Introduction

The stage in which the final feature is achieved in ceramic tiles is sintering. Sintering is carried out in roller kilns because it is suitable for industrial production and the temperature can be evenly distributed in the furnace [1]. The basic characteristics expected from ceramic roller include excellent porous structure together with high temperature, thermal shock resistance, bending, and compressive strength. The presence of all these properties together improves the service life of the roller and reduces maintenance costs. The composition and firing regime of the rollers cause different crystalline phases to form, thus determining the characteristics of the finished roller. Kiln rollers mostly have mullite (3Al2O3·2SiO2) or alumina-mullite structures. The high creep strength of mullite at high temperatures and the high strength of alumina are the main factors for the use of both as kiln roller materials.

In roller kilns, the tiles are in direct contact with the roller and move on the kiln rollers. The firing regime and the atmosphere of the kiln change the chemical composition of the rollers and cause them to break. Kiln rollers are subject to destruction and structural deterioration due to the reactions of alkaline earth (magnesium and calcium) and alkali (potassium and sodium) elements and small amounts of iron, zinc, and chromium elements. While alkaline elements are dispersed in the roller structure due to their melting properties and form a glassy phase, other elements are attached to the roller surface and generally form calcium silicoaluminates and spinel phase structures. These changes in the structure also change the technical properties of the roller. The roller, which has undergone a chemical structure change, is broken and separated into waste due to heat and mechanical stresses during the roller change in the furnace [2].

Many research and development studies are carried out to protect nature and ensure sustainable production by recycling waste materials generated during production. There is also a lot of literature on the use of waste in the ceramics industry [3,4,5]. Elmahgary et al. [6] examined alternative tile formulations using three types of waste (waste sludge, roller waste powder, and cyclone dust) generated in ceramic tile production. In the study, these wastes were used in different proportions in the ceramic floor tile composition and the tests performed on these tiles have revealed that the bodies developed with the addition of waste at certain rates meet the ceramic floor tile standards. In a study on waste-related wall tiles, red waste mud, which is a waste of Konya Seydisehir aluminum plant, was used instead of kaolin in the body recipe, and its effect on physical properties was examined. It was found that when red mud waste was used instead of kaolin at a rate of 30% in the body recipe, there was an increase in dry and firing strength values [7]. Youssef and Ghazal [8] conducted a study on the use of kiln roller waste dust on the floor and porcelain tiles. It has been determined that these powders can be used up to 10% in the floor tile body composition. It has been stated that the water absorption and total porosity values within the porcelain tile body are similar to the standard tiles, but studies should be done to adjust the firing shrinkage [8]. In the study, on the usability of chromite processing powder waste, it was stated that it is possible to use chromite waste as an alternative colorant in colored ceramic glazes [9]. Bahtli and Erdem studied using foundry waste sand for investment casting instead of clay and kaolin in porcelain tile formulations. They found that using the foundry waste increases the density, firing shrinkage, and firing strength values of porcelain tile [10]. Roushdy [11] aimed to reuse roller kiln and ceramic tile sludge waste powder in ceramic floor tile body composition. In the study, optimum floor tile properties could be obtained with the compositions 1% roller kiln waste, 35% ceramic tile sludge, 64% floor tile mixture or 2% roller kiln waste, 24% ceramic tile sludge, 74% floor tile mixture.

Recycling of waste in previous studies contributes to the decrease in natural raw material supply, reduction of production costs, and conscious consumption of natural resources due to the use of waste as an alternative raw material [12,13,14,15,16,17].

When the literature is examined, different wastes are evaluated as raw materials and generally used within the body. In this study, the use of ceramic kiln roller waste in porcelain matte-opaque glaze composition is examined. The oxides in the ceramic glazes cover the surface in a glassy or partially glassy form in the face of heat and prevent abrasion, scratching, etc. It increases the surface resistance against chemical and mechanical effects. At the same time, the glaze layer gives the surface a decorative feature and ensures easy cleaning [18]. For this reason, it has been investigated whether the kiln rolls are used instead of aluminum oxide and quartz raw materials in porcelain tile matte-opaque glaze.

