Contents lists available at ScienceDirect

Ceramics International

journal homepage: www.elsevier.com/locate/ceramint

Microstructural and mechanical investigation of high alumina refractorycastables containing nano-titania

Ahmad Reza Abbasiana,⁎, Najmeh Omidvar-Askaryb

a Department of Materials Engineering, Faculty of Engineering, University of Sistan and Baluchestan, Zahedan, IranbDepartment of Materials Engineering, Naghshe Jahan Institute of Higher Education, Isfahan, Iran

A R T I C L E I N F O

Keywords:Alumina refractoryCastableHiboniteNano-titaniaSinteringAl2O3-CaO-TiO2

A B S T R A C T

In this study, mechanical and structural properties of two high alumina refractory castables, i.e. without nano-titania and with 0.4 wt% TiO2, are investigated. Bulk density, apparent porosity, cold compressive strength, coldmodulus of rupture, permanent linear changes, hot modulus of rupture and shock resistance tests were per-formed on the samples. The microstructure and phase analyses were conducted by field emission scanningelectron microscope equipped with EDX analyzer and X-ray diffraction, respectively. Results showed that hi-bonite phase formation in refractory is dependent on the temperature which could be decreased by addition ofnano-titania, leading to the improvement of refractory strength. On the other hand, the decrease of hot modulusof rupture and shock resistance was observed for the refractory castable containing nano-titania. Also,Ca3Ti8Al12O37 phase was produced after firing at 1000 °C in the presence of nano-titania. Probable mechanismsare proposed for phase and microstructural changes in high alumina refractory castable containing nano-titania.

1. Introduction

Persistent developments in high-temperature industries stiffly relyon the quality of their refractory materials. High alumina refractoriesare produced as extremely consumed compositions due to outstandingfeatures at high temperatures including stability in oxidizing and re-ducing atmospheres, considerable slag and metal corrosion resistance,remarkable abrasion resistance and excellent refractoriness [1–8].Today, everybody is talking about nano-materials, even advertisementsfor consumer products use the prefix “nano” as a keyword for specialfeatures [9]. Nano-materials either refer to materials with nano-scaledimensions or bulk materials containing nano-sized particles. Nano-particles are defined as particles of less than 100 nm in diameter. Thesematerials often demonstrate characteristics which are surprisingly dif-ferent from the same materials with larger dimensions [10]. Recently,trend on development of alumina refractories is focused on using ofnano-materials. Nano-materials improve several properties of re-fractory products. It was found that colloidal silica (a stable dispersionof nano-sized particles of amorphous SiO2) is suitable as a calcium-freebinder for high alumina ultra-low cement refractory castables [11,12].Yaghoubi et al. [12] reported that the refractory castables containingoptimum amount of colloidal silica show a remarkable increase in bothfluidity and mechanical strength in dried and sintered states. It was alsofounded that nano-SiO2 particles increase the rate of needle-shaped

mullite formation during sintering at 1400 °C. Mukhopadhyay et al.[13] incorporated 8.0 wt% nano-MgAl2O4 in a high alumina-based lowcement castables. They showed the improvement of thermal shock andslag resistance behavior of the refractory in corrosive environments. Animprovement of the mechanical strength both at room and high tem-peratures as well as an improvement of the thermal shock resistance ofAl2O3–C refractory by using nano-scaled powders of MgAl2O4, aluminasheets and carbon nanotubes has been reported by Roungos and Ane-ziris [14]. Ultra-low cement castables based on tabular alumina wereprepared with nano-ZrO2 [15]. It was found that addition of 3 wt%nano-ZrO2 particles can increase the modulus of rupture values of re-fractory castables by up to 50%. The influence of additions of nano-TiO2, nano-ZrO2, nano-SiO2 and nano-MgO into alumina based mate-rials for refractory applications has also been investigated [16]. Thegeneration of a micro-crack network after sintering due to the forma-tion and decomposition of aluminum titanate (Al2TiO5) before and afterthermal shock treatment leads to higher thermal shock resistance.

TiO2 nanoparticles are commercially available. This nano-materialhas been employed in various applications including selective synthesisof organic compounds for solar cells, splitting of water for green-energyhydrogen production, photocatalysts for removal of organic and in-organic pollutants, photo-killing of pathogenic organisms and air pur-ification [17] as well as their very important applications as additive inrefractories. It was observed that the addition of 0.5 wt% nano-TiO2

https://doi.org/10.1016/j.ceramint.2018.09.165Received 27 July 2018; Received in revised form 11 September 2018; Accepted 16 September 2018

⁎ Correspondence to: Department of Materials Engineering, University of Sistan and Baluchestan, P.O. Box 98155-987, University Blvd, Zahedan, Iran.E-mail address: abbasian@eng.usb.ac.ir (A.R. Abbasian).

Ceramics International 45 (2019) 287–298

Available online 17 September 20180272-8842/ © 2018 Elsevier Ltd and Techna Group S.r.l. All rights reserved.

T

into silica brick composition leads to highest amount of tridymite phaseformation [18]. It was reported that the incorporation of nano-TiO2 intotypical aluminosilicate refractory bricks leads to the improved me-chanical characteristics with respect to the typical aluminosilicates,presumably because of a better compaction during the raw materials'mixing stage [19]. The mullite-bonded alumina-silicon carbide re-fractories in the presence of nano-TiO2 up to 1 wt% showed highermechanical strength because of increasing of mullite phase and highergrowth of its needle-like grains [20]. Tontrup used nano-TiO2 as sin-tering aid in 99% alumina refractory bricks [21]. By using only 0.4%TiO2 nanoparticles, the firing temperature was reduced from 1730 °C to1500 °C and the cold crushing strength (CCS) increased from 8.5MPa to13.8 MPa [21]. The self-flow type of high-alumina low-cement re-fractory castables containing 0–1wt% nano-TiO2 particles are also in-vestigated [22]. The results indicated significant effect of nano-TiO2 onthe flowability characteristics, phase composition, physical and me-chanical properties of alumina castables. With increase of nano-TiO2

particles, the flowability and workability of refractory castable tend todecrease. With addition of 0.5 wt% nano-TiO2 to the refractory castablecomposition, the mechanical strength of refractory in all firing tem-peratures tends to increase. Addition of 1 wt% nano-titania can de-crease the mechanical strength of refractory castable after firing.

In this research, the effect of nano-titania on mechanical andstructural characteristics of high alumina castables is comprehensivelyinvestigated. Apparent porosity, bulk density, cold crushing strength,cold and hot modulus of rupture, permanent linear change, thermalshock resistance as well as phase and morphological analyses are per-formed for high alumina castable without nano-titania (i.e. pure ca-stable) and refractory castable containing 0.4 wt% nano-TiO2 (i.e.0.4TiO2 castable).

2. Experimental

2.1. Raw materials

Source of supply, purity, density and content of raw materials usedfor the preparation of the refractory castables are listed in Table 1.Tabular alumina was used with four particle size distributions, i.e.3–6mm, 1–3mm, 0–1mm and less than 45 µm. The nano-TiO2 powderwith anatase crystal structure as additive has mean particle size andspecific surface area of 21 nm and 50m2/g, respectively. Fig. 1 shows atransmission electron microscopy (TEM) image of the nano-TiO2

powder provided by the supplier. Dolapix FF 26, which is a poly-carbonic acid, and citric acid were used as dispersing agent and settingcontroller of refractory castables, respectively. 0.4 wt% nano-TiO2 wasselected, because higher amounts of nano-TiO2 significantly diminishesthe flowability of the refractory castables.

