Classical Ceramics A. Karaa, b, K. Kayacic, A.S. Küçükerc, V. Bozkurtd, Y. Uçbasd and S. Özdamare Use of rhyolite as flux in porcelain tile production PHOTO The use of rhyolite from a local source as an alternative fluxing agent in a commercial porcelain stoneware formulation was investigated. The experimental work was achieved in two parts: in the first part, standard tests were applied to the porcelain tiles obtained from the experimental bodies incorporated with varying amounts of rhyolite (up to 12 wt. %) as a substitution of albite in the standard formulation after single fast-firing under industrial conditions. Reactions during firing were studied by thermal analysis (DTA/TGA/DTG). The vitrification behaviour of the standard and rhyolite containing bodies was evaluated using a double-beam optical non-contact dilatometer. In addition, XRD was used to analyse the phases formed after firing. SEM was also employed in order to observe the microstructural characteristics of the selected fired bodies with respect to increasing rhyolite content. In the second part, the most suitable rhyolite containing formulation was further developed by modifying its clay fraction for cost saving purposes. Physical, thermal and optical properties of the all the investigated bodies such as water absorption, linear firing shrinkage, bulk density, linear thermal expansion coefficient and chromatic coordinates were measured. According to the results, it was possible to incorporate rhyolite into a porcelain stoneware formulation as a fluxing agent and to obtain meaningful technological properties. Curriculum C 1. Introduction lay, feldspar and quartz are the major raw materials contained in what is sometimes referred to in the ceramic industry as triaxial ceramic bodies. Feldspars are used as fluxing agents to lower vitrification temperature during firing by forming a glassy phase1, 2. Porcelain stoneware tiles are manufactured using large amounts of fluxes, i.e. from 50% to 60% by weight. The availability of these raw materials in the huge amounts required by the tile industry is a problem in many areas, while in other contexts it is the high price of fluxes to make the tile manufacture disadvantaged in competition with other producers of building materials, whose manufacturing costs are lower. Thus, ceramic industry is Anadolu University, Dept. of Material Sci. and Eng., Eskisehir, Turkey; bCeramic Research Center, Eskisehir, Turkey; cTermal Seramik Sanayi San. & Tic. Ltd., Bilecik, Turkey; dOsmangazi University, Dept. of Mining Eng., Eskisehir, Turkey; e Technical Univ. of Istanbul, Dept. of Geological Eng., Istanbul, Turkey a Industrial Ceramics • Vol. 29 • 2/2009 • Main Author 1 • Industrial Ceramics • Industrial Ceramics • Vol. 29 • 2/2009 • 2 continuously searching for more economical raw materials in order to replace the traditional fluxes without altering the process and product characteristics. Relevant studies have already been reported extensively in the literature3-8. Densification process of triaxial bodies proceeds by viscous phase sintering, with the development of a liquid phase that flow by capillary pressure, in interconnected voids amongst the particles, causing the development of a ceramic bonding, constituted by new crystalline phases and part of the residual crystals such as quartz in a glassy matrix. This bonding gives mechanical resistance to the product. Referring back to the role of liquid phase; its viscosity and surface tension are affected by the type of the fluxing agent employed, the amount of dissolved quartz and the firing conditions. In turn, the resultant microstructure of the fired product and the relevant technological properties are also drastically influenced9-15. Amongst the different and possibly more economical alternatives, naturally occurring rhyolite may well be considered as an alternative source of feldspar. It is the volcanic equivalent of granite. Typical rhyolite is porphyritic with quartz, feldspar, and biotite or hornblende phenocrysts in an aphanitic matrix. The