Experimental

Two different kiln roller waste samples (RW1 and RW2) were collected from production kilns in the NG Kutahya Seramik factory to determine the suitable roller waste to be used in matte glaze trials. RW1 roller waste samples were collected from the firing zone of the floor tile production kiln, and RW2 roller waste samples were taken from the firing zone of the porcelain tile production kiln. Chemical and mineralogical analysis of two different roll waste samples were carried out with Spectro Xrf Xepos X-ray spectrometer and Rigaku brand Miniflex 600 model X-ray diffractometer and are given in Table 1 and Fig. 1, respectively. According to the analysis results a suitable roller waste was selected for the porcelain tile matte-opaque glaze recipe and glaze recipes were prepared with this kiln roller waste. The porcelain tile matt-opaque glaze composition used in the production of NG Kutahya Seramik Factory is taken as a reference and accepted as the standard recipe (STD). Roller waste contains mainly quartz and alumina (Table 1). For this reason, it has been used as an alternative to quartz and alumina in the recipes. The standard glaze recipe contains alumina and quartz at a rate of 12%. Roller wastes were used instead of alumina and quartz raw materials in at a rate of 5% and 9%. Recipes were developed to obtain the same Seger ratios as the standard glaze. Recipes are indicated as PG1, and PG2 respectively. Additionally, the standard glaze recipe contains 3% zircon as an opacifier. In order to see the effect of roller waste as an opacifier, the zircon was completely removed and roller waste was used instead of zircon in a 3% ratio in PG3 composition. The Seger ratios of recipes are given in Table 3. The raw materials and different waste kiln roller samples were obtained from the NG Kütahya Seramik factory. The kiln roller samples, which were separated into waste, were crushed and ground into powder in an alumina ball mill. The chemical analysis of the raw materials used in the porcelain tile matte glazes and different waste kiln roller samples were determined with the Spectro Xrf Xepos X-ray spectrometer and are given in Table 2. Phase compositions of materials were investigated by X-ray diffraction (XRD) using a Rigaku Miniflex 600 system in the 2θ range of 10°–70° and Ni-filtered Cu Kα1 radiation (λ = 0.15406 nm). Differential thermal analysis (DTA) and thermogravimetric analysis (TG) of standard matt-opaque glaze were carried out simultaneously using Netzsch STA 449 F3. The raw materials which were weighed according to their ratios in the glaze recipe were ground in a laboratory alumina ball mill with a capacity of 300 g until a 45 µ sieve analysis was 0.5–1%. The density of the prepared glazes was measured with a pycnometer in the range of 1500–1550 g L−1. The viscosity behavior of the recipes was measured between 70 and 75 s with a 4 mm diameter Ford cup. The glazes were sieved from the ball mill through a 125 µm sieve and were applied with a pistol on the 15 cm × 15 cm engobed porcelain tile. The glazed tiles were sintered at 1190 °C for 44 min in the NG Kutahya Seramik factory industrial kiln. Mineralogical analysis (XRD) of the developed and standard recipes was made with a Rigaku Miniflex 600 model. The gloss values of matte-opaque glazes were determined with the KSJ MG268-F2 gloss meter device, and the thermal expansion coefficients were determined with a dilatometer (Netzsch DIL 402 (10–7/°C–400 °C)). The surface roughness of glazes was measured with the Time brand 3110 (TR110) model surface roughness device. Color measurements (L, a, b) of matte glazes were carried out by using a spectrophotometer (Konica Minolta CM 600D). The sintering behaviors of glazes were characterized by using the Misura ODHT HSM 1600/80 heating microscope. The surface wear test was performed on the standard and developed samples with the Gabbrielli ISO-4 Abrasimeter surface wear device according to the TS EN ISO 10545-7 standard.

Table 1 Chemical compositions of kiln roller wastes
Full size table
Fig. 1
figure 1

X-Ray diffraction analysis of RW1 and RW2 kiln roller wastes, M: mullite, C: corundum, Cr: cristobalite

Full size image
Table 2 Chemical compositions of raw materials
Full size table
Table 3 Seger ratios of glaze recipes
Full size table

