Corrosion Resistance of Alumina Chromia Ceramic Materials Against Molten Slag

Full text

(1)Materials Transactions, Vol. 43, No. 10 (2002) pp. 2561 to 2567 c 2002 The Japan Institute of Metals. Corrosion Resistance of Alumina-Chromia Ceramic Materials Against Molten Slag Takehiko Hirata1, ∗ , Tatsuo Morimoto1 , Akio Deguchi2 and Nozomu Uchida3 1. Advanced Technology Research Center, Mitsubishi Heavy Industries, Ltd., Yokohama 236-8515, Japan Yokohama Research and Development Center, Mitsubishi Heavy Industries, Ltd., Yokohama 231-8715, Japan 3 Nagaoka University of Technology, Nagaoka 940-2188, Japan 2. Improvement of the corrosion resistance of refractories is required critically, since Al2 O3 based conventional refractories used for incineration at the power generation plants are seriously corroded by the molten oxide mixture. In this study, the effect of the addition of Cr2 O3 to Al2 O3 on the corrosion resistance, and the corrosion behavior of Cr2 O3 –Al2 O3 system are investigated in the molten oxide. Cr2 O3 –Al2 O3 ceramic materials with TiO2 as a sintering additive were used for the corrosion tests. Specimens were soaked into the SiO2 –CaO–Al2 O3 based molten oxide and rotated. The corrosion rate of the material is controlled by the diffusion of components of Cr2 O3 –Al2 O3 through the boundary layer. The corrosion resistance was improved by addition of Cr2 O3 . The largest improvement was achieved when the addition of Cr2 O3 was 81.6 mol%. It can be explained by the difference of diffusion rate of Cr3+ and Al3+ . (Received April 12, 2002; Accepted August 12, 2002) Keywords: corrosion, aluminum oxide, chromium oxide, molten slag, boundary layer, diffusion. 1. Introduction It is important to improve the corrosion resistance of the refractory materials to achieve durability of the combustion chamber in the incineration furnace or the power plant, since the refractories in the combustion chambers are corroded extensively by the oxide melt originating from the cinders.1–3) In the steel production, the corrosion resistant refractories have been developed as furnace liner materials against the oxide melt.4, 5) The latest ones are advanced non-oxide type refractories such as SiC type or MgO–C type.6, 7) However, these materials cannot be used in the strong oxidizing atmosphere including the combustion chamber. Thus the candidate material for the combustion chamber refractories need to be selected from the oxides. MgO based materials have also been used in the steel production field.4) However, MgO based materials have a poor thermal shock resistance. So it is difficult to use these materials in the combustion chamber, since the temperature in the chamber may change greatly and abruptly. Al2 O3 ceramics has been used as the combustion chamber refractories.8) Although corrosion is serious especially at the melt-gas boundaries, Al2 O3 is the prime candidate for the refractory because of its superior characteristics i.e. easy to produce, easy to handle, and less expensive. The corrosion resistance of Al2 O3 is higher for slag with particularly low basicity than that of MgO based materials.9) The dissolution mechanism of Al2 O3 into the molten oxide has been studied extensively.10–13) Cooper and Kingery10) studied the dissolution mechanism of Al2 O3 single crystal into CaO–SiO2 –Al2 O3 molten oxide. They concluded that the dissolution rate was controlled by the diffusion of Al2 O3 into the molten oxide. Kobayashi and Oyama11, 12) also studied the dissolution mechanisms of Al2 O3 , and ZrSiO4 into FeO based oxide mixture by the rotating dissolution method. They showed that the dissolution rate was linearly proportional to the corrosion time and that the activation energy of the dis∗ Yokohama. Research and Development Center, Mitsubishi Heavy Industries, Ltd., Yokohama 231-8715, Japan.. solution reactions was 230 kJ/mol for Al2 O3 and 334 kJ/mol for ZrSiO4 . They concluded that the rate determining process was the diffusion of oxide components through the boundary layer in the molten oxide. Ueda13) studied the dissolution rate of the Al2 O3 sintered compact into CaO–SiO2 –Al2 O3 based slag. It indicates that the speed of the Al2 O3 dissolution is controlled by diffusion of the Al2 O3 . Nevertheless, most of above researches treated the monolithic Al2 O3 materials, and researches for the corrosion of the Al2 O3 based multi-componet system were insufficient. Examining the multi-component materials, Cr2 O3 can be the first candidate for the second component of the Al2 O3 . Since the diffusion rate of Cr ion through the molten oxides is slower than that of other