Materials Chemistry
Forming of Self-Sustaining Ce0.9Gd0.1O1.95 Film by Aqueous Tape Casting Method
2021 Volume 62 Issue 6 Pages 770-773
Details
2021 Volume 62 Issue 6 Pages 770-773
Self-sustaining Ce0.9Gd0.1O1.95 (GDC) film was fabricated by aqueous tape casting method. An aqueous tape casting of slurry was performed using poly acrylic acid (PAA) as dispersant, poly carboxylate ammonium (PCA) as binder, poly ethylene glycol (PEG) as plasticize, and deionized water as solvent. The rheology of the slurries was evaluated with cone-plate viscometer. The conditions for preparing stable slurries were studied, regarding with solid concentration and ball-milling period, and were optimized by viscosity measurements and aging experiments. Green tapes with thickness in the range of 65 to 130 µm were prepared by conditions; solid concentration of 25–26 vol% and gate height of 0.2–0.3 mm with moving rate of 45 cm min−1. Conventional sintering techniques were available for densification, and semi-transparent GDC film of 47–95 µm thickness was successfully fabricated by heating a green sheet at 1500°C in air.
This Paper was Originally Published in Japanese in J. Soc. Mater. Sci., Japan 68 (2019) 543–548. The caption is slightly changed.
Cerium oxide (CeO2) has been applied and further expected in various application fields, for example, the ultraviolet absorbers, the color materials, the oxygen storage ability carrier of the catalyst for automobile exhaust purification, the gas separating device and the fuel cell for a low-temperature operation as a solid electrolyte.1–11) When its attention is paid to use to the high ionic conductors in solid oxide fuel cell, the film of an electrolyte and/or electrode material, including formation techniques of thin films on substrates, will become important. The self-standing films using the tape cast method is one of the essential components in these devices.12,13) Although the thick film formation technique commonly uses an organic solvent, the problem of the solvent treatment has still left from a viewpoint of volatile organic compounds (VOC) regulation in the future industry. The aqueous slurry is better in the tape cast for forming thick film, because of taking such atmospheric environment preservation into consideration. It is also an important issue to examine the film forming parameters using aqueous system instead of an organic solvent in a process design of the ceramic.14–17)
Several papers of the aqueous tape cast method have been reported, however, there still remain several problems of examination toward practical CeO2-base electrode. Densification of Ce0.9Gd0.1O1.95 film was examined by a tape cast using a doctor blade technique and ethanol which is of relatively low toxicity.18) However, the achievement of improved Gd-doped CeO2 solid electrolyte membrane using only water as a solvent is still rare,19–21) regarding with research of shaping of the thickness of less than 100 µm. In this work, transparent Ce0.9Gd0.1O1.95 (GDC) thick film was fabricated by aqueous tape casting and the slurry composition was optimized. The preparation of slurry which used water as a solvent, its rheology characteristics and a film-forming behavior were examined and then a fabrication of self-sustaining film with more thinly than a thickness of 100 µm was firstly achieved.
The starting materials are follows; Ce0.9Gd0.1O1.95 (GDC) powder (Anan Chemicals, mean particle-diameter 0.24 µm, and specific surface area 30 m2g−1),22,23) a binder of polyacrylic acid (PAA, average molecular weight 2.5 × 105, Wako Pure Chemical), a dispersant of ammonium salt, polycarboxylate ammonium (PCA, the Toagosei A6114, average molecular weight 10,000) and a plasticizer polyethylene glycol (PEG, Wako Pure Chemical).
In forming experiments, the solid content of GDC powder was set from 21 to 29 vol% in slurry. Both the binder and the plasticizer of 10 mass% versus water, the dispersant of 5 mass% versus GDC, 100 ml water and GDC powder were input a bottle. Ball-mill procedure was done in a polyethylene container with 10 mm diameter plastic balls and mixing time was changed from 1 h to 24 h.
Molding operation of thick film was performed using a doctor-blade equipment (DP-150, Tsugawa machine) after degassing treatment of slurry in vacuum. The green sheets were pulled out on the carrier film of 100 mm width and under the gate height of 0.1 to 0.5 mm at the tape feeding speed of 22.5 to 45 cm min−1 from a slurry tank at room temperature. Then solidification was carried out in atmosphere for a day. After making a sheet with suitable length, it was removed from a carrier film and cut to films several centimeters square in size, and then they were put onto an alumina plate. The sheets were put in a furnace and heated slowly for degreasing, and then further heated by a heating rate of 5°C min−1 and kept at 1500°C for 5 h in air.22,23) No sintering aid was used in this sintering experiment.
