Effect of TiO 2 in Fine Zircon Sintering and Properties

: The effect of the TiO 2 addition in the ceramic processing of dense zircon materials from zircon fine powders was established. The addition of TiO 2 (5-10 wt%) permitted to obtain dense ceramics at lower temperatures (100-150 o C below), with comparable mechanical behavior. The thermochemical processes were described after a multi-technique experimental approach, which included a sintering analysis, powder X-Ray diffraction analysis (XRD), scanning electronic microscopy (SEM) and Vickers hardness of the polished dense obtained ceramics. After 1400 o C heating programs, the added TiO 2 acts as a sintering aid with no important chemical reactions, and presented improved mechanical behavior in comparison with pure zircon ceramics. On the other side, in samples fired at 1500 o C, TiO 2 partially (≈50 %) reacts with zircon, forming ZrTiO 4 , while the formed SiO 2 goes to the grain boundaries. Samples with 5 wt% TiO 2 present better mechanical behavior than the ones with 10 wt%. The performed mechanical characterization indicates the merits of the material processed by this inexpensive processing route. Developed density, hardness (≈10 GPa) and fracture toughness (≈2 MPa.m -1/2 ) are comparable with the best figures reported.


Introduction
It is well known that zircon (ZrSiO 4 ) materials are good refractories for high temperature uses, due to zircon's chemical stability, low thermal conductivity, low thermal expansion and corrosion resistance, among others properties [1][2][3][4][5]. Zircon materials are used in glass and steel industries [5,6] especially because of pure zircon undergoes no major structural changes below 1680 o C, temperature at which it dissociates into zirconia and silica. The decomposition temperature decreases with the presence of impurities or mechanical pretreatments [1]. Due to their high refractoriness, zircon powders are difficult to sinter. To overcome this drawback, some additives are usually employed as sintering aids; the typical additives include, for example, refractory clays and titania (TiO 2 ) [7]. Although beneficial for sintering, sometimes the incorporation of these additives results in the loss of some of zircon's distinctive properties.
Generally, dense zircon ceramics without additives are processed by advance sintering techniques such as hot pressing [8,9] or spark plasma sintering (SPS) [10,11]. On the other side, it has been reported that the mechanochemical activation by high energy ball milling is a suitable pretreatment for raw materials in the obtention of dense ceramics. Its goal is to activate the chemical and physical processes by incorporating surface energy powders that are usually nanosized, thus effectively achieving homogeneous mixtures of powders, even if they differ in particle size [12][13][14]. Sintering is one of these processes activated by high energy milling.
The first attempt to study the mechanical activation of zircon was reported by Motoi [15]. It was demonstrated that prolonged ball milling not only causes particle size reduction but also amorphizes the mineral, which leads to partial decomposition into ZrO 2 and SiO 2 and can also enhance solubility in fluorhidric acid.
The milling effect on the zircon alumina reaction sintering for obtaining mullite zirconia refractory composites was also studied [16,17].
In particular, the mechanical behavior of this kind of material is of technological interest, and is related to the microstructure, phase composition and sintering grade. Suarez et al. reported the effect of colloidal processing optimization by slip casting of highly concentrated aqueous dispersions on the sintering of dense zircon ceramics [18]. In other work, Rendtorff et al. studied the effect of milling for a nonconventional advance sintering method (SPS) and the mechanical properties of dense zircon ceramics were reviewed and compared [10].
In a prior work we presented a systematic study of the effect of high energy milling treatment and the subsequent direct sintering of pure zircon fine powder [14]. Although there weren`t changes in the morphology of the activated powders, observed by SEM, they sintered after thermal treatment at 200 o C below the un-milled powders [14]. XRD allowed the identification of an incipient partial dissociation of zirconium silicate after long term (60-120 minutes) high energy milling treatments. This slight dissociation was also observed after the thermal treatments of the studied powders.
Zirconium titanate (ZrTiO 4 ) is commonly used as a dielectric in microwave devices due to its high permittivity to microwave frequencies [22]. ZrTiO 4 has been proposed for structural applications at severe thermomechanical conditions [20,21], and in composite structural ceramics [23,24]. The introduction of zirconia particles imbibed in structural ceramics has proven to result in several reinforcement mechanisms [25,26]. This strategy has been proposed and studied for zircon based materials [11,27].
To our knowledge, no systematic TiO 2 addition study has been carried out. The main objective of this work is to establish the effect of the TiO 2 addition in the ceramic processing of dense zircon ceramics, from fine zircon powders. For this, a systematic formulationprocessing-properties study was carried out. The high energy milling pretreatment has proven to enhance sintering processes [14], which was assessed as well. Particularly sinter parameters, developed phases, microstructure and mechanical properties were studied and correlated.

