CHAPTER INTRODUCTION

Transcription

1 CHAPTER 4 EFFECT OF ALKALINE EARTH METAL AND TRANSITION METAL DOPANTS ON THE STRUCTURAL, OPTICAL AND ELECTRONIC PROPERTIES OF YTTRIUM STABILIZED ZIRCONIA NANOPARTICLES 4.1 INTRODUCTION The tetragonal and cubic phases could exist when doped with alkaline earth metal oxides (MgO, CaO, etc.) or by transition metal oxides (Y 2 O 3, etc.) or by rare earth metal oxides (CeO 2, etc.) at room temperature (Bechepeche et al. 1999). According to Iwasaki et al. (1992), the mechanical strength of stabilized zirconia is improved without lowering the ion conductivity by providing a solid electrolyte comprising stabilized zirconia and a metal oxide dispersed within grains or grain boundaries of stabilized zirconia. Stabilization of zirconia in particular phase is recognized as the stabilized zirconia. It can be obtained by addition of about 5-10 mol%, particularly about 8 mol% of a stabilizer such as yttrium, cerium, calcium or magnesium. With the stabilized zirconia, the first group metal oxide can be added which do not generally form solid solution with the host and the second group metal oxide that forms the solid solution can also be added as well

2 without changing their physical properties and chemical stability. For example, alumina, chromia and mullite can be easily added to stabilized zirconia which forms as a composite oxide and a second group metal oxide such as magnesia, barium oxide and calcia could form solid solution with stabilized zirconia. When the second group oxide is used as the dopant, the metal oxide may be in the form of separate grains which are dispersed within the stabilized zirconia grains depending upon the preparation conditions. Nevertheless, such a metal oxide may be partially solid dissolved, particularly around the stabilized zirconia grains. The limitation of the metal oxide in mol% in the stabilized zirconia is preferably up to about 30%, more preferably 0.01 to 20% and specifically 0.1 to 5% (Iwasaki et al. 1992). Addition of a lower concentration of metal oxide improves the strength of the stabilized zirconia whereas a greater concentration of metal oxide in stabilized zirconia lowers the ionic conductivity significantly. Manganese ion stabilizes zirconia in the cubic phase and moreover it delays the cubic to tetragonal (ct) phase transformation responsible for slow conductivity decay (>1000 h) even at high temperature of about C (Herle and Vasquez 2004). Theoretical (Ostanin et al. 2007, Jia et al. 2009) and experimental research (Clavel et al. 2008, Yu et al. 2008, Zippel et al. 2010, Srivastava et al. 2011) in Mn doped YSZ is however limited and is mostly oriented in exploiting the magnetic properties especially room temperature ferromagnetism for spintronics application. For Mn-stabilized zirconia special magnetic properties were recently predicted depending on the number of oxygen vacancies (Ostanin et al. 2007). Cu/ZrO 2 materials were proposed by Velu et al. (2000), Liu et al. (2002), Matter et al. (2004) and Fisher et al. (1999) as catalysts in the process of oxidative steam reforming of methanol. However, to the best of the author knowledge there are no reports

3 available analyzing the effect of these transition metal ions with yttria stabilized zirconia. Here, in the present study it is believed that the well stabilized cubic YSZ can be formed by adding these dopants. Ni doped YSZ is widely used as anode material in SOFC (Park et al., 2011). A number of researchers have addressed the interaction of YSZ with small additions of NiO, most of the research work has been focused on phase stabilization (Kondo et al., 2003), aging (Mori et al. 2003) and electrical properties (Kondo et al. 2003, Linderoth et al. 2001, Herle and Vasquez 2004). It has been reported by Kondo et al. (2003) that addition of nickel enhances the ionic conductivity of zirconia. Moreover, Herle and Vasquez (2004) have proved that lower concentration of NiO is advantageous as it can significantly reduce the sintering temperature. ZnO is well known not only as a semiconductor but also as a probable oxygen-ion conductor due to enrichment of oxygen vacancies at higher temperature. It has been reported that ZnO could be used as an effective sintering aid and an optimal scavenger for grain boundary in yttria-stabilized zirconia. Small addition of ZnO is found to be effective in reducing the sintering temperature and promoting the densification rate of the ceramics. The 5.0 wt% ZnO-doped YSZ has ~96% relative density, as compared to ~89% relative density for the undoped sample. The total conductivity of 8YSZ was evidently increased by doping small amount of ZnO. At intermediate temperature (~300 C), the maximum enhancement of grain boundary conductivity was observed with 5.0 wt% ZnO dopant (Liu and Lao 2006). In this case also there are no more related reports available. Here, in the present study the 5 mol% Zn has been used to stabilize the cubic phase of YSZ.