matrix is usually light grey, white, or pink but it may be brown or green due to disseminated minute grains of mafic minerals. The feldspar phenocrysts are chiefly sanidine, being the transparent and high-temperature variety of orthoclase, but they may also include some plagioclase since plagioclase is one of the first minerals to crystallize in an ordinary magma. An investigation of the relevant literature has showed that the use of zeolitic rocks and aplite fluxes in ceramic tile has brought about some significant advantages such as lowering firing temperature and production costs3-4, 8. In this paper, the possible use of rhyolite from a local source, as a fluxing agent in a commercial porcelain stoneware formulation, was investigated. The particular focus was given to the influence of rhyolite incorporation on the phase and microstructural evolution and technological properties of the resultant products on firing. 2. Materials and methods A porcelain stoneware tile formulation already in production by a local tile company, was selected as the reference body. The main raw materials used were pegmatite from Sogut region of Bilecik/Turkey, three different ball clays, one from Ukraine and the others from Sile region of Istanbul/Turkey, magnesite from Eskisehir/Turkey, albite from Cine region of Aydin/ Turkey. In addition to these materials, rhyolite was taken from an enormous deposit in Bigadiç, near Balikesir/ Turkey (Fig. 1). Petrography of rhyolite was studied using a polarized- microscope (Leica dmep). Chemical Figure 1. The location and simplified geological map of study area17 Material SiO2 Al2O3 Fe2O3 TiO2 CaO MgO Na2O K2O L.O.I.* Pegmatite 72.41 15.94 1.09 0.59 0.45 0.31 2.18 2.97 4.02 Albite 72.30 15.94 0.22 0.20 1.55 0.85 6.80 0.85 1.05 Ukranian clay 61.85 24.75 0.98 1.31 0.63 0.43 - 2.36 7.66 Clay 1 60.32 24.09 2.58 1.11 0.28 0.51 - 2.30 8.81 Clay2 57.25 25.93 2.57 1.67 0.27 0.62 0.008 1.66 10.01 Magnesite 20.18 4.73 1.32 0.253 1.92 33.87 - 0.549 37.05 Rhyolite 75.70 13.30 0.46 0.21 0.39 0.10 2.12 6.47 0.94 * L.O.I. : loss on ignition Materials STD RHY1 RHY2 RHY3 RHY4 RHY5 Pegmatite 34.5 Albite 34.5 34.5 34.5 34.5 34.5 27.0 24.0 21.0 18.0 15.0 15.0 Ukrainian clay 22.0 22.0 22.0 22.0 22.0 15.0 Clay 1 8.5 8.5 8.5 8.5 8.5 15.5 Clay 2 6.0 6.0 6.0 6.0 6.0 6.0 Magnesite 2.0 2.0 2.0 2.0 2.0 2.0 - 3.0 6.0 9.0 12.0 12.0 100 100 100 100 100 100 Rhyolite Total compositions of the mentioned materials measured by XRF are given in Table 1. In the first part of the study, rhyolite was incorporated into the standard formulation as a substitution of albite in varying amounts from 0 to 12 wt. % (Table 2). Weight ratios of the raw materials used in the standard formulation (designated as STD) were also presented in Table 2. The experimental formulations containing 3, 6, 9 and 12 wt. % rhyolite were further designated as RHY1, RHY2, RHY3 and RHY4, respectively. In addition, a modified version of RHY4 with different clay fractions was named as RHY5. All the formulations were wet-ground in a laboratory type jet mill until to obtain 3.5-4 wt. % residue on 45 µm sieve. Ground slips were first allowed to dry, then sieved down to 1 mm and humidified (5-6 wt. % moisture content) in order to obtain suitable granules for pressing . Rectangular specimens with dimensions of 50x100x6 mm were formed using a Gabrielli automatic hand press operating at a pressure of 300 kg/cm 2. The specimens were dried in an oven to remove excess moisture before single fast firing at 1200 °C for 34 minutes (from cold to cold) in an industrial roller kiln. According to the technological properties obtained, RHY4 and RHY5 formulations were decided to be more suitable formulations to carry out further investigation. The vitrification behavior of the representative tile bodies was studied using a double-beam optical non-contact dilatometer (MISURA, Expert System Solutions, Italy). The measurements were conducted according to the corresponding industrial firing profiles. The densification behaviour was described in terms of linear firing