Journal of Advanced Ceramics Journal of Advanced Ceramics —A novel type of ceramic material for thermal —A novel type of ceramic material for thermal barrier coatings barrier coatings Ce 1  x Sm x O 2  x /2 —A novel type of ceramic material for thermal

at Springerlink.com Abstract: In this study, Ce 1  x Sm x O 2  x /2 ceramics were synthesized by sol–gel route and solid state sintering method. The phase structure was analyzed by X-ray diffraction (XRD), Fourier transform infrared (FTIR) spectroscopy, and Raman spectroscopy. The morphologies of the synthesized powders and the corresponding bulk samples were observed using scanning electron microscopy (SEM). Their thermal diffusivities and thermal expansion coefficients were measured by the laser-flash method and the pushing-rod method, respectively. Results show that pure Ce 1  x Sm x O 2  x /2 powders with single fluorite structure are synthesized successfully, and their microstructures of the corresponding bulk samples are very dense. With the increase of Sm 2 O 3 content, their thermal expansion coefficients decrease due to the higher electro-negativity of Sm 3+ ions as compared with that of Ce 4+ ions. Their thermal conductivities at 1000 ℃ lie in the range of 1.62–2.02 W/(m·K) due to the phonon scattering caused by the substituted atoms and oxygen vacancies. The Ce 1  x Sm x O 2  x /2 ceramics can be used as


Introduction
Thermal barrier coatings (TBCs) are advanced material system which is always used to protect underlying metallic components in turbine engines from damage caused by corrosive hot gas or high temperature [1][2][3].
In the system of thermal barrier coatings, the ceramic top coat shields the underlying material from heat, and the metallic bond protects the substrate against high temperature degradation and improves adherence of the top coat [4]. Because of the excellent thermophysical properties, 7-8 wt% yttria-stabilized zirconia (YSZ) ceramic has been widely employed as the top coat material by the current commercial thermal barrier coatings in high temperature turbine components. However, the thermal insulation ability and working lifetime of the YSZ thermal barrier coating can be injured severely for long-term application above 1200 ℃ due to its inherent phase transformation and enhanced sintering [5,6]. Therefore, it is very urgent to develop alternatives to YSZ for advanced TBC applications.
The excellent ceramic candidates for TBCs must possess a few important performances, such as low thermal conductivity, appropriate thermal expansion,  good phase stability at high temperature, low sintering rate, high melting point, chemical inertness, and good adherence to the metal substrate [7]. However, ceramic materials matching all the requirements are still very rare in light of the current standard. Now, low thermal conductivity and appropriate thermal expansion coefficient have been regarded as the primary selection criterions of the ceramic materials for TBC applications. In recent years, ceramic oxides with pyrochlore structure or defect fluorite structure have been widely studied [8][9][10][11]. Except for the A 2 B 2 O 7 -type (A = rare earth element, B = Zr, Ce, Hf, Sn) oxides [1][2][3][4][5][6][7][8][9][10], the cerium oxides with fluorite structure have recently attracted extensive attention due to a diversity of applications, such as conversion catalysts for selective hydrogenation of unsatured compounds, catalysts for three-way automobile exhaust systems, abrasives for chemical polishing slurries, gates for metal-oxide semiconductor devices, and luminescent materials for violet/blue fluorescence [12][13][14]. Now, the rare earth doped CeO 2 (RE 2 O 3 -CeO 2 ) have also been considered to be new materials for TBCs and solid oxide fuel cells due to the excellent electrical, mechanical, and thermophysical properties [15][16][17]. For example, Cao et al. [18] studied the thermal conductivity and thermal expansion coefficient of La 2 Ce 2 O 7 . Patwe et al. [19] reported the lattice thermal expansion of Gd 2 Ce x Zr 2x O 7 . Zhang et al. investigated the thermophysical properties of (Sm 1x Gd x ) 2 Ce 2 O 7 [20] and (Sm 1x Dy x ) 2 Ce 2 O 7 [21]. Zha et al. [22] found that the electrical conductivities of Ce 1x Gd x O 2x/2 (GDC) and Ce 1x Sm x O 2x/2 (SDC) at 700 ℃ are almost equal to the value of YSZ at 1000 ℃. Compared with pure doped ceria oxide (DCO) electrolyte, the DCO-chloride or DCO-carbonate composite electrolyte not only has much higher ionic conductivity, but also shows higher ionic transference number at intermediate temperature range [23,24], and these electrolytes also have good chemical stability [25].
Although thermophysical properties of a few rare earth stabilized CeO 2 have been reported by some researchers, the present reports about rare earth stabilized CeO 2 applications for TBCs are still not systemic. Therefore, investigation of the thermophysical properties of rare earth stabilized CeO 2 is still of notable significance. Previous works have discussed the electrical conductivity of Ce 1x Sm x O 2x/2 system, but did not deal with thermophysical properties of Ce 1x Sm x O 2x/2 oxides. In the present study, Ce 1x Sm x O 2x/2 oxides were synthesized by sol-gel method and pressureless sintering technology, and the phase composition and thermophysical properties of Ce 1x Sm x O 2x/2 (x = 0.1, 0.3, and 0.5) oxides were evaluated.