4 For SOFC application it is important to stabilize the cubic structure, in particularly, at nanoscale. In recent years, research has been directed towards stabilized nano-zirconia at room temperature by various dopants inorder to obtain phase stability over time as well as to improve the ionic conductivity by increasing the oxygen vacancies. This demands a clear understanding of the influence of various dopants on the structural and electronic band structure of the stabilized zirconia specifically at nanoscale as their properties are entirely different from the bulk counter-part. However there is lack of research directed on such concept of material science which will pay way for better understanding of the factors that impulse the phase stabilization and ionic conductivity. This chapter clearly depicts the influence of various dopants on the structural, optical and electronic properties of yttria stabilized zirconia and effective stabilization of cubic phase at room temperature by alkali earth metal oxide and transition metal oxide dopants. 4.2 EXPERIMENTAL SYNTHESIS AR grade Zirconium oxychloride (Himedia) and Yttrium nitrate (Himedia) were used as the precursor. The precursors used for the doping were AR grade Barium nitrate (Merck), Magnesium nitrate (Merck), Manganese nitrate (Merck), Nickel nitrate (Merck), Copper nitrate (Merck) and Zinc nitrate (Merck). Appropriate amount of the precursor salts in the molar ratio Zr:Y:M::0.87:0.08:0.05 (where 5 mol% of M refers to the cationic dopant added) were dissolved in required amount of double distilled water by stirring. Oxalic acid was added drop-wise to the above mixed solution under vigorous stirring until the solution turns pale white gel in the case of Ba, Mg and Zn as dopant whereas a light brownish gel, light greenish gel and bluish

5 gel in the case of Mn, Ni and Cu respectively as dopant. Stirring was continued till the gel gets dissolved and forms a solution. Precipitation of the precursor solution by drop-wise addition of liquid ammonia was done to maintain the ph of the solution at 8. Ultrasonication was further continued for 45 min and the temperature of the solution was maintained at room temperature by immersing it in a cold water bath. The precipitate was aged for 12 h at room temperature and then the supernatant water was decanted. The obtained precipitate was again subjected to ultrasonication for another 15 min by maintaining the temperature at room temperature by using a cold water bath. The precipitate was then washed with water and ethanol several times and dried in hot air oven at 100 C. The dried precipitate was ground to fine powder using agate mortar and pestle. All the synthesized powders were calcined at 700 C for 2 h. The powders doped with alkali metal were subjected to prolonged heat treatment for 8 h at 700 C. For convenience the samples are named as Mg-YSZ, Ba-YSZ, Mn-YSZ, Ni-YSZ, Cu-YSZ and Zn-YSZ. 4.3 RESULTS AND DISCUSSION Structural Characterization The identification of phases in the synthesized zirconia powder sample was carried out by X-ray powder diffraction studies using Cu K radiation (1.54 Å), and comparing the interplanar distances and intensity values with those of the corresponding standard peaks using JCPDS files. The crystallite size of the nano-sized zirconia was evaluated from X-ray powder

6 diffraction data using Scherrer formula (Cullity and Stock 2001) for the most intense peak (1 1 1) plane of zirconia crystal using the equation 2.1. It is evident from the XRD patterns shown in Figure 4.1 that YSZ shows poor crystallinity when doped with alkali earth metal ions such as Mg and Ba. In order to obtain well crystalline features the calcination time was further extended to 8 h and there the well defined high intensity diffraction peaks can be seen as in Figure 4.2. The observed peak position and their relative intensity are indexed to cubic phase of zirconia and shows increase in crystallite size as well. It can be concluded that when doping with Ba in YSZ, there is larger crystallinity delay that requires prolonged calcination time when compared to pure YSZ and transition metal doped YSZ as well. This may be due to the large ionic radius of Ba 2+ (142 pm) which suppress the grain growth rate in shorter calcination time. Therefore, it requires more time to get complete crystallization. As a result of the metal ion incorporation, strain is induced in the lattice due to the difference in their ionic radii which reduces the particle size as observed by the peak broadening and poor crystallinity indicating quantum confinement. At further continuation of heat treatment, the lattice strain gets relieved to some extent thereby improving the crystallinity of the material.

