Sc

Sa

b

a

ARRA

KLPPP

1

tmccImosGtt

mthhsipm

ef

0d

Materials Chemistry and Physics 132 (2012) 309– 315

Contents lists available at SciVerse ScienceDirect

Materials Chemistry and Physics

j ourna l ho me pag e: www.elsev ier .com/ locate /matchemphys

ynthesis and characterization of lanthanum aluminate powders via a polymeromplexing plus combustion route

huai Lia,b,∗, Bill Bergmanb, Zhe Zhaoa,b

Department of Materials and Environmental Chemistry, Arrhenius Laboratory, Stockholm University, SE-10691, SwedenDepartment of Materials Science and Engineering, School of Industrial Engineering and Management, Royal Institute of Technology, SE-10044, Sweden

r t i c l e i n f o

rticle history:eceived 16 January 2011eceived in revised form 31 July 2011ccepted 13 November 2011

a b s t r a c t

Lanthanum aluminate powders were prepared by a simple polymer complexing plus combustion methodusing PVA or PEG as complexing agent and fuel. The influence of different polymers on phase purity,powder morphology and sintering performance were investigated. Trace amount impurity La2O3 exists

eywords:anthanum aluminateerovskiteVAEG

in the PEG powder, but it could be eliminated after high temperature sintering. The pure phase LaAlO3 canbe easily obtained in PVA powders calcined at 950 ◦C even severe aggregation always exists. PEG showsadvantages over PVA in terms of the densification and microstructure control during sintering process.The high relative density of 97.0% and homogeneous fine microstructure with grain size < 3 �m can beobtained in the PEG-derived sample sintered at 1600 ◦C for 5 h. To obtain better quality LaAlO3 powdersthrough combustion route, PEG is preferred over PVA.

. Introduction

Lanthanum aluminate (LaAlO3) has gained much attention inhe last decade for its dielectric properties and perfect lattice

atching to many materials with perovskite structure. LaAlO3eramics can be used as resonators in microwave filter appli-ation due to its favourable microwave dielectric properties [1].n addition, this material offers small lattice and thermal mis-

atches with perovskite materials, and thus it has emerged asne of the most favoured high-temperature superconducting (HTS)ubstrates or buffer layers for depositing ferroelectric films [2–4].enerally, the performance of polycrystalline LaAlO3 ceramics in

he above-mentioned applications is dependent on the microstruc-ure defects, e.g. residual porosity, impurity, etc.

As the basis of making LaAlO3 ceramic, LaAlO3 powder is nor-ally synthesized by solid state mixing and reaction method in

he temperature range of 1500–1700 ◦C [5,6]. Although the methodas the advantage of simplicity, it requires repeated and prolongedeat treatment and grinding process. The synthesized powders mayuffer the drawbacks of chemical inhomogeneity, low sinterabil-

ty with large particle size, and undesired impurities. In order torepare high performance LaAlO3 ceramics, several wet chemicalethods have been developed to overcome the drawbacks in solid

∗ Corresponding author at: Department of Energy Materials and Technology, Gen-ral Research Institute for Nonferrous Metals, 100088, China. Tel.: +86 10 82241238;ax: +86 10 82241294.

E-mail address: shuail@kth.se (S. Li).

254-0584/$ – see front matter © 2011 Elsevier B.V. All rights reserved.oi:10.1016/j.matchemphys.2011.11.019

© 2011 Elsevier B.V. All rights reserved.