metal ions,8) and Cr2 O3 forms a complete solid solution with Al2 O3 . Of course it is known14) that pure Cr2 O3 also reacts with high basicity slag, resulting in generation of toxic Cr6+ in the sintered body or slag. However, formation of solid solution with Al2 O3 reduces the risk of Cr6+ generation. It is expected that a high corrosion resistant material will be developed. The aim of this study is to develop a more stable ceramic material in molten slag. The corrosion behavior of the candidate material Al2 O3 –Cr2 O3 ceramics was investigated in CaO–SiO2 –Al2 O3 slag, especially as a function of Cr2 O3 content in the specimen and the basicity and viscosity of the slag. 2. Experimental Procedures 2.1 Chromia-alumina ceramics The starting powders used were Cr2 O3 (α-Cr2 O3 ; purity, 99.4%; average particle size, 0.3 µm), Al2 O3 (α-Al2 O3 ; purity, 99.997%; average particle size, 0.25 µm), and TiO2 (purity, 99.99%; average particle size, 0.27 µm). These powders were mixed with the ball mill for 360 ks using buthanol as solvent. The composition of the mixed powders are shown in Table 1. The powder-buthanol mixture was dried in a rotary evaporator. The derived powder was pressed into the disks of about 70 mm in diameter and 10 mm in thickness. These com-.

(2) 2562. T. Hirata, T. Morimoto, A. Deguchi and N. Uchida Table 1 Composition of the specimen. Content (mol%). Cr2 O3. Al2 O3. Cr2 O3 content for Al2 O3 plus Cr2 O3 (mol%). TiO2. 48. 48. 67.2. 28.8. 81.6. 14.4. 85. 0. 100. 96. 50 4. 70. Table 2 Chemical composition of the oxide mixture for corrosion test. Content (mass%) SiO2. Al2 O3. FeO. CaO. MgO. 37.6. 16.3. 5.6. 37.6. 2.9. Fig. 1 Schematic diagram of test equipment for corrosion test.. pacts were sintered at the temperature of 1773 K for 10.8 ks in a vacuum furnace. The sintered bodies have relative densities > 95%. Cylindrical specimens for the corrosion experiment were cut out from the sintered bodies. The dimension of the specimen is 5 mm in diameter and 50 mm in length. The roughness of the specimens surface was about 20 µm. 2.2 Corrosion test 2.2.1 Relationship between the corrosion depth and Cr2 O3 content in Al2 O3 –Cr2 O3 ceramics The corrosive oxide mixture was composed of SiO2 , CaO, Al2 O3 , MgO and FeO; detailed compositions were shown in Table 2. Figure 1 shows the corrosion test equipment. Specimens were set at the bottom of the rod made of high-purity Al2 O3 with 10 mm in diameter and 500 mm in length. Forty grams of the oxide mixture was set in the platinum crucible of 44 mm in diameter and 47 mm in height. Specimens were soaked into the molten oxide at the specified temperatures, then rotated. Rotation rates measured by a tachometer were 1.66, 3.33 and 6.66 s−1 . Rotation rates were stable during the holding time. Corrosion time was 5.4, 9, 10.8, 18 or 27 ks. After the corrosion test, the specimens were taken out from the molten oxide. The specimen after the corrosion test showed a shape as shown in Fig. 2. Corrosion at the boundary area between the molten oxide and the vapor phase was rather severe. As the rotation speed increases, the soaked area was uniformly corroded. The specimen was cut at the interval of 2 mm perpendicular to the rotating axis. To eliminate influence at the. Fig. 2 Corrosion profile of specimen. (Testing temperature: 1873 K, Holding time: 9 ks). boundary, 4 mm from the boundary was cut off. The cross section was polished. Diameters of the corroded specimens were measured by an optical microscope. The depth of corrosion damage was defined as the difference of the radii calculated from diameters before and after the corrosion tests. The value of the depth of corrosion was determined from the average of 5 or more measurement. 2.2.2 Relationship between the corrosion depth and CaO content in the molten slag The composition of the specimen was 81.6 mol%Cr2 O3 – 14.4 mol%Al2 O3 –4 mol%TiO2 . The corrosive oxide mixture was composed of SiO2 , CaO and Al2 O3 . The CaO contents in the slag were 20, 40, 50 and 60 mol%. Detailed compositions were shown in Fig. 3. Specimens were soaked into the molten slag at 1873 K, then rotated. Rotation rates measured by a tachometer, were 3.33 s−1 . Rotation rates were stable during the holding time. Corrosion time was 3.6 ks. After the corrosion test, the specimens were taken out from the molten oxide. 3. Results and Discussion 3.1 Relationship between the corrosion depth and Cr2 O3 content in Al2 O3 –Cr2 O3 ceramics Figure 4 shows the relationship between the corrosion depth and the corrosion time for the specimens of 48 mol%Cr2 O3 –48 mol%Al2 O3 –4 mol%TiO2 . The corrosion.

