Formation of Aluminum Titanate with Small Additions of MgO and SiO₂

Abstract

The formation of aluminum titanate was investigated by isothermal treatments of samples obtained from equimolar mixtures of alumina and titania, containing small amounts of silica and magnesia. Results of differential thermal analysis and Rietveld refinements of data collected by X-ray powder diffraction (XRPD) showed that additions of silica in amounts used in this work did not influence the formation of aluminum titanate. However, the presence of magnesia favored the formation of aluminum titanate in two steps, first one by incorporating Mg²⁺ into Al₂TiO₅ lattice during its initial formation, and the second one by accelerating the Al₂TiO₅ formation, contributing to large quantities of this phase. MgO doped samples have also developed a more suitable microstructure for stabilizing of Al₂TiO₅, what make them promising for applications such as thermal barriers, internal combustion engines and support material for catalyst.

Keywords:
aluminum titanate; Rietveld; magnesia; silica

1 Introduction

Aluminum titanate ceramics (Al₂TiO₅) are promising materials for insulating applications in automotive industry, such as engine components, portliners, manifolds and linings of turbochargers^1,2. Also, this ceramic is suitable as catalyst carriers for purification of fume produced by cars, containers and tubes for storing or conveying melted steel as well as protective tube for thermocouples. This is because of their low thermal expansion coefficient (0.2 - 1x10^-6 K^-1), low Young’s modulus, high melting point (~1860^oC), low thermal conductivity (0.9 – 1.5 Wm^-1K^-1) and outstanding thermal shock resistance which can reach up to 500 Wm^-11-4.

However, a pronounced anisotropy in thermal expansion coefficient, typical of aluminum titanate, induces severe microcracking during the cooling process that takes to damage mechanical properties of the final sintered material ^1,5. The anisotropic behavior is due to the crystal structure of the Al₂TiO₅ which is isomorphous with pseudobrookite (Fe₂TiO₅), crystallizing in the orthorhombic space group Cmcm, with a theoretical density of 3.70 g/cm³.

Another problem associated with the formation of Al₂TiO₅ which has limited its application field is the eutetoid-like decomposition to α-Al₂O₃ and TiO₂ (rutile) within temperature range from 900 to 1280^oC⁶. This reaction accompanied by an 11% molar increase contributes to a microcracking process and consequently to deterioration of the ceramic component^7,8.

The referred limitations are usually solved using composites with alumina or mullite, or else additives to stabilize the structure and avoid decomposition reactions^3,7,9,10. Studies reported that additives can increase the formation of aluminum titanate which leads to thermodynamic stability, as well as an improving in sintering¹¹. Although there are some researches about the formation of aluminum titanate doped with different substances^2,12-16, most are related to the presence of large amounts of additives with conflicting results. This encouraged us to investigate the effect of small contents of MgO and SiO₂ on formation of Al₂TiO₅.

Moreover, additives were selected considering the small difference between the size of its cation and aluminum ionic radius (Al³⁺=0,054nm, Si⁴⁺=0,041nm and Mg²⁺=0,071nm) in order to avoid large distortion in aluminum titanate crystal and damaging the thermal expansion of the material¹⁷.

2 Experimental procedure

The raw materials used in this study were α- Al₂O₃ (A1000 SG, Alcoa), TiO₂ (rutile, Merck), MgO (Sigma-Aldrich, 99.99% trace metals basis) and SiO₂ (fumed silica, Aerosil –Degussa). Titania and alumina powders were separately ground in a ball mill for 16 hours. Afterwards, equimolar mixtures of alumina and titania, containing 0-1.0 wt-% silica or magnesia (as listed in Table 1) were ground in an attritor mill at 250 rpm for 2 hours using isopropanol as liquid medium.

Table 1
Additives in the equimolar mixture of alumina and titania.

sample	Silica (wt. %)	Magnesia (wt. %)
ATP	-	-
AT25S	0.25	-
AT50S	0.5	-
AT100S	1	-
AT25M	-	0.25
AT50M	-	0.5
AT100M	-	1

After grinding process, the compositions were dried in a rot evaporator. Samples were submitted to differential thermal analysis (DTA) using air atmosphere to 1400ºC, at a heating rate of 10ºC/min, (Netzsch, DSC 404 F3 Pegasus) so that possible reactions involved from 1250 to 1400^oC could be explored.