state reaction route. In order to have small particle size for betterdensification, low heat treatment temperature is always regardedas the key point. Taspinar and Tas [7] reported a successful con-version of precursor to pure LaAlO3 powder at 750 ◦C for 16 husing the self propagating combustion, however heavily agglom-erated particles were obtained. Different fuels were reported toprepared LaAlO3 powder by using self propagating combustion,such as urea [7,8], sucrose [9], or citric acid [10], etc. Li [11] and Kuo[12] implemented the same co-precipitation method, where thereaction between diluted ammonia and nitrate salts leads to fineprecursors. The nanocrystalline LaAlO3 powders with a particle sizeless than 50 nm were obtained by calcination at 700 ◦C. Tian et al.[13] reported the preparation of spherical LaAlO3 nanoparticles viaa reverse microemulsion process. The well defined monodipseredspherical nanoparticles are about 50 nm after calcination at 800 ◦Cfor 2 h. Pure and well developed LaAlO3 cubic-shaped particleswith size < 3 �m were even successfully synthesized at temper-ature of 630 ◦C using a molten salt method, which so far is thelowest synthesis temperature [14]. Considering the application forceramic production, the highest relative density of LaAlO3 ceramicsobtained by conventional sintering is 98.7%, which is still not per-fect for the applications as substrates and microwave componentswhere zero porosity is highly preferred [12]. There is still a need tofind simple and reliable methods to prepare fine and homogeneouspowder with excellent sintering properties.

The polymer complexing plus combustion route is regarded as asimple and cost-effective gel method [15]. The method is character-ized with stabilization of cations in liquid through either chemicalbonding between functional groups in polymer chains and cations

3 ry and Physics 132 (2012) 309– 315

oshapppp[pppsLabfsnppscgtrabp

2

iA5

lPmTTasaa

ctic1t(

wTpi5

Xawi

80706050403020

-LaAlO3

Inte

nsity

2θ

a

80706050403020

-LaAlO3

-La2O3

Inte

nsity

2θ

b

10 S. Li et al. / Materials Chemist

r steric entrapment of cations into the polymer network. Thistabilization process prevents precipitation and thus ensures aomogeneous mixing of reagents at atomic level. Logically, it can benticipated that different polymer should lead to different powderroperties due to the different structure configuration and com-lexing efficiency. Various complex oxides have been successfullyrepared with this method using water soluble polymers, such asoly(vinyl alcohol) (PVA) [16,17] or poly(ethylene glycol) (PEG)18–20]. But the effect of polymer was rarely discussed for theurpose of getting high quality powders. Adak and Pramanik [21]repared LaAlO3 powder by using PVA as complexing agent. Phaseure LaAlO3 was obtained by the heat treatment of the precur-or powder at 650 ◦C. Kakihana et al. [22] also succeeded with pureaAlO3 powder with large surface area of 13–16 m2 g−1 at 750 ◦C by

similar combustion method based on the in situ polyesterificationetween citric acid and ethylene glycol. The lack of successful storyor 100% dense LaAlO3 ceramic indicates that the more detailedtudy for better powder through solution combustion route is stillecessary. As one important factor in combustion method, the com-lexing polymers must play important roles for the final powderroperties. Simply considering the available complexing sites iningle PVA and PEG chain, PVA is more efficient to combine moreations in a single molecule. But there is still no research to clearlyive a guideline how the different polymers will influence the syn-hesis and properties of the final complex oxides. In this study, weeport a comparative study of LaAlO3 powders by using PVA or PEGs complexing agent. The differences between powders preparedy these two complexing agents are compared in terms of phaseurity, morphology and sintering performance.

. Experimental

The powder was synthesized using the following start-ng materials, supplied by Alfa Aesar: La(NO3)3·6H2O (99.9%),l(NO3)3·9H2O (98%), PVA (98–99%, molecular weight of7,000–66,000) and PEG (molecular weight of 1500).

The PVA-metal nitrate precursor for LaAlO3 was prepared as fol-ows. The PVA solution was made by dissolving weighed amount ofVA in distilled water and stirring on a hot plate at 150 ◦C. Stoichio-etric amounts of nitrate salts were then added to the PVA solution.

he molar ratio of PVA monomers to the total metal cations was 2:1.he PVA-metal nitrate solution was stirred for 1 h at room temper-ture. Then the solution was kept at 80 ◦C to evaporate water whiletirring for 5 h. No precipitation was observed as water evaporated,nd finally a crisp gel was obtained. The obtained gel was groundednd calcined at 950 ◦C for 6 h (10 ◦C min−1 heating rate).

