Partial Reduction of NiAl 2 O 4 - Mechanism and the Influence of Doping
- PDF / 2,301,996 Bytes
- 6 Pages / 414.72 x 648 pts Page_size
- 102 Downloads / 169 Views
microstructure at the reduction front of an undoped NiA1204 reduced at 1300 0C for 2 hours at an oxygen activity of 10-12-3. Ni particles are present along spinel grain boundaries. The "defect spinel" metastable phase(s) surrounding the Ni particles were identified to have an oxygen-ion sublattice similar to that of the spinel matrix, by analysis of the electron diffraction patterns. Bassoul etc.[4] pointed out that one of the "defect spinel" phases, NiAll0016, likely to be present here, had a monoclinic periodic antiphase boundary structure based on the spinel structure. The Ni + 'defect spinel" layer located between the unreduced spinel and the Ni + a-A1203 region was -20pm thick. Fig. 2 shows a TEM image of the microstructure of an undoped spinel specimen reduced at 1300'C for 4 hours at an oxygen activity of 10-12.3. Large irregular-shaped Ni particles (1-4pm) are present along the alumina grain boundaries. Small equiaxed Ni particles (20-250nm), precipitated within some of the alumina grains, were usually associated with dislocations. Small pores (25-200nm) were observed in some alumina grains.
Ni N
Defect spinel
Fig. 1. The reduction front of the undoped
Fig. 2. The undoped spinel specimen
hours at an oxygen activity of 10-12-3
activity of 10-12.3
spinel specimen reduced at 1300PC for 2
reduced at 1300'(C for 4 hours at an oxygen
XRD results showed that Ni and "defect spinel" phases were present in the undoped spinel specimen reduced at 1100TC for less than 24 hours at an oxygen activity of 10-14.8. The structure of the "defect spinel" was similar to the monoclinic B phase (NiAlI0016) [4]. Ni + O-A1203 were present in a specimen reduced under the same conditions for 24-48 hours. For specimens reduced for 144-168 hours, a-A1203 + Ni were present. Doping with MgO and Cr203 had very little effect on the partial reduction reaction. Before reduction, both dopants were completely dissolved within the spinel phase. After reduction, both dopants remained in the ceramic phase. Due to their limited solubility in both the spinel and the a-A120 3, most of the Y20 3 was present in the form of a garnet phase (YAG) with composition of Y3A150 12 (3Y20 3 :5A120 3), before and after reduction. The YAG particles (0.5-0.8jpm) were distributed uniformly throughout the entire specimen. Although these YAG particles served as additional nucleation sites for the Ni, the microstructure after partial reduction was similar to that of the undoped specimen. Fig. 3 shows a TEM image of the microstructure close to the reduction front of a 1 mole% TiO2doped specimen reduced at 1300*C for 2 hours at an oxygen activity of 10-12.3. A very narrow "defect spinel" layer (approximately 50-100nm thick) is present between the unreduced spinel and the Ni-a-Al 203 mixture. Fig. 4 is a TEM image of the reduced region of the same specimen, far from the reduction front, where both Ni and alumina grains are equiaxed in shape and uniform in size. No small pores were observed within the alumina grains. Though Ti was completely dissolved in the spin
Data Loading...
