4-1 Results of SEM, XRD and TEM
In the three-series products, BaO and BaCO3 were used as the barium precursors and TiO2 with various morphologies were used as the titanium precursors.
The reaction time and temperature were all the same in the three-series synthesis processes, which were 1 h and 700 C, respectively. The SEM images shown in Figures 4.1 and 4.2 demonstrate that the shape of the products are related to that of the titanium precursors and the dissolution rate of both barium and titanium precursors extremely. The XRD patterns shown in Figure 4.3 verify that there are no detectable impurity peaks in the three-series products and each of them is tetragonal structure by fitting the overlapping (002) and (200) diffraction peaks. The high-resolution transmission electron microscope (HRTEM) image and selected-area electron diffraction (SAED) pattern shown in Figure 4.4 manifest that BT-3 was grown along the [001] direction and the lattice fringe shows the lattice constant a is about 4.01 Å.
35
Figure 4.1 SEM images and powder XRD patterns of the titanium precursors: (a) SEM image of the spherical TiO2, (b) SEM image of the rod-shaped K2Ti4O9, (c) SEM image of the rod-shaped TiO2, and (d) Powder XRD patterns of the rod-shaped K2Ti4O9 and rod-shaped TiO2
36
Figure 4.2 SEM images of the three-series products: (a) BT-1, (b) BT-2, and (c) BT-3
37
Figure 4.3 Powder XRD patterns of the three-series products and JCPDS Card: (a) BT-1, (b) BT-2, (c) BT-3, and (d) JCPDS Card #81-2203
38
Figure 4.4 TEM image, HRTEM image with SAED pattern, and EDS spectrum of BT-3: (a) TEM image, (b) HRTEM image with SAED pattern, and (c) EDS spectrum
39
obtained the spherical BaTiO3; the rod-shaped TiO2 obtained the rod-shaped BaTiO3. Hence, we can change the shape of the products by changing the shape of the precursors effectively. We can also change the size of the precursors to obtain the products with different sizes. Therefore, the larger the radii of the spherical TiO2
and rod-shaped TiO2 are, the larger the radii of the spherical BaTiO3 and rod-shaped BaTiO3 are, and vice versa.
4-3 Comparison between BT-2 and BT-3
A prodigious result is that as we used BaO as our barium precursor that reacted with the rod-shaped TiO2 in the molten salt, we obtained BT-2 in contradiction with the result of BT-3. Because of the “break-up” of the rod-shaped TiO2 when we heated it individually in NaCl-KCl flux at 700 C for 1 h and the unchangeableness of BT-3 when we heated BT-3 in NaCl-KCl for another 1 h at 700 C, the shape change from rod-shape to cube-shape in BT-2 might be ascribed to the “break-up” of the rod-shaped TiO2 before or during the reaction.
4-4 Refinement by using two-peak fit
40
We obtained the lattice constants a and c of BT-1, BT-2 and BT-3 in Table 4.1 by two-peak fit to the overlapping (002) and (200) diffraction peaks of Figure 4.5.
The tetragonalities (c/a ratio) of BT-2 and BT-3, which are 1.00503 and 1.00443 respectively, are larger than that of the BT-1, which is 1.00366. All of their c/a ratios are larger than 1, which means!the unit cells of the three-series products are all in tetragonal structure, and obviously lower than that of bulk, which is 1.01128.
This phenomenon may be attributed to the surface relaxation45. Since the particle sizes in a-direction of BT-1, BT-2 and BT-3 are all at the nanoscale and smaller than that of bulk, they will experience a surface relaxation and thus increase the a values.
On the other hand, because the particle sizes in c-direction of both BT-1 and BT-2 are at the nanoscale and of BT-3 is at the microscale, which is similar to the microparticles (bulk), the behavior of the particle size in c-direction of BT-3 will be much like that of bulk and thus has the c value comparably closer to bulk, which is 4.035 Å, and larger than that of BT-1 and BT-2. Consequently, because all the three-series products have the larger a values and smaller c values than bulk has, the c/a ratios of BT-1, BT-2 and BT-3 will smaller than that of bulk. Furthermore, by comparing the separation between the (200) and (002) peaks with the c/a ratio, the closer the c/a value to unity, the shorter the separation between the two peaks is.