2.2. Sample preparation

The particle size distribution of the refractory castables was ad-justed to a theoretical continuous curve based on the Dinger and Funkmodel [23] with distribution modulus equal to 0.28, which was suitablefor vibration castable [24].

Raw materials were weighed (each batch 4.5 kg) and dry mixed for4min in a Eirich mixer. Then water was added and the wet mixingcontinued for another 4min to form the castable. The water content forall refractory castables was 5.3 wt% based on dry weight of raw ma-terials. samples was vibro-casted into steel molds using a vibrating tableat frequency of 60 Hz for 1min. Samples in the molds were wrapped inplastic sheets and cured for 24 h at room temperature (24–26 °C). Thecured samples were demolded and kept at room temperature for an-other 24 h before drying at 110 °C. Cured samples were then fired inlaboratory electrical furnace at 1000, 1350 and 1550 °C with soakingtime of 6 h. Two high alumina castables are prepared in this in-vestigation. The first one was without nano-TiO2 (i.e. pure castable)while the second one contained 0.4 wt% nano-TiO2 (hereafter called0.4TiO2 castable). 0.13 wt% and 0.20 wt% dispersant was used for pureand 0.4TiO2 castables, respectively. The amount of dispersant was ad-justed relative to the amount of the nanomaterial used in the for-mulation. Higher amounts of dispersing agent are needed for dispersionof nanoparticles because of their high surface area.

2.3. Characterization methods

Apparent porosity (A.P) and bulk density (B.D) were measured bythe Archimedes procedure according to ASTM C20-00(2015)[25].

Cold crushing strength (CCS) of the samples were measured ac-cording to the ASTM C133-97 [26] in a compressive tester. Eq. (1) wasused to determine the CCS of the refractory castables with size of70mm×70mm×70mm.

=CCS FAmax

0 (1)

where, Fmax is the load at which fracture occurs (kg) and A0 is the areaof the sample (cm2).

Modulus of rupture (MOR) of the refractory castables was measuredat two cold and hot states according to DIN EN 993-7 [27]. MOR is themaximum stress that a rectangular test piece (160mm×40mm×40mm) can withstand when it is bent in a three-point bending device. Thetest specimen was kept horizontally in a support having two edges andthen load was applied on the sample uniformly. The failure load (Fmax )of the specimen was used for the calculation of MOR according to thefollowing formula:

⎛⎝

⎞⎠

= × ⋅⋅

MORkg

cmF L

b h32

s2

max2 (2)

where Ls is the distance between bearing edges (cm); b is the width of

Table 1Raw materials used for the preparation of high alumina castable.

Raw materials Source Purity (wt%) Density (g/cm3)

Content (wt%)

Tabularalumina

Alteo, France > 99.5 3.50 91

Reactivealumina

PFR 20, Alteo,France

> 99.8 3.92 7

High aluminacement

Secar 71, Kerneos,China

(Al2O3)> 70 2.93 2

Nano-titania Aeroxide Co. (P25) > 99.5 0 or 0.4dolapix FF 26 Zschimmer &

Schwarz, Germany> 99 0.13 or 0.2

Citric acid C6H8O7, Panreac,Spain

Pure 1.66 0.01

Fig. 1. TEM image of nano-TiO2 powder provided by the supplier.

A.R. Abbasian, N. Omidvar-Askary Ceramics International 45 (2019) 287–298

288

specimen (cm); h is the thickness of the specimen (cm). Cold modulus ofrupture (CMOR) of the specimens was measured using a universaltesting machine (SANTAM, Iran) at ambient temperature. Hot modulusof rupture (HMOR) of the samples was evaluated after pre-firing at1350 and 1550 °C. Final temperature for the measurement of themodulus of rupture was 1350 °C with a heating rate of 5 °C/min. HMORtests were done in air atmosphere by Netzsch apparatus.

Permanent linear change (PLC) of refractory castables (70mm×70mm×70mm) fired at various temperatures was measured ac-cording to ASTM C 179-14 standard [28].

The thermal shock resistance of refractory castables was determinedusing DIN 51068 standard quench tests [29] in which the material isheated and cooled subsequently and the number of heating and coolingcycles that a refractory can withstand prior to failure is taken as itsthermal shock resistance. The quantification was done by the number ofcycles (up to 30 cycles) to withstand such temperature. Cylindricalspecimens (50mm×50mm) were heated at 950 °C for 30min andthen suddenly cooled down to ambient condition by quenching in waterfor 10min. In addition, the number of cycles before creation of anycrack in the specimen was reported as the spalling resistance.

The phase transformation during firing of refractory castables wasdetermined by X-ray diffraction (XRD). For XRD analysis, the refractorycastables were crushed and ground to very fine powders using an agatemortar. The XRD patterns were recorded on a unit (Philips PW1800,Netherland) using Cukα radiation at 40 kV and 30mA. The obtaineddiffraction patterns in the 2 theta range of 10° to 90 ° were analyzedusing X-pert Highscore software.

The microstructure analysis of samples coated with gold was per-formed using field emission scanning electron microscope (FE-SEM,MIRA3 TESCAN-XMU) equipped with an electron dispersive X-rayspectroscopy (EDX) detector.

3. Results and discussion

3.1. Phase analysis

Fig. 2 shows XRD pattern obtained for pure castable after drying andfiring at different temperatures. As shown in Fig. 2a, the main peaksobtained for refractory castable after drying at 110 °C correspond toAl2O3 (JCPDS No. 01-075-1862). Also some peaks with lower intensityincluding Al(OH)3 (JCPDS No. 01-074-1775) and Ca3Al2O6·(H2O)6(JCPDS No. 01-071-0735) are also observed which are formed by hy-dration of CA and CA2 in alumina cement [30]. Because of the lowconcentration of alumina cement (2 wt%) used in the refractory ca-stable, the amount of hydrated phases are comparatively low, so theirpeak intensity is very lower than that of alumina.

The main XRD peaks (Fig. 2b) obtained for pure castable after firingat 1000 °C correspond to alumina (JCPDS No. 01-075-1862) togetherwith low-intensity peaks corresponding to Ca3Al2O6 phase (JCPDS No.00-038-1429). Wang et al. [31] similarly showed that Al(OH)3 andCa3Al2O6(H2O)6 phases are formed in alumina castable containing10 wt% high alumina cement without any microsilica after drying at110 °C. They also reported that after firing the refractory castable at1000 °C, Ca7Al12O33 and CaAl2O6 phases are formed in the structure.They observed that Al(OH)3 and Ca3Al2O6·(H2O)6 phases are dehy-drated to AlO(OH) and Ca12Al14O33 phases at temperatures between230 and 300 °C according to reactions (3) and (4), respectively.

→ +Al(OH) 2AlO(OH) 2H O3 2 (3)

→ + +7Ca Al O .6H O Ca Al O 9CaO 42H O3 2 6 2 12 14 33 2 (4)

At temperatures about 650 °C, Al2O3 forms by reaction (5)

→ +2AlO(OH) Al O H O2 3 2 (5)

And at temperatures about 1000 °C, CaAl2O4 phase is partiallyformed by reaction (6). Reaction (6) completes at 1030–1120 °C.

Fig. 2. XRD patterns obtained for pure castable after drying at 110 °C (a) andafter firing at different temperatures of 1000 °C (b), 1350 °C (b) and 1550 °C (d).