7 Figure 4.1 XRD patterns of (a) Mg-YSZ and (b) Ba-YSZ calcined at 700 C for 2 h Figure 4.2 XRD patterns of (a) Mg-YSZ and (b) Ba-YSZ calcined at 700 C for 8 h

8 Figure 4.3 XRD patterns of (a) YSZ (b) Mn-YSZ (c) Ni-YSZ (d) CuYSZ and (e) Zn-YSZ calcined at 700 C for 2 h Figure 4.4 Extended view of high intensity (111) plane in the XRD patterns of (a) YSZ (b) Mn-YSZ (c) Ni-YSZ (d) Cu-YSZ and (e) Zn-YSZ calcined at 700 C for 2 h

9 For the 5 mol% transition metal doped YSZ all the diffraction peaks could be indexed to cubic symmetry with a peak shift to higher 2 values. The XRD patterns of YSZ doped with 5 mol% of transition metal oxide shows better crystallinity compared to alkali earth metal doped YSZ when calcined for shorter time of 2 h. The Ni-YSZ shows better crystallinity compared to other samples. The crystallite size and crystallinity of the material depends on the dopant added. The peak broadening observed is due to the non uniform strain which confirms the presence of nanometric YSZ. It can be noted that the phase formed is predominantly cubic due to incorporation of dopants. Moreover, the diffraction planes of the cubic and tetragonal phase are almost nearer to each other and hence clear determination of diffraction planes in the case of broader peaks are difficult. Therefore, it cannot be suggested that the material is purely cubic. It may be a combination of cubic and tetragonal phases (pseudocubic phase). This is clearly shown in Figure 4.3 where the standard JCPDS data and for cubic (red lines) and tetragonal (black lines) phase respectively are given for comparison with the experimental data. The intensity of the diffraction peak decreases with increasing vacancy concentrations which are consistent with the PL results. The (111) peak broadens with increasing stacking fault and the location of the (111) peak is shifted to higher angles. In the extremely small crystals, the greatest shift occurs when the stacking fault is located in the center of the cube (Makinson et al. 2000). The lattice constant of stabilized ZrO 2 varies depending on the amount and the nature of the stabilizing element (Schubert et al. 2009). As the particle size is very small the fraction of atoms in the surface layer is large and

10 these atoms are less strongly bonded to their neighbours than atoms in the bulk. As a result, the unit-cell dimensions at the surface are larger than in the core. The unit-cell parameters are an average measurement, based on the average interatomic distances. Thus, smaller particle size leads to larger average interatomic distances and hence the unit-cell parameter would be larger compared to the bulk. According to Esposito et al. (2011), the structural properties depend markedly on the precursor and on the synthesis procedure. Esposito et al. (2010) has proposed the concept of crystallization delay in the Cu doped Zirconia. This is caused by the incorporation of dopant metal ions into the ZrO 2 lattice i.e. they occupy the position of Zr 4+ ions. Based on their assumption, the same hypothesis was followed in the present case. With the ionic radius of the dopant the variation in lattice parameter was observed as summarized in the Table 4.1.

11 Table 4.1 Variation of lattice parameters of zirconia with respect to various metal ion dopants Sample name c-zro 2 JCPDS c-zro 2 Tsunekawa al., (2003) et c-zro 2 Manna et al., (2010) Mean Crystallite size D (nm) Strain x 10-4 Dislocation density x lines/m 2 Lattice constant a (Å) Volume V (Å 3 ) bulk ~ YSZ Mg-YSZ (700 C, 2 h) Mg-YSZ (700 C, 8 h) Ba-YSZ (700 C, 2 h) Ba-YSZ (700 C, 8 h) Mn-YSZ (700 C, 2 h) Ni-YSZ (700 C, 2 h) Cu-YSZ (700 C, 2 h) Zn-YSZ (700 C, 2 h)