(3) Corrosion Resistance of Alumina-Chromia Ceramic Materials Against Molten Slag. 2563. CaO. 80. 20. 60. 40. 40. 60. 80. Al2O3. 20. 20. 40. 60. SiO2. 80. Fig. 3 Chemical composition of CaO–SiO2 –Al2 O3 based molten oxide mixture for corrosion test (mol%).. Fig. 6 Relationship between the depth of corrosion and the holding time at various temperatures. (Cr2 O3 content: 48 mol%, Rotation rate: 3.33 s−1 ).. Log Rate Constant , K /s -1. -3. E=333kJ · mol-1 -4. -5. -6 5.05. Fig. 4 Relationship between the depth of corrosion and the holding time for the 48 mol%Cr2 O3 –48 mol%Al2 O3 –4 mol%TiO2 specimen. (Testing temperature: 1873 K).. 5.10. 5.15. 5.20. 5.25. Reciprocal Temperature, T. -1. 5.30 -4. 5.35. -1. /10 K. Fig. 7 Relationship between the inverse of temperature and the rate constants. (Cr2 O3 content: 48 mol%, Rotation rate: 1.66 s−1 ).. Log Depth of Corrosion, L /mm. 0 - 0.1 - 0.2 - 0.3 1 - 0.4. 2. - 0.5 - 0.6 - 0.7. 0. 0.2. 0.4. 0.6. 0.8. 1.0. 1.2. Log Rotation Rate, r / s-1. Fig. 5 Relationship between the depth of corrosion and the rotation rate. (Cr2 O3 content: 48 mol%, Testing temperature:1873 K, Holding time: 9 ks).. depth is proportional to the corrosion time under all rotation conditions. At the same corrosion time, the corrosion depth increases with the increasing rotation rate. It means that the corrosion depth increases with the increasing velocity of the slag flow around the specimen. Figure 5 shows the relationship between the depth of damage and the rotation rate of. the specimen. The depth is proportional to the square root of the rotation rate. Figure 6 shows the relationship between the depth of damage and the holding time at various temperatures. When the temperature increases, the depth of damage increases. Figure 7 shows the relationship between the rate constant and the reciprocal temperature. The activation energy of the corrosion is estimated to be 333 kJ·mol−1 . Figure 8 shows the schematic diagram of the corrosion mechanism of oxide ceramics. The dissolution process of the oxides ceramics can be controlled by ➀ diffusion of the oxide components through the boundary layer of slag existing at the front of the surface of the oxides, or ➁ the dissolution rate of the oxides into the slag. If diffusion through the boundary layer is the dominant controlling process, the dissolution rate would is reciprocally proportional to the thickness of the boundary layer, which decreases in the manner reciprocally proportional to the square root of rotation rate of the specimen rod. Therefore, the depth of damage is proportional to the square root of the rotation rate. On the other hand, if the solution reaction at the surface of the oxides is the controlling process, the rate of reaction is independent of the thickness of the boundary layer. Thus, the depth of damage would not depend on the rotation rate.15–18) Consequently, the solution process in this study was controlled by.

