Characterization of Si3N4-Al interface after corrosion tests

Caracterização da interface Si3N4-Al após testes de corrosão

Characterization of Si₃N₄-Al interface after corrosion tests

Caracterização da interface Si₃N₄-Al após testes de corrosão

C. dos Santos^I; S. Ribeiro^I; K. Strecker^I; C. R. M. da Silva^II

^IDepartamento de Engenharia de Materiais, DEMAR-FAENQUIL, Polo Urbo-Industrial, Gleba AI-6, Lorena, SP, 12600-000

^IICentro Técnico Aeroespacial, Divisão de Materiais, AMR-CTA, Pça. Marechal do Ar Eduardo Gomes 50, S. J. Campos, SP, 12228-904. [e-mail]

ABSTRACT

Silicon nitride is a covalent ceramic material of high corrosion resistance and mechanical stability at elevated temperatures. Due to these properties, its use in metallurgical processes, such as the casting of alloys, is increasing. Therefore, the characterization of the interface between Si₃N₄ and the casted metal is of great importance to investigate possible interactions, which might deteriorate the ceramic mould or contaminate the metal. In this work, the use of Si₃N₄ as crucible material for Al-casting has been studied, by investigating the corrosion attack of liquid Al at a temperature of 1150 ºC during 30 days in air. The interface was characterized by X-ray diffraction, scanning electron microscopy and energy dispersive spectroscopy. It has been found that due to superficial oxidation two oxide layers form – SiO₂ on Si₃N₄ and Al₂O₃ on Al – which effectively hinder further reactions under the conditions studied, confering high corrosion resistance to the Si₃N₄ crucible.

Keywords: silicon nitride, corrosion behavior, scanning electron microscopy, X-ray diffraction.

RESUMO

Nitreto de silício (Si₃N₄) é um material cerâmico covalente de elevada resistência à corrosão e estabilidade mecânica em temperaturas elevadas. Devido a essas propriedades, sua utilização em processos metalúrgicos, tais como em fundição de ligas metálicas, é crescente. Desta forma, a caracterização da interface entre Si₃N₄ e metais fundidos é de grande interesse para investigar possíveis interações as quais poderão deteriorar o material cerâmico e/ou contaminar o metal. Nesse trabalho o uso de Si₃N₄ como material base de cadinhos para fundição de alumínio foi estudado, pela investigação do ataque corrosivo de Al líquido a 1150 ºC durante 30 dias, ao ar. A interface foi caracterizada por difração de raios X, microscopia eletrônica de varredura e espectroscopia de energia dispersiva. É encontrado que devido à oxidação superficial dois óxidos se formam - SiO₂ no Si₃N₄ e Al₂O₃ no Al - os quais evitam possíveis reações sob as condições estudadas, conferindo alta resistência à corrosão aos cadinhos de Si₃N₄.

Palavras-chave: nitreto de silício, comportamento de corrosão, microscopia eletrônica de varredura, difração de raios X.

INTRODUCTION

Silicon nitride (Si₃N₄) is used as a structural ceramic material with interesting characteristics for high temperature applications, such as wear and corrosion resistance [1-4]. One of the many fields of application as structural ceramic is in metallurgical processes as refractory material, thermocouple protection or casting valve [5, 6]. The corrosion behavior of this material has been mainly studied on metal/ceramic junctions under inert atmospheres [7-11], thus avoiding oxidation common in industrial processes.

The phase identification by mass contrast, using emission of backscattered electrons in SEM, is an important tool in situations where the phases present have an accentuated difference of their atomic number. Even so, it should be noticed that the use of X-ray diffraction analysis for the final phase identification is indispensable.

In the Al/Si₃N₄ system, due to the adherence of Al₂O₃ originating from the aluminum surface, on the surface of Si₃N₄, the difficulty of chemical attack of this oxide and the small thickness of the interfacial layer formed, it is extremely difficult to affirm if new phases are formed at the interface. Thus, it is necessary to ally those characterization techniques to the chemical analysis of the energy dispersive spectroscopy (EDS), in order to identify new phases formed during the tests and to determine possible chemical reactions and the mechanisms that act during corrosion.

The purpose of this work is to demonstrate that the characterization techniques association is an effective tool to characterize ceramic-metal interfaces.

EXPERIMENTAL

Materials

Crucibles with high relative density (98%) were produced using a ceramic powder mixture composed of 86 vol.% of a-Si₃N₄ (UBE Industries-Japan) and 14 vol.% of a-Al₂O₃ (BAIKALOX- Germany) and a mixed concentrate of yttrium and rare earth oxides, CTR₂O₃ (FAENQUIL-DEMAR-Brazil, obtained by alcaline fusion of the mineral xenotime, precipitation by oxalic acid and subsequent calcination), in the stoichiometry of the intergranular phase Y₃Al₅O₁₂. This mixture was milled for 6 h (1000 rpm) and compacted by isostatic pressing (100 MPa). After this step the compacts were sintered under 1.5 MPa of N₂ atmosphere at a heating rate of 10 ºC/min up to 1800 ºC, with an isothermal holding time of 30 min and subsequently heating up to 1900 ºC with an isothermal holding time of 2 h. After sintering, the samples were submitted to a heat treatment for devitrification of the intergranular phase at 1400 ºC during 24 h, under 0.1 MPa of N₂ atmosphere [1, 12].