A similar preparation route was used for PEG-metal nitrate pre-ursor. Stoichiometric amounts of nitrate salts were dissolved inhe PEG solution. The PEG solution was prepared by adding PEGn distilled water, with 2:1 molar ratio of PEG monomers to metalations. The resulting PEG-metal nitrate solution was stirred for

hr at room temperature and then evaporated at 80 ◦C to obtainhe crisp gel. The crisp gel was calcined in air at 950 ◦C for 2 h10 ◦C min−1 heating rate).

The as-calcined powders were milled for 2 h using planetary millith zirconia balls (Ø5 mm) and isopropanol as the milling media.

hese powders were dry-pressed into pellets with a compactionressure of 200 MPa. Green pellets were sintered at temperatures

n the range of 1450–1600 ◦C for 5 h in air with a heating rate of◦C min−1.

The phases in the samples were investigated using powder

-ray diffraction (XRD) (X’Pert Pro, PANalytical) with Cu K�1 radi-tion (45 kV, 40 mA). Scans were taken in the 2� range of 20–80◦

ith a step size of 0.017◦. The morphology of powders was exam-ned by transmission electron microscope (TEM) (JEM–3010, JEOL).

Fig. 1. XRD patterns of 950 ◦C calcined LAO powders: (a) PVA powder; (b) PEGpowder.

Microstructure of the sintered pellets was studied by scanning elec-tron microscope (SEM) (JSM–7000, JEOL). To avoid the electroniccharging, pellets were carbon coated. The density of sintered spec-imen was measured using the Archimedes method.

3. Results and discussions

The X-ray diffraction patterns of the calcined powders areshown in Fig. 1. The PVA powder is essentially phase pure after cal-cination at 950 ◦C for 6 h. As shown in Fig. 1a, no secondary phasecan be found in this PVA powder, and all the diffraction peaks canbe indexed with perovskite LaAlO3 phase (JCPDS 31-0022) whichexhibits a rhombohedral structure with the space group of R3c[23]. The PEG sample shows mainly perovskite LaAlO3 phase aftercalcination. However, small amount of secondary phase, La2O3, isdetected as shown in Fig. 1b. The fraction of La2O3 is estimatedto be 6% by a semi quantitative calculation using the most inten-sive peaks of phases. It is believed that a homogeneous mixing ofcations in the precursor is of great importance to enhance homo-geneity and thus the phase purity in synthesized powders. The XRDresults illustrate a different stabilization capacity of metal cationsfor these two polymers. Fig. 2 shows the chemical structures of PVAand PEG. Apparently, due to the sufficient complexing sites alongthe molecule chain, PVA can guarantee a chemical homogeneityin the solution during the process, and thus results in phase pure

calcined powder. By contrary, each PEG molecule only providestwo ends as the chemical bonding sites for cations. The cationsstabilized by PEG in gel should be mainly attributed to the stericentrapment. The present result indicated that steric entrapment

S. Li et al. / Materials Chemistry and Physics 132 (2012) 309– 315 311

OHa

O

nH OH

b

in19TsAafit

Piddwtoitc

F

n

Fig. 2. Chemical structures of complexing polymers: (a) PVA; (b) PEG.

s less efficient than the normal chemical bonding for a homoge-eous cation gelation. However, all the sintered samples (1450 ◦C,500 ◦C and 1600 ◦C) from both PEG and PVA powders calcined at50 ◦C only show single LaAlO3 phase in the XRD patterns (Fig. 3).he minor impurity of La2O3 in the PEG powder can be further con-umed to form LaAlO3 during the high temperature heat treatment.lthough the PEG method suffers from the shortcoming of a tracemount of impurity after calcination, both methods can success-ully produce single phase sintered pellets. Therefore, the impurityn LaAlO3 powders does not influence the final sintered samples inerms of phase purity.