Results and discussion

Characterization of raw materials and standard glaze

The chemical analysis and XRD patterns of two different kiln roller wastes are given in Table 1 and Fig. 1, respectively. Both of the two kiln rollers contain mainly Al2O3 and SiO2. The amount of Al2O3 in the RW1 roller waste is significantly lower compared to the RW2 roller waste, while the amount of SiO2 is significantly higher. This result is also confirmed by the XRD analysis of the roller wastes. Mullite and corundum are the main crystalline phases determined in roller waste. When the peak intensities were compared in the XRD graph, it was determined that the amount of corundum crystal phase in RW2 roller waste was significantly higher. Small amounts of other oxides are present due to the contact of the rollers with the ceramic tiles. In the XRD mineralogical analysis of the RW1 roller sample taken, it was determined that there is a different cristobalite (SiO2) phase. Therefore, RW1 waste was not used in matte glaze recipes as it would reduce opacity. The matte appearance in glazes is an effect caused by the presence of crystals dispersed in the glaze. Because of the crystals embedded in the glaze, the incident light is refracted and scattered. Therefore matte glazes are often also more or less opaque. Opacity is depending on the degree of crystallization and the refractive index of crystals. The most efficient crystals for opaque glazes are those with higher refractive indices than those of the glass matrix. Some of these crystals are corundum Al2O3 (n = 1.77), spinel MgAI2O4 (n = 1.72), diopside CaMgSi2O6 (n = 1.67), mullite Al6Si3O18 (n = 1.64), wollastonite CaSiO3 (n = 1.63) and even anorthite CaAl2Si2O8 (n = 1.58). The refractive index of cristobalite is very close to the refractive index of the glassy phase and is about 1.5 [19]. According to the XRD analysis, especially corundum and mullite peak intensities of RW2 roller waste are higher. This shows that the roller waste has more corundum and mullite phases. Consequently, RW2 roller waste was determined to be used in matte-opaque glaze recipes.

Thermal analysis of the standard glaze was made by TG–DTA curves and is given in Fig. 2. The TG–DTA curve of standard glaze showed that three decompositions occurred at temperatures 518.0 °C, 784.5 °C, and 799.0 °C, with a total mass loss of 33.7%. The first decomposition occurred at 518.0 °C and corresponded to kaolin to metakaolin decomposition [20]. The mass loss was 2.16%. Dolomite decomposition occurred between 650 and 850 °C in two main stages. In the first stage, dolomite decomposes into calcite, periclase, and carbon dioxide. In the second stage, the decomposition of calcite to calcium oxide and carbon dioxide occurred at 799.0 °C. The total mass loss was 7.63%. An endothermic peak at about 574.0 °C is attributed to α to β-quartz transition. The exothermic peak seen at 910.5 °C can be explained as anorthite and diopside crystallization according to the XRD results in Fig. 5. Anorthite and diopside predominantly crystallize in the range of 800–990 °C [21,22,23]. Since diopside and anorthite crystallize in similar temperature ranges, it is estimated that a single crystallization peak is observed. The thermal behavior, decomposition, and crystallization temperatures, of standard glaze in this study are consistent with the XRD analysis (Fig. 5) and results previously reported in the literature. With the increase in temperature, first diopside and then anorthite decomposes. When the temperature increases to 1086.0 °C, the diopside is partially dissolved. Then anorthite crystals begin to melt at 1133.0 °C. No further transformation was observed up to 1200 °C.

Fig. 2
figure 2

TG–DTA curves of standard matte-opaque glaze

Full size image
Fig. 3
figure 3

Glazed surfaces after wear resistance test

Full size image
Fig. 4
figure 4

Heating microscopy results of Std, PG-1, PG-2, and PG-3 glazes

Full size image
Fig. 5
figure 5

X-Ray diffraction analysis of the Std, PG-1, PG-2, and PG-3, Q: quartz, Zr: zircon, D: diopside, A: anorthite

Full size image

The chemical analysis of the raw materials used in porcelain tile matte-opaque glaze recipes are given in Table 2 and the Seger ratios of the recipes are given in Table 3. In PG1 and PG2 glazes, RW2 kiln roller waste was used instead of alumina and quartz, with the standard glaze Seger ratios not changing significantly. Roll waste was used at the ratio of 5% in the PG1 and 9% in the PG2 glaze recipe. Roller waste mainly contains SiO2 and Al2O3, but also some impurities such as alkali oxides (Na2O, K2O, CaO, MgO) and Fe2O3 (Table 1). Therefore, the alkali oxide ratios of the glaze recipe slightly increased according to the standard with the use of roll waste instead of quartz and alumina. Additionally, 3% of RW2 roll waste was used instead of zircon in the PG3 glaze recipe to determine the effect of roll waste on opacity.

Optical and physical properties

Color values (L*, a, b) of glazes are given in Table 4. The results showed that the whiteness value (L*) of the glazes decreased by increasing the kiln roller waste amount in all glaze compositions. Although roller waste is used at the lowest ratio (3%) in the PG3 glaze composition, it is noteworthy that the decrease in the whiteness (L) degree of the glaze is the highest. The use of roller waste instead of zircon in the PG3 glaze caused this significantly decreasing effect on whiteness. In matte-opaque glaze compositions, the opacity is mainly related to the refractive index difference between the crystals and the glassy phase. For this reason, zircon is often used as an opacifier in ceramic glazes with its high refractive index (1.94) [24, 25]. The whiteness value of PG1 and PG2 glazes was also lower than the standard glaze.