12 The lowering of the lattice parameters values of the samples compared to those of pure zirconia were an unquestionable proof of the dopant ion zirconium replacement. The decrease in unit cell parameters of cubic phase is supported by the shift in 2 to higher values. The ionic radii of Mn 2+, Ni 2+, Cu 2+ and Zn 2+ are 67, 44, 57 and 60 pm which are smaller than that of Zr 4+ with ionic radii 84 pm (Shannon 1974). Therefore substitution of transition metal dopant ion into zirconia lattice leads to lattice volume shrinkage. The metal ions used as dopants are generally of smaller size and lower valence (than Zr 4+ ions) that may result in a decrease in the unit cell volume and generation of positive holes with lattice defects (oxygen vacancies). The typical FTIR spectra of alkali earth metal and transition metal doped Yttria stabilized Zirconia are shown in Figure 4.5 and Figure 4.6 respectively which are considered as the finger print of the material. The broad peak exhibited in the FTIR Spectra in the range cm -1 corresponds to the stretching vibration of physically adsorbed OH with the metal ion on the surface (Truffault et al. 2010). The band located around 1400 cm -1 and 1625 cm -1 represents bending vibration of water molecules (Phoka et al. 2009, Truffault et al. 2010, Wang et al. 2012). The peak centered around cm -1 can be attributed to the coupling effect of stretching and bending vibration of OH groups (Sarkar et al. 2007). The stretching frequency of metal-oxygen (M-O) band is found below 600 cm -1. It has been predicted by Srinivasan et al. (2010) that the M-O absorption bands become broader in the FTIR spectra as the particle size decreases due the enhanced surface effects.

13 Figure 4.5 FTIR spectra of (a) Mg-YSZ and (b) Ba-YSZ calcined at 700 C for 8 h Figure 4.6 FTIR spectra of (a) Mn-YSZ (b) Ni-YSZ (c) Cu-YSZ and (d) Zn-YSZ calcined at 700 C for 2 h

14 4.3.2 Optical and Electronic Properties Figure 4.7 and 4.8 shows the diffuse reflectance spectra of alkali earth metal ion doped yttrium stabilized zirconia and transition metal ion doped yttrium stabilized zirconia respectively. The energy band gap is determined from UV Diffuse reflectance data by transforming it into a function of reflectance as proposed by Kubelka-Munk. The Kubelka-Munk plot for determining band gap energy is shown is Figure 4.9 and 4.10 for the alkali earth metal doped yttrium stabilized zirconia and transition metal doped yttrium stabilized zirconia respectively. KM plot is plotted with the n th power of product of function of reflectance F(R) and photonic energy (E g = h) against the photonic energy. Since Zirconia is considered as a direct band gap semiconductor the value of n is taken as 2 for allowed transitions (Joy et al., 2012). The energy band gap is found out by extrapolating the linear portion of the graph to the X-axis. The formation of defects, such as oxygen vacancies, lead to reflectance at lower energies due to the presence of donor levels located inside the forbidden band (Manna et al. 2010). The oxygen vacancies in ZrO 2 crystals can induce the formation of new energy levels in the band gap region. The observed red-shift of the cut-off wavelength could be due to oxygen vacancy that is in fact responsible for lowering the band gap energy. Among all the transition metal doped YSZ, Mn-YSZ shows poor reflectance which is due to the lower optical features of Mn by nature. Comparing the band gap energy of the metal ion doped YSZ tabulated in Table 4.2, Ba-YSZ and Zn- YSZ show comparatively high energy band gap which is due to quantum

15 confinement effect as supported by the PL results and smaller nanosize of the crystallites as deduced from the XRD results. Figure 4.7 Diffuse Reflectance Spectra of (a) Mg-YSZ and (b) Ba-YSZ calcined at 700 C for 2 h Figure 4.8 Diffuse Reflectance Spectra of (a) Mn-YSZ (b) Ni-YSZ (c) Cu- YSZ (d) Zn-YSZ and (e) YSZ calcined at 700 C for 2 h

16 Figure 4.9 KM plot of (a) Mg-YSZ and (b) Ba-YSZ calcined at 700 C for 2h Table 4.2 Variation of energy band gap of zirconia with respect to alkali metal and transition metal ion dopants Sample Name Energy band gap E g (ev) YSZ 5.1 Mg-YSZ 4.85 Ba-YSZ 5.08 Mn-YSZ 4.85 Ni-YSZ 4.93 Cu-YSZ 5.00 Zn-YSZ 5.06

17 Figure 4.10 KM plot of (a) Mn-YSZ (b) Ni-YSZ (c) Cu-YSZ (d) Zn-YSZ and (e) YSZ calcined at 700 C for 2 h

18 Figure 4.11 Photoluminescence Spectra of (a) YSZ (b) Mg-YSZ and (c) Ba-YSZ Figure 4.12 Photoluminescence Spectra of (a) Mn-YSZ, (b) Ni-YSZ, (c) Cu-YSZ and (d) Zn-YSZ