(4) 2564. T. Hirata, T. Morimoto, A. Deguchi and N. Uchida. Fig. 8 Schematic diagram of the corrosion mechanism of oxide ceramic materials.. Depth of Damage , L /mm. 1. 0.8. 0.6. 0.4. 0.2. 0 48. 67.2. 81.6. Cr2O3 Content. 96. (mol ). Fig. 9 Relationship between the depth of corrosion and the Cr2 O3 content of the specimen. (Testing temperature: 1873 K, Holding time: 9 ks, Rotation rate: 3.33 s−1 ).. the diffusion of the oxide components through the boundary layer. Shoji et al.19) conclude that the depth of damage is proportional to the power of 0.6 to 0.8 of the velocity of the molten oxide. Cooper et al.15) show that the depth of damage is proportional to the power of 0.5 of the rotation rate. K. Goda20) indicates that if the flow of the molten oxide is in a boundary layer, the depth of damage is proportional to the power of 0.5 of the rotation rate. The diffusion rate of Cr ion through the boundary layer is considered to be slower than that of other metal ions.8) So, addition of Cr2 O3 to Al2 O3 is expected to decrease the corrosion rate of Al2 O3 ceramics. Figure 9 shows the relationship between the corrosion depth and the Cr2 O3 content. Corrosion tests were carried out at 1873 K for 9 ks at 3.33 s−1 . The corrosion depth depended. Fig. 10 Scanning electron micrographs of cross section of specimens after the corrosion test.. on the Cr2 O3 content. The depth of damage was decreased with increasing Cr2 O3 content, reaching the minimum level at 81.6 mol%Cr2 O3 –14.4 mol%Al2 O3 –4mol%TiO2 . Figure 10 shows the scanning electron micrographs of cross section of specimens after the corrosion test. From Fig. 10, the corrosion damage zone was observed at the surface of specimens. The thickness of damage zone decreases with increasing Cr2 O3 content. Figure 11 shows the distribution of the elements of slag and specimen in the 48 mol%Cr2 O3 –Al2 O3 –4 mol%TiO2 ceramics. In the damage zone, the condensation of Mg and Fe was observed. The decrease of Al concentration was also observed..

(5) Corrosion Resistance of Alumina-Chromia Ceramic Materials Against Molten Slag. Fig. 11. 2565. Distribution of elements in the 48 mol%Cr2 O3 –Al2 O3 –4 mol%TiO2 ceramics after the corrosion test.. 3.2 Relationship between the corrosion depth and CaO content in the molten slag As mentioned above, the corrosion process of the Al2 O3 – Cr2 O3 ceramic materials is controlled by diffusion of the solute, i.e. Al or Cr ion through the molten oxide mixture. The viscosity and solubility of the slag may influence the diffusion rate through the boundary layer.21–23) In this study, to clarify influence of viscosity and solubility of slag to the depth of corrosion, the relationship between CaO content and the corrosion rate was investigated. The relationship between CaO content and the depth of corrosion of Cr2 O3 –Al2 O3 ceramics are shown in Fig. 12. The depth of corrosion increases rapidly with increasing CaO content in the slag. Since the properties of slag such as the viscosity and solubility change with changing CaO content, the corrosion rate is considered to change with CaO content in the slag. Figure 13 shows the relationship between the viscosity of slag and CaO content.24) The viscosity decreases rapidly with. increasing CaO content. Therefore, the depth of corrosion seemed to increase with increasing CaO content in the slag. The solubility of component of specimen to the molten oxide affects the driving force of diffusion through the boundary layer. Figure 14 shows the relationship between the CaO content in the slag and the solubility of Al2 O3 and Cr2 O3 .25–28) From Fig. 14, the solubility of Cr2 O3 is smaller than that of Al2 O3 in the slag. Since the solubility of Al2 O3 and Cr2 O3 increases with increasing CaO content, the corrosion rate of Cr2 O3 –Al2 O3 increases with increasing CaO content. Rapp29) showed that the solubility of specimen in the molten salt depends on the basisity gap between the specimen and salt. The basicity of molten oxide, Al2 O3 and Cr2 O3 are shown in Fig. 15. The bacisity is expressed with B-parameter, which is proposed by Morinaga.30) According to the definition, the B parameter is estimated as follows; B = n i Bi. (1).