Materials and Experimental Procedures 2.1 Materials and methods
A commercial zircon (ZrSiO 4 ) powder (Kreutzonit Super Extra Weiß, Mahlwerke Helmut Kreutz GmbH, Germany, 0.8 µm) and anatase (TiO 2 ) powder (Cicarelli, Argentina) were used (Z and T respectively). Two powder mixtures were prepared adding 5 and 10 wt% of additive (ZT). They were mixed in a planetary mill (HEBM) (Pullverizer 7 Premium Line, Fritsch Co., Ltd., Germany) at 350 rpm for 3 min for homogenization of the unmilled mixtures, and at 850 rpm for 60 min for the milled ones (Table I). The same treatment was performed for pure zircon powder (Z). Zirconia jars of 85 ml were used and 60 g of zirconia balls (10 mm diameter) as mill and milling media respectively; 6 g of powder were used in each batch, dispersed in ethanol [11,14,28].
X-ray diffraction (XRD-Bruker D2 phaser Cu-Kα incident radiation at 30 kV to 10 mA) performed at 2θ between 15 and 80° was used to evaluate the effect of the milling time on the crystalline phases. Rietveld refinement was performed in order to follow the possible partial zircon dissociation and zirconium titanate formation [28][29][30].
Disc-shaped samples of 15 mm diameter were prepared from dried and screened (mesh 100) powders; these were first uniaxially pressed at 25 MPa and then isostatically pressed at 100 MPa [31,32]. Afterwards, samples were fired at 1300, 1400 and 1500 o C in an electric furnace. The employed heating and cooling rate were both, in all cases, 5 o C.min -1 with a holding time of 2 hours (Table I).
Tab. I Powders mixtures and obtained ceramic samples according to additive employed, milling time and sintering temperature.
To assess the densification of the composite the theoretical density (D Theo ) of the composite was estimated through the following Equation: where V i is the volume fraction of each crystalline phase (i) evaluated by XRD-Rietveld quantification and D i is the theoretical density of each phase.
Developed microstructures were evaluated by scanning electron microscopy (SEM-JEOL JCM-6000, Japan). Finally, Vickers hardness was evaluated (Buhhler Indentament 1100, USA) by, at least, ten indents under 3 Kg load and 15 second dwelling time [14]. Fracture toughness (K IC ) was determined by measuring the cracks generated by indentation via Evans and Charles method [34], using the formula proposed showed in Equation 2.
where H is Vickers hardness; a is the semi diagonal of the indentation and c is the distance from the center of the indentation to the tip of the crack.

Powder milling pretreatment, effect of the TiO 2 addition
The effect of milling on pure zircon fine powders has been previously studied [14]; showing a notorious enhancement of the sinterability of these powders accompanied by a slight zircon dissociation (Equation 3) and the absence of important changes in grain morphology or grain size distribution. It can be presumed that the performed milling process (in ethanol) improves the homogenization of these fine powder mixtures.
Figs. 1a and 1b compare the XRD patterns of the powders before and after the HEBM with 5 and 10 wt% of additive respectively (ZT) with the milled powder of zircon Z-60. The ZT unmilled mixtures show only zircon and anatase (TiO 2 ) phases [35]. No broadening of the peaks is observed, showing that the pretreatment did not affect the phases crystallinity. At the same time, the inset in this figure shows the main peaks of the zirconia phase (tetragonal and/or cubic) [28], showing a clear evidence of the incipient silicate dissociation in the milled powders according the Equation 3.   Table II shows the milling and sintering temperature effect on the textural properties, theoretical density (D Theo ), apparent density, open porosity and linear shrinkage of the obtained zircon samples. The open porosity is the most illustrative sintering parameter.

Thermal treatment of the powder mixtures 3.2.1. Sintering parameters: Shrinkage, apparent density and open porosity as a function of the processing variables
In general, HEBM improved sintering for all samples when compared to its unmilled par (at same sintering temperature and TiO 2 addition). Also, TiO 2 addition resulted in lower porosity values/enhanced sintering when compared with samples at same sintering temperature and no TiO 2 .