Experiment
In the current investigation, Sm 2 O 3 powders (Rare-Chem Hi-Tech Co. Ltd., Guangdong, China; purity ≥ 99.9%) and Ce(NO 3 )·6H 2 O (Zibo Huantuo Chemical Co. Ltd.; analytical pure) were chosen as the raw materials. Before weighting the raw powders, the samarium oxide powders were firstly calcined at 800 ℃ for 2 h to remove the adsorptive water and carbon dioxide in air, and then weighted samarium oxide powders were dissolved in diluting nitric acid. Ce(NO 3 )·6H 2 O was dissolved in distilled water and all solutions were mixed with constant stirring. Subsequently, the pH value of the mixed solution was adjusted to 6 by adding ammonia hydroxide drop wise. At the same time, ethylene glycol was put into the resultant solution, and the mole ratio of ethylene glycol to cerium was 1.8:1. The mixed solution was then continuously evaporated on a water bath till a viscous liquid was obtained, and the viscous liquid was heated at 130 ℃ using air oven till a porous solid mass was obtained. The obtained porous solid mass was ground in an agate mortar and activated at 800 ℃ for 2 h in a muffle oven. At the end, the achieved powders were isostatically cold pressed into pellets at 100 MPa, and the pellets were pressureless sintered at 1600 ℃ for 10 h in air to fabricate dense bulk samples.
An X-ray diffractometer (XRD, D8advance Bruker) with Ni filtered Cu K radiation (0.1542 nm) was used to analyze the phase structure of the synthesized powders and the corresponding bulk samples. The infrared spectra and Raman patterns of the synthesized powders were recorded by a Fourier transform infrared (FTIR) spectrometer (Nicolet 380) and a laser Raman spectrometer (Renishaw inVia-Reflex), respectively. A scanning electron microscope (SEM, Quanta-250, FEI) was selected to observe the microstructure of the synthesized powders and the corresponding bulk samples.
A high temperature dilatometer (Model Netzsch DIL 402C/7, Germany) was utilized to measure the thermal expansion coefficients (TECs) of bulk samples in the temperature range of 20-1200 ℃. The thermal diffusivity measurement () in 200-1000 ℃ was carried out using laser-flash method (Model LFA1000, Linseis, Germany) in an argon atmosphere. The specific heat capacity (c p ) from 20 to 1200 ℃ was calculated using Neumann-Kopp rule in light of the reference specific heat values of Sm 2 O 3 and CeO 2 [26]. The actual bulk density () of the sintered samples was measured by the Archimedes drainage method at room temperature. The thermal conductivity (k) of Ce 1x Sm x O 2x/2 oxides was achieved by Eq. (1), and the actual thermal conductivity (k 0 ) was computed using Eq.
(2) [21] in order to eliminate the influence of porosity () on thermal conductivity of bulk samples.
3 Results and discussion

1 Characterization about powders
The X-ray diffraction patterns of the synthesized powders are displayed in Fig. 1 Figure 3 reveals the FTIR spectra of Ce 1x Sm x O 2x/2 powders calcined at 800 ℃ for 2 h in the wave number range of 500-4000 cm 1 . Obviously, several typical infrared absorption bands can be found at about   570-590, 620-680, 1620-1640, and 3400-3500 cm 1 .
Another absorption band at about 1010 cm 1 can be observed in FTIR pattern of Ce 0.5 Sm 0.5 O 1.75 , which can be attributed to the little residual Sm 2 O 3 powders [27]. The absorption band near the 3400-3500 cm 1 is the evidence of water molecules contained in the powders [28], and the band located at 1620-1640 cm 1 represents another vibration of the water molecules [29]. The bands near 570-590 and 620-680 cm 1 are the typical absorption peaks of CeO 2 , and the variation of intensity and wave number of these two infrared bands can be attributed to the doping of Sm 2 O 3 [30].
The typical Raman patterns of Ce 1x Sm x O 2x/2 powders are plotted in Fig. 4 together with the data of micron-size CeO 2 . In the case of micron-size CeO 2 , the main peak at 461.49 cm 1 can be attributed to the F 2g Raman band from the space group Fm3m of cubic fluorite structure [31,32]. With the increasing content of Sm 2 O 3 , the width of the main Raman band enhances clearly, which means that a large number of oxygen vacancies are created [31,33]. Furthermore, the main peak of Ce 0.5 Sm 0. 5 [35], and the additional low intensity Raman bands around 250.65 and 375.22 cm 1 are usually assigned to the presence of extrinsic oxygen vacancies generated into the ceria lattice improving diffusion rate of bulk oxygen after samarium addition [36].
The micro-morphology of the Ce 1x Sm x O 2x/2 powders is displayed in Fig. 5. Obviously, the synthesized Ce 1x Sm x O 2x/2 powders exhibit a certain agglomeration. Ce 0.7 Sm 0.3 O 1.85 and Ce 0.5 Sm 0.5 O 1.75 have a size of about 50 nm; however, the average particle size of Ce 0.9 Sm 0.1 O 1.95 is only about 15 nm. The average particle size obtained from SEM is consistent with the analytical results of XRD.