(6) 2566. T. Hirata, T. Morimoto, A. Deguchi and N. Uchida 250. 80 TiO2. 200. Al2O3. Solubility (mass ). Depth of Corrosion, D /. m. 60 150. 100. 40. MgO 20. 50. Cr2O3 0. 0. 10. 20. 30. 40. 50. 60. 70. 80. 90. 0. 100. 30. CaO Content (mol ). 35. 40. 45. 50. 55. 60. CaO Content (mol ). Fig. 12 The relationship between the depth of corrosion and CaO content in the CaO–SiO2 –Al2 O3 based molten oxide.. Fig. 14 The relationship between the solubility and CaO content in the CaO–SiO2 based molten oxide. (Temperature: 1873 K).. 50 1. 40 0.8. Basicity (B-parameter). Viscosity (Pa s). 30. 20. 0.6. 0.4 Cr2O3. 10 0.2. Al2O3. 0 40. 60. CaO Content(mol. 80. 100. Fig. 13 The relationship between the viscosity and CaO content in the CaO–(75SiO2 –25Al2 O3 ) based molten oxide. (Temperature: 1873 K, Viscosity: Pa·s).. Bi =. 0 0. ). 40. 60. 80. 100. CaO Content (mol ). Fig. 15 The relationship between the basicity and CaO content in the CaO–(75SiO2 –25Al2 O3 ) based molten oxide. (The basicity gap is indicated with B-parameter).. (ri + 1.40)2 − 0.405 Zi × 2. (2) 1.023 where, n i is the fraction of cation i against the number of all cations, ri is the radius and Z i is the valence of the cation i. This parameter estimates the degree of localization of the lone pair electron. This degree of localization could express the basicity of the molten oxide with ionic bonds. From Fig. 15, the bacisity increases with increasing CaO content in the molten slag. Figure 16 shows the relationship between the CaO content in the slag and basisity gap between the slag and the components of specimen, Cr2 O3 and Al2 O3 . The basisity gap between Cr2 O3 and slag is smaller than that between Al2 O3 and slag, then the solubility of Cr2 O3 becomes smaller compared to that of Al2 O3 . As a result, the corrosion rate of Cr2 O3 seems to be smaller than that of Al2 O3 . In Fig. 16, the relationship between the CaO content and the basicity gaps between the slag and TiO2 and MgO are indicated. Also, the dependence of the solubility of TiO2 and MgO on the CaO content are shown in Fig. 14. From Figs. 14. 20. 0.5 TiO2 0.4 Al2O3 B. 20. 0.3 Basisity gap,. 0. Cr2O3. 0.2. MgO. 0.1. 0 30. 35. 40. 45. 50. CaO Content (mol. 55. 60. 65. ). Fig. 16 The relationship between the basicity gap and CaO content in the CaO–(75SiO2 –25Al2 O3 ) based molten oxide. (The basicity gap, ∆B, between the components of specimens and the oxide is indicated with B-parameter)..