Fig. 2.
Correlation between the sintering parameters, apparent Density (g/cm 3 ) as a function of Shrinkage (%) of the studied zircon based materials.
After 1300 o C treatments pure samples do not sinter, sintering occurs after equivalent thermal treatments in the TiO 2 added samples; particularly 10 wt% additions result in a slight higher sintering than the 5 wt% addition. After 1400 o C treatments, the pure sample presents a certain grade of sintering enhanced by the milling pretreatment and the additive samples are fully sintered. Moreover, after 1500 o C treatments the only not fully sintered sample is the unmilled pure zircon sample (Z-0-15), which presents 8 % porosity.
It could be stated that the addition of TiO 2 , reduces sintering temperatures and allows the possibility of obtaining zircon ceramics with similar textural properties than pure ones. Achieved densities are in all the cases below pure zircon's theoretical density (4.56 g.cm -3 ), being these differences explained in term of the presence of lighter phases (TiO 2 , SiO 2 ), close porosity, and partial dissociation. Nevertheless, density is expectably correlated with shrinkage and open porosity, as shown in Fig. 2, thus proving that, any of them could be employed as sintering parameter.

Crystalline phases developed after the heating treatment
The crystalline phases determination after the different sintering treatments was performed by XRD. Fig. 3a compares XRD patterns of sintered ceramics under thermal treatments of 1400 and 1500 o C from zircon powders with and without milling (Z-60 and Z-0 respectively). The inset in this figure shows the main peaks of the zirconia phase (monoclinic, tetragonal and/or cubic) evidencing higher zirconium silicate dissociation in the sintered ceramic at 1500 o C obtained from milled powder.     Table III shows the quantification of crystalline phases (zircon, rutile and srilankite) by Rietveld method for all sintered ceramics. Samples sintered at 1500 o C show the presence of srilankite phase. No crystalline silica phases (quartz, cristobalite or tridymite) were detected; presumably, the produced SiO 2 is forming in the grain boundaries as a glassy phase. The resulting values of the R wp parameters were in all cases adequate and below 17.  5 shows the crystalline phases after the processing route employed as a function of the three different variables that were explored: milling pretreatment [14]; TiO 2 addition (0-10 wt%), and sintering temperature (1300-1500 o C). Samples fired at 1300 o C were not completely sintered (see Fig. 5); on the other side, samples fired at 1400 and 1500 o C are fully sintered. In all cases, the principal crystalline phase is zirconium silicate, which is accompanied by the described phases. For samples fired at 1300 and 1400 o C, the quantified (unreacted) TiO 2 in the sintered samples is similar to the initially incorporated in the formulations: ≈ 5 and ≈ 10 wt%, and the milling pretreatment effect is not relevant. For samples fired at 1500 o C the described thermochemical processes (Equations 4 and 5), were observed with a partial advance (≈ 50 %), and unreacted TiO 2 was observed in both ZT5 and ZT10 samples with and without milling treatment.

Developed microstructure by scanning electron microscopy (SEM)
The processes of pure zircon materials in an equivalent route were previously described [14]. We intend to describe the effect of the TiO 2 addition, milling pretreatment and sintering maximum temperature. SEM images of the gold coated polished surfaces were analyzed (Figs 6 and 7). Dense homogenous microstructures were observed in all ceramics developed.
The effect of the milling pretreatment and sintering temperature on the 5 wt% added samples is analyzed in Fig. 6. It shows the ZT5-0-14, ZT5-60-14, ZT5-0-15 and ZT5-60-15 sintered samples. Images "a" and "c" belong to ceramics obtained from unmilled powders, and images "b" and "d" from the milled ones. Dense homogeneous microstructures are generally obtained by the chosen processing route, the TiO 2 addition assure, in all the cases, the correct sintering of zircon powders. Fine grain size and rounded grains are observed. This is explained by the previous milling treatment; some grain growth is observed as well, comparing grain size with starting powders diameters. The isostatic pressing results in high but not necessarily full compaction of the fine powders Because of this, some micropores (some microns in size) can be observed in all the microstructures. As observed by XRD, TiO 2 incorporation resulted in a partial ZrSiO 4 dissociation, followed by ZrTiO 4 formation. This results in the formation of a SiO 2 based glassy phase that imbibe the silicate grains, observed in the grain border as a darker phase in Fig. 7a, as in similar systems. ZrTiO 4 grains appear in a lighter gray color marked with an S. No important differences were observed with or without the milling treatment at 1500 o C.
Compared with the ceramics obtained from zircon without additives using the same heat treatment and milling time, an improved microstructure is observed (porosity near null and density near 4.2-4.4 g.cm -3 ) [14] with the addition of TiO 2 in the studied proportions. The developed microstructure allowed the evaluation of the Vickers hardness. As it can be seen, all materials present expected micro pores (closed apparently) considering the processing route; and the open porosity was shown to be close to zero.   Fig. 6c and 6d (ZT5-0-15, ZT5-60-15) but now compared with the ZT10-0-15, ZT10-60-15 with 10 wt% of Titania addition. The images demonstrate that a high sintering was achieved for all ceramics.
Detected phases are in concordance with XRD, analysis (Table III). Dense zircon (Z) materials with the presence of some pores (P) as well as some imbibed srilankite (S) grains and glassy (G) grain boundaries can be observed (phases were corroborated by EDS punctual analysis). This fact is a merit of the processing route. Several larger zircon rounded grains are observed (diameters over 5 µm), which is expected to occur at higher temperatures (1500 o C).  Table I for labels). , the fracture toughness of some of materials studied is determined, which are presented in Table  V. The morphology of the observed cracks presents low tortuosity. Hardness, can only be evaluated in pore free microstructures, with indents sufficiently larger than the material microstructure [36]. Fig. 9 shows the Hv values of the sintered materials as a function of the relative density (%). A direct correlation can be observed in the studied range. Particularly, it can be assessed that dense materials fired at 1400 o C from milled powders present the highest hardness, being this value remarkably higher than the one of samples fired at 1500 o C. This material is well densified but the TiO 2 present did not react (see Fig. 5).