In order to investigate the reactions taking place in the compositions shown in Table 1, uniaxially (50 MPa) and cold isostatically pressed (200 MPa) pellets were subjected to isothermal treatments at air, being quickly introduced into a tubular furnace which was previously stabilized at 1250, 1300 and 1400^oC, respectively. The pellets were held in respective temperatures for one hour and quenched in water so that a rapid cooling would be ensured to preserve, at room temperature, the aluminum titanate phase which tends to decompose in α-Al₂O₃ and TiO₂ during slow cooling processes. After that, samples were dried at 110° C for 24 hours.

Pulverized heat-treated pellets were analyzed by X-ray powder diffraction (D5000 Siemens/Brucker diffractometer equipped with copper tube and scintillation detector) in a range from 15 to 100^o (2θ) with step of 0.02^o (2θ) and 5s/pass. Phases quantitative analysis and computations of lattice parameters were carried out by the Rietveld method using Topas Academic¹⁸ as computer program.

Scanning electron microscopy (Philips XL30 scanning electron microscope) was performed in fractured samples to evaluate the microstructure after the heat treatments.

3 Results and Discussion

Figure 1 shows DTA curves of the studied compositions. There is an endothermic peak between 1300 and 1360°C, which can be associated with the solid state reaction of formation of aluminum titanate. This reaction occurred at lower temperatures when MgO was present, showing that MgO anticipates the formation of Al₂TiO₅ compared to other evaluated compositions. This behavior should be related to a prior reaction with Mg²⁺ ions represented by another endothermic event found at around 1250°C in AT50M and AT100M compositions. Instead, additions of SiO₂showed no influence on this temperature, since the reactions in compositions with SiO₂ (AT50S and AT100S) overlap with that of the composition without additive (ATP).

Figure 1
Differential thermal analysis (DTA) curves of samples from 1200 to 1400 °C.

X-ray powder diffraction analysis (Figure 2) of isothermally heat-treated samples (1250, 1300 and 1400 for 1 hour) show the evolution of the formation of aluminum titanate. Quantitative analysis with Rietveld refinements confirms the efficiency of MgO additions to Al₂TiO₅ formation. Plots of Al₂TiO₅, Al₂O₃ and TiO₂ contents in function to amount of additive (MgO or SiO₂) and temperature of isothermal heat treatment (Figure 3), indicates that Al₂TiO₅ phase was formed in all investigated samples, except in pure sample (ATP) and in that with 0.25 wt-% SiO₂ (AT25S), both treated at 1250^oC. Although this temperature did not favor the formation of Al₂TiO₅, MgO already proved to be an effective additive for obtaining the studied ceramic.

Figure 2
X-ray powder diffraction (XRPD) spectra taken from samples obtained by heat treatments at 1250, 1300 and 1400 for 1 hour. (a) ATP, (b) AT100M and (c) AT100S. T is aluminum titanate, C is α-alumina and R is rutile.

Figure 3
Al₂TiO₅ (a and b), Al₂O₃ (c) and TiO₂ (d) contents calculated by the Rietveld method as a function of the additive content and heat treatment temperature. ATS is the abbreviation for SiO₂ doped samples; ATM is the abbreviation for MgO doped samples; 1250, 1300 and 1400 are the heat treatment temperatures (^oC).

The greater efficiency of MgO with respect to SiO₂ is quite pronounced when the heat treatment was performed at 1300°C (see Figures 3a and 3b). Just as an example, the sample containing 1 wt-% MgO reached 81.72 wt-% Al₂TiO₅, while other doped with 1 wt-% SiO₂promoted formation of 22.26 wt-% Al₂TiO₅. It means a difference of three times in amount of formed Al₂TiO₅.

At 1400^oC, despite the addition of MgO has promoted formation of Al₂TiO₅ in higher amounts than the addition of SiO₂, the influence of additive concentration is very slight. In other words, the effect of temperature exceeded that of additives when the heat treatment proceeded at 1400^oC.

