Subscribe Now Subscribe Today
Research Article

Effect of Different Concentrations of Titanium Oxide (TiO2) on the Crystallization Behavior of Li2O-Al2O3-SiO2 Glasses Prepared from Local Raw Materials

Omar A. Al-Harbi, Esmat M.A. Hamzawy and M. Mujtaba Khan

Lithium Aluminum Silicate (LAS) glasses produced from local raw materials (white silica sand and clay) and Li2CO3 were nucleated by different concentrations of TiO2 to transform into glass-ceramics. Studies were performed using Differential Thermal Analysis (DTA), X-Ray Diffraction (XRD), Scanning Electron Microscopy (SEM) and Dilatometer. Addition of TiO2 affected phase evolution, morphology and formation of β-spodumene along with brookite and Al2TiO5 phases. The SEM micrographs of the heat-treated glass samples also reflected bulk crystallization. However, softness of crystal edges took place on samples subjected to muliti-stage heat-treatment. This may attributed to partial melting of the pre-developed phases. The CTE values were low to medium with the formation of β-spodumene and brookite, but were high with the formation of Al2TiO5 and amorphous phasesin addition to both the β-spodumene and brookite phases. The values of low, medium and high Coefficient of Thermal Expansion (CTE) ranged between 3.02-49.672x10-7 °C-1 in the double and -2.171-31.737x10-7 °C-1 in the multi-stage heat treatments with in a temperature range of 25-300°C.

Related Articles in ASCI
Similar Articles in this Journal
Search in Google Scholar
View Citation
Report Citation

  How to cite this article:

Omar A. Al-Harbi, Esmat M.A. Hamzawy and M. Mujtaba Khan, 2009. Effect of Different Concentrations of Titanium Oxide (TiO2) on the Crystallization Behavior of Li2O-Al2O3-SiO2 Glasses Prepared from Local Raw Materials. Journal of Applied Sciences, 9: 2981-2986.

DOI: 10.3923/jas.2009.2981.2986



Although, lithium-aluminum-silicates based glass-ceramics have been studied for decades but still it remains a challenge for scientists to improve or modify its properties. The Li2O-Al2O3-SiO2 glass-ceramics of low thermal expansion coefficient are considered as primary requirements of many industrial sectors. The characteristics of the glass-ceramics depend on the nature of crystalline phases and the microstructures. Earlier investigators have successfully utilized primarily raw materials to prepare different type of glass-ceramics by using locally available raw materials and produced lithium aluminum silicates based glass-ceramics of low coefficient of thermal expansion (Omar et al., 1986; Khater, 1987; El-Shennawi et al., 1998; Hamzawy, 1992; Kauffman and Dyk, 1993; Idris and Khater, 2004; Al-Harbi and Khan, 2008). Stookey (1960) reported that TiO2 is the most common nucleating agent in the glass-ceramics. It is also recognized as one of the best nucleating agents in the fabrication of Li2O-Al2O3-SiO2 (LAS) glass-ceramics by conventional melting and crystallization process (Bash, 1995).Barry et al.(1970) reported that TiO2 acts as a surfacial active agent and increases the nucleation rate in the Li2O-Al2O3-SiO2 system.

Titania nucleated glass-ceramics in the Li2O-Al2O3-SiO2 type having a thermal expansion coefficients less than 15×10-7 °C and in some cases approaching zero, were produced (McMillan, 1979). The amount of TiO2 in various types of glass-ceramics ranged from 2-20% on weight basis (McMillan, 1979). When used in large amounts, TiO2 becomes a major part of glass composition. Zdaniewski(1978) suggested that TiO2 decreases the viscosity of the base glasses at high temperature which is favorable for nucleation and growth of the main crystalline phase. The TiO2 is soluble in a wide range of molten glasses. But during cooling or subsequent reheating, large numbers of submicroscopic particles are precipitated and these apparently assist the development of major crystal phases from the glass (McMillan, 1979). The objective of the present study is to determine the possible role of TiO2 in the Li2O-Al2O3-SiO2 glass on characteristics such as thermal behavior, crystallization, microstructures and the coefficient of thermal expansion.