The size and morphology of the LaAlO3 powders derived fromVA and PEG are illustrated by both SEM and TEM images shownn Figs. 4 and 5. The two polymers lead to very different pow-er morphologies. The PVA powder shows a typical 2-D sheet-likeendrite structure consisted of fine particles connected each otherith no obvious boundary contrast (Figs. 4a and 5a). The size of

he dendrites is widely distributed in the range of several to tensf micrometers. The single-crystal nature of such dendrite sheet

s further demonstrated by the continuous lattice fringe betweenhe two selected particles (Fig. 5b). The measured spacing of therystallographic planes is 2.65 A, which corresponds to spacing of

807060504030202θ

1450oC

1600oCa -LaAlO3

1500oC

Inte

nsity

80706050403020

1450oC

2θ

1500oC

Inte

nsity

1600oCb -LaAlO3

ig. 3. XRD patterns of sintered LAO specimens: (a) PVA pellets; (b) PEG pellets.

Fig. 4. Typical morphology of powders calcined at 950 ◦C observed in SEM: (a) PVApowder; (b) PEG powder.

(1 1 0), 2.68 A, of perovskite LaAlO3. In contrast to PVA powder, thePEG powder has a rather regular shape and is weakly agglomer-ated as normal (Fig. 4b). The TEM image of PEG powder showsclear boundaries between particles (Fig. 5c). The dendrite struc-ture cannot be found in PEG powder after intended search underTEM. However, both PVG and PEG powders have an average particlesize of 100–150 nm if the single branch of dendrite is regarded asone particle. For powder compaction and sintering, the sheet-likedendrite structure potentially will introduce extra difficulties forpores’ removement.

Due to the high local temperature in the combustion process,sintered necks between particles can be anticipated [10], but thesingle-crystal dendrite structure cannot be simply attributed tosuch a high temperature characteristic. Actually, the local temper-atures in the selected PVA and PEG combustion systems shouldbe similar. The dendrite structure in PVA powder might be linkedwith other mechanisms. Generally, morphology of powder par-ticles is independent of chemical composition, and more linkedto the gel characteristics [13,24,25]. Costa et al. [25] observedthe different powder morphology of scandia-stabilized zirconiaby polymeric precursor and polyacrylamide techniques. The bigmolecular size and long chains of complexes of metal cations frompolymerization and esterification between citric acid and ethyleneglycol in the polymeric technique result in severe agglomerates

with irregular shapes. However, the polyacrylamide technique pro-duces soft agglomerates consisting of loss spherical particles, sincethe resultant complex has shorter chains than the one producedby polymeric technique. In this work, PVA provides a continuous

312 S. Li et al. / Materials Chemistry and

Fig. 5. Typical morphology of powders calcined at 950 ◦C observed in TEM: (a) PVAp

cpTciitocptne

owder; (b) enlarged area indicated in (a); (c) PEG powder.

hemical bonding network positions of hydroxyl groups along theolymer chain. The distance between cations is extremely small.herefore, it is possible to obtain a strong driving force to form thehain-shaped or dendrite structure when the precursor is calcinedn air. The intrinsic entangled conformation of PVA will further facil-tate the formation of a sheet-like network in the final powder. Onhe contrary, each PEG molecule only provides two hydroxyl groupsn the two ends for the chemical bonding between the polymer andations. Most of the cations are stabilized by a physical entrapment

rocess in the entangled PEG chains. The larger distance betweenhe cations offers less possibility to have a close contact betweenewly formed oxide crystallites during the heat treatment. Durant al. [26] reported a stronger agglomeration in PVA powder than

Physics 132 (2012) 309– 315

PEG powder in the synthesis of CeO2 when the same polymercomplexing route was used; the green pellets prepared from PVApowders were subjected to obvious density gradients during thedry-press process, which leads to the inferior densification.