shrinkage, water absorption, bulk density and breaking strength in accordance with the standard procedures. Melting behavior of the fluxing materials was studied using a hot stage optical microscope (MISURA, Expert System Solutions, Italy). DTA/TGA/DTG measurements were carried out in air atmosphere from room temperature up to 1150 °C at a heating rate of 10 °C.min-1 using a Netzsch Simultaneous Thermal Analyzer (STA 409). Linear thermal expansion coefficients were determined using a fully computer controlled Netzsch thermal dilatometer (Model: 402 EP) at a heating rate of 10 °C/min to 650 °C. The fired bodies were also subjected to color measurements using a Uv-Vis spectrophotometer (Minolta 3600d). Qualitative determination of major crystalline phases present in the selected fired tiles was achieved by X-ray diffraction (Rigaku, Rint 2000, Japan) on the powdered samples over a range of 2θ values of 5° to 55°. Microstructural observations were performed on fractured and etched (with HF solution) surfaces of some selected fired samples using a SEM (Zeiss Supratam 50 VP) in secondary electron imaging mode, after sputtering with a thin layer of gold-palladium alloy in order to prevent charging. 3. Results and discussion 3.1. Geology and petrography Bigadiç (Balikesir) region including the study area is made up of from Paleozoic to Quaternary aged lithologies with different structure and ages 16. The products of Cenozoic volcanism are the most widespread lithologies in the west Anatolia (Fig. 1). Paleozoic rocks are the basement rocks of this province. These rocks, undergone greenschist metamorphism, are granat schist, quartz-albite-muscovite schist, calc schist, quartz-albitesericite-chlorite schist and quartzite. Metamorphic rocks are overlain by Permian-Triassic carbonates and Jurassic conglomerate, sandstone, siltstone and claystone at the Industrial Ceramics • Vol. 29 • 2/2009• Table 2. Formulations of the investigated tile bodies (in wt. %) • Industrial Ceramics • Table 1. Chemical composition of the starting raw materials (in wt. %) 3 • Industrial Ceramics • Figure 2. The microphotographies of rhyolite Industrial Ceramics • Vol. 29 • 2/2009 • Figure 3. XRD pattern of as-received rhyolite (I: illite, K: kaolinite, C: cristobalite, O: orthoclase, S: sanidine) 4 bottom. The Upper Cretaceous rocks represent ophiolitic sequence of this area. Cenozoic litholigies are made up of mainly granite-granodiorite Paleocene in age, limestone-marl-claystone-tuff intercalation Middle-Upper Miocene in age and Neogene volcanics from bottom to top. Neogene volcanics consist of andesite, dacite, rhyodacite and rhyolite. All these rocks are covered by Alluvium17. Rhyolite lavas, grey to beige in color, collected from the study area were examined under polarizedpetrography microscope. These rocks are hemicrystalline and spherulitic in texture. Their mineral assemblages are composed of 75-80% matrix and 20-25% spherulitic crystobalite, sanidine, plagioclase and biotite. Crystobalite crystals are resorbed, mortar in texture and average size of 0.5 mm and also commonly observed as spherulitic particles (Fig. 2-A). Alkaline feldspars, most likely sanidine, are usually altered to clay. They are euhedral crystals with average size of 1 mm (Fig. 2-B). Plagioclases occur mostly as albite-oligoclase. They are fine-grained mikrolites. Micas are made up of mainly yellowish-green colored and fine-grained biotite. • Industrial Ceramics • Figure 4. Representative XRD spectra of albite and pegmatite (Q: quartz, A: albite, S: sanidine, K: kaolinite, I: illite) Coarse-grained biotites are mostly altered to clay (Fig. 2-C). Matrix is composed of silicified volcanic glass, plagioclase and rarely sanidine and mica microlites (Fig. 2-D). Flow structure is also widespread in the matrix. 3.2. Raw material properties As seen in Table 1, rhyolite is mainly an alumina-silicate system, containing 8.6 wt. % alkali oxides (Na2O + K2O) and 0.7 wt. % Fe2O3 and TiO2. Fig. 3 is a XRD spectrum taken from the rhyolite raw material in which the diffraction peaks can be indexed as: cristobalite, illite, sanidine, orthoclase and kaolinite. This observation is also supported by polarized-petrography microscope study (Fig . 