The study was carried out at Geology Research Laboratory, National Center for Water Research (NCWR), King Abdulaziz City for Science and Technology (KACST), Riyadh during 2008.

Table 1: Chemical composition of the starting raw materials

Table 2: Chemical composition of the glass batches in oxide form (weight basis %)
Added/100 g glass oxide

Preparation of base glass samples: Locally available white silica-sand and clay with Li2CO3 (Commercial Grade Reagents) and TiO2 were used as the starting materials. The TiO2 was added at the rate of 0, 3, 6, 9, 12 and 15 g per 100 g of glass oxide. A base glass batch with sequential additions of TiO2 concentrations was prepared for the experiment. The chemical analysis of the raw materials used in the experiment is shown in Table 1. The chemical composition of prepared glasses is shown in Table 2. The prepared base glass batch was mixed well by the Ball Mill, then transferred into sintered aluminum crucibles and placed in the Molybdenum silicide furnace at temperatures ranging from 1350-1450°C for 2 h. The glass melt was stirred at an interval of half an hour for homogenization during the whole process. A bubble free glass melt was poured into preheated steel casting moulds as rods and patties. Later on, the glass melt was transferred to a preheated muffle furnace at 600°C for annealing.

Sample characterization: About 30 mg of the powdered sample (between 25-50 μg grain sizes) was used for DTA analysis by A Shimadzu DTG-60H micro Differential Thermo Analyzer. The Al2O3 powder was used as a reference material. A heating rate of 10°C min-1 was maintained for all the DTA tests. A double and a multi-stage heat-treatment schedule were followed the DTA temperature peaks. The glass samples were heat-treated between 780-1050°C for 2 h in the double and multi-stage schedule. The powdered samples were tested to identify the crystallized phases using Mini Flex of Rigaku-Japan, adopting Ni-filtered Cu-radiation. The JCPDS diffraction file Cards (2001) were used as reference data for the interpretation of X-ray patterns obtained in this study.

The microstructures and the crystal growth of some heat-treated gold coated glasses were examined by SEM using Joel-JSM-5800, Japan. The dimensional changes occurring with the change of temperature, were measured by A NETZECH DIL 402 PC-Germany Instrument. The CTE of prepared samples was measured with a dimension of 0.5×0.5×2.0 cm3 at a heating rate of 5°C min-1.


The DTA data showed that the crystallization temperature of LA-3 base glass (without nucleating agents) were at 918 and 963°C and the intensity of peak was also high (Table 1). The effect of batch composition on the crystallization of glass showed that the presence of small amount of TiO2 in batch composition changed the rate and phase of crystallization as well as the viscosity of the resultant glass melt. Similar to TiO2, Al2O3 also reduced the melting point and improved crystallization. The Al2O3 ion can be four or six-co-ordinate with oxygen giving rise to tetrahedral AlO4 or octahedral AlO6 groups. The tetrahedral group can replace SiO4 tetrahedral in silicate lattices to give the different arrangements. Most of the earlier workers (Das and Douglas, 1967; Omar et al., 1971; Shennawi et al., 1991) agreed that Al2O3 as well as the SiO2 ion were effective in increasing the viscosity of the glass melts by forming polymer units between AlO4 and SiO2 tetrahedral groups.

Weak transition temperatures (TG) were observed in the fabricated glasses ranging from 709 to 740°C (Fig. 1). The endothermic peaks characters were almost identical in the temperature of 709-740°C. The low intensity of Tg peak was due to slow rate of glass melt cooling (Mazurin and Gankin, 2007). Other than LA3-T1 and LA3-T2 samples, the exothermic peaks were sharp in the temperature range of 914-916°C. However, in comparison to base glass the intensity of peaks was proportional to the increasing concentration of TiO2 compared to LA3, LA3T-3 and LA3-T5 samples (Fig. 1). In comparison to the parent glass, the exothermic peak temperatures either decreased or increased according to the TiO2 ratios. In LA3-T1 (containing ~ 3% TiO2) and LA3-T2 (containing ~ 6% TiO2) glasses, the decrease in exothermic temperature took place by about 50 and 100°C, respectively (Fig. 1). This is in agreement with El-Shennawi et al. (1998), who considered that addition of TiO2 acted as a nucleating agent. But when the concentration of TiO2 increased to 9% (as in LA3-T3) and 12% (LA3-T4) and 15% (LA3-T5) in glass samples, the exothermic peaks were almost identical to the base glass (LA3). McMillan (1979)used TiO2 in various types of glass-ceramics ranging from 2-20% (on weight basis) and observed that when TiO2 is used in high concentrations, it becomes a major part of the glass composition.