A.R. Abbasian, N. Omidvar-Askary Ceramics International 45 (2019) 287–298

289

+ →Ca Al O 5Al O CaAl O .12 14 33 2 3 2 4 (6)

[31]XRD results also confirm the presence of Ca3Al2O6 in addition to the

main alumina phase. It seems that gibbsite (Al(OH)3) is removed byreactions (3) and (5), however, removal of structural water fromCa3Al2O6·(H2O)6 and formation of Ca3Al2O6 directly take place by re-action (7) instead of reactions (4) and (6). Therefore, Ca3Al2O6 phaseformed instead of CaAl2O4 phase that was also reported by Wang et al.[31].

→ +Ca Al O . (H O) Ca Al O 6H O3 2 6 2 6 3 2 6 2 (7)

XRD pattern for pure castable after firing at 1350 °C is shown inFig. 2c. The main peak again corresponds to alumina (JCPDS No. 01-010-0173). Peaks with relatively lower intensities are also observedwhich are related to CaAl4O7 (JCPDS No. 01-76-0706) and Ca5Al6O14

(JCPDS No. 00-003-0149). These phases may be produced through thefollowing reaction:

+ → +2Ca Al O 3Al O Ca Al O CaAl O3 2 6 2 3 5 6 14 4 7 (8)

Wang et al. [31] also demonstrated the formation of CaAl4O7

through reaction of calcium aluminate phases, formed at low tem-peratures, together with fine grain alumina at temperatures in the rangeof 1120–1340 °C according to the following reaction:

+ →CaAl O Al O CaAl O2 4 2 3 4 7 (9)

As can be seen in Fig. 2d, by increasing the firing temperature to1550 °C, peaks with relatively high intensity corresponding to hibonitephase (CaAl12O19) (JCPDS No. 00-033-0253) are appeared in additionto main alumina peaks (JCPDS No. 01-010-0173). The hibonite phaseforms by the following reactions:

+ →CaAl O 4Al O CaAl O4 7 2 3 12 19 (10)

+ →Ca Al O 27Al O 5CaAl O5 6 14 2 3 12 19 (11)

Formation of hibonite phase at temperatures above 1400 °C throughreaction (10) was also reported by Wang et al. [31].

XRD pattern for 0.4TiO2 castable after drying at 110 °C is shown inFig. 3a which confirms that the main phase in the refractory castable isalumina (JCPDS No. 01-075-1862) with relatively low-intensity peakscorresponding to Al(OH)3 (JCPDS No. 01-085-0611) andCa3Al2O6·(H2O)6 (JCPDS No. 01-079-1286). So, from phase point ofview, there is no difference between the XRD patterns of 0.4TiO2 andpure castables. It should be mentioned that TiO2 peaks are not identi-fiable in Fig. 3a because of their relatively low intensity.

Fig. 3b shows the XRD pattern obtained for 0.4TiO2 castable afterfiring at 1000 °C. It is obvious from the figure that increasing the firingtemperature has significantly affected the phase composition of therefractory castable containing nano-titania. Peaks corresponding toCa3Ti8Al12O37 (JCPDS No. 00-037-1232) are also observed in additionto the main alumina phase (JCPDS No. 01-075-1862).

Gibbs free energies for the formation of ternary and quaternarycompositions from binary compounds (CaO, TiO2 and Al2O3) are listedin Table 2. Using these thermodynamic data, Jacob and Rajitha [32]depicted ternary phase diagrams of CaO-TiO2-Al2O3 system. The iso-thermal section of the diagram at 927 °C is illustrated in Fig. 4. Thediagram (Fig. 4a) is valid in the temperature range of 646–1110 °C. Itcan be seen from the Fig. 4a that Ca3Ti8Al12O37 is a stable compound inCaO-TiO2-Al2O3 system. Table 3 shows reactions and chemical poten-tials of CaO calculated by Jacob and Rajitha [32]. According to thetable, it is evident that the least chemical potential difference corre-sponds to the formation of Ca3Ti8Al12O37 phase. Consequently, it ispossible for reaction (12) to occur at about 1000 °C

+ + →Ca Al O 8TiO 5Al O Ca Ti Al O3 2 6 2 2 3 3 8 12 37 (12)

Fig. 3c shows the XRD pattern obtained for 0.4TiO2 castable afterfiring at 1350 °C. By increasing the firing temperature to 1350 °C,

Fig. 3. XRD patterns obtained for 0.4TiO2 castable after drying at 110 °C (a)and firing at different temperatures of 1000 °C (b), 1350 °C (c) and 1550 °C (d).

A.R. Abbasian, N. Omidvar-Askary Ceramics International 45 (2019) 287–298

290

hibonite phase (CaAl12O19) (JCPDS No. 00-038-0470) is detected be-sides main alumina phase, while peaks corresponding to Ca3Ti8Al2O37

phase are disappeared. The isothermal section of CaO-TiO2-Al2O3

system at 927 °C is shown in Fig. 4b which is valid for temperatures inthe range of 1223–1431 °C. According to Fig. 7b [35], reaction (13)occurs at 1350 °C in refractory castables containing nano-titaniathrough which Ca3Ti8Al12O37 phase transforms to tialite (Al2TiO5) andperovskite (CaTiO3) phases. In addition, unreacted Ca3Al2O6 phasefrom reaction (12), transforms to CaAl4O7 by reaction (14) and thenhibonite phase form by reaction (15).

→ + +Ca Ti Al O Al O 3CaTiO 5Al TiO3 8 12 37 2 3 3 2 5 (13)

+ →Ca Al O 5Al O 3CaAl O3 2 6 2 3 4 7 (14)

+ →CaAl O 4Al O CaAl O4 7 2 3 12 19 (15)

It has shown that tialite (Al2TiO5) decomposes to its constituents byreaction (16) during cooling, so its detection is not possible in XRDpattern [33,34].

→ +Al TiO Al O TiO2 5 2 3 2 (16)

Badiee and Otroj [22] reported the formation of CaTiO3 phase inalumina castable containing 1 wt% nano-titania after firing at 1250 °C,while they couldn’t identify CaTiO3 phase in the refractory castablecontaining 0.5 wt% nano-titania. They found that by increasing theamount of nano-titania in alumina castable fired at 1250 °C, CaTiO3

phase increases and CaAl12O19 phase diminishes. Their results are inagreement with the results presented in the current study. It seems thatincreasing the amount of nano-titania results in more Ca3Ti8Al12O37

phase at 1000 °C and consumption of Ca3Al2O6 according to reaction(12). Consequently, consumption of Ca3Al2O6 causes less production ofCaAl4O7, hence less hibonite phase produces via reactions (14) and(15). On the other hand, by increasing the amount of Ca3Ti8Al12O37, thepotential of CaTiO3 formation by reaction (13) also increases.

Fig. 3d shows the XRD pattern obtained for 0.4TiO2 castable afterfiring at 1550 °C. It can be seen that by increasing the firing tempera-ture to 1550 °C, the relative intensity of hibonite to alumina peaks alsoincreases which confirms that the amount of hibonite phase has beenincreased [35]. A weak peak corresponding to CaTiO3 (JCPDS No. 00-040-0043) is also observed at 2θ=46.46. It is probable that tialite andhibonite phases are responsible for the formation of more CaTiO3

phase. After firing the alumina castable at 1550 °C, Badiee and Otroj[22] observed that more CaTiO3 and CaAl12O19 phases are formed byincreasing the amount of nano-titania. They reported that the increaseof CaAl12O19 phase may be related to its formation at lower tempera-tures. But they didn’t mention any reason for the increase of CaTiO3

phase. It appears from reaction (17) that by increasing the firing tem-perature, perovskite is produced by the reaction of hibonite with tialite.