Mechanical properties, Hardness and fracture toughness
Expectably, not fully sintered sample (Z-0-14) presents lower hardness, which restricts the actual application of this material as structural ceramics. The addition of new phases might affect the mechanical behavior of structural ceramics [25], as zirconia grains are involved in several toughening mechanisms [25,26]. Apparently, ZrTiO 4 formation do not improves hardness of the zircon materials (TiO 2 added samples fired at 1500 o C).
The obtained values were compared with other dense zircon materials in Table IV. The achieved values correspond to dense zircon ceramics, and adequate for structural applications coupled with high refractoriness and special resistance to several severe chemical environments [37,38] like other zircon based ceramics (Table IV). However, the achieved mechanical behavior is slightly below materials processes by more sophisticated routes like SPS [39,40] or sol-gel derived materials [41]. While the first presents equipment and geometry restrictions, the second one presents, due to the usage of chemical precursors, environmental and scalability constraints, sustaining the merit of the present processing route. Fig. 9. Vickers hardness as a function of the relative density of the materials, for different sintering conditions, milling pretreatment and TiO 2 content.
Finally, Table V shows the Fracture Toughness (K IC ) of the zircon-titania based ceramics. Typically, fracture toughness of zircon ceramics is ranged between 2.0 and 3.5 MPa.m −0.5 [10]. The achieved values are in the upper limit of the reported values. Apparently, materials fired at 1500 o C, fully dense and with a partial ZrTiO 4 formation presents slightly higher fracture toughness values. The multiphasic microstructure obtained might explain the toughening of the material [25,26].
Hardness and toughness values obtained show the merits of the materials processed in the present study. While resulting density is comparable with the best figures reported, hardness and fracture toughness achieved were as high as any pure dense zircon material.

Sample
Sintering

Conclusion
The effect of the TiO 2 addition in the ceramic processing of dense zircon ceramics from fine zircon powders was established. For this, a systematic formulation -processingproperties study was carried out. The sintering parameters were evaluated; their effect in the sintering was established. Both milling and TiO 2 addition (up to 10 wt.%) enhance de zircon fine powders consolidation.
The thermochemical processes were described in terms of the processing variables: formulation and heating programs; homogenous dense microstructures were achieved.
The addition of TiO 2 permitted to obtain dense ceramics at lower temperatures around 100-150 o C below, with comparable mechanical behavior, without employing any sophisticated processing route, like sol-gel derived ceramics or electric current activated sintering strategies like SPS. The thermochemical processes were described. After 1400 o C heating programs, added TiO 2 acts as a sintering aid with no important chemical reactions, these samples presented the better mechanical behavior. On the other side in samples fired at 1500 o C, TiO 2 partially (≈ 50 %) reacts with zircon forming ZrTiO 4 , the formed SiO 2 goes to the grain boundaries.
Remarkably, samples with 5 wt% TiO 2 present better mechanical behavior than the 10 wt% samples. The mechanical characterization shows the merits of the material processed in the present study. While resulting density is comparable with the best figures reported; the hardness and fracture toughness achieved were as high as any pure dense zircon material.