As a result of the isotropic particle structure of BT-1 and BT-2, they would arrange randomly without dominant orientation on the substrate. Thus, the intensity ratio between the (200) and (002) XRD peaks would close to 1:2 due to the (200) and (020) peaks are inseparable shown in Figures 4.5(a) and (b). On the contrary, because BT-3 has the anisotropic particle structure, its (002) planes will lay on the substrate with the higher probability rather than standing erectly on the substrate.
Therefore, the intensity of the (200) peak would much higher than that of the (002) peak in XRD measurement as shown in Figure 4.5(c).
41
Table 4.1 The refinement results of BT-1, BT-2 and BT-3
Label a(b) (Å) c (Å) c/a Vol. (Å3)
JCPDS (#81-2203) 3.990 4.035 1.01128 64.238
BT-1 4.00671 4.02138 100366 67.558
BT-2 4.00244 4.02256 1.00503 64.488
BT-3 4.01117 4.02892 1.00443 64.823
Figure 4.5 Two-peak fit of the powder XRD patterns at (200) and (002) peaks of the three-series products: (a) BT-1, (b) BT-2, and (c) BT-3
42
4-5 Proposed mechanism of molten-salt synthesis
Based on the above observations, the dissolution rate of the reactants in the molten salt plays an important role that affects the reaction rate and the morphologies of the products critically. For one thing, if both reactants are soluble in the molten salt, dissolution-precipitation mechanism may be the dominant mechanism—the product would be readily synthesized via precipitation from the salt containing the dissolved reactants—throughout the synthesis of BaTiO346; for another, if one reactant is much more soluble than the other, in-situ transformation mechanism will be the dominant mechanism. The more soluble one would dissolve into the salt at the early stage of the reaction, then diffuse onto the surfaces of the less soluble reactant and finally react to form the product. According to this argument, we can infer that the dissolution rate in the molten salt of BaCO3 using in BT-3 is faster than BaO using in BT-2 because the rod-shaped TiO2 can preserve the original morphology in BT-3 but break up into several cube-shape in BT-2.
Being more precise to our synthetic mechanism of BaTiO3 in the molten salt, the proposed synthesis schematic diagrams are provided and shown in Figure 4.6.
The following are the details of the main steps: When the mixture of BaCO3, TiO2
and NaCl-KCl flux was put in the combustion boat at room temperature, the precursors existed in the solid form like the first illustration in the upper flow of the proposed schematic diagram (the dissolution rate of the barium precursor is faster than that of the titania). When we increased the temperature at a constant rate to the melting point (ca. 657 C) of NaCl-KCl flux, NaCl-KCl flux will transform into molten salt and act as a solvent where BaCO3 and TiO2 act as solutes dissolved inside.
Upon further heating, because of the faster dissolution rate of BaCO3 comparing with TiO2, it would dissociate and dissolve into the molten salt priorly, then disperse
43
readily to the periphery of TiO2 and then cover the entire surface of TiO2 to form the BT-shell which was able to hinder the rod-shaped TiO2 from the break-up process.
Finally, BaCO3 would diffuse through the BT-shell and react with TiO2 completely after heating. Thus, we derived the rod-shaped BaTiO3, which had the same morphology as the used TiO2.
In the second flow of the proposed schematic diagram (the dissolution rate of the barium precursor is slower than that of the titanium precursor), the mixture of BaO, TiO2 and NaCl-KCl flux would experience the same processes like the first flow before the transition of NaCl-KCl flux to the molten salt. Because the dissolution rate of BaO was lower than that of BaCO3, the rod shaped TiO2 would have the chance to break up into several fragments from the grain boundaries in the molten salt before BaO dissociated and dissolved into the molten salt and formed a BT-shell outside the surface of the rod-shaped TiO2. As a result of the break-up process of the rod-shaped TiO2 before the formation of BT-shell, we finally derived the cube-shaped BaTiO3 that had high quality without impurity peaks in the XRD pattern.
44
Figure 4.6 Proposed schematic diagram of the synthetic mechanism of BaTiO3 in the molten salt: the upper flow is due to the faster dissolution rate of the barium precursor than that of the titanium precursor and the lower flow is due to the slower dissolution rate of the barium precursor than that of the titanium precursor.
45