In order to further investigate the powder performance in thedensification and microstructure evolution process during sinter-ing, ball milling was implemented to break up the agglomeratesin the as-calcined powders. The morphologies of PVA and PEGpowders after ball milling are shown in Fig. 6. After 2 h millingtreatment, most large-area sheet-like dendrite structures in thePVA powder disappeared. However, some rod-shaped particle andsmall dendrite structure still exist, which implies that the millingprocess is only efficient to break the big size sheet network butnot possible to totally eliminate the dendrite nature in PVA pow-der (Fig. 6a and b). Meanwhile, PEG powder particles are reducedto smaller size with the weak agglomerations (Fig. 6c and d). Aftermilling, the particle size in PEG powder is slightly smaller than thatin PVA powder.

To compare the sintering performance of these as-preparedpowders, green pellets were sintered at 1450 ◦C, 1500 ◦C and1600 ◦C for 5 h. The sintering behaviour of the two powders is com-pared in Fig. 7, where the relative density is plotted as a functionof sintering temperature. The theoretical density of 6.52 g cm−3 isused for calculating the relative density. The PVA samples showa sharp increase in relative density when the sintering tempera-ture increases from 1450 ◦C to 1500 ◦C. However, the densificationis rather limited after 1500 ◦C, with a minor increase when thetemperature goes up to 1600 ◦C, where the relative density onlyincreases from 96.5% to 96.8%. While for PEG samples, the rela-tive density increases with increasing sintering temperature andthis implies a continuous densification process. The highest rel-ative density of 97.0% is obtained for the PEG pellet sintered at1600 ◦C for 5 h. The density of LaAlO3 pellet produced in this studyis comparable to those best results reported by other researchers[11,12,27]. The relative densities of the PVA sample and PEG sam-ple sintered at 1600 ◦C are quite similar. This indicates that theexistence of small amount of La2O3 in the PEG powder is not sodetrimental for the densification process. Nevertheless, it is stillhighly preferred to remove such a minor impurity to improvethe stoichiometry precision. This is under further investigation byimproving the gel-forming process in the PEG method. Consider-ing the slightly smaller particle size in PEG powder, it is quite logicto expect a higher driving force and thus a higher relative den-sity of sintered samples at low temperature of 1450 ◦C. However,a reversed sequence for the densities of PVA and PEG 1500 ◦C sin-tered samples is observed in Fig. 7, which suggests that other factorsexcept for the starting particle size need to be considered for sucha change.

Fig. 8 shows the SEM images of the fracture surface of the sin-tered pellets. Inter-granular fracture mode dominates in all theLaAlO3 pellets. With the increase of sintering temperature, bothPVA and PEG samples show an increase of grain size. Although par-ticle size is comparable in PVA and PEG powders (Fig. 4), the grainsizes of the sintered PVA samples are much larger than those of PEGsamples. Even at a higher temperature of 1600 ◦C, the grain size inthe PEG pellet is still between 2 and 3 �m while the grain size in thePVA sample is larger than 6 �m. Intergranular pores can be foundin all samples, but the characteristics of the pores are rather dif-ferent in PEG and PVA samples. Firstly, a noticeable difference inthe pore size development can be easily identified in all the sam-ples shown in Fig. 8. For samples sintered at 1500 ◦C and 1600 ◦Crespectively, the pore size in PEG pellets is much smaller than that

in PVA pellets. There is no obvious change of the maximum poresize in the PEG pellets (indicated by the arrows in Fig. 8) when thesintering temperature increases from 1500 ◦C to 1600 ◦C, while thePVA pellets show definitive increase in both average pore size and