+ → +CaAl O Al TiO 7Al O CaTiO12 19 2 5 2 3 3 (17)

Fig. 2d shows that hibonite phase is produced in pure castable afterfiring at 1550 °C. However, as shown in Fig. 3c, nano-titania has asignificant effect on the phase formation in refractory castables. Itcauses hibonite phase to be formed at lower temperatures (i.e. 1350 °C).In other words, nano-titania accelerates the formation of hibonitephase. These observations confirm that the formation of hibonite phasedepends on the temperature, and nano-titania can decrease this tem-perature.

Our results are in good agreement with [22,36] which demonstratedthat nano-titania acts as mineralizer and decreases the formation tem-perature of hibonite phase. Two mechanisms are supposed for the

Table 2Gibbs free energies for the formation of ternary and quaternary compositionsfrom binary compounds (CaO, TiO2, Al2O3) [32].

No. compound free energy, KJ mol−1 temperature range, K

1 CaAl12O19 − 17430–37.2 T(± 1500) 923–19002 CaAl4O7 − 16400–26.8 T (± 2500) 923–19003 CaAl2O4 − 18120–18.62 T (± 1500) 923–19004 Ca12Al14O33 − 86100–205.1 T (± 17500) 923–16005 Ca3Al2O6 − 17000–32.0 T (± 1500) 923–18006 Ca3Ti2O7 − 164217–16.838 T (± 185) 900–12507 Ca4Ti3O10 − 243473–25.758 T (± 275) 900–12508 CaTiO3 − 80140–6.302 T (± 85) 900–12509 Ca3Ti8Al12O37 − 248474–15.706 T (± 70) 900–1250

Fig. 4. Isothermal sections of CaO-TiO2-Al2O3 system at 927 °C. The diagram(a) and (b) are valid in the temperature range of 646–1110 °C and1223–1431 °C, respectively.

Table 3Reactions and chemical potentials of CaO at 927 °C in CaO-TiO2-Al2O3 system[32].

No. defined reactions, ΔμCaO ηTi/(ηTi + ηAl) phasecomposition

ΔμCaO,KJ mol−1

1 6Al2O3 + 8TiO2

+ 3CaO→Ca3Ti8Al12O37

0, 1, 0.4 89.107

2 TiO2 +CaO→CaTiO3 1, 1 87.7023 Ca3Ti8Al12O37 + 5CaO→8CaTiO3

+ 6Al2O3

0.4, 1, 0 − 86.859

4 6Al2O3 +CaO→CaAl12O19 0, 0 − 62.0705 CaAl12O19 +2CaO→3CaAl4O7 0, 0 − 41.8056 CaAl4O7 +CaO→2CaAl2O4 0, 0 − 32.3687 3CaTiO3 +CaO→Ca4Ti3O10 1, 1 − 11.2758 7CaAl2O4 + 5CaO→Ca12Al14O33 0, 0 − 9.7949 Ca12Al14O33 +9CaO→7Ca3Al2O6 0, 0 − 6.17610 2Ca4Ti3O10 +CaO→3Ca3Ti2O7 1, 1 − 4.504

A.R. Abbasian, N. Omidvar-Askary Ceramics International 45 (2019) 287–298

291

mineralizing effect of nano-titania. Badiee and Otroj [22] believed thatnano-titania acts as preferred sites for hibonite nucleation. However,some researchers suggested that the mineralizing effect of nano-titaniaarises from the formation of liquid phase [36]. Braulio and Pandolfelli[36] used different additives including magnesium fluoride, titania,zirconia, magnesium borate and borosilicate in alumina-magnesiumcastables with cement bonding. They found that borosilicate and zir-conia had no effect on reduction of hibonite formation temperaturewhile titania showed the most temperature reduction. So, in Al2O3-CaOsystem, additives which formed liquid phase could reduce hiboniteformation temperature. Borosilicate doesn’t lower this temperature,although it forms liquid phase. They showed that borosilicate formsunstable liquid phase, while other additives form stable liquid phase[36].

According to the above discussion, it seems that the theory proposedby Badiee and Otroj [22] may not be true; because zirconia couldn’tdecrease the hibonite formation temperature. So, the second theory onthe effect of nano-titania on the formation of liquid phase is more re-levant.

The hibonite phase formed in the presence of liquid phase shouldhave planar morphology [37–39]. FESEM images presented in Section3.4 show that the hibonite phase formed in 0.4TiO2 castable at 1350 °Chas planar morphology while the hibonite phase formed at 1550 °C inpure castable has both planar and equiaxed morphologies. So, it can beconcluded that the presence of trace impurities in pure castable is re-sponsible for the formation of liquid phase at 1550 °C. This liquid tendsto selectively wet low energy grain boundaries and forms planar hi-bonite phase. On the other hand, equiaxed hibonite phase form throughnucleation and growth mechanism in grain boundaries which are notcompletely wetted by the liquid phase.

In refractory castable containing nano-titania, Ca3Ti8Al12O37 phaseforms at 1000 °C. It seems that the formation and transformation ofCa3Ti8Al12O37 is the main mechanism for the formation of planar hi-bonite. Most likely, Ca3Ti8Al12O37 wets the grains. Badiee and Otroj[22] suggested that hibonite forms at 1250 °C in alumina castablescontaining 0.5 wt% nano-titania which is in accordance with the me-chanisms proposed in the present study.

3.2. Apparent porosity and bulk density

The effect of nano-titania on the bulk density and apparent porosityof the refractory castables as a function of firing temperature is shownin Fig. 5a and b, respectively. It can be seen from the figure that theaddition of nano-titania increases the porosity, so decreasing the den-sity of the crude castable (before firing). As compounds containingnano-titania have low flowability, bubbles trapped within the structureincrease the porosity, so decreasing the density.

It can be seen that, at 1000 °C, the porosity of both pure and 0.4TiO2

castables are increased after firing. The increase of porosity is slightlyhigher for pure castable in comparison with 0.4TiO2. The density ofpure castable is also decreased, while that of 0.4TiO2 has not changedsignificantly. This can be due to the high-temperature decomposition ofhydrated phases which have been formed during the setting of aluminacement. According to the phase analysis, formation of a quaternarycompound with the composition of Ca3Ti8Al12O37 (density of 3.88 g/cm3) instead of Ca3Al2O6 (density of 3.03 g/cm3) may be the reason forthe slight decrease in density in 0.4TiO2 castable despite the relativeincrease of the porosity.

Results show that rising of the firing temperature to 1350 °C en-hances the densification process, resulting in lower apparent porosityand so, higher bulk density of pure castable. At 1350 °C, the porosity of0.4TiO2 castable decreases which may be due to the formation of newphase(s) in the sample. The expansion of these phases can compensatefor the contraction arises from the densification process. The phasetransformations are discussed in detail in Section 3.1. In 0.4TiO2 ca-stable, formation of hibonite with the chemical composition of

CaAl12O19 is responsible for the structural expansion [22] which de-creases the density.

