Introduction

Urbanization and industrialisation generate huge amounts of waste materials [1] whose storage may directly affect the environment [2, 3]. Cement industries are among the most polluting facilities, and according to the central pollution control board (CPCB), they come under the red category where highly toxic and carcinogenic pollutants are emitted from the cement kiln [4]. Therefore, developing new technologies aimed at reusing and converting these industrial wastes into useful materials is of high interest, being economically important for protecting the environment and contributing to the sustainable development of the industry [5, 6].

Recent statistics from 2017 indicate that there are 28 cement plants in Egypt, with a total capacity of 80 MTPA. Bypass cement dust (BCD), which is a by-product of manufacturing Portland cement, represents approximately 8% of the production capacity [7]. It is a fine-grained material that is collected from exhaust gases by electrostatic precipitators during the calcination process. This large amount of industrial waste may cause significant environmental and ecological damage; therefore, it should be treated properly and the utilization of this waste of high importance [8].

Glass-ceramics produced from various industrial waste materials is a promising idea and has been applied successfully to crystallize fundamental crystalline phases from the glass state [9]. Some of these materials have become commercial products, successfully applied as insulators, floor elements and wall tiles [10]. The production process uses specific routes and conditions such as heat treatment, temperature and dwelling time [1]. The main parameters that affect the properties and determine the further application of the obtained glass-ceramics are the chemical composition and morphology of these wastes.

Several researchers have been working on incorporating industrial waste materials into glass-ceramics for various structural applications such as adsorbents, lightweight structural components, filters and thermal protection systems [11]. Kayali [12] prepared ceramic bricks based on 100% fly ash as the solid ingredient using the same equipment that is commonly used in the clay brick industry. The specimens were first compressed in moulds to achieve the desired shape and then fired at 1000–1300 °C. The produced ceramic bricks were 28% lighter than standard clay bricks and showed excellent mechanical properties. The specimens showed a compressive strength of 43 MPa and tensile strength of 10.3 MPa, which is almost three times higher than that of standard clay bricks [12].

Basegio et al. [13] successfully utilized tannery sludge mixed with different proportions of clay. Batches were shaped in the mould by applying hydraulic pressure and then the bricks were sintered at 1000, 1100 and 1180 °C. The authors found that increasing the sludge ratio resulted in increased porosity and water absorption. On the other hand, by increasing sintering temperature, water absorption and porosity were decreased considerably. A higher firing temperature and a lower amount of sludge resulted in a ceramic brick with a bending strength of 25 MPa and the greatest dry density [13].

Many researchers have focused on the production of porous glass–ceramics based on industrial waste for special applications, such as sound and thermal insulation and lightweight aggregate in concrete. Acoustic and thermal insulators with appropriate microstructural properties present a great sustainable opportunity since using industrial waste contributes to the reduction of CO2 released to the atmosphere by 3.1 kg per kg of the manufactured material [14, 15]. Garcia et al. [16] successfully produced porous material for acoustic insulation based on a mix of slag with other wastes.

Another promising application of glass-ceramics based on waste was applied by Albertini et al. [17]. They obtained a porous glass–ceramics made of residues to eliminate dead zones in chemical–biological reactors. Chakartnarodom and Ineure [18] successfully produced foamed glass bricks based on a mix of glass waste with fly ash and rice husk, by applying 1% sodium silicate as a binder and 10% calcium carbonate as a foaming agent. The manufacturing process involved firing at various temperatures up to 900 °C for 1 h. The rectangular bricks were 30 cm × 30 cm × 7 cm. The measured compressive strength and density of the foamed glass were 59.9 kg/cm2 and 421 kg/m3, respectively [18].