Table 4 Optical and physical properties of glaze compositions
Full size table

The matte and glossy appearance of the glazes is explained by the amount of specular reflectance from the glaze surface. When a surface is less than smooth, some of the light reflected specularly from the surface is reflected at angles other than the incident light, because the surface was not flat at the point of interaction [26]. In addition, the diminished glossiness observed in glazes containing added alumina can be elucidated by the refractory characteristics of crystalline phases, as they function as a structural framework that impedes the process of densification. When the glossiness and surface roughness values are evaluated, it is observed that the surface roughness of these glazes decreases while the glossiness increases. The gloss value increases with an increase of 5% and 9% by mass in the addition of roller waste. Surface roughness is indicated by Ra which defines the arithmetic average of the absolute values of the profile height deviations from the mean line [27]. The surface roughness of the glazes was expressed by the parameter Ra and ranged from 0.5 to 5. In glazes, matte appearance is directly related to surface roughness and as the surface roughness increases, the matte effect increases. The main factor that affects the surface roughness is the tendency of the glaze to crystallize. In addition to that, the morphology of the crystals also affects the surface roughness [27, 28]. When the peak intensities of anorthite and diopside crystals are compared in the normalized XRD graph given in Fig. 5, it is observed that the peak intensities of anorthite and diopside crystals decreased when roller waste was used instead of alumina and quartz in the glaze composition. It can therefore be concluded that the use of roller waste in glaze compositions limits the crystallization of anorthite and diopside. Thus, the gloss and surface roughness values obtained in the glazes developed with roller waste were justified with XRD results.

In Fig. 3, the surface image of the glazes after the wear resistance test is given. Surface abrasion values of roller waste-added glazes showed similar results to the standard. According to the results of the abrasion test performed at 6000 cycles, it was determined that the abrasion resistance of all glazes was PEI 4. It can therefore be concluded that all glazes can be classified as glass–ceramic coatings with high abrasion resistance. The wear resistance of ceramic glazes is the result of a variety of factors, such as micro-hardness and roughness. But it also depends on the amount and type of crystalline phase and the presence of a glassy phase [24, 29]. It is seen that there is no change in the wear resistance values of the roller waste-added glazes. The PG2 glaze with the lowest surface roughness was characterized by the same abrasion resistance as the standard glaze with the highest surface roughness among the tested glazes. This shows that there is no significant decrease in the amount of crystallization to change the wear resistance with the use of roller waste in glaze compositions.

Thermal properties

Heating microscopy analysis results of glazes are given in Fig. 4 and characteristic points are given in Table 5. Although the heating microscope curves of the standard and developed glaze samples were obtained close to each other, it was determined that the glaze softened a little more with the increase in the use of roller waste instead of quartz and alumina in the glaze recipe. In PG1 and PG2 glaze recipes, roller waste was used instead of alumina and quartz, which are hard raw materials with high melting temperatures. The melting points of alumina and quartz are very high, which are over 2000 °C and 1750 °C. In addition, there are alkali oxides such as Na2O, K2O, CaO, MgO, and Fe2O3, which facilitate the melting of glaze, in the roller waste composition. Hence, the melting behavior of PG1 and PG2 glazes was slightly softer than the standard. In the PG3 glaze, the roller waste used instead of zircon hardened the glaze by causing the development of anorthite and diopside crystal phases (Fig. 5). These melting behavior analysis results also confirm the surface glossiness values of the glazes. PG1 and PG2 glazes, which show softer melting behavior, have higher gloss values compared to the standard glaze, while the harder PG3 glaze is more matte.

Table 5 Thermal expansion coefficients of glazes
Full size table

Thermal expansion coefficients (α) of the glazes which are shown in Table 5 were determined between 100 and 800 °C. According to the results, thermal expansion coefficient values decreased after the roller waste was used in glaze recipes. The increase in the amount of waste used especially in PG1 and PG 2 glazes significantly decreased the coefficient of thermal expansion. The low thermal expansion feature of the roller waste-added glazes is due to its least amount of crystalline phases such as quartz and diopside (Fig. 5), which increase the thermal expansion coefficient. Table 5 shows that the decrease in the amount of quartz and diopside crystals reduced the thermal expansion of ceramic glazes. Due to its extremely low thermal expansion, fused quartz has been widely used to reduce the thermal expansion of glazes. On the other hand, crystalline quartz significantly increases the coefficient of thermal expansion [30]. Diopside crystals also have a higher coefficient of thermal expansion than anorthite crystals [31,32,33].