2.2 Characterization of slurry and materialsThe viscosity of slurry was measured by a cone-plate type viscometer using a viscoelasticity equipment (MR-500, Rheology Co.). The shear rate was changed from 0.1 to 300 s−1 under the condition; cone angle ψ of 2.042 degree, cone diameter r of 3.999 cm, torque of 2.000 kg cm−1 and load of 5.000 kg. When the conical rotary moment M (dyn cm) is applied, the viscosity η (Pa s) can be measured by rotating at angular velocity Ω (rad s−1), according to eq. (1).
| \begin{equation} \eta = \frac{\text{3M}}{2\pi r^{3}} \cdot \frac{\psi}{\Omega} \end{equation} | (1) |
Scanning electron microscopy (SEM) images on surface and cross-section of films were obtained by field-emission scanning electron microscopy (FESEM; Jeol JSM6100) at 15 keV. Thickness in film products was measured on formed sheets and sintered ceramics by a micrometer instrument and the FESEM image. Powder X-ray diffraction (XRD) patterns were recorded using a diffractometer (Rigaku Rint2200) with Cu Kα radiation at 15 kV and 30 mA. Slurry and formed sheets were visually observed and photographically recorded.
The influence of the flow properties on solid concentration was first examined for production of slurry suitable for tape forming. The solid concentration was changed and the relationship between viscosity and shear rate was measured. Ball-mill mixing time in the present work was set on 6 h. Figure 1 shows a plot of viscosity versus shear rate of slurry with various solid concentrations between 21 and 29 vol%. The viscosity was still large in the range where a shear rate is small. On the other hand, the viscosity became small on the condition that a shear rate was large, and also depending on the solid concentration. If sheet drawing velocity is made quicker, it will be expected that shaping itself is possible using thin slurry with lower concentration. However, a series of rates from a low speed until high-speed should be achieved for the practical operation of shaping. In slurry with small solid concentration, for example, the slurry often may flow out of a gate naturally when it is put in the tank of equipment at the time of shaping. Especially, this may happen with less solid concentration than 23 vol%. Therefore, it was considered in this work that slurry near 25 vol% was suitable from the necessity of changing in the moderate viscous range for forming procedure.
A plot of viscosity versus shear rate of slurry with various solid concentrations. ■; 21 vol%, □; 23 vol% and ●; 25 vol% and ○; 27 vol% and ▲; 29 vol%.
The change of viscosity in ball-mill time was investigated using 25 vol% slurry. Figure 2 shows the relationship between viscosity and shear rate. The viscosity decreased suddenly between early 2 h and 4 h when mixing time become long, and hardly changed in the range of milling time for 6–8 h. The coagulation is ordinary state in starting raw powders, and the linking among GDC powders should dissociate with a ball-milling procedure. It is thought that the dispersed degree become higher and viscosity fell in connection with milling time. Moreover, chains of macromolecule in binder should be cut in part by mechanical destruction. These are effective in viscosity-falling control and ball-mill time is considered to have attained them change regularly by 6 h in this experiment.
A plot of viscosity versus shear rate of 25 vol% slurry prepared with various ball-milling periods. ◇; 2 h, □; 4 h, ○; 6 h, △; 8 h.
The instable change of slurry causes trouble in stock and cannot fabricate the whole sheet at uniform concentration and thickness. Then, in order to know the stability of slurry, the sedimentation test at the room temperature in the air was performed. Figure 3 shows the photograph in the stability test on a prepared slurry (25 vol%) with soaking period from 1 min to 24 h in atmosphere. After 24 h, powder has suspended similarly well in slurry and there is no sedimentation, thus, the time-dependent change was not recognized. It was expected that this slurry was stable enough and the influence of the concentration distribution on the whole sheet would not appear. Thus, the slurry with stability sufficient by the final examination has been produced by selecting the present condition.
Stability of a prepared slurry (25 vol%) with soaking period in atmosphere. Left to right; 1 min, 30 min, 2 h, 5 h, and 24 h.
In order to control the thickness of the sheet to fabricate, the gate height (gap width of a doctor-blade) was changed with 0.1, 0.2, 0.3, 0.4 and 0.5 mm and the shaping experiment was conducted. The thickness of the green sheet, acquired to gate height, was shown in Fig. 4. Since slurry did not come out when gate height was 0.1 mm, sheet forming was not able to be carried out. Moreover, although it was possible to have pulled out slurry in the case where gate height was 0.4 mm and 0.5 mm, many cracks went into the sheet after desiccation. Therefore, it was thought that both the gate heights of 0.2 and 0.3 mm were suitable. Distinction of the success or failure in moldability test was shown in Fig. 4 (proper conditions surrounded with a circle).