Glass-ceramic materials present outstanding characteristics in composition, which make them favourable for weather and corrosion-resistant applications in advanced technology electronics and construction materials, unlike the traditional preparation of glass-ceramics from pure chemicals, which highly increases the cost of the produced materials. This study is based on primarily industrial wastes, which are considered very economically important [19,20,21]. The main aim of this study is to produce lightweight glass–ceramics by the direct sintering of bypass cement dust mixed with glass waste without using any external foaming agents and determining the optimum composition that results in the lowest density together with good mechanical properties.

Batch preparation

The raw materials used in this study were bypass cement waste and glass waste. The bypass cement waste was bought from the Masr Banisuif cement factory in Egypt and analysed by X-ray fluorescence spectroscopy. The major composition of the cement waste was calcia, silica and alumina with some other alkalis acting as a softening agent, as shown in Table 1. The major composition of glass waste was silica, sodium carbonates and calcium carbonates.

Table 1 Results of chemical analyses of glass waste and cement waste
Full size table

The materials (bypass cement waste and glass waste) were crushed using a ball mill to obtain a very fine powder that passed through a 75 μm mesh. Five batches, 20 bp, 30 bp, 35 bp, 40 bp and 50 bp, were prepared based on cement waste and glass wastes. The numbers indicate the weight percentage of the cement waste; the remaining volume is glass waste. For example, sample 20 bp contains 20% cement waste and 80% glass waste. Batch percentages with corresponding chemical compositions are shown in Table 2.

Table 2 Batch percentages (wt%) and chemical composition of investigated glass-ceramics
Full size table

After weighing the batches, they were mixed with a few drops of carboxymethyl cellulose to act as a binder and prevent rupturing. Subsequently, the samples were compressed by a uniaxial dry press into 10 cm × 10 cm × 1 cm disks with two parallel surfaces to avoid irregularity in thickness. Subsequently, the samples were moved to a preheated muffle furnace at 300 °C for 1 h to eliminate water and make the sample dry and ready for sintering without any cracks or rupturing.

Heat treatment

Thermal treatment of the samples was applied in a muffle furnace from room temperature to 1000 °C at 50 °C intervals and a 10 °C/min heating rate. One-hour soaking was applied when the furnace reached 1000 °C. The soaking time was based on the optimum time needed to induce the crystallization process in this kind of glass system as reported by Francis and Abdel Rahman [22]. After soaking, the furnace was switched off and the samples were kept inside to cool down to room temperature. On average, it took approximately 12 h to ensure complete cooling of the samples.

Experimental methodology

X-ray fluorescence

An X-ray fluorescence spectrometer, model “OPTIM’X WDXRF” was used for the chemical analysis of the glass and cement waste. The wastes were first crushed and the powder produced was sieved under 75 μm, then fed into the apparatus with 15 g from each waste to perform the analysis properly.

Differential scanning calorimetry

Differential scanning calorimetry (DSC) scans of about 300 mg of powdered glass–ceramic samples were applied with Al2O3 powder as a reference material using an analyser. The analysis was applied at a heating rate of 10 °C/min and sensitivity of 8 μV/cm for all the DSC runs that were performed in ambient conditions and a pure helium flow of 1.6 l/h. This methodology was used for determining the material heat capacity.

X-ray diffraction analysis

All samples obtained from fired batches were finely ground and scanned using an X-ray diffractometer, Empyreal, at room temperature using Cu kα radiation (λ = 1.5405 Å) and a scanning speed of 0.003°/min from 10° to 60° of 2Ɵ (Bragg angle) and 45 kV/40 mA as the applied power to detect the crystallized phases.

Scanning electron microscopy

The microstructure of the crystalline samples was studied using a scanning electron microscope (SEM), model LV-SEM Hitachi S-3400N. Chemical composition was measured with an EDS analyser.

Thermal expansion

Linear thermal expansion coefficient measurements of the glass-ceramics were carried out using a dilatometer, type NETZSCH DIL 402C. Twenty-millimeter long rectangular samples were heated up at 5 °C/min and the elongation on the chart paper was recorded. The thermal expansion was measured from 25 to 800 °C.