The effect of MgO additions on formation of Al₂TiO₅ most likely involves the endothermic reaction detected in the DTA curves (Figure 1) for MgO doped samples (AT50M and AT100M). This reaction can be associated with the altered values found for the Al₂TiO₅ lattice parameters of MgO doped samples treated at 1250^oC (Table 2). Ishitsuka et al.¹⁹ studying Al₂TiO₅ synthesized with Mg(OH)₂, observed changes in lattice parameters of Al₂TiO₅ phase assigned to formation of solid solutions of Al_2(1-x)Mg_xTi_1+xO₅, at low temperatures. Some works related the presence of a transitional phase during the formation of Al₂TiO₅, however conflicting opinions yet are discussing in view of the presence of this phase, as well as its composition and effect on the formation of Al₂TiO₅^12,14,20.

Table 2
Al₂TiO₅ lattice parameters obtained by Rietveld refinements.

Identification	a (Å)	b (Å)	c (Å)	Cell vol. (Å3)
AT25M-1250oC	3.59462(55)	9.5020(23)	9.7300(24)	332.340(13)
AT25M-1300oC	3.58728(18)	9.45799(56)	9.68367(61)	328.552(33)
AT25M-1400oC	3.588785(85)	9.44434(21)	9.65933(21)	327.390(13)
AT50M-1250oC	3.60200(37)	9.5054(15)	9.7299(16)	333.138(82)
AT50M-1300oC	3.58971(13)	9.45307(38)	9.66753(42)	328.056(23)
AT50M-1400oC	3.589806(86)	9.44414(22)	9.65758(23)	327.417(13)
AT100M-1250oC	3.59642(28)	9.5156(10)	9.7300(11)	332.982(58)
AT100M-1300oC	3.59229(11)	9.44913(29)	9.66158(31)	327.953(18)
AT100M-1400oC	3.592480(75)	9.44984(21)	9.66067(21)	327.964(12)

Altered values found for the Al₂TiO₅ lattice parameters of MgO doped samples treated at 1250^oC (Table 2) suggest that aluminum titanate is formed in two steps, just characterized by the endothermic reactions. In the first step, Mg²⁺should be incorporated into Al₂TiO₅ lattice during its formation, producing a solid solution that behaved as a nucleating agent to accelerate the second stage of Al₂TiO₅ formation. It is means that the efficiency of the additive is related to the Mg²⁺ incorporation into the structure of Al₂TiO₅.

Besides, Al₂O₃ and rutile phases were detected in the samples (Figures 2 and 3), indicating that the solid state reaction for formation of Al₂TiO₅ could not be concluded in any conditions used here. Although high amounts of Al₂TiO₅ have been formed during heat treatments at 1400^oC, Al₂O₃ and TiO₂ were also found as residual phases.

The quantitative phase analysis clearly demonstrates that larger amounts of alumina remained as residual phase in all samples (Figure 3c), indicating that rutile is preferably consumed to form Al₂TiO₅. These results reveal that an equimolar mixture of Al₂O₃ and TiO₂ does not contribute to complete the formation of Al₂TiO₅ by means of solid state reaction.

Figure 4 shows micrographs of samples with fractured surfaces. By this figure, it is observed that both additions result in inhibition of grain growth of Al₂TiO₅, and MgO additions are more efficient in this action compared to SiO₂ additions. This suggests that MgO is also more efficient in stabilization of Al₂TiO₅phase and consequently for reducing of microcracks. This because the phenomena are related to grain size, i.e., Al₂TiO₅ becomes stable at room temperature, below a critical grain size. Also, below a certain critical grain size (which is not necessarily the same), the stresses caused during cooling become insufficient for formation of microcracks^21-23. This fact controls thermal and mechanical properties of the material^22,24,25.

Figure 4
Scanning electron micrographs of the fracture surface of the samples treated at 1400^oC for 1 hour. (a) ATP, (b) AT25S (c), AT50S, (d) AT100S, (e) AT25M, (f) AT50M and (g) AT100M.