S. Li et al. / Materials Chemistry and Physics 132 (2012) 309– 315 313

g: (a)

matscptp

Fig. 6. SEM and TEM morphology of powders after 2 h ball millin

aximum pore size in Fig. 8. Secondly, for PVA samples, there is clear difference in the pore surface when the sintering tempera-ure increases from 1500 ◦C to 1600 ◦C, where the well developedurface fringes can be found in the 1600 ◦C samples. Such a fringe

an be only formed by strong surface diffusion during the sinteringrocess. Thus, the PVA samples undergo an extraordinary final sin-ering stage where surface diffusion is as active as other diffusionrocesses. If only the surface energy is considered, the concaved

16001500

94

96

98

Rel

ativ

e de

nsity

, %

Sintering temperature, oC

PVA PEG

Fig. 7. Change of relative density as a function of sintering temperature.

PVA powder; (b) PVA powder; (c) PEG powder; (d) PEG powder.

grain surface around the pores found in PVA samples should intro-duce a tendency of pore shrinkage. But the further pore growthindicates that more factors need to be considered. Based on thesolid state sintering theory, the numbers of grains around the porecan be expressed as a function of pore to grain size ratio. The poregrowth will occur when the pore size is lager than a critical sizeif the dihedral angle is kept constant. The large pore size formedat lower temperatures should be the reason for the pore growthat 1600 ◦C in PVA samples. Once the surface diffusion dominatesthe mass transport in the sintering process, there will be no furtherdensification and only substantial grain growth happens in poly-crystalline solids. The surface diffusion may dominate the masstransport in the PVA sample during sintering process at temper-atures > 1500 ◦C. In Fig. 7, it can be seen that only trace increaseof the relative density from 1500 ◦C is achieved for the PVA sam-ples, while the density of PEG samples shows a continuous increasewith the sintering temperature. Furthermore, the big difference ingrain size between PEG and PVA pellets should be attributed to theactive surface diffusion in the PVA samples sintered at high temper-atures. Such a difference in the diffusion mechanism is highly linkedwith the morphology of the starting powder. PVA powder showsdendrite structure which will introduce bigger pore size in greencompact and also will lead to difficulties in removing big poresthrough grain boundary diffusion during sintering. Therefore, the

pore size in PVA pellets is much larger than that in PEG pellets, evena similar density was obtained at 1600 ◦C. This further indicates thatPEG pellet can be obtained a higher density if the starting powdercan be improved in the future. Although phase purity in the PEG

314 S. Li et al. / Materials Chemistry and Physics 132 (2012) 309– 315

et, 150

ppiiipc

itsp

4

tfapdt9pstl

[

[[

Fig. 8. Microstructure of the fracture surface of sintered pellets: (a) PVA pell

owder synthesis is of worse performance, the normal sinteringrocess without active surface diffusion process in the final sinter-

ng stage gives PEG powder strong advantages over PVA powder,f the densification and grain size control is considered. Therefore,t can be concluded that the combustion route using PEG as com-lexing agent can provide better densification and microstructureontrol.

Considering the similarity between the different complex oxidesn combustion process, where the agglomeration and particle struc-ure are not yet seriously considered, the results obtained in thistudy indicate that polymer selection can be critical for makingowders suitable for the final applications.

. Conclusions

A simple polymer complexing and combustion method is usedo prepare nanocrystalline LaAlO3 powders. The influence of dif-erent polymers on the phase purity, morphology of productnd sintering performance is investigated using different com-lexing agents, namely PVA and PEG. Small amount of La2O3 isetected in the PEG powder, but it could be eliminated after sin-ering. Although the PVA powder is phase pure after calcination at50 ◦C, it contains more seriously aggregated particles. The mor-

hology of the starting powder has dramatic influence on theintering performance in aspects of densification and microstruc-ure control. PEG shows strong advantages over PVA due to muchess degree of agglomeration state in the synthesized powder.

[[[[[

0 ◦C; (b) PVA pellet, 1600 ◦C; (c) PEG pellet, 1500 ◦C; (d) PEG pellet, 1600 ◦C.