Porosity decreases by continuing the densification process at highertemperatures (i.e. 1550 °C) resulting in higher density for both re-fractory castables. Here again, the amount of the porosity reduction ishigher in 0.4TiO2 castable.

3.3. Permanent linear changes (PLC)

Fig. 6 shows the permanent linear changes (PLC) of pure and0.4TiO2 castables at three different firing temperatures, i.e. 1000, 1350

Fig. 5. The effect of 0.4 wt% nano-titania on bulk density (a) and porosity (b) ofalumina castables after firing at different temperatures.

Fig. 6. The effect of 0.4 wt% nano-titania on permanent linear changes (PLC) ofalumina castables after firing at different temperatures.

A.R. Abbasian, N. Omidvar-Askary Ceramics International 45 (2019) 287–298

292

and 1550 °C. Both samples show expansion due to the formation of newphases at 1000 °C. It can be seen that the expansion of 0.4TiO2 sample isless than that of pure castable which is attributed to the nature and typeof these new phases. The difference is also noticeable at 1350 °C atwhich new phases are formed in 0.4TiO2 castable, causing more ex-pansion than pure castable (see XRD results in Section 3.1). Besides,0.4TiO2 castable shows about 0.6% contraction at 1550 °C due to theprogress of densification process. On the other hand, pure castableshows positive volume stability despite the decrease of contraction. Theresult of competition between the contraction (due to densificationprocess) and the expansion (due to the formation of new phases) in purecastable will be positive volume stability.

3.4. Structural examinations

Figs. 7, 8 shows the microstructure of the pure and 0.4TiO2 ca-stables fired at 1350 °C, respectively. The microstructure of pure ca-stable (Fig. 7a) is relatively uniform; however, two distinct regions canbe distinguished in 0.4TiO2 castable (Fig. 8a). Higher magnificationimages of region 1 are shown in Fig. 8b while those of region 2 areillustrated in Fig. 8c. As can be seen, region 1 is composed of inter-connected grains without grain boundary porosity which implies sin-tering of grains in 0.4TiO2 castable, while region 2 mainly consists ofinterlocked planar particles. The XRD pattern shown in Fig. 3c and theEDX analysis (not shown) demonstrated that these particles are hibo-nite phase and result in a porous matrix as shown in Fig. 8c.

Fig. 7b shows the higher magnification FESEM image of pure ca-stable microstructure shown in Fig. 7a. In this microstructure, partiallyconnected particles with high intraparticle porosity are observed, im-plying initial stages of sintering process [40–42]. Comparing Fig. 7bwith Fig. 8b reveal that nano-titania not only accelerates the densifi-cation process, but also increases the grain growth rate. The graingrowth of alumina grains in the presence of TiO2 usually occurs at finalstages of sintering [41,43,44]. So, one can claim according to the por-osity distribution that the microstructure shown in Fig. 8b implies thefinal stage of the sintering process. Titania (TiO2) is an aliovalent ad-ditive and its cationic capacity is different from that of the host. Sodensification of alumina at lower sintering temperatures in the presenceof TiO2 results from the change of diffusion mechanism as well as thelimited dissolution of Ti+4 cations as a substitute for Al+3 cations inalumina lattice. This substitution results in the formation of atomicvacancies in alumina hexagonal structure, providing more paths forbulk diffusion [45]. The structure shown in Fig. 8b demonstrates theacceleration of densification process in the presence of nano-titania.

The microstructure of pure castable fired at 1550 °C is shown inFig. 9a and b which is composed of weak connected particles with in-terconnected porosity. Contrary to refractory castables containingnano-titania, there is no evidence of grain growth, implying inter-mediate stages of sintering process [41]. It is also observed that nu-cleation of equiaxed hibonite grains has occurred around aluminagrains, although some hibonite grains are in the form of planar crystals.Despite the lack of a detectable liquid phase in the structure of thedesigned compositions at high temperatures, the generation of theelongated hibonite grains might be favored by the effective contactbetween CaO (formed due to the cement hydrates and presented in rawmaterials) and fine alumina [46]. It seems that the sintering tempera-ture or time has not been sufficient for densification and removal ofporosities. In addition, according to [47], the planar hibonite phase,which is produced in Al2O3-CaO system, makes Al2O3 matrix porousand inhibits densification.

Consequently, it is possible to increase alumina density by low-temperature sintering in the presence of nano-titania. This phenomenonhas also been reported in previous works [45,48–50], however, basedon the results obtained in the current study, the enhancement of den-sification at lower temperatures in the presence of nano-titania is moreintense than what had been reported previously. In other words, thestructure of 0.4TiO2 castable fired at 1350 °C (Fig. 8b) is in the finalstage of sintering, whereas increasing of firing temperature from 1350to 1550 °C has not been sufficient for complete sintering of pure ca-stable (Fig. 9a). So, a different sintering mechanism should be involvedin refractory castables containing nano-titania.

Three possible driving forces which decrease the free energy of thesystem are particle surface curvature, external applied pressure andchemical reaction. If the chemical reaction can be used to increase thedensification, it can provide the required driving force for sintering[41,51].

Gibbs free energy for the formation of hibonite (CaAl12O19) andcalcium di-aluminate (CaAl4O7) phases can be obtained from Eqs. (18),(19) [52]. Eq. (20) is consistent with reaction (15) and its Gibbs freeenergy is derived from the difference of Gibbs free energies of reactions(18) and (19):

+ → = − −ΔCaO 6Al O CaAl O G 17430 37.2T2 3 12 190 (18)

[52]

+ → = − −ΔCaO 2Al O CaAl O G 16400 26.8T2 3 4 70 (19)

[52]

Fig. 7. FESEM images in different magnifications obtained for pure castable fired at 1350 °C.

A.R. Abbasian, N. Omidvar-Askary Ceramics International 45 (2019) 287–298

293

+ → = − −ΔCaAl O 4Al O CaAl O G 1030 10.4T4 7 2 3 12 190 (20)

The required free energy for the formation of hibonite at 1350 °Caccording to reaction (20) is 15,070 KJmol−1 which is much higherthan the driving force of an applied stress. Formation of CaAl4O7 andCaAl12O19 phases in 0.4TiO2 castable at 1350 °C involves 13.6 and3.01 vol% expansion, respectively [53,54]. This expansion exerts pres-sure on the adjacent regions serve as nucleation sites for the formationof hibonite phase. This pressure acts as an external pressure which isapplied in processes such as hot pressing. So, it seems that the pressuredeveloped by the expansion together with the energy developed byreaction (20) provide extra driving force for sintering of alumina par-ticles.

Fig. 9c and d shows FESEM images of two distinct structures in0.4TiO2 castable fired at 1550 °C. A complete-sintered structure withsome intraparticle porosity is shown in Fig. 9c which can be consideredas full sintering of the structures [41] corresponding to Fig. 8b. How-ever, the alumina and hibonite grains (Fig. 9d and e) have becomerounded by increasing the sintering temperature from 1350 to 1550 °C.This results in less interlocking of the grains, which finally decreasesmechanical strength of the refractory castable. So, the increase ofstrength due to the decrease of porosity in 0.4TiO2 castable at 1550 °C isbalanced by the decrease of strength due to less interlocking of thegrains. Additionally, as shown in Fig. 9f, CaTiO3 phase deposited insome grain boundaries acts as a sintering aid and causes an integrated

structure [55,56].