Microhardness

Indentation microhardness was determined for the obtained samples of glass-ceramics. Vicker’s hardness values were investigated with a Zwick 3212002/00 Tester, under a 0.1 kg load. A 10-s loading time was fixed for all crystalline samples.

Bending strength

Bending strength was evaluated in three-point bending tests of the glass-ceramic samples with unpolished, as-produced test pieces using a universal testing machine, INSTRON 4469, with a crosshead speed of 2 mm/min. The results obtained after multiple measurements were averaged.

Density and porosity

Density and porosity were measured using the Archimedes method, where water absorption was calculated as follows: water absorption percentage = (WS − WD/WD) × 100, where WD is the weight of the dry sample, WS was the weight of sample saturated with water. Total porosity was estimated as the apparent density/relative density ratio.

Results and discussion

Table 2 shows that the major constituents of all samples are silica and calcia, with a total percentage of roughly 75%. While adding bypass cement dust from 20–50 pb, the ratio of silica decreased from 61.16 to 45.36% while the ratio of calcia increased from 18.95 to 30.13%. This was because major constituent of bypass cement dust is calcia, so this change in chemical composition was accompanied by changes in physical, chemical and mechanical properties.

Figure 1 shows the X-ray diffraction analysis (XRD) of the obtained glass-ceramics. The major crystalline phase for all specimens is wollastonite, which may be attributed to the high percentage of silica and calcia found in the samples. Moreover, 1000 °C is high enough to successfully crystallize the wollastonite phase. It seems that almost all peaks in XRD are related to low-temperature monoclinic wollastonite β-CaSiO3 of maximum relative intensity (100%) at a diffraction angle of 2Ɵ equal to 29.898° (2.98 A°). The unreacted silica was working as a matrix of the glass and did not form any internal phases. These results are in line with Obeid [23], who prepared wollastonite based on silica and limestone with pure chemicals. The formation of wollastonite was summarized by the following chemical equation: CaCO3 + SiO2 → CaSiO3 + CO2.

Fig. 1
figure 1

X-ray diffraction of all specimens after heat treated at 1000 °C for 1 h

Full size image

In this study, bypass cement dust, rich in calcia, can be reacted with the glass wastes that are rich in silica to form a wollastonite phase (CaSiO3), where the nominal composition of wollastonite is 48.28 wt% CaO and 51.72 wt% SiO2, according to Table 2. For example, sample 20 bp contains 18.95 CaO + 61.16 SiO2. The percentage of wollastonite theoretically formed in a sample can be approximated by dividing the ratio of the calcia by the ratio of nominal composition, such as 18.95/48.28 = 39%. To calculate the ratio of silica consumed to form a wollastonite phase in sample 20 bp, the nominal composition can be multiplied by 39% as follows: (39/100) × 51.72 = 20.17%. Thus, the weight percentage of silica and calcia consumed to form wollastonite in sample 20 bp is 20.17% and 18.95%, respectively. Moreover, the residual silica can easily form the glass matrix, and can be calculated as follows:

$$ \mathrm{The}\ \mathrm{residual}\ \mathrm{silica}=\mathrm{weight}\ \mathrm{of}\ \mathrm{silica}\ \mathrm{in}\ \mathrm{sample}\ 20\mathrm{pb}\hbox{---} \mathrm{weight}\ \mathrm{of}\ \mathrm{silica}\ \mathrm{consumed}\ \mathrm{to}\ \mathrm{form}\ \mathrm{wollastonite}\ 61.16\%-20.17\%=40.99\%. $$

Based on the previous calculation, it is clearly shown that by increasing calcia, the tendency for forming wollastonite increased too, which can be confirmed by XRD. Figure 1 shows that the intensity and explicitly of the wollastonite peaks for 35 bp is much higher than 20 bp and 30 bp, although for 40 bp and 50 bp, the wollastonite ratio decreased again due to the decreasing silica ratio.