Fig. 1: DTA traces of the glasses

Barry et al. (1970) reported that TiO2 acts as a surfacial active agent and increases the nucleation rate in the Li2O-Al2O3-SiO2 system. Morsi and El-Shennawi(1983) indicated that little amount of TiO2 was more effective to enter into network forming positions than into network modifying positions. Sandstorm et al.(1980) reported that >4 g TiO2 probably acts in reverse. However, in the present study, the DTA traces showed that LA3 base glass and other samples having TiO2 concentration of 6% and above have almost identical exothermic peaks. These exothermic peaks were associated with the crystallized lithium-aluminum-silicate phases.

β-Spodumene, brookite and aluminum titanium oxide (Al2TiO5) phases were developed through out the different heat-treatment schedule (Fig. 2, 3). In both the double and multiple-stage heat-treatments of Li2O-Al2O3-SiO2-TiO2 system, there was no significant difference in the crystallization of β-spodumene and brookite phases (Table 3). However, in the multi-stage heat-treatment, the crystallization of aluminum titanium oxide (Al2TiO5) as well as an increase in the intensity of XRD peaks took place (Fig. 2, 3). In the Li2O-Al2O3-SiO2 system, the literature proved that the crystallization of titnate is an independent phase (Bae et al., 2005) here, while others assured the formation of Al2TiO5 as a primary or micro-separated phase with TiO2 acting as a nucleating agent (4 mole %) (Beall and Duke, 1983).

Fig. 2: XRD patterns for the glasses heat-treated in double stage double-stage heat treatment, S: β-spodumene, B: Brookite

Table 3: Development of crystalline phases in double and multi-stage heat-treatments
AlTit: Al2TiO5

The heat-treatment of the present glasses and the base glass containing parent TiO2 showed that low ratios of TiO2-containing samples acted mainly as nucleating agent. On the contrary, high ratio of nucleating agent not only acted as a catalyst but also as dependance developed phases along with the main phases. The double heat treatment in the case of LA3-T1 and LA3-T2 samples indicate that the crystallization does not take place below 900°C and resulted in an amorphous phases.

Fig. 3:
XRD patterns for the glasses heat-treated in multi-stage heat treatments, S: β-spodumene, B: Brookite, At: Aluminium titanium oxide Al2TiO5

Almost homogenous bulk crystalline glass-ceramics of major β-spodumene with brookite and may with Al2TiO5 were developed (Fig. 4a-d). The examination of SEM photographs showed the roughness and softness of the crystal edges that characterized the double and multi-stage glass-ceramic samples, respectively (Fig. 4). This may be attributed to the increase of temperature that lead to partial melting and an increase in glass quantity and frequently the softening of grain edges. However, the major β-spodumene may be observed as subhedral tetragonal crystals and might be modified into subhedral six-sided crystals. The presence of TiO2 in both forms (as brookite and Al2TiO5) helped in the bulk crystallization of glass. As also, the mechanism showed the development of these phases as early as glass in glass phase's separation or formation of such heterogeneous TiO2 containing phases and at high temperature epitaxial crystallization took place.

The Coefficient of Thermal Expansion (CTE) in the present glass-ceramic samples were affected by heat-treatment, developed phases, size and orientation of grains and residual glasses (Table 4, Fig. 5, 6). Through the heat-treatment schedule, the developed crystalline phases were β-spodumene with brookite and Al2TiO5 that gave CTE values between 3.02-49.672x10-7 °C-1 and 22.709-58.356x10-7 °C-1 in the double stage and between -2.171-31.737 and 16.986-33.150x10-7 °C-1 in multi-stages determined over the temperature ranges from 20-300°C and 25-500°C (Table 4).