3.5. Strength of refractory castables

It is well known that strength varies exponentially with porosity andincreases with decreasing the porosity [57]. Fig. 10 shows CCS, CMORand HMOR variations of pure and 0.4TiO2 castables as a function offiring temperature. The CSS (Fig. 10a) increase in 0.4TiO2 castable,despite its higher porosity (Fig. 5b), may be attributed to the dispersionof fine nano-titania particles in the matrix of the castable, resulting inhigher strength of 0.4TiO2 sample. Results show that after firing at1000 °C, the CCS of both refractory castables has been improved despitetheir higher porosity than their raw counterparts. In pure castable, thiscan be related to the formation of new phases on sub-micron particlesafter dehydration of hydrated phases in high-alumina cement, while in0.4TiO2 castable, this improvement can be related to the formation ofCa3Ti8Al12O37 phase which bonds to the matrix, hence increases theCCS. The higher strength of 0.4TiO2 castable in comparison with that ofpure castable at this temperature may be associated with its relativelylower porosity. According to the results, nano-titania results in en-hancement of CCS in all firing temperatures. Even at 1350 °C, the CCSof 0.4TiO2 castable increases despite increasing of its apparent porosity.It seems that the strength of the matrix enhances due to the formationof hibonite phase, resulting in improvement of the CCS.

Fig. 8. FESEM images obtained for 0.4TiO2 castable fired at 1350 °C.

A.R. Abbasian, N. Omidvar-Askary Ceramics International 45 (2019) 287–298

294

Hibonite has a significant effect on mechanical properties of therefractory castable. Although planar particles develop porosity in thestructure, their locking to the matrix results in the enhancement of thestrength at 1350 °C. It is reported that the formation of elongated grainsor secondary phases in the microstructure improves mechanical

properties due to the mechanisms known as deviation or bridging ofcracks [58]. It is reported that interlocked planar grains act as barriersfor crack propagation [38,39,59].

Although the porosity increases, the higher strength of the matrixcompensates for the stress concentration caused by this increment.

Fig. 9. FESEM images obtained for refractory castables fired at 1550 °C in different magnifications. (a-b) pure and (c-f) 0.4TiO2 castables.

A.R. Abbasian, N. Omidvar-Askary Ceramics International 45 (2019) 287–298

295

Significant enhancement of mechanical properties was noticed at1550 °C which may be arised from the formation and reinforcement ofceramic bonds as well as considerable decrease of porosity after sin-tering.

Variation of cold modulus of rupture (CMOR) for pure and 0.4TiO2

castables as a function of firing temperature is illustrated in Fig. 10b.The variation of CMOR is similar to that of CCS except for the samplefired at 1550 °C. So, the same reasons can be considered for variation ofCMOR.

In contrast to CCS, the CMOR of 0.4TiO2 castable after firing at1550 °C has been decreased in comparison with pure castable. This canbe attributed to structural changes. Failure in compressive test stemsfrom the growth and coalescence of defects following by the dissocia-tion and breakdown of the sample. So, contrary to bending strength,compressive strength depends mainly on total defects, but not on thelargest and sharpest crack [60]. As a result, pure castable has lowercompressive strength than 0.4TiO2 castable because of its relativelyhigher porosity. On the other hand, it seems that the microstructure of

0.4TiO2 castable is more sensitive to the loading condition in bendingtests. Further investigations are needed to find out the reasons of suchsensitivity.

The HMOR results obtained for pure and 0.4TiO2 castables fired at1350 °C and 1550 °C are shown in Fig. 10c. It can be seen that thepresence of nano-titania at both firing temperatures leads to improve-ment of HMOR. However, the HMOR of both refractory castables de-creases by increasing the firing temperature.

Hot bending strength of refractory castables at temperatures above1250 °C has been modeled by Adam-Gibbs theory [61] in which hotbending strength (σ) is proportional to viscosity at high temperatures(Eq. (21)):

= +σ A BT T C

lnln( / ) (21)

where T is thermomechanical temperature, and A, B and C are con-stants.

The presence of impurities, particularly in high-alumina refractorycastables, leads to the formation of more glassy phase and reduction ofviscosity at high temperatures [62]. On the other hand, formation ofglassy phase has significant effect on hot bending strength of high-alumina refractory castables [61]. In particular, at temperatures above1250 °C, hot bending strength obviously decreases due to the viscousflow of liquid phase. So it seems that at high temperatures, liquid phaseforms in both refractory castables and remains in the structure as aglassy phase which results in decreasing of measured high-temperaturestrength of refractory castables. XRD and microstructural observationsindicated the presence of CaTiO3 phase in grain boundaries of 0.4TiO2

castable which becomes mushy at 1350 °C and can be responsible forthe strength reduction [63].

Fine grain microsilica in alumina castables causes formation of lowmelting phases (such as CAS2 and C2AS) and considerable reduction ofhigh-temperature strength [64]. In pure castable, however, the high-temperature strength reduces although no additive (particularly mi-crosilica) is used in raw materials. There is also no evidence of lowmelting phases in this sample. So, it seems that little amounts of im-purities in raw materials lead to the formation of liquid phase at1550 °C which finally transforms to glassy phase. There is always littleamount of silica as impurity in alumina raw material which, along withother impurities, can produce liquid phases at high temperatures [63].It has been observed that the liquid tends to selectively wet low energyboundaries, while leaving other boundaries unwetted [63].

3.6. Thermal shock resistance

Results obtained for thermal shock resistance of pure and 0.4TiO2

castables fired at 1350 °C and 1550 °C are listed in Table 4. It can beseen from the table that both refractory castables fired at 1350 °C arenot failed after 30 cycles. Pure castable fired at 1550 °C also withstands30 cycles while 0.4TiO2 castable fired at 1550 °C has broken after 24cycles. In addition, both 0.4TiO2 castables (fired at 1350 and 1550 °C)have cracked after 5 cycles, while the cracking in both pure castablesoccurs after 9 cycles. By increasing the firing temperature of 0.4TiO2

castables from 1350 to 1550 °C, the thermal shock resistance of the

Fig. 10. The effect of 0.4 wt% nano-titania on CCS (a), CMOR (b) and HMOR(c) of alumina castables after firing at different temperatures.

Table 4The effect of 0.4 wt% nano-titania on thermal shock resistance of alumina ca-stables after firing at 1350 and 1550 °C.

Firing temperature (°C) 1350 1550

Sample pure 0.4TiO2 pure 0.4TiO2

Thermal shock resistancea 30 24 30 30Spalling resistanceb 5 5 9 9

a Number of cycles prior to failure.b Number of cycles before any cracking.

A.R. Abbasian, N. Omidvar-Askary Ceramics International 45 (2019) 287–298

296

samples has decreased which may be due to the presence of CaTiO3

phase in some grain boundaries which results in the failure of thesamples after 24 cycles.

4. Conclusions

This work aimed at studding the effect of nano-titania addition tohigh alumina castables on the microstructural characteristics and me-chanical properties. The obtained results can be summarized as follows:

1) Nano-titania addition decreased the flowability of the crude castableand consequently, decreased the density due to enhanced porosity.

2) Hibonite phase was formed at lower temperatures due to the mi-neralizing effect of nano-titania. The formation and transformationof Ca3Ti8Al12O37 after firing at 1000 °C is the main mechanism forthe development of planar hibonite in the nano-titania containingrefractory castable.

3) Hibonite formation at lower temperatures exerts extra pressure onalumina during sintering which finally enhances the mechanicalstrength of refractory castable containing nano-titania.