Figure 2 shows a scanning electron micrograph of sample 20 bp. The formation of a rectangular wollastonite phase among the glass matrix can be observed at 1000 °C. In addition, the formation of closed and open pores consolidated with the wollastonite phase was also noticed. The microstructure shows a large size appropriate grain growth without the precipitation of any secondary phases. EDS revealed that the composition of the crystal has a predominantly silicon (Si) content followed in abundance by oxygen (O), calcium (Ca) and sodium (Na). The majority of the microstructure is closed pores covered by wollastonite crystals. The quantity of closed pores is higher than open pores, which may be this attributed to the excess of residual silica after forming the wollastonite. This causes evaporation of bubbles. Although the microstructure has many pores, the matrix of the glass phase is highly rigid, as seen in Fig. 2. This probably reflects good mechanical properties. Moreover, forming closed pores has a good advantage of slightly absorbing water, unlike open pores which absorb water aggressively. Therefore, this type of microstructure can be applied in specific applications such as salt purification process and thermal insulation, as reported by Sulhadi [24].

Fig. 2
figure 2

Scanning electron micrographs and EDS of 20 bp after heat treated at 1000 °C for 1 h

Full size image

Figure 3 illustrates the microstructure of sample 30 bp, where the quantity of open pores increased because the volume of silica matrix decreased gradually by adding cement waste, which directly affects the density and porosity of the sample. Figure 3 also shows that many wollastonite crystals formed with irregular distribution along the matrix, and because the ratio of cement waste increased gradually, it also directly emphasized the increasing percentage of wollastonite formation. Irregular-shape grains and a rough morphology with a slight grain growth can be noticed. This is related to the incomplete reaction between silica and calcia. EDS shows the main components of the main crystal formed are silica and calcia, where the ratio of calcia increased sharply due to the addition of 30% cement waste, which is highly rich in calcia.

Fig. 3
figure 3

Scanning electron micrograph and EDS of 30 bp after heat treated at 1000 °C for 1 h

Full size image

Figure 4 demonstrates the SEM micrographs for batch 35 bp. It is obvious that the ratio of open pores rapidly increased compared to samples 20 bp and 30 bp. This may be attributed to the highest percentage of calcia that joined and reacted with silica to form wollastonite. Therefore, a high percentage of silica was consumed to form wollastonite and left a small portion of silica to hold the matrix. This made it easy for gases to escape from the matrix and form many open and closed pores. EDS also confirmed a higher percentage of calcia incorporated to form wollastonite in comparison with the other samples.

Fig. 4
figure 4

Scanning electron micrograph and EDS of 35 bp after heat treated at 1000 °C for 1 h

Full size image

Figure 5 represents the microstructure of sample 40 bp, which was composed of 40% cement waste and 60% glass waste. In comparison with sample 35 bp, the texture is completely changed. Although there is excess calcia found in cement waste that was supposed to react with silica sufficiently, it was noticed that the ratio of calcia that joined with silica to form wollastonite is lower than in the rest of the samples. This may be attributed to decreasing silica content by decreasing the ratio of glass waste. Hence, most silica incorporated in the glass matrix without reacting with calcia and those make the matrix very rigid and have a great influence on preventing gases from escaping the matrix and, consequently, it showed the lowest degree of porosity. It was easy to form closed pores rather than open pores which need a high degree of energy to make gases get out successfully.

Fig. 5
figure 5

Scanning electron micrograph and EDS of 40 bp after heat treated at 1000 °C for 1 h

Full size image

Figure 6 shows the micrograph and EDS of sample 50 bp. In this composition, the ratio of calcia joined to form wollastonite decreased again but, in this case, it is proportional to the decreasing silica ratio, and wollastonite crystal formation increased. Moreover, the remaining unreacted calcia incorporated in the matrix and compensated for the deficiency of silica, which had a great influence on initiating gases escaping again and left many open pores that directly affect the density of the samples. Moreover, the crystals were distributed in an unorganized manner along the structure.

