (O3/UV-SZVI/Kau) 100% removal of both nitroaromatic compounds and residual ozone. On the other hand, photo-ozonation alone could only partially remove these components, with 38% DNT removal and 3.2 mg/L residual O3 observed, proving the importance of integration with the catalytic matrix for more effective treatment of the studied effluent.
 Keywords: DNT degradation, explosives industry effluent, Scrap zero-valent iron.

Strain Hardening of Polypropylene Microfiber Reinforced Composite Based on Alkali-Activated Slag Matrix.

A comparative study of the fracture features, strength and deformation properties of pseudo strain-hardening composites based on alkali-activated slag and Portland cement matrices with polypropylene microfiber was carried out. Correlations between their compositions and characteristics of stress–strain diagrams under tension in bending with an additional determination of acoustic emission parameters were determined. An average strength alkali-activated slag matrix with compressive strength of 40 MPa and a high-strength Portland cement matrix with compressive strength of 70 MPa were used. The matrix compositions were selected for high filling the composites with polypropylene microfiber in the amount of 5%-vol. and 3.5%-vol. ensuring the workability at the low water-to-binder ratios of 0.22 and 0.3 for Portland cement and alkali-activated slag matrices, respectively. Deformation diagrams were obtained for all studied compositions. Peaks in the number of acoustic signals in alkali-activated slag composites were observed only in the strain-softening zone. Graphs of dependence of the rate of acoustic events occurrence in samples from the start of the test experimentally prove that this method of non-destructive testing can be used to monitor structures based on strain-hardening composites.

The impediment and promotion effects and mechanisms of lactates on the hydration of supersulfated cements - Aiming at a performance enhancement

Supersulfated cement (SSC) is a promising low-carbon binding material, however, its low hydration degree highly limits the performance and application of SSC products. In this work, aiming at a property enhancement, sodium lactate was added into mixtures of SSC, mainly composed of the hemihydrate gypsum and high-alumina slag. The experimental and theoretical results indicated that the hydration process of SSC could be finely controlled through lactates for two opposite mechanisms. At the early stage, the presence of lactates decreases the release of hydration heat and delays the hydration of SSC, which is beneficial to maintain the workability. In the long term, the molecular dynamics simulations suggest that lactates stably adsorb on the slag surface and strongly attract Ca ions, gradually weakening the structural integrity of geopolymer components in slag matrix. As a result, the dissolution of slag is accelerated, and thus more hydration product and higher mechanical properties are obtained for SSC. The findings in this study shed new lights on hydration mechanisms of SSC and help broaden the application of this low-carbon binding materials.

Microstructure and Porosity Evolution During the Reduction, Softening and Melting of Iron-Bearing Materials

The performance of the blast furnace is strongly affected by the position and thickness of the cohesive zone, which is largely influenced by the high-temperature properties of the iron-bearing materials. During its reduction, softening and melting, ferrous materials undergo major microstructure changes and its understanding is essential to develop new raw-materials, technologies, and models. In this study, the behavior of reduction, softening and melting of a lump ore, an acid pellet, and a sinter was characterized by softening and melting (S&M) experiments. After that, to access the samples’ structural transformations, interrupted S&M tests were carried out up to four different conditions based on contraction and pressure loss levels. The obtained products were characterized according to its density (true and apparent), porosity (open, closed and total), phase composition by X-ray diffraction and microstructure (reflected light microscopy and electron scanning microscopy). From the S&M test results, three main regions of reduction were characterized, namely: solid/gas reduction, reduction retardation, and melting reduction. In the solid/gas reduction, samples open porosity increased, with reduction following the shrinking core model. On the region of reduction retardation, a sharp decrease in open porosity was identified together with the diminishing of the reduction rate, which occurred due to the iron shell porosity being clogged due to the slag transfer from the particles’ cores to its periphery. At the melting exudation region, reduction retardation ceased and exudation of the ferrous slag lead to a peak of reduction. The lower the reduction degree of the samples at this stage, the higher the consumption of carbon. Furthermore, at 10 pct contraction, a pseudo-globular wustite structure interspersed with slag was observed for the pellet and sinter cores. At 50 pct contraction, the previous structure coalesced to form a globular shape wustite in a well-connected slag matrix.