4) The cold compressive strength of the refractory castable containingnano-titania fired at 1550 °C was increased while its cold bendingstrength was decreased.

5) The hot bending strength and thermal shock resistance of the re-fractory castable containing nano-titania were decreased due to theformation of CaTiO3 phase deposited in alumina grain boundaries.

Acknowledgements

The authors would like to acknowledge the support of Nab CeramKhavar Co., Teheran, Iran, by providing the dispersant materials used inthis research work. Also, the authors would like to thank NuclearScience and Technology Research Institute (NSTRI) for supporting thiswork. The authors acknowledge Amir Reza Gardeshzadeh for proof-reading the manuscript.

References

[1] A. Surendranathan, An Introduction to Ceramics and Refractories, CRC Press, 2014.[2] S. Banerjee, Monolithic Refractories: A Comprehensive Handbook, World Scientific,

1998.[3] H. Walker, Handbook of Refractory Practice, Harbison-Walker Refractories

Company, 2005, pp. 7–8.[4] M. Rigaud, Corrosion of refractories and ceramics, in: R.W. Revie (Ed.), Uhlig's

Corrosion Handbook, Third edition, John Wiley & Sons, 2011, pp. 385–398.[5] H. Warlimont, Ceramics, in: W. Martienssen, H. Warlimont (Eds.), Springer

Handbook of Condensed Matter and Materials Data, Springer Berlin Heidelberg,Berlin, Heidelberg, 2005, pp. 431–476.

[6] F. Sarhadi, M. Shafiee Afarani, D. Mohebbi-Kalhori, M. Shayesteh, Fabrication ofalumina porous scaffolds with aligned oriented pores for bone tissue engineeringapplications, Appl. Phys. A 122 (2016) 390.

[7] M. Shafiee Afarani, A. Samimi, E. Bahadori Yekta, Synthesis of alumina granules byhigh shear mixer granulator: processing and sintering, Powder Technol. 237 (2013)32–40.

[8] P. Gehre, C. Wenzel, C.G. Aneziris, Investigation of shaped alumina based re-fractories used in slagging gasifiers, Ceram. Int. 37 (2011) 1701–1704.

[9] D. Vollath, Nanomaterials: An Introduction to Synthesis, Properties andApplication, Second ed., Wiley-Vch, 2013.

[10] A.M. Rashad, A comprehensive overview about the effect of nano-SiO2 on someproperties of traditional cementitious materials and alkali-activated fly ash, Constr.Build. Mater. 52 (2014) 437–464.

[11] R.D. dos Anjos, M.R. Ismael, I.R. de Oliveira, V.C. Pandolfelli, Workability andsetting parameters evaluation of colloidal silica bonded refractory suspensions,Ceram. Int. 34 (2008) 165–171.

[12] H. Yaghoubi, H. Sarpoolaky, F. Golestanifard, A. Souri, Influence of nano silica onproperties and microstructure of high alumina ultra-low cement refractory ca-stables, Iran. J. Mater. Sci. Eng. 9 (2012) 50–58.

[13] S. Mukhopadhyay, P. Pal, B. Nag, P. Jana, Influence of gel-derived nanocrystallinespinel in a high alumina castable: Part 2, Ceram. Int. 33 (2007) 175–186.

[14] V. Roungos, C.G. Aneziris, Improved thermal shock performance of Al2O3–C re-fractories due to nanoscaled additives, Ceram. Int. 38 (2012) 919–927.

[15] C. Gogtas, H.F. Lopez, K. Sobolev, Effect of nano-YSZ and nano-ZrO2 additions onthe strength and toughness behavior of self-flowing alumina castables, Ceram. Int.42 (2016) 1847–1855.

[16] S. Dudczig, D. Veres, C.G. Aneziris, E. Skiera, R.W. Steinbrech, Nano- and micro-metre additions of SiO2, ZrO2 and TiO2 in fine grained alumina refractory ceramicsfor improved thermal shock performance, Ceram. Int. 38 (2012) 2011–2019.

[17] E. Khaksar, M. Shafiee Afarani, A. Samimi, In situ solvothermal crystallization ofTiO2 nanostructure on alumina granules for photocatalytic wastewater treatment, J.Mater. Eng. Perform. 23 (2014) 92–100.

[18] P. Manivasakan, V. Rajendran, P.R. Rauta, B.B. Sahu, P. Sahu, B.K. Panda,S. Valiyaveettill, S. Jegadesan, Effect of TiO2 nanoparticles on properties of silicarefractory, J. Am. Ceram. Soc. 93 (2010) 2236–2243.

[19] C. Carlucci, F. Conciauro, B.F. Scremin, A.G. Antico, M. Muscogiuri, T. Sibillano,C. Giannini, E. Filippo, C. Lorusso, P.M. Congedo, G. Ciccarella, Properties of alu-minosilicate refractories with synthesized boron-modified TiO2 nanocrystals,Nanomater. Nanotechnol. 5 (2015) 1–7.

[20] N. Zakeri, S. Otroj, M.R. Saeri, Study of nano-titania addition on mechanicalstrength of mullite-bonded alumina-silicon carbide nano-composites, J. Adv. Mater.Eng. 34 (2015) 49–58.

[21] Á.N.C. Lima, D.K. Topolski, Nanomaterials for applications in refractory materials,in: C.P. Bergmann, aM.Jd Andrade (Eds.), Nanostructured Materials for EngineeringApplications, Springer-Verlag, 2011, pp. 133–139.

[22] S.H. Badiee, S. Otroj, Effect of nano-titania addition on the properties of high-alu-mina low-cement self-flowing refractory castables, Ceram.-Silikáty 55 (2011)319–325.

[23] J.E. Funk, D.R. Dinger, Derivation of the Dinger-Funk Particle Size DistributionEquation, Predictive Process Control of Crowded Particulate Suspensions: Appliedto Ceramic Manufacturing, Springer, US, Boston, MA, 1994, pp. 75–83.

[24] A. Abbasian, M. Rahimipour, H. Nouranian, A. Salardini, M. Amin, Effect of de-flocculants on microsilica containing ultra low cement Al2O3-SiC refractory ca-stable, Ind. Ceram. 30 (2010) 113–119.

[25] A. C20-00(2015), Standard test methods for apparent porosity, water absorption,apparent specific gravity, and bulk density of burned refractory brick and shapes byboiling water, ASTM International, West Conshohocken, PA, 2015.

[26] A. C133-97(2015), Standard test methods for cold crushing strength and modulus ofrupture of refractories, ASTM International, West Conshohocken, PA, 2015.

[27] D.E. 993-7, Methods of test for dense shaped refractory products – Part 7:Determination of modulus of rupture at elevated temperatures, DIN Standard, 1998.

[28] A. C179-14, Standard test method for drying and firing linear change of refractoryplastic and ramming mix specimens, ASTM International, West Conshohocken, PA,2014.

[29] D. 51068:2008, Testing of ceramic raw and basic materials - Determination of re-sistance to thermal shock – water quenching method for refractory bricks, DINStandard, 2008.

[30] A.P. Luz, V.C. Pandolfelli, Halting the calcium aluminate cement hydration process,Ceram. Int. 37 (2011) 3789–3793.

[31] Y. Wang, X. Li, B. Zhu, P. Chen, Microstructure evolution during the heating processand its effect on the elastic properties of CAC-bonded alumina castables, Ceram. Int.42 (2016) 11355–11362.