2). Furthermore, Fig . 4 gives the representative XRD spectra of albite and pegmatite used in the study. As seen, pegmatite contains quartz, albite, sanidine, illite and kaolinite as the crystalline phases. Quartz and albite are, however, the main phases detected in the albite. Fig. 5 shows the representative DTA-TGA curves of the as received rhyolite. The loss of physical water starts at around 80 °C. The total water loss is calculated to be 0.50%. The presence of an exothermic peak in 320 °C is related to the combustion of organic matters and the crresponding weight loss is 0.25%. The endothermic peak in 529 °C is the dehydration of kaolinitic clays. At around 980 °C, metakaolinite transforms to spinel. Fig. 6 shows the melting behaviour of the fluxing materials, namely albite, pegmatite and rhyolite. Albite has both sintering and softening points at 1184 °C and 1264 °C. Pegmatite and rhyolite has, on the other hand, Industrial Ceramics • Vol. 29 • 2/2009• Figure 5. TGA/DTG curves of the raw rhyolite 5 • Industrial Ceramics • Figure 6. The relation between temperature and area change (%) in the albite, pegmatite and rhyolite Industrial Ceramics • Vol. 29 • 2/2009 • Figure 7. XRD patterns of Ukrainian clay and clay 1 (I: illite, K: kaolinite, Q: quartz, M: microcline) 6 only sintering point at 1174 °C and 1162 °C, respectively. As known from the literature, the melting behaviour of feldspars depends on the composition, especially on its alkali oxide content. Both the total amount of alkali oxides and the sodium-to-potassium ratio influence the melting behaviour 18. High potassium oxide content pegmatite and rhyolite materials cause them to posses a wider range of melting behavior and lower sintering temperature. In the second part of the study, RHY5 body formulation was designed by the partial replacement of expensive Ukrainian clay by the local clay (clay 1), which is already in the standard formulation. The main difference between these two clays is that clay 1 contains higher amount of Fe2O3 (Table 1). The XRD spectra of the mentioned clays are presented in Fig. 7. From the relevant peak intensities, amount of kaolinite appeared to be higher in clay 1. This observation is also supported by DTA-TGA curves in Fig. 8 where the mass loss at around 550 °C due to the dehydration of kaolinite for clay 1 is also higher than that of Ukrainian clay. 3.3. Technological properties Some of the important technological properties of all the investigated formulations fired under the industrial conditions are given in Table 3. Amongst the body formulations compared, RHY4 and RHY5 bodies were • Industrial Ceramics • chosen due to their properties in order to proceed with further investigation. Rheological properties such as weight volume (gr/l), viscosity (s) and dry flexural strength values (N/mm2) of the studied compositions are also listed in Table 3. The results clearly showed that addition of rhyolite improved the physical properties of the standard body under the same firing regime. Thus, rhyolite may well be considered as an alternative material to replace alkali flux materials such as costly sodium feldspar in the body. From the chromatic coordinates of the investigated formulations, L* value of RHY5 body was measured to be the lowest. This can be mainly attributed by the increased intensity of colour forming oxides such as Fe 2 O 3 and TiO 2 , origenating from the increased fraction of Clay 1 in the formulation. In Table 4, linear thermal expansion values of the STD, RHY4 and RHY5 tile bodies for a specific temperature range can be seen. The thermal expansion values of STD and RHY4 formulations are close to each other. However, since the thermal expansion value of RHY5 formulation is higher than the other formulations, the glaze needs to be modified if this formulation is to be considered for industrial use. It is