Fig. 4:
SEM micrographs of LA3-base-LA3-T5 glasses heat-treated in double and multi-stage heat-treatment, (a) LA3-base 780°C/2 h-916°C/3 hL, (b) LA3-base 780°C/2 h-916°C/3 h-970°C/3 h, (c) LA3-T5 780°C /2 h-916°C/3 h and (d) LA3-T5 780°C/2 h-916°C/3 h-970/°C 3 h

Fig. 5: CTE values of glasses in double stage heat treatment

Fig. 6: CTE values of glasses in multi-stage heat treatment

Table 4: Mean CTE values of glasses heat-treated in double and multi-stage treatment
Β-Sp: β-spodumene, Br: brookite, AT: Al2TiO5

The CTE of β-spodumene (Li2O-Al2O3-4SiO2) was about 0.9x10-6 K-1 at a temperature range between 20-1000°C (McMillan, 1979), whereas the data on brookite or Al2TiO5 is limited. Shimazu et al. (2008) studied Al2TiO5 based glass ceramics with very low CTE values because of micro-cracks developed at grain boundaries induced by the high anisotropy of the thermal expansion along -3.0, +11.8 and +21.8 [mm/(mimic) ×106] for its three crystallographic axes. In the present study, the glass-ceramics were compact samples and CTE values were different from those obtained in the earlier investigated materials.

The LA3 base glass subjected to crystallization at double and multi-stage recorded the lowest CTE values between 8.699-22.988 and -2.171-16.986×10-7 °C-1 at a temperature range of 25-500°C, respectively. This suggests that the crystallization of β-spodumene along with brookite gave the lowest CTE values. Furthermore, the incorporation of TiO2 and the formation of amorphous phases after double heat treatment increased the CTE values up to 49.672 (25-300°C) and 58.356×10-7 °C-1 (25-500°C), respectively. This may be ascribed to the increase of glassy phase as well as in the double stage heat-treatment that are crystalline free samples, which were characterized by the formation of Al2TiO5 phase in addition to β-spodumene and brookite. The CTE values of high TiO2 containing samples (LA3-T3 and LA-3-T5) along with the formation of β-spodumene and brookite were medium ranging from 3.020-20.279×10-7 °C-1 (25-300°C) and 22.709-58.356 (25-500°C) in the double stage, whereas it ranged from 14.392-20.612 (25-300°C) and 25.205-28.362×10-7 °C-1 (25-500°C) in multi-stage heat-treatment.

In general, according to the developed phases, the CTE values were low to medium in TiO2 free addition in presence of β-spodumene and brookite. Whereas, the CTE values were high with the formation of Al2TiO5 in addition to β-spodumene and brookite.


The characteristics of glass-ceramics in the Li2O-Al2O3-SiO2(TiO2) system depended on the sequential addition of TiO2 and the heat-treatment parameter crystalline. The β-spodumene, brookite and Al2TiO5 were the developed phases during different heat treatment schedule. The low TiO2 ratios in the samples acted mainly as a nucleating agent. But at high TiO2 ratios, it acted as a nucleating agent and as a dependent phase that can be considered along with the developed main phases. Bulk crystalline materials were developed during glass fabrication process. The softness of the crystal edges in the multistage heat-treatment was due to the increase of glassy phase due to partial melting of the pre-developed phases. The CTE values were low to medium with the formation of β-spodumene and brookite, but were high with the formation of Al2TiO5 and amorphous phasesin addition to both the β-spodumene and brookite phases.


This study was supported by King Abdulaziz City for Science and Technology (KACST) under Grant No. 24-01 which is gratefully acknowledged.

Al-Harbi, O.A. and M.M. Khan, 2008. Investigation and development of local raw materials for glass and glass-ceramics. Final Report, KACST, Riyadh, Project No. BM-41-1-123.