The thermal expansion coefficient of glazes is one of the most important features to be considered while designing the composition. The thermal expansion coefficient of the glaze must be compatible with the body. Otherwise, it may cause various deformation and cracking problems in porcelain tiles. Since compressive stress is required in ceramic tile production, the thermal expansion coefficient of the glaze is lower than the body. However, if the thermal expansion of the glaze is too low, concavity occurs [33]. The thermal expansion coefficient of roller waste-added glazes is quite lower than standard and thus may cause less convexity deformation in porcelain tiles.

XRD analysis

XRD patterns of glazes are given in Fig. 5. Anorthite, diopside, quartz, and zircon crystalline phases were determined in all glaze compositions. Zircon and anorthite are the main crystalline phases, whereas the diopside and quartz are the secondary crystalline phases. Although glazes have very close Seger ratios, it can be observed that anorthite and diopside crystallization decreases with the use of roller waste in PG1 and PG2 glazes. When the peak intensities of the quartz phase are compared, it is seen that the quartz phase is less in recipes PG1 and PG2. This shows that the use of roll waste increases the quartz solubility in the glaze. The result of X-ray diffraction verified the high gloss, low surface roughness and low thermal expansion coefficient values of PG1 and PG2 glazes. On the other hand, in the PG3 recipe where the roller waste is used instead of zircon, it is seen that the zircon phase decreases significantly, while the development of anorthite and diopside phases increases. Despite this increase in anorthite and diopside crystallization, it can be deduced that roller waste used instead of zircon in the glazes led to a decrease in opacity. This is due to the high opacifying effect of zircon in glazes.

Scanning electron images of glazed surfaces of all samples and EDX analysis are shown in Fig. 6. The microstructure of matte-opaque glazes is less amorphous phase and more compact with the presence of crystals. Diopside (D), anorthite (A), and zircon (Z) crystals and glassy phases are observed in all glazes. EDX point analysis confirms the results of XRD results, indicating the presence of diopside and anorthite, the chemical composition of the crystalline phases identified in the SEM images. The EDX analysis given in Fig. 6 shows the elemental composition for the diopside and anorthite phases identified in the glazes. In the SEM studies, bar-shaped anorthite crystals are observed as clearly darker gray areas. EDX analysis in Fig. 6 confirmed that the clustered-shaped crystals belong to the diopside. White and acicular structures typical of zircon crystals are also observed in the SEM images with lengths ranging from 1 to 5 μm. The zircon crystals phase is mainly responsible for the excellent opacity of glazes.

Fig. 6
figure 6

SEM images and EDX analysis of STD, PG-1, PG-2, and PG-3 glazes

Full size image

Conclusions

In this study, the use of kiln roller wastes in a porcelain tile matte-opaque glaze recipe was investigated with an environmentalist approach and it was aimed to contribute to sustainable production. In accordance with the results of the chemical analysis and phase analysis of the waste, the roller waste was used instead of alumina, quartz, and zircon in standard glaze composition. Experimental results showed that when roller waste is used instead of alumina and quartz in matte-opaque glaze composition, it decreases the crystallization development of anorthite and diopside. Thus, the usage of kiln roller waste increases the gloss value and thermal expansion coefficient of the glaze and decreases its whiteness. However, the most significant decrease in the whiteness value of the glaze was obtained when the roller waste was used instead of zircon. Therefore, it will not be appropriate to use roller waste instead of zircon in matte glaze compositions where opacity is required. Moreover, when it is used instead of zircon, it has been revealed that it improves anorthite and diopside crystallization, and as a result, the surface roughness of the glaze increases, and the matte appearance increases. According to the results, when roller waste was used at a rate of 5%, the acceptable values to the standard glaze technical properties were obtained. When all the results were evaluated, it was concluded that it’s possible to use kiln roller waste up to 5% ratio in porcelain tile matte-opaque glaze compositions with also some optimizations to be made in the Seger ratios of the glaze composition. Considering the use of roller waste instead of alumina and quartz, working with lower SiO2/Al2O3 ratios in matte-opaque glaze compositions will help prevent the loss of matte appearance in the glaze. Optimizations to increase the CaO/MgO ratio in glaze recipes will also increase the opacity of the glaze. In this way, the use of roller waste as a raw material in the composition of the porcelain tile glazes is important in terms of reducing the cost of the glaze recipe, as well as recycling the waste into production and protecting the environment.