A plot of film thickness versus gate height in doctor blade process, where successful forming can be achieved under the condition of 0.2–0.3 mm in gate height.
Figure 5 shows the optical photographs of 100 mm-width sheets which were formed with the gate height of 0.4 mm and 0.3 mm. The practical width can be compared and shown as the underlying green sheet which was used in fabrication. No crack was observed over the whole by the sheet lengthened in the case of gate height 0.3 mm in neither the width direction nor the length direction. On the other hand, in gate height 0.4 mm, many cracks were seen in a starting part and near the termination side. In our previous experience, the contraction by desiccation is solvable by increasing the fraction of a binder by lowering solid concentration of slurry and lowering viscosity. However, in this experiment, solid concentration was decided to carry out by high-solid slurry with smaller amount of organic additives as much as possible. It was observed that sheet-thickness was greatly dependent on gate height rather than sheet feeding speed from the above examination. The thickness of a sheet becomes large when gate height is enlarged with every feeding speed. Therefore, it is appropriate that sheet can be fabricated by a feeding speed of 45 cm min−1 and by 0.2 mm and 0.3 mm for gate height.
Photographs of formed sheets of 100 mm in width with gate height of 0.4 mm (upper) and 0.3 mm (lower).
Figure 6 shows the cross-section SEM images of dried green sheets produced on the condition of gate height 0.2 mm and 0.3 mm and dense ceramic films followed by the sintering procedure at 1500°C. The thickness of the green film was 65 µm and 132 µm when gate height was 0.2 mm and 0.3 mm respectively. Moreover, the thickness of the sintered sheets was 47 µm and 95 µm, respectively. Shrinkage toward the thickness direction between a green film and a sintered film, which was estimated from the SEM image, was 28% (0.2 mm) and 31% (0.3 mm), respectively. Although gate height greatly influenced sheet thickness, the relationship between thickness and height was not described by simple proportionality relation.
SEM images of cross-section of Ce0.9Gd0.1O1.95 films which were dried (a), (c) and then sintered at 1500°C (b), (d), with gate height of 0.2 mm (a), (b) and 0.3 mm (c), (d). Each scale bar represents 50 µm (a), (b) and 100 µm (c), (d).
Since the contraction rate in thickness direction was approximately 30%, it was expected that large contraction should start also into a sheet side. The sheet after sintering actually produced the cracks in several centimeters. The sintered compact of the sheet by gate height 0.2 mm became weak thinly with 47 µm thickness when handling nature was taken account. The sintered compact of the sheet fabricated by gate height 0.3 mm was strong enough, and the sheet of several tens mm2 area was able to take it out as a stable sintered compact. In this experiment, if handling of after-sintering material is taken into consideration, the conditions of gate height 0.3 mm or more can be concluded that it is suitable for thick film production with strength. On the other hand, it is expectable using this slurry for the sheet forming of less than 50 µm thickness and sintering operation to be sufficiently possible by equipment and handling if we could apply higher controllability with advanced machines. By X-ray diffraction analysis it was confirmed that each sintered compact consists of a CeO2 phase.
Finally, the photograph of a sheet sintered at 1500°C is shown in Fig. 7. It turns out that the sintered Ce0.9Gd0.1O1.95 film is transparent to some extent, and Japanese characters (that means “cerium oxide”) under a film are visible. The apparent density of a sintered film was 6.95 g cm−3. Since the sintered compact fully became dense and the pores extremely decreased, it is possible that the scattered reflection of light decreased and translucency was acquired. The ceramic thick film with this composition will be a material suitable for use of the solid electrolyte of a fuel cell and a gas separating device.
Photograph of transparent Ce0.9Gd0.1O1.95 sheet (a gage is cm-scale).
The tape cast method was applied to examine the forming process of CeO2 base thick film. The optimum conditions of sheet forming were examined on whole process from slurry preparation to sintering of films. Although it was under the comparatively narrow condition, it was found out that the sintered compact of a good self-sustaining film could be obtained. The main point of production conditions is summarized as follows.
This work was supported by a Grant-in-Aid for Scientific Research (No. 17H03100) from the Japan Society for the Promotion of Science (JSPS) and the Project of Creation of Life Innovation Materials for Interdisciplinary and International Researcher Development of the Ministry of Education, Culture, Sports, Science and Technology, Japan.