19 The origin of photoluminescence in zirconia is discussed in detail in Section 3.3. All the undoped and doped YSZ show almost similar PL emission. They show three prominent peaks in the UV, violet and green region of the electro-magnetic spectrum with emission wavelength centered around 361 nm, 400 nm and 493 nm corresponding to excitation at 293 nm as shown in Figure 4.11 and The intense zirconia emission peak at 361 nm in the ZrO 2 calcined in air can be due to the ionized oxygen vacancies (F+ and F centers) from the conduction band (Joy et al. 2011). Here UV emission can arise as a result of the radiative recombination of a photogenerated hole with an electron occupying the oxygen vacancy. It is reported by Lai et al. (2000) that there is formation of surface traps in the ZrO 2 leading to electron transitions, even with a small amount of excitation energy. The PL emission band observed at 400 nm might be due to the transitions from the surface trap states in the conduction band to lower energy levels close to the valance band. This is in agreement with the results reported in the literature (Neppolian et al. 2007, Lai et al. 2000). A very broad emission peak in the green region is seen in the case of BaYSZ and ZnYSZ which is due to quantum confinement effect of reduced particle size and the presence of more oxygen vacancy in the material. The broad band may be due to the inhomogeneous broadening from a distribution of surface or defect states because of the narrow particle size distribution (Joo et al. 2003). The presence of a certain type of dopant may inhibit the luminescence of other centers; here Zn-YSZ acts as a "killer".

20 4.3.3 Morphological Characterization Figure 4.13 FESEM image and corresponding EDAX spectrum of (a) Mg- YSZ and (b) Ba-YSZ The FESEM images of the transition metal doped YSZ (Figure 4.14) reveal agglomerated uniform sized spherical shaped nanoparticles whereas alkaline earth metal doped YSZ (Figure 4.13) exhibit irregular shape and non-uniform particle size distribution which is due to the low crystallinity. Further, the severe agglomerations of the particles are due to the change in surface energy considerably by the addition of dopants. The corresponding EDAX spectrum shows the composition of the elements present in the sample. The atomic ratio of Zr:Y:M (where M is the dopant metal ion) obtained from the EDAX analysis agrees well with the initial composition taken for synthesis of the nanomaterial.

21 Figure 4.14 FESEM images of (a) Mn-YSZ, (b) Cu-YSZ (c) Ni-YSZ with corresponding EDAX spectrum and (d) Zn-YSZ with corresponding EDAX spectrum 4.4 CONCLUSION The dopants influence the crystal growth and phase of the material. Moreover, it can be inferred that crystallite size reduction can be made by addition of such dopants as it suppresses grain growth and crystallization. Usually a very smaller size of the particle (<10 nm) reduces crystallinity of the sample and the material exhibits amorphous nature. All the synthesized samples are nanosized <20 nm as determined from XRD analysis by using Scherrer formula. All the metal ion doped YSZ are predominantly cubic

22 phased. Crystallization delay was observed in Ba-YSZ compared to YSZ and other metal ion doped YSZ. For transition metal doped YSZ peak shifts to higher 2 value marked by decrease in lattice constant as a result of incorporation of metal ions into the YSZ lattice. Since diffraction peaks of tetragonal and cubic phase are closer to each other determination and confirmation of the phase in the case of broader peaks is impossible and still higher sophisticated analysis is required to refine the peak to get a improved report and absolute phase determination. Variation of intensity of the diffraction peaks was observed associated with the vacancy concentrations which are consistent with the PL results. FTIR confirms the formation of the material. Red shift in reflectance spectra was observed due to formation of new energy levels in the forbidden gap. Band gap energy of various metal ion doped YSZ have been determined by using KM plot. Ba-YSZ and Zn-YSZ show high values of energy band gap which is due to quantum confinement effect. From the photoluminescence spectra, the observed UV emission evidences the presence of of F-centres in the sample. Ba-YSZ and Zn-YSZ shows a very broad peak in the green region due to quantum confinement of reduced particle size and the presence of oxygen vacancy. This result was in good agreement with the determined higher band gap values. The surface morphology of the transition metal doped YSZ was observed to compose agglomerated almost uniform sized spherical nanoparticles whereas alkaline earth metal doped YSZ shows agglomerated irregular shaped non-uniform sized particles in the nano regime.