Arsenic removal from copper slag matrix by high temperature sulfide-reduction-volatilization

Arsenic contamination has been a major problem in copper slag utilization. Arsenic is easily incorporated into the silicate-based matrix, making the arsenic difficult to volatilize. In this study, pyrite was selected to depolymerize the matrix structure and volatilize the glassy arsenic by sulfide-reduction-volatilization reaction. The optimum technological parameters and mechanism of glassy arsenic volatilization by pyrite were further studied. The optimum operating parameters for glassy arsenic volatilization by pyrite were determined to be a temperature of 1200 °C, a holding time of 60 min, a heating rate of 5 °C/min, a basicity of 0.3, and a pyrite addition content of 15%. The arsenic volatilization ratio reached 80.9% under these experimental conditions. Besides, the mechanism of glassy arsenic volatilization was elucidated by XRD, XPS, FTIR, and SEM analyses. These results indicate that, with the increase in temperature, the pyrite decomposes to generate a variety of sulfur-based reducing substances (FeS, FeS1-x, S2(g)). Through “oxygen capture reaction”, these sulfur-based reducing substances depolymerize the bridging oxygen structure from the glass former ([AsO4], [FeO4], and [SiO4]) by the conversion of (Q2 +Q3)→(Q0 +Q1) and result in the precipitation of glass former ([AsO4], [FeO4] and [SiO4]) combining with the nearby cation. In this process, the glassy arsenic is released by the glass network and participates in reductive volatilization reaction with sulfur-based reducing substances, converting the glassy arsenic with high thermal stability to volatile arsenic oxide and arsenic sulfide. These findings provide a theoretical support for the in situ volatilization of arsenic in copper smelting and centralized control of arsenic contamination.

Effects of pozzolanic reaction on carbonation degree and strength of steel slag compacts containing zeolite

In this paper, a novel method, zeolite partially replacing steel slag, was proposed to futher improve the properties (compressive strength and carbonation degree) of steel slag compacts futher after CO2 curing. Inductively coupled plasma (ICP) results illustrated that the Ca, and, Fe ion concentrations in the steel slag matrix substituted by 5 wt% zeolite (water-to-solid ratio of 10) were 78.19 mg/L and 4.09 mg/L, 19.3% and 99.0% higher than that of pure steel slag after 1 d of hydration curing respectively. The decreased crystallite size of C4AF and increase in Fe ion concentration illustrates that zeolite with pozzolanic properties could react with CaO-FeOX matrix components such as brownmillerite (C4AF), to produce C-S-H gel, which possesses higher solubility than the CaO-FeOX matrix, to improve the Ca ion concentration. In terms of the compressive strength and carbonation degree results, the optimum zeolite substitution ratios were 5 wt% and 15 wt%, respectively. After 1 d hydration and 2 h of carbonation curing, the compressive strength of the steel slag compacts with 5 wt% zeolite (CSZ5) was 14% higher than that of pure steel slag compacts (CSZ0). With respect to carbonation degree, the steel slag substituted with 15 wt% zeolite (CSZ15) had a 10.2% higher carbonation level than that of CSZ0.

Preparation of high-strength biochar composite briquette for blast furnace ironmaking

In this research, using iron-oxide fines (average size: 2.5 μm) and biochar fines (average size: 50.0 μm), the biochar composite briquette (BCB) for blast furnace (BF) application was prepared by cold briquetting followed by heat treatment. The preparing conditions were optimized regarding its cold crushing strength. Anti-pulverization capability, reaction development, and structure evolution of the optimally-designed BCB under simulated BF conditions were then examined. Results of optimizing BCB preparation conditions showed that a heating temperature of 1073 K was optimal for preparing the BCB. The optimally-designed BCB contained 11.10 wt.% carbon, 72.21 wt.% Fe3O4, 11.25 wt.% FeO, and 0.77 wt.% Fe, 6.44 wt.% gangue, and had a cold crushing strength of 1800 N/briquette. Results of BCB behavior under simulated BF conditions showed that the cold crushing strength after partial reaction of the BCB ranged from 1500 N/briquette to 5500 N/briquette and its maximum volume shrinkage degree was 0.45. The high anti-pulverization capability of the BCB was supported by the slag matrix or the iron network. Under the simulated BF conditions, the BCB underwent five stages of reduction by atmosphere, partial self-reduction and reduction by atmosphere, full self-reduction, partial self-reduction and gasification by atmosphere, and gasification by atmosphere. It is inferred from the experimental findings that, by charging the BCB in BF, an increase of top gas utilization efficiency could be realized, and a favorable influence on lowering the temperature level of the thermal reserve zone could be obtained.