High density of 97.0% and homogeneous fine microstructure withgrain size < 3 �m can be obtained for the PEG pellet after sin-tering at 1600 ◦C for 5 h. It is more important to get powderswith desirable morphology and less agglomeration when dif-ferent polymers can be selected for the synthesis of complexoxides.

References

[1] I.M. Reaney, D. Iddle, J. Am. Ceram. Soc. 89 (2006) 2063.[2] B.E. Park, H. Ishiwara, Appl. Phys. Lett. 82 (2003) 1197.[3] J.A. Smith, M.J. Cima, N. Sonnenberg, IEEE Trans. Appl. Supercond. 9 (1999)

1531.[4] J.M. Phillips, J. Appl. Phys. 79 (1996) 1829.[5] J.A. Kilner, P. Barrow, R.J. Brook, M.J. Norgett, J. Powder Sources 3 (1978)

67.[6] T.L. Nguyen, M. Dokiya, S.R. Wang, H. Tagawa, T. Hashimoto, Solid State Ionics

130 (2000) 229.[7] E. Taspinar, A.C. Tas, J. Am. Ceram. Soc. 81 (1997) 133.[8] V. Singha, S. Watanabe, T.K.G. Rao, J.F.D. Chubaci, H. Kwak, Solid State Sci. 13

(2011) 66.[9] Z. Negahdaria, A. Saberia, M. Willert-Porada, J. Alloys Compd. 485 (2009)

367.10] J. Chandradass, M. Balasubramanian, K.H. Kim, Mater. Man. Proc. 25 (2011)

1449.11] W. Li, M.W. Zhuo, J.L. Shi, Mater. Lett. 58 (2004) 365.12] C.L. Kuo, C.L. Wang, T.Y. Chen, G.J. Chen, I.M. Hung, C.J. Shih, K.Z. Fung, J. Alloys

Compd. 440 (2007) 367.

13] Z. Tian, W. Huang, Y. Liang, Ceram. Int. 35 (2009) 661.14] Z.S. Li, S.W. Zhang, W.E. Lee, J. Euro. Ceram. Soc. 27 (2007) 3201.15] M. Kakihana, J. Sol–Gel Sci. Technol. 6 (1996) 7.16] M.A. Gulgun, M.H. Nguyen, W.M. Kriven, J. Am. Ceram. Soc. 82 (1999) 556.17] S.J. Lee, A.B. Elizabeth, W.M. Kriven, J. Am. Ceram. Soc. 82 (1999) 2049.

ry and

[[

[

[[

[[

S. Li et al. / Materials Chemist

18] W.S. Lee, T. Isobe, M. Senna, Adv. Powder Technol. 13 (2002) 43.19] H.B. Park, H.J. Kweon, Y.S. Hong, S.J. Kim, K. Kim, J. Mater. Sci. 32 (1997)

57.20] X. Li, H.B. Zhang, F. Chi, S.J. Li, B.K. Xu, M.Y. Zhao, Mater. Sci. Eng. B18 (1993)

209.21] A.K. Adak, P. Pramanik, Mater. Lett. 30 (1997) 269.22] M. Kakihana, T. Okubo, J. Alloys Compd. 266 (1998) 129.

[[

[

Physics 132 (2012) 309– 315 315

23] S. Bueble, K. Knorr, E. Brecht, W.W. Schmahl, Surf. Sci. 400 (1998) 345.24] A. Sin, P. Odier, Adv. Mater. 12 (2000) 649.

25] G.C.C. Costa, R. Muccillo, J. Alloys Compd. 503 (2010) 474.26] P. Duran, F. Capel, D. Gutierrez, J. Tartaj, C. Moure, J. Euro. Ceram. Soc. 22 (2002)

1711.27] R. Elsebrock, C. Makovicka, P. Meuffels, R. Waser, J. Electroceram. 10 (2003)

193.