2 Characterization of bulk samples
The XRD patterns of the sintered Ce 1x Sm x O 2x/2 samples are plotted in Fig. 6. Obviously, the X-ray diffraction patterns for bulk samples are very close to those displayed in Fig. 1, which means that the densified samples still remain the single fluorite structure. From Fig. 6, the weak peaks near 30° and 32° in the XRD pattern of Ce 0.5 Sm 0.5 O 1.75 disappear, which means that the residual Sm 2 O 3 also enters the lattice of CeO 2 in the procedure of sintering. It can be observed clearly from Fig. 7 that the grain size of these bulk ceramics is inhomogenous, and the average grain size is several micrometers. The obtained bulk samples have dense microstructure; however, some apparent pores can still be seen in Fig. 7. Their relative densities determined by actual density and theoretical density in sequence are 93.7%, 92.8%, and 95.6%; the grain  boundaries are very clean and no other phases can be found in these interfaces.

3 Thermal expansion coefficients
The dilatometric measurement data of Ce 1x Sm x O 2x/2 ceramics with calibration are presented in Fig. 8. Clearly, the typical linear thermal expansion property can be noted in the measuring temperature range of 20-1200 ℃, which also means that there is no phase transformation occurred in the measuring temperature range. In order to minimize the mismatch between the ceramic layer and the metal substrate, a high thermal expansion coefficient for ceramics of TBCs is required. The temperature dependence of the thermal expansion coefficient of Ce 1x Sm x O 2x/2 ceramics is exhibited in Fig. 9, together with the data of 8YSZ which were measured in the former research of the authors. As shown in Fig. 9, the thermal expansion shows an increasing temperature tendency owing to the increasing atomic spacing at high temperatures. From Fig. 9, the thermal expansion coefficient of Ce 1x Sm x O 2x/2 decreases gradually with increasing Sm 2 O 3 content, and Ce 0.5 Sm 0.5 O 1.75 has the lowest thermal expansion coefficient, which is still higher than that of 8YSZ. It is well known that the thermal expansion has close relationship with the ionic bond strength, and the ionic bond strength is affected by the electro-negativity of cations composing the crystal expressed as the following equation [

4 Thermal conductivity
Based on the specific heat values of CeO 2 and Sm 2 O 3 , the computed specific heat capacities of Ce 1x Sm x O 2x/2 ceramics according to the Neumann-Kopp rule are plotted in Fig. 10. Obviously, the specific heat capacity of Ce 1x Sm x O 2x/2 ceramics increases with the increasing temperature, and decreases with Sm 2 O 3 content at identical temperatures. The dependence of thermal diffusivity of Ce 1x Sm x O 2x/2 ceramics on temperature is shown in Fig. 11, and the data displayed in Fig. 11 are average values of every three measurements at identical temperature. It can be observed clearly that the thermal diffusivities decrease gradually with increasing temperature in the present measuring temperature range, which shows a typical phonon thermal conduction mechanism.
In light of the values of thermal diffusivity, density, and specific heat capacity of Ce 1x Sm x O 2x/2 ceramics, the final values of thermal conductivity are plotted in Fig. 12. It can be noted that the thermal conductivity is inversely proportional to the increasing temperature in the current temperature range, and the thermal conductivities of Ce 1x Sm x O 2x/2 ceramics decrease obviously with the increase of Sm 2 O 3 content. However, the thermal conductivity of Ce 0.5 Sm 0.5 O 1.75 is slightly higher than that of Ce 0.7 Sm 0.3 O 1.85 . According to the phonon thermal conduction theory, the thermal conductivity in electrical insulation solids is proportional to the mean free path of phonon. The