[32] K.T. Jacob, G. Rajitha, Gibbs energy of formation of Ca3Ti8Al12O37 and phase re-lations and chemical potentials in the system Al2O3-TiO2-CaO, J. Phase EquilibriaDiffus. 33 (2012) 293–302.

[33] T. Kolodiazhnyi, G. Annino, M. Spreitzer, T. Taniguchi, R. Freer, F. Azough,A. Panariello, W. Fitzpatrick, Development of Al2O3–TiO2 composite ceramics forhigh-power millimeter-wave applications, Acta Mater. 57 (2009) 3402–3409.

[34] I.-M. Low, Z. Oo, In situ diffraction study of self-recovery in aluminum titanate, J.Am. Ceram. Soc. 91 (2008) 1027–1029.

[35] M. Fuhrer, A. Hey, W.E. Lee, Microstructural evolution in self-forming spinel/cal-cium aluminate-bonded castable refractories, J. Eur. Ceram. Soc. 18 (1998)813–820.

[36] M.A.L. Braulio, V.C. Pandolfelli, Tailoring the microstructure of cement-bondedalumina–magnesia refractory castables, J. Am. Ceram. Soc. 93 (2010) 2981–2985.

[37] L. An, H.M. Chan, K.K. Soni, Control of calcium hexaluminate grain morphology inin-situ toughened ceramic composites, J. Mater. Sci. 31 (1996) 3223–3229.

[38] T. Nagaoka, S. Kanzaki, Y. Yamaoka, Mechanical properties of hot-pressed calciumhexaluminate ceramics, J. Mater. Sci. Lett. 9 (1990) 219–221.

[39] P.G. De. La Iglesia, O. García-Moreno, R. Torrecillas, J.L. Menéndez, Influence ofdifferent parameters on calcium hexaluminate reaction sintering by Spark Plasma,Ceram. Int. 38 (2012) 5325–5332.

[40] A.R. Abbasian, M.R. Rahimipour, Z. Hamnabard, Initial Sintering Kinetics ofLithium Meta Titanate at Constant Rates of Heating, Iranian Journal of MaterialsScience & Engineering, 10 (2013) 44–53.

[41] M.N. Rahaman, Ceramic Processing and Sintering, second ed., Marcel Dekker Inc.,New York, 2003.

[42] C. Falamaki, M.S. Afarani, A. Aghaie, Initial sintering stage pore growth mechanismapplied to the manufacture of ceramic membrane supports, J. Eur. Ceram. Soc. 24(2004) 2285–2292.

[43] C.-J. Wang, C.-Y. Huang, Effect of TiO2 addition on the sintering behavior, hardnessand fracture toughness of an ultrafine alumina, Mater. Sci. Eng.: A 492 (2008)306–310.

[44] D.S. Horn, G.L. Messing, Anisotropic grain growth in TiO2-doped alumina, Mater.Sci. Eng.: A 195 (1995) 169–178.

[45] R.D. Bagley, I.B. Cutler, D.L. Johnson, Effect of TiO2 on initial sintering of Al2O3, J.Am. Ceram. Soc. 53 (1970) 136–141.

[46] A.P. Luz, L.B. Consoni, C. Pagliosa, C.G. Aneziris, V.C. Pandolfelli, Sintering effectof calcium carbonate in high-alumina refractory castables, Ceram. Int. 44 (2018)10486–10497.

[47] S.J. Wu, L.C.D. Jonghe, M.N. Rahaman, Subcutectic densification and second-phaseformation in Al2O3-CaO, J. Am. Ceram. Soc. 68 (1985) 385–388.

A.R. Abbasian, N. Omidvar-Askary Ceramics International 45 (2019) 287–298

297

[48] R.J. Brook, Effect of TiO2 on the initial sintering of Al2O3, J. Am. Ceram. Soc. 55(1972) (114-114).

[49] K. Hamano, Effects of TiO2 on sintering of alumina ceramics, J. Ceram. Soc. Jpn. 94(1986) 505.

[50] T. Ikegami, K. Kotani, K. Eguchi, Some roles of MgO and TiO2 in densification of asinterable alumina, J. Am. Ceram. Soc. 70 (1987) 885–890.

[51] A.R. Abbasian, M.R. Rahimipour, Z. Hamnabard, Phase transformation duringsintering of Li2TiO3 nanocrystallites synthesised by hydrothermal method, MicroNano Lett. Inst. Eng. Technol. (2016) 822–824.

[52] M. Allibert, C. Chatillon, K. Jacob, R. Lourtau, Mass‐spectrometric and electro-chemical studies of thermodynamic properties of liquid and solid phases in thesystem CaO‐Al2O3, J. Am. Ceram. Soc. 64 (1981) 307–314.

[53] M. Braulio, M. Rigaud, A. Buhr, C. Parr, V. Pandolfelli, Spinel-containing alumina-based refractory castables, Ceram. Int. 37 (2011) 1705–1724.

[54] W. Yuan, Q. Zhu, C. Deng, H.-X. Zhu, Effects of SnO2 addition on the properties ofalumina-magnesia refractory castables, New J. Glass Ceram. 5 (2015) 1–7.

[55] D. Teodosiev, V. Petkov, R. Valov, J. Georgiev, M. Selecka, V. Ghica, Compositematerial on Al2O3 basis coated with vitreous carbon for medical needs, PowderMetall. Progress. 11 (2011) 326–329.

[56] L.S. Oliveira, D.X. Gouveia, M.A.S. Silva, A.S.B. Sombra, Study of the performanceof dielectric resonator antennas based on the matrix composite of Al2O3–CaTiO3,

Microw. Opt. Technol. Lett. 57 (2015) 963–969.[57] R.L. Coble, W.D. Kingery, Effect of porosity on physical properties of sintered alu-

mina, J. Am. Ceram. Soc. 39 (1956) 377–385.[58] Z.D.I. Sktani, A.Z.A. Azhar, M.M. Ratnam, Z.A. Ahmad, The influence of in-situ

formation of hibonite on the properties of zirconia toughened alumina (ZTA)composites, Ceram. Int. 40 (2014) 6211–6217.

[59] C. Domınguez, J. Chevalier, R. Torrecillas, L. Gremillard, G. Fantozzi,Thermomechanical properties and fracture mechanisms of calcium hexaluminate, J.Eur. Ceram. Soc. 21 (2001) 907–917.

[60] M. Bengisu, Engineering Ceramics, Springer, 2001.[61] W. Yuan, Q. Zhu, C. Deng, H. Zhu, The effect of the glass phase on the modulus of

rupture of high-alumina refractories, J. Ceram. Sci. Technol. 6 (2015) 215–220.[62] A. Terzić, L. P, A. Milutinović-Nikolić, Influence of the phase composition of re-

fractory materials on creep, Sci. Sinter. 38 (2006) 255–263.[63] C.A. Bateman, S.J. Bennison, M.P. Harmer, Mechanism for the role of magnesia in

the sintering of alumina containing small amounts of a liquid phase, J. Am. Ceram.Soc. 72 (1989) 1241–1244.

[64] Q. Wang, Y. Li, N. Xu, R. Xiang, S. Li, Microstructure and phase evolution of cement-bonded high-alumina refractory castables for porous purging plugs, J. Ceram. Sci.Technol. 9 (2018) 1–6.

A.R. Abbasian, N. Omidvar-Askary Ceramics International 45 (2019) 287–298

298