Bae, S.J., U.K. Kang, O. Dymshits, A. Shashkin, M. Tsenter and A. Zhilin, 2005. Raman spectroscopy study of phase transformations in Titania-containing lithium aluminosilicate glasses doped with CaO. J. Non-Crystallizat. Solid, 351: 2969-2978.
CrossRef  |  Direct Link  |  

Barry, T.J., D. Clinton, L. A.J. Lay, R.A. Mercer and R.P. Miller, 1970. The crystallization of glasses based on eutectic compositions in the systems Li2O-Al2O3-SiO2: Part 2: Lithium metasilicate-β-eucryptite. Mater. Sci., 5: 117-126.

Bash, H., 1995. Low Thermal Expansion Glass Ceramics. Springer, Germany, pp: 223.

Beall, G.H. and D.A. Duke, 1983. Glass-Ceramic Technology. In: Glass Science and Technology, Uhlmann, D.R. and N.J. Krieidl (Eds.). Academic Press, Orlando.

Das, C.R. and R.M. Douglas, 1967. Reactions between water and glass, part III. Phys. Chem. Glass, 85: 178-178.

El-Shennawi, A.W.A., M.M. Morsi, G.A. Khater and S.A.M. Abdel-Hameed, 1998. Thermodynamic investigation of crystallization behavior of pyroxene basalts based glasses. J. Thermal Anal., 15: 553-560.
Direct Link  |  

Hamzawy, E.M.A., 1992. Catalyzed Crystallization of some glasses based on nepheline-syenite. M.Sc. Thesis. Cairo University, pp: 130.

Idris, M.H. and G.A. Khater, 2004. Effect of glass composition and heat treatment Parameters on crystallization behavior in the Li2O-BaO-Al2O3-SiO2 system. Phys. Chem. Glass, 45: 141-145.
Direct Link  |  

Kauffman, R. and D.V. Dyk, 1993. Feldspar, Geology of Industrial Rocks and Minerals. Dover Publications, New York, pp: 473-481.

Khater, G.A., 1987. The use of some Egyptian clay for the production of glass ceramic materials. M.Sc. Thesis. Ain Shams University.

Mazurin, O.V. and V. Yu-Gankin, 2007. Glass transition temperature: Problem of measurements and analysis of the existing data. Proceedings of the International Congress on Glass, Strabburg France.

McMillan, P.W., 1979. Glass Ceramics. 2nd Edn., Academic Press, London.

Morsi, M. and A.W. A. EL-Shennawi, 1984. Some physical properties of silicate glasses containing TiO2 in relation to their structure. Phys. Chem. Glass, 25: 64-68.
Direct Link  |  

Omar, A.A. and A.W.A. El-Shennawi, 1971. Crystallization of some mole ton Egyptian basaltic rocks and corresponding glasses. United Arab Republic J. Geol., 15: 65-73.

Omar, A.A., S.M. Salman and M.Y. Mahmood, 1986. Phase relation in the diopside anorthite-akermanite system. Ceramic Int., 12: 53-59.
CrossRef  |  

Sandstorm, D.R., F.W. Lytle, P.S.P., W.B. Gregor, J. Wong and P. Schultz, 1980. Influence of the amount of Titania on the texture and structure of Titania supported on silica. J. Non-Crystallizat. Solids, 41: 201-201.

Shennawi, A.W.A., A.A. Omar and G.A. Khater, 1991. Crystallization of celsian polymorphous in some alkaline earth aluminosilicate glasses. Glass Technol., 32: 124-131.

Shimazu, T., M. N. Isu, T. Miura, K.O. Ogawa, H. Maeda and E.H. Ishida, 2008. Plastic deformation of ductile ceramics in the Al2TiO5-MgTi2O5 system. Mater. Sci. Eng. A, 487: 340-346.
CrossRef  |  Direct Link  |  

Stookey, S.D., 1959. Catalyzed crystallization of glass in theory and practice. Ind. Eng. Chem., 51: 805-808.
CrossRef  |  Direct Link  |  

Zdaniewski, W.A., 1978. Microstructure and kinetics of crystallization of MgO-Al2O3-SiO2 glass-ceramics. J. Am. Ceramic Soc., 61: 199-204.
CrossRef  |  Direct Link  |  

©  2019 Science Alert. All Rights Reserved