4 Conclusions

Additions of MgO provided prior formation of Al₂TiO₅ owing to a further endothermic reaction detected at low temperature. The presence of two endothermic reactions in MgO doped samples as well as results of XRPD and Rietveld refinement led to conclude that the formation of Al₂TiO₅occurred in two stages. The first stage was associated with formation of an aluminum titanate solid solution containing Mg²⁺, and the second reaction was related to actual formation of Al₂TiO₅. Also, small contents of MgO shown to be promising in the stabilization of Al₂TiO₅, whereas they promoted the development of a microstructure consisting of small grains.

In contrast, additions of silica do not promote expressive results related to the formation of Al₂TiO₅.

5. References

• ¹
  Kim JI, Zhao F, Gong J, Lee KS, Han IS, Kuk WS. Thermal durability of aluminum titanate-mullite composites with high thermal shock resistance. Journal of Ceramic Processing Research. 2003;4(2):71-70.
• ²
  Buscaglia V, Nanni P, Battilana G, Aliprandi G, Carry C. Reaction sintering of aluminium titanate: I-Effect of MgO addition. Journal of the European Ceramic Society. 1994;13(5):411-417. DOI: 10.1016/0955-2219(94)90018-3
  » https://doi.org/10.1016/0955-2219(94)90018-3
• ³
  Buscaglia V, Nanni P, Leoni M, Thermodynamics and kinetics of decomposition of AlTiO. ₂₅-base ceramicsKey Engineering Materials. 1997;132-136:810-813.
• ⁴
  Nagano M, Nagashima S, Maeda H, Kato A. Sintering behavior of AlTiO. ₂₅ base ceramics and their thermal propertiesCeramics International. 1999;25(8):681-687.
• ⁵
  Tsetsekov A. A comparison study of tialite ceramics doped with various oxide materials and tialite-mullite composites: microstructural, thermal and mechanical properties. Journal of European Ceramic Society. 2005;25(4):335-348. doi:10.1016/j.jeurceramsoc.2004.03.024
  » https://doi.org/10.1016/j.jeurceramsoc.2004.03.024
• ⁶
  Low IM, Oo Z, O’Connor BH. Effect of atmospheres on the thermal stability of aluminium titanate. Physica B. 2006;385-386(Part 1):502-504. doi:10.1016/j.physb.2006.05.256
  » https://doi.org/10.1016/j.physb.2006.05.256
• ⁷
  Ibrahim DM, Mostafa AA. MgO stabilized tialite prepared by urea formaldehyde polymeric route. British Ceramic Transactions. 1999;98(4):182-186. DOI:10.1179/096797899680417
  » https://doi.org/10.1179/096797899680417
• ⁸
  Thomas HAJ, Stevens R, Gilbart E. Effect of zirconia additions on the reaction sintering of aluminium titanate. Journal of Materials Science. 1991;26:3613-3616.
• ⁹
  Ananthakumar S, Jayasankar M, Warrier KG. Microstructural, mechanical and thermal characterisation of sol–gel-derived aluminium titanate–mullite ceramic composites. Acta Materialia. 2006;54:2965-2973.
• ¹⁰
  Perera FH, Pajares A, Melendez JJ. Strength of aluminium titanate/mullite composites containing thermal stabilizers. Journal the European Ceramic Society. 2011;31(9):1695-1701. doi:10.1016/j.jeurceramsoc.2011.03.012
  » https://doi.org/10.1016/j.jeurceramsoc.2011.03.012
• ¹¹
  Preda M, Crisan IM, Jitiann A, Crisan D, Zaharescu M. Reaction mechanisms of tialite formation in the presence of mineralizers. Journal of Optoelectronics and Advanced Materials. 2000;2(5):563-568. http://joam.inoe.ro/arhiva/pdf2_5/Preda.pdf
  » http://joam.inoe.ro/arhiva/pdf2_5/Preda.pdf
• ¹²
  Buscaglia V, Caracciolo F, Leoni M, Nanni P, Viviani M, Lemaitre J. Synthesis, sintering and expansion of Al_0.8MgTi: a low-thermal-expansion material resistant to thermal decomposition._0.61.6O₅Journal of Materials Science. 1997;32(24):6525-6531.