Flexural strengthening of reinforced concrete beams using fabric reinforced Alkali-Activated Slag matrix

Old Reinforced Concrete (RC) buildings are facing different degrees of structural deterioration and require proper strengthening to enhance their structural performance as well as to extend their life span. Fabric reinforced Alkali-Activated Slag (AAS) matrix is proposed to strengthen RC beams in this study. Seven RC beams with and without strengthening were prepared and tested under four-point bending. Test results indicate that use of AAS matrix as replacement for conventional cement-based matrix can change the failure mode of the strengthened beams from end-debonding of strengthening layer to slippage combined with rupture of fabric. The AAS-based strengthening strategy is able to enhance the loading capacity and flexural stiffness of RC beams, as well as reduce the strain of tensile reinforcements. Except for specimens that failed in premature debonding, increasing the fabric amount in the strengthening scheme improves the loading capacity of beams. In an optimal case, the yielding and ultimate loads of the strengthened beams are enhanced by 22.2% and 26.4%, respectively. Moreover, an analytical model was developed to predict the characteristic loads of the fabric reinforced AAS matrix strengthened beams. It shows that the analytical model could overestimate the yielding and ultimate loads of the strengthened beams, probably due to slippage and reduced synergistic effect of fabric bundles in the strengthening system. Based on that, two efficiency factors of 0.35 and 0.25, taking account of the area of effective fabric, are obtained and recommended to estimate the yielding and ultimate loads of fabric reinforced AAS matrix-strengthened beams, respectively.

The slagging propensity of Zn-rich scrap tyre ash in reducing environments consisting of CO and CO2 and its promotion of the slagging of Si-rich coal ash

This study includes a comprehensive investigation on the ash slagging characteristics of zinc (Zn)-rich tyre ash and its interaction with silica (Si)-rich coal ash blended at different ratios in a 5% CO in CO2 reducing environment. Such an environment is expected to occur near the burner in a combustor, as well as serves as a fundamental study on the slagging propensity upon the effect of CO, which is a principal reducing agent in a gasifier. An approach consisting of experimental studies using modified inclined plane (M-IP) methodology, advanced analytical techniques including the scanning electron microscopy (SEM) and x-ray photoelectron (XPS) spectroscopy, and thermodynamic equilibrium calculations were used to elaborate the slagging propensities of the single and blended ashes. Furthermore, the fate of Zn within tyre ash and its potential implications on the slagging properties were evaluated. Tyre ash easily melts to liquid slag upon increasing temperature compared to the silica-rich coal ash. Zinc does not evaporate but rather remains within tyre slag mostly as a hemimorphite (Zn4(Si2O7)(OH))-like coordination structure attributing to the characteristic blue colour of tyre slag. The low ionic potential value of 2.7 for Zn2+ is characteristic of basic ions that have a greater tendency to depolymerise the Si-rich ash slag matrix and acts as a fluxing agent in ash slag. Moreover, it was proven that Zn2+ plays a synergistic role with Ca2+ in enhancing the flowability of the Si-rich slag remarkably. Furthermore, for the pure tyre slag, Zn was found to preferentially deposit at the corundum (Al2O3) substrate interface and even penetrate/react with the porous corundum, causing a strong corrosion on the substrate. However, blending of tyre ash with coal ash alleviates this problem by a preferred scavenging of Zn into the entire Si matrix, which in turn protects the corundum plate.

Assessment of factors influencing the elution of chromium from ferrochromium slag using factorial design of experiment

The solid waste material ferrochromium slag from the metallurgical industries encounters with disposal problem because of the presence of chromium. From the chemical composition study, the waste slag was observed to contain sizeable amount of environmentally harmful residual chromium. Chromium speciation in the slag was observed to exist as mostly immobile trivalent state. But there may be possibility of its conversion to leachable hexavalent chromium and its subsequent release from slag matrix causing chromium pollution problems. Elution of chromium from slag was studied under adverse environmental situations. The importance of factors influencing chromium elution from the slag samples were assessed using full factorial design analysis and accordingly regression equations were developed for elution of both hexavalent and total chromium. The significant physico-chemical factors affecting elution of chromium from slag matrix are obtained to be pH, size of particle, time of contact and their combinations.

Inhibition Kinetics of Chromium Leaching by Calcite Coating on the Surface of Stainless Steel Slag via the Gas–Solid Accelerated Carbonation Process

Chromium leaching from stainless steel slag should be dealt with when utilized them as construction materials, an original idea of increasing calcite coating compactness in slag matrix is proposed for solving this problem. The paper originally proposes a carbonation strategy balancing carbon dioxide sequestration efficiency and chromium leaching inhibition kinetics through forming calcite coating. 48–75 μm stainless steel slag is carbonated at 300 °C and 0.4 MPa for 120 min, carbonated slag and fresh slag are applied for chromium leaching tests, aiming to discuss the inhibition kinetics of calcite coating combining with XRD, SEM–EDS, FT-IR and EMPA analysis. The results show that 7.9% valorization degree can be achieved for stainless steel slag under the conditions, and a thin layer of calcite coating is formed in slag matrix. Diffusion rates of chromium are in the range from 1.32 × 10−11 to 2.12 × 10–11 min−1 and from 0.617 × 10−11 to 1.47 × 10−11 min−1 for fresh and carbonated slag, respectively. Inhibition efficiency of calcite coating on diffusion rate is about 30% to 53%, and diffusion active energies for fresh slag and carbonated slag are 6.84 kJ/mol and 12.51 kJ/mol, respectively, which represents that carbonation strategy is indeed valid for inhibiting chromium leaching, when stainless steel slag is utilized as construction materials or road-based aggregates.