Fig. 6
figure 6

Scanning electron micrograph and EDS of 50 bp after heat treated at 1000 °C for 1 h

Full size image

Figure 7 shows the measurements results of the coefficient of thermal expansion for samples 35 bp and 50 bp, where the recorded glass transition was 473 and 484 °C, respectively. This can be explained by increasing the cement waste from 35 to 50%, with the accompanying decreasing ratio of alkali that is directly affected by increasing the softening point where a deficiency of alkali inhibits glass formation at low temperature. That is why the softening temperature increased from 517 to 529 °C, to overcome the deficiency of alkali for samples 35 bp and 50 bp, respectively. Therefore, the coefficient of thermal expansion slightly decreased from 12.0 × 10−6 to 11.6 × 10−6 for 35 bp and 50 bp, respectively. The thermal expansion curves of the studied glasses are mostly like those characteristics of maximum silicate glasses. Where, generally, glasses with a lower expansion coefficient have higher transition and softening temperatures and vice-versa. The dilatometric Tg and Ts temperatures of the studied glasses showed a reverse behaviour when compared with the expansion coefficient behaviour, i.e. they increased with increasing CaO content. The heat capacity for sample 50 bp was also lower than the heat capacity for 35 bp due to increases in the glass phase in the latter sample, which acted as a barrier to heat transfer.

Fig. 7
figure 7

Thermal expansion coefficient and heat capacity of samples 35 bp and 50 bp

Full size image

The average values of density and porosity for the obtained glass-ceramics were calculated and are shown in Table 3. The results can be predicted by following the change of the microstructure as a result of the change of the chemical composition.

Table 3 Physical properties of investigated glass-ceramics
Full size table

Figure 8 shows the relationship between chemical composition, density and total porosity. It is clearly shown that by increasing the cement waste ratio from 20 to 35%, the density reached its lowest value (1.3 g/cm3), which is due to the highly porous structure corresponding to sample 35 bp. The total porosity designed for sample 35 bp is 56%; then, the porosity decreased gradually to 33% and 31% for samples 40 bp and 50 bp, respectively. Therefore, the density sharply increased due to the corresponding open pores decreased in both sample 40 bp and 50 bp. The net shape for the obtained glass-ceramics for sample 30 bp is shown in Fig. 9.

Fig. 8
figure 8

Effect of cement waste percentage on the density and porosity of the obtained glass-ceramics

Full size image
Fig. 9
figure 9

The obtained glass-ceramic sample (30 bp)

Full size image

Table 4 shows the microhardness and bending strength for glass-ceramic samples. The results show that sample 50 bp exhibits the highest microhardness. This result could be explained by more dispersed wollastonite and iron-enriched dark phases. However, the best mechanical results (high hardness and flexural strength) were found for sample 30 bp. This could be explained by the evenly homogenous distribution of the wollastonite phase, which increased the strength of the glass matrix. The lowest mechanical properties were achieved for sample 20 bp with microhardness and flexural strength values of 364.2 HV and 16.33 MPa. This was due to the lack of wollastonite formation and distribution in the glass matrix.

Table 4 Microhardness and flexural strength of samples
Full size table

Conclusions and future work

Cement waste with different proportions from 20 to 50% was mixed with glass waste to form lightweight glass–ceramics by applying a powder technology method. The characterization of the obtained glass-ceramics can be summarized as follows:

  • XRD shows that the major crystalline phase for all specimens is wollastonite;

  • The coefficient of thermal expansion is slightly decreased by increasing the cement waste;

  • The highest porosity and lowest density, with values 56% and 1.3 g/cm3, respectively, were found for sample 35pb (35% cement waste + 65% glass waste);

  • The most promising mechanical properties in terms of flexural strength and Young’s modulus were obtained for samples 30pb and 50pb. These batches are recommended for further investigation regarding potential structural application.