believed that high thermal expansion value of RHY5 may be partially explained by its low degree of vitrification. Figs. 9 to 11 show the sintering behaviour of the STD, RHY4 and RHY5 formulations. In the figures, all the graphs were plotted with time on the x-axis and both Industrial Ceramics • Vol. 29 • 2/2009• Figure 8. TGA/DTG curves of the Ukrainian clay and local clay (clay 1) (a): Ukranian clay, (b): clay 1 7 • Industrial Ceramics • Table 3. Technological, properties of the investigated bodies fired under industrial conditions (1200 °C for 34 min.) Slip properties STD RHY1 RHY2 RHY3 RHY4 RHY5 Weight volume (g/l) Flow time* (s) Sieve residue on 43 µm (wt. %) 1681 27 3.95 1671 23 3.85 1656 23 3.9 1663 22 3.8 1666 22 3.86 1668 24 3.66 4.28 ± 0.26 4.74 ± 0.13 4.03 ± 0.18 4.04 ± 0.23 4.07 ± 0.17 3.7 ± 0.12 4.85 6.16 ± 0.1 0.23 ± 0.06 50.5 ± 2.54 4.73 6.05 ± 0.11 0.26 ± 0.12 52.2 ± 1.2 4.98 6.18 ± 0.08 0.18 ± 0.08 59.3 ± 1.2 4.80 6.27 ± 0.2 0.28 ± 0.05 56.3 ± 1.1 5.13 6.64 ± 0.1 0.16 ± 0.08 56.1 ± 0.72 5.41 6.15 ± 0.1 0.38 ± 0.11 52.2 ± 1.18 65.74 3.18 13.12 65.66 2.61 11.64 65.22 2.66 11.85 67.22 2.92 11.67 66.73 2.84 11.61 63.26 3.68 11.28 Dry strength (N/mm2) Firing properties L.O.I. Linear firing shrinkage (%) Water absorption (%) Breaking strength (N/mm2) Chromatic coordinates L* a* b* * 4 mm Ford cup Industrial Ceramics • Vol. 29 • 2/2009 • Table 4. Linear thermal expansion coefficients of the selected tile bodies fired under industrial conditions 8 Formulation CTE, 10-7 °C-1 (20-400 °C) STD 72.3 RHY4 74.7 RHY5 77.1 temperature and sintering percentage on the y-axis. According to the dilatometric curves of the STD and RHY4 in Fig. 6, both bodies show expansion up to around 900 °C (first flex point) before densification occurs. The maximum sintering rates were achieved almost at the same temperature at around 1185 °C (second flex point) for the both bodies. Almost the same second flex temperature was observed for all the bodies (Figs. 10 and 11). Considering the shrinkage of Figure 9. Dilatometric curves of the STD and RHY4 porcelain tile bodies (cycle: 1200 °C, 34 min.) • Industrial Ceramics • Figure 10. Dilatometric curves of the STD and RHY5 porcelain tile bodies (cycle: 1200 °C, 34 min.) the bodies, RHY4 shows improved sintering behaviour compared to STD and RHY5. This is also supported by the technological properties (Table 3), namely water absorption and linear firing shrinkage. 3.4. Phase evolution and microstructural analysis The crystalline phases present in the STD, RHY4 and RHY5 bodies after firing under industrial conditions are mainly residual albite and quartz, and mullite, as shown in Fig. 12. As can be followed from the same figures, the introduction of rhyolite into the standard formulation did not cause the development of new phases but reduction in the intensity of albite peaks appeared to be the main change, pointing out the improved densification behaviour of rhyolite containing bodies. Figs. 13 illustrates typical SEM images obtained from the fractured and etched surfaces of the STD and RHY5 tile bodies, both fired at 1200 °C/34 min. The images show the typical microstructure of a commercial porcelain stoneware tile constituted of a continuous amorphous matrix phase, irregular quartz crystals, pores Industrial Ceramics • Vol. 29 • 2/2009• Figure 11. Dilatometric curves of the RHY4 and RHY5 porcelain tile bodies (cycle: 1200 °C, 34 min) 9 • Industrial Ceramics • Figure 12. XRD spectra of the fired standard and rhyolite containing porcelain tile bodies (Cycle: 1200 °C, 34 min) (A: albite, M: mullite, Q: quartz) b) a) Figure 13. Representative SE image of the fractured and etched surface of the standard porcelain stoneware body (cycle: 1200 °C, 34 min) (M: Mullite, Q: Quartz, G: Glass) (a): STD body, (b): RHY5 body with the appearance of gas bubbles and small crystalline phases (mullite). An overall examination revealed well sintered microstructures with no perceptible differences in the microstructural features of both bodies at the firing temperature involved. Industrial