The influence of superabsorbent polymer on the properties of alkali-activated slag pastes

This paper discusses the interaction between superabsorbent polymer (SAP) and alkali-activated slag pastes and the influence of SAP on the properties of the pastes. SAP showed higher absorption in pore solutions extracted from alkali-activated slag pastes compared to that extracted from portland cement paste. Hydrogel desorption in alkali-activated slag matrix activated with Na2CO3 was noticeably delayed compared to alkali-activated slag matrix activated with NaOH matrix or portland cement matrix; this is attributed to a significant delay in the solid skeleton development in this matrix. Addition of SAP was shown to be effective in reducing autogenous shrinkage in alkali-activated slag pastes. A general reduction in compressive strength and electrical resistivity was observed in alkali-activated slag pastes when SAP was added; macrovoid formation and increased overall water/binder are possible reasons for this reduction, respectively.

Xinjiang lignite ash slagging and flow under the weak reducing environment at high temperatures – Slag viscosity and its variation with ash type and addition of clay

In this study, five different ashes derived from Xinjiang lignite were tested for slagging at 1300–1400 °C in a weak reducing environment (1% CO in nitrogen), at different exposure time from 10 min to 2 h. The apparent viscosities for different slags were determined by using a modified inclined plane (M-IP) method with an inclination of 25° based on the slag travel length per unit mass. In particular, the variation on slag viscosity upon the blending of additives including clay and MgO was examined in detail. As have been confirmed, the slags derived from the original basic and neutral ashes start to flow from 1400 °C onwards. The viscosities calculated based on the M-IP method at 1400 °C are in the range from ∼1 Pa·s for neutral ash to ∼5 Pa·s for the basic ash slags. For both neutral and acidic slags, their viscosities calculated by the M-IP method show a good agreement with some of the existing empirical models that have been validated for high-rank coal slags. Upon the addition of 10 wt% clay, the viscosity of basic ash at 1300 °C was reduced to 1.53–1.71 Pa·s, demonstrating a comparable flow-ability and viscosity to that of the original ash at 1400 °C. The addition of MgO promotes the slag flow-ability and reduces the slag viscosity to a relatively small extent in comparison to clay. Instead of melting with other elements in the ash matrix to form slag, the discrete Mg grains are observed in the slag matrix. It could even partially react with the alumina plate. The precipitation of Mg may promote the slag flow-ability by alleviating the slag penetration into the alumina plane and the resultant friction between them. Additionally, the viscosities for both the original slags and those blended with clay decrease quickly upon the increase of the exposure time, substantiating a non-linear slagging propensity and hence, probably non-Newtonian property. A minimum exposure time of 40 min is essential to ensure these slags to settle down and reach their equilibrium viscosities in the M-IP method.

Investigation of the structure of CaO-Al2O3-SiO2 melts as a basis for the development of new agglomerated welding fluxes and industrial refractories

A significant number of welding fluxes and industrial ceramics are produced in the Ukraine. The production of these materials is sufficient for both domestic needs and significant export. It is assumed that the ternary CaO-Al2O3-SiO2 system may become the basis for the development of agglomerated welding fluxes and technical ceramics. Three samples of ternary CaO-Al2O3-SiO2 system were studied by means of high-temperature X-ray diffraction above the melting point: 23.3CaO- 14.7Al2O3-62.0SiO2 wt. % (sample 1, eutectic), 9.8CaO-19.8Al2O3-70.4SiO2 wt. % (sample 2, eutectic), 15.6 CaO- 36.5Al2O3-47.9SiO2 wt. % (sample 3). Experimental scattering intensity curves, structure factors and the radial distribution function of atoms were obtained. The structural parameters of short-range order were calculated using melt structure models obtained by Reverse Monte Carlo method. The existence of microparticle associates of the mullite or sillimanite type immersed in the slag matrix is assumed. It is assumed that the calcium ions coordinated aroundthese associates are surrounded by oxygen melt anions. Such a structure can be considered as a colloidal solution. The microparticle of mullite is the core of the aggregate that is surrounded by calcium and oxygen ions from the molten medium.