• ¹³
  Jiang L, Chen X, Hang G, Meng Y. Effect of additives on properties of aluminium titanate ceramics. Tranactions of Nonferrous Metal Society of China. 2011;21(7):1574-1579. doi:10.1016/S1003-6326(11)60899-6
  » https://doi.org/10.1016/S1003-6326(11)60899-6
• ¹⁴
  Korim T. Effect of Mg and Fe. ²⁺³⁺ ions on formation mechanism of aluminum titanateCeramics International. 2009;35:1671-1679.
• ¹⁵
  Yoleva A, Djambazov S, Arsenov D, Hristov V. Effect of SiO addition on thermal hysteresis of aluminum titanate. ₂Journal of the University of Chemical Technology and Metallurgy. 2010; 45(3):269-274. http://dl.uctm.edu/journal/node/j2010-3/6_Albena_Yoleva_269-274.pdf
  » http://dl.uctm.edu/journal/node/j2010-3/6_Albena_Yoleva_269-274.pdf
• ¹⁶
  Thomas H, Stevens R. Alumina Titanate - A Literature review. Part. 2 Engineering properties and thermal stability. British Ceramics Transactions Journal. 1989;88:184-190.
• ¹⁷
  Arenas IB. Reactive sintering of aluminum titanate. In: Lakshmanan A, editor. Sintering of Ceramics – New Emerging Techniques [Internet]. Rijeka, Croatia: InTech; 2012. p.505-523. [cited 2015 Jan 30]. Available from: http://www.intechopen.com/books/sintering-of-ceramics-new-emerging-techniques/reactive-sintering-ofaluminum-titanate
  » http://www.intechopen.com/books/sintering-of-ceramics-new-emerging-techniques/reactive-sintering-ofaluminum-titanate
• ¹⁸
  TOPAS-Academic version 5 [Internet]. Brisbane, Australia: Coelho Software; 2012. [cited 2013 Apr 17]. Available from: http://www.topas-academic.net
  » http://www.topas-academic.net
• ¹⁹
  Ishitsuka M, Sato T, Endo T, Shimada M. Synthesis of thermal stability of aluminum titanate solid solutions. Journal of the American Ceramic Society . 1987;70(2):69-71.
• ²⁰
  Huang YX, Senos AM, Baptista RJ. Synthesis of aluminium titanate powder by different processing methods. In: Fourth Euro-Ceramics Conference. Italy: Faenza; 1995. p. 187-194.
• ²¹
  Takahashi M, Fukuda M, Kukuda H, Yoko T. Preparation, structure, and properties of thermally and mechanically improved aluminium titanate ceramics doped with alkali feldspar. Journal of the American Ceramic Society. 2002;85(12):3025-3030. DOI: 10.1111/j.1151-2916.2002.tb00573.x
  » https://doi.org/10.1111/j.1151-2916.2002.tb00573.x
• ²²
  Giordano L, Viviani M, Bottino C, Buscaglia MT, Buscaglia V, Nanni P. Microstructure and thermal expansion of Al₂TiO-MgTiO solid solutions obtained by reaction sintering. ₅₅Journal of the European Ceramic Society. 2002;22(11):1811-1822.
• ²³
  Ohya Y, Nakagawa Z, Hamano K. Crack healing and bending strength of aluminum titanate ceramics at high temperature. Journal of the American Ceramic Society. 1988;71:C232-C233. DOI: 10.1111/j.1151-2916.1988.tb05064.x
  » https://doi.org/10.1111/j.1151-2916.1988.tb05064.x
• ²⁴
  Parker FJ, Rice RW. Correlation between grain size and thermal expansion for aluminum titanate materials. Journal of the American Ceramic Society. 1989;72:2364-2366. DOI: 10.1111/j.1151-2916.1989.tb06091.x
  » https://doi.org/10.1111/j.1151-2916.1989.tb06091.x
• ²⁵
  Kuszyk JA, Bradt RC. Influence of grain size on effects of thermal expansion anisotropy in MgTiO _25.Journal of the American Ceramic Society. 1973;56:420-423.

Publication Dates

• Publication in this collection
  26 Feb 2016
• Date of issue
  Mar-Apr 2016

History

• Received
  25 Aug 2015
• Reviewed
  06 Nov 2015
• Accepted
  23 Dec 2015