(7) Corrosion Resistance of Alumina-Chromia Ceramic Materials Against Molten Slag. and 16, the basicity gap and the solubility show the same dependence on CaO content. Therefore, this basicity gap seems to be a useful index that expresses the solubility of oxide in the slag. From Figs. 14 and 16, MgO seems to show the high corrosion resistance for the high basicity slag. 4. Conclusion The corrosion mechanism of Al2 O3 –Cr2 O3 materials in CaO–Al2 O3 –SiO2 slags was studied by a rotating rod method to improve the corrosion resistance of Al2 O3 against the molten slag by adding Cr2 O3 . The results are as follows: (1) The corrosion resistance of Al2 O3 base ceramics is improved by adding Cr2 O3 . (2) 81.6 mol%Cr2 O3 –14.4 mol%Al2 O3 –4 mol%TiO2 ceramic material shows the highest corrosion resistance. (3) As a result of the rotation corrosion test, the depth of corrosion was found to be proportional to the square root of the rotation rate. (4) The corrosion of Al2 O3 –Cr2 O3 ceramic materials is controlled by diffusion of components of Al2 O3 –Cr2 O3 through the boundary layer in the slag. (5) The depth of corrosion is decreased with increasing Cr2 O3 content. (6) The basicity gap between the slag and specimen reflects the solubility of the specimen. REFERENCES 1) H. Yamamoto, H. Tsunoda, H. Motomura and S. Ono: Mitsubishi Juko Giho 25 (1988) 230–233. 2) S. Nishikawa: Mitsubishi Juko Giho 36 (1999) 122–125. 3) R. L. Jones: Surface and coatings technology 39/40 (1989) 89–96. 4) K. Sugita: Tetsu to Hagane 65 (1979) 1462–1474. 5) K. Ichikawa: Bulletin of the Ceramic Society of Japan 30 (1995) 906– 910.. 2567. 6) K. Sugita: Refractories for Iron and Steelmaking, (Chijin Shokan Co., Ltd., 1995) pp. 147–153. 7) K. Sugita: Refractories for Iron and Steelmaking, (Chijin Shokan Co., Ltd., 1995) pp. 254–261. 8) M. Nakamura: Proceedings of the 7th international symposium on ultra-high temperature materials ’97 in Tajimi (The committee meeting of the 7 th international symposium on ultra-high temperature materials, 1997) pp. 73–84. 9) K. Sugita: Refractories for Iron and Steelmaking, (ChijinShokan Co., Ltd., 1995) pp. 95–101. 10) B. N. Samaddar, A. R. Cooper and W. D. Kingery: J. Am. Ceram. Soc. 47 (1964) 249–254. 11) H. Kobayashi and T. Oyama: J. of Ceram. Soc. of Japan 83 (1975) 97– 102. 12) H. Kobayashi and T.Oyama: J. of Ceram. Soc. of Japan 82 (1974) 40– 47. 13) K. Ueda: J. Japan Inst. Metals 63 (1999) 989–993. 14) A. Yamaguchi: Bulletin of the Ceram. Soc. of Japan 15 (1980) 254–262. 15) A. R. Cooper and W. D. Kingery: J. Am. Ceram. Soc. 47 (1964) 37–43. 16) S. Taira, K. Nakashima and K. Mori: ISIJ Inter. 33 (1993) 116–123. 17) S. Uchida, R. Suzuki and M. Tokuda: J. Japan Inst. Metals 60 (1996) 826–833. 18) M. Cable: Corrosion of advanced ceramics, (1994) 285–296. 19) Y. Shoji, S. Uchida and T. Ariga: Metall. Trans. 1313 (1982) 439–445. 20) K. Goda: Taikabutsu 49 (1997) 701–707. 21) A. Yamaguchi: Taikabutsu 44 (1992) 168–174. 22) T. Miyaji, B. Sakamoto, E. Kudo: Taikabutsu 52 (2000) p. 532. 23) B. Sakamoto: Taikabutsu 53 (2001) p. 246. 24) G. Urbain, F. Cambier, M. Deletter and M. R. Anseau: Trans. J. Br. Ceram. Soc. 80 (1981) p. 139–141. 25) R. C. Devries, R. Roy and E. F. Osborn: J. Am. Ceram. Soc. 38 (1955) 161. 26) E. F. Osborn and A. Muan: Phase Equilibrium Diagrams of Oxide Systems, Plate 1., published by the Am. Ceram. Soc. and the Edward Orton, Jr., Ceramic Foundation (1960). 27) E. F. Osborn and A. Muan: “Phase Equilibrium Diagrams of Oxide Systems”, Plate 2., published by the Am. Ceram. Soc. and the Edward Orton, Jr., Ceramic Foundation (1960). 28) F. P. Glasser and E. F. Osborn: J. Am. Ceram. Soc. 41 (1958) 362. 29) R. A. Rapp: Hot Corrosion of Materials, in Selected Topics in High Temperature Chemistry, (Elsevier, NY, 1989) pp. 291–329. 30) K. Morinaga, H. Yoshida and H. Takebe: J. Am. Ceram. Soc. 77 (1994) 3113–3118..

(8)

No results found