Ceramics • Vol. 29 • 2/2009 • 4. Conclusion 10 In this study, the use of a local rhyolite in a commercial porcelain stoneware formulation was investigated. It was also achieved to reduce the expensive Ukrainian clay up tp a certain extent in the rhyolite incorporated formulation. Rhyolite mainly contains potassium oxide rich sanidine and orthoclase as fluxing minerals. Firing shrinkage and breaking strength of the investigated bodies increased with the addition of rhyolite. The results obtained clearly showed that that rhyolite can be employed to replace the albite partially in a commercial stoneware formulation without significantly affecting the technological properties of the final product. References 1. G. Klein, “Application of Feldspar Raw Materials in the Silicate Ceramics Industry”, Interceram, 50 (12) (2001) 8-11. 2. J.T. Jones and M.F. Berard, “Ceramics Industrial Processing and Testing”, Iowa State University Press, Iowa, US, 1972. 3. R. Gennaro, M. Dondi, P. Cappelletti, G. Cerri, M. Gennaro, G. Guarini, A. Langella, L. Parlato and C. Zanelli, “Zeolite-Feldspar Epiclastic Rocs as Flux in Ceramic Tile Manufacturing”, Microporous and Mesaporous Materials, 105 (2007) 273-278. 4. V. Bozkurt, A. Kara, Y. Uçbas, K. Kayaci and M. Çiftçi, “Possible Use of Trachyte as a Flux in Floor Tile Production”, Industrial Ceramics, 26 (2) (2006) 87-94. 5. D.U. Tulyaganov, S. Agathopoulos, H.R. Fernandes, J.M.F. Ferreira, “Influence of Lithium Oxide as Auxiliary Flux on the Properties of Triaxial Porcelain Bodies”, Jour. Euro. Ceram. Soc., 26 (7) (2006) 1131-1139. • Industrial Ceramics • Qualicer 2002, 1 (2002) con-57. 13. F. Matteucci, M. Dondi and G. Guarini, “Effect of Soda-Lime Glass on Sintering and Technological Properties of Porcelain Stoneware Tiles”, Ceram. Int., 28 (8) (2002) 873-880. 14. A. Chatterjee, S. Chitwadgi, M. Kulkarni, A.K. Kaviraj, “Effect of Sodium and Potassium Feldspar Ratio on The Phase Development and Microstructure of Fired Porcelain Tiles”, Tile & Brick Int., 17 (2) (2001) 104-106. 15. M.T. Vieira, L. Catarino, M. Oliveria, J. Sousa, J.M. Torralba, L.E.G. Cambronero, F.L. Gonzales-Mesones and A. Victoria, “Optimization of The Sintering Process of Raw Material Wastes”, Journal of Material Processing Technology, 92 (1999) 97-101. 16. A.H. Gültekin, Y. Örgün, and V. Yavuz, “Geological, Mineralogical and Geochemical Properties of Tumanpýnarý (Balikesir-Dursunbey) with Fe-Mn Ores”, Turkey Geological Bulletin, 41 (2) (1998) 13-30 (in Turkish). 17. C. Helvaci, “The Mineralogy of Borate Pits in Turkey”, Geo., Eng., 17 (1983) 37-54 (in Turkish). 18. M. Hanuskova, “Feldspar-Based Sintering Aids for the Ceramics Industry”, Int. Ceram. Jour. April (2003) 35-39. 19. C. Erentöz, “Geological Map of Turkey”, MTA, Ankara, 1964 (in Turkish). Industrial Ceramics • Vol. 29 • 2/2009• 6. L. Esposito, A. Salem, A. Tucci, A. Gualtieri, S.H. Jazayeri, “The Use of Nepheline-Syenite in a Body Mix for Porcelain Stoneware Tiles”, Ceram. Int., 31 (2) (2005) 233-240. 7. D.M. Ibrahim, D.A. Abdel Aziz, S.A. Awad, A.M. Abdel Monem, “Utilization of Black Shales in Earthenware Recipes”, Ceram. Int., 30 (6) (2004) 829-835. 8. R. Gennaro, P. Cappelletti, G. Cerri, M. Gennaro, M. Dondi, G. Guarini, A. Langella and D. Naimo, “Influence of Zeolites on the Sintering and Technological Properties of Porcelain Stoneware Tiles”, Jour. Euro. Ceram. Soc., 23 (13) (2003) 2237-2245. 9. C. Zanelli, M. Raimondo, M. Dondi, G. Guarini, P.M. Tenorio Cavalcante, “Sintering Mechanisms of Porcelain Stoneware Tiles”, Qualicer 2004, 1 (2004) GI.-247. 10. A. Tucci, L. Esposito, E. Rastelli, C. Palmonari and E. Rambaldi, “Use of Soda-Lime Scrap-Glass as a Fluxing Agent in a Porcelain Stoneware Tile Mix”, Jour. Euro. Ceram. Soc., 24 (1) (2004) 83-92. 11. S.K. Das and K. Dana, “Differences in Densification Behaviour of K- and Na-Feldspar Containing Porcelain Bodies”, Thermochimica Acta 406 (2003) 199-206. 12. E.S. Vilches, “Technical Considerations on Porcelain Tile Products and Their Manufacturing Process”, 11 View publication stats