Corrosion Tests

In the sintered crucibles, aluminum bars were melted and treated during 30 d at 1150 ºC in air. The objective of these tests was to verify the behavior of the dense Si₃N₄ ceramics in similar conditions to industrial aluminum melting, at temperatures where the chemical reactions between Si₃N₄ and aluminum are favored thermodynamically [7, 9]. For the accomplishment of the tests, the aluminum bars had their surfaces cleaned by pickling with a NaOH solution (10%) during 20 min. The crucibles with aluminum were put into an electric furnace, in air. After testing, the crucibles were cut and the cross-section embedded in phenolic resin for the metallographic preparation by grinding and polishing to a 1 mm finish.

After the corrosion tests, the internal surface of the crucibles was analyzed by X-ray diffraction (XRD) with the objective to identify phases in the system, using Cu_ka radiation, at 0.05 degree/s and 2q between 10º and 80º.

The microstructural evaluation of the Si₃N₄/Al interface of the samples was examined using mixed emission of backscattered electrons (BSE) and secondary electrons (SE) [13] in a LEO 1450VP Scanning Electron Microscope, and chemical analysis by energy dispersive spectroscopy (EDS) in the contact region.

RESULTS AND DISCUSSION

Microstructural Analysis

The crucibles were cut and the cross sections were ground and polished as described previously. Fig. 1 presents the microstructures of the cross sections of crucibles tested during 30 d, at 1150 ºC, focusing on the contact area between aluminum and the Si₃N₄ crucibles.

The presence of three well defined phases is observed, being the central phase, probably, the result of the corrosion between aluminum and Si₃N₄. It is noticed, that the central film, of darker tonality, is uniform and with a thickness of approximately 10 mm. By this phase tonality, it is supposed that the atoms composing this phase have smaller atomic numbers than the one of the Si₃N₄ with intergranular phase (Z_interface < Z_{Si3N4+Al203+CTR203}), since darker phases are due to smaller atomic numbers [13].

Chemical analysis by energy dispersive spectroscopy (EDS) was used to identify the elements and the quantity of each element in the region analyzed, besides verifying spectrum differences among the three areas analyzed, with prominence for the central area. Fig. 2 presents the microstructure of the contact region Al/ Si₃N₄, indicating the areas analyzed by EDS.

Based on Fig. 2, EDS measurements were carried out and the results in form of EDS spectra are presented in Fig. 3.

In region 1 of Fig. 3, EDS analysis shows the presence of Al confirming that this is the region where aluminum was deposited into the crucibles for the corrosion tests. The Au peak in the spectrum is due to the recovering used in the samples metalization. The identification of only Al in region 1 is insufficient to affirm that only pure aluminum exists in this area. The elements O and N are not identified with clarity by EDS analysis [13]. Thus, also in this area oxides such as Al₂O₃ can exist. This hypothesis is reinforced by the fact that aluminum is oxidized easily in contact with the atmosphere, even at room temperature [14].

Region 3, shown in Fig. 2, corresponds to the Si₃N₄ crucibles. The presence of the Si is observed, as justified by the chemical composition of Si₃N₄. The presence of the Al in this region is due to the intergranular phase formed during sintering, since a relatively large amount of Al₂O₃ has been used as additive. The other oxide used as additive was Y₂O₃ whose element Y possesses identical relative position to Si, thus difficulting the identification of these elements.

Fig. 3.2 shows the spectrum corresponding to the interface. A large signal is observed corresponding to Si and a small signal to Al. Comparatively, it is observed that the intensity of the Al peak is much smaller than that of the region 3.

This takes us to believe that a different phase is present in that region, composed for the major part of Si, and due to the tonality of this phase, and this new formed phase possesses Z < Z_Si3N4. Besides, the Al can be some residual content or some interference during the scan, since, in spite of the EDS analysis has been executed pontually, this kind of analysis detects chemical elements present in the interaction volume, i.e. in an area of 1.5 mm on the surface and 3 mm depth of the sample.

X-ray Diffraction

The X-ray diffraction technique was used to complement the corrosion interface analysis in the samples tested during 30 d at 1150 ºC. Two different regions were analyzed after the tests: 1) the surface of the crucibles subjected to oxidation without contact with aluminum during the corrosion tests and; 2) the region in contact with aluminum during the corrosion test. Fig. 4 presents a layout with the regions analyzed by XRD, microstructure of the surface oxidation and the diffraction patterns corresponding to the analyzed regions.

The SEM images of this region present typically SiO₂ in the form cristoballite [15-17]. This microstructure originates from the Si₃N₄ oxidation that is thermodynamically favored at this temperature and atmosphere [15, 16]. The X-ray diffraction pattern of this region confirms this supposition. The presence of SiO₂ and Si₂N₂O is verified, demonstrating that the Si₃N₄ crucibles have been oxidized, according to proposed reactions (equations A and B) [16].

The second observed region presented a dark and adherent film. During cooling, the solid Al formed shrinks more than the Si₃N₄ crucible due to the higher thermal expansion coefficient and can be easily retrieved. This Si₃N₄ surface was analyzed by XRD. In region 2 the diffraction pattern shows a degree of underground noise, characteristic for the presence of amorphous phases, and Al₂O₃ in several structures, denominated transitory aluminas that are intermediary phases for the crystallization of a-Al₂O₃ [14]. This indicates that, in the internal surfaces of the crucibles, aluminum has been oxidized forming the transitory phases of Al₂O₃. The Si₃N₄ peaks identified in this diffraction pattern correspond to the crucible material.

Final discussion

Gathering the information generated by the various characterization techniques, we propose a theory to describe the process of corrosion in the Si₃N₄ crucibles: the crucibles were heated under oxidizing atmosphere and solid aluminum bars were deposited, with superficial Al₂O₃ already present. This solid and resistant Al₂O₃ film represents a barrier for the contact between liquid Al and solid Si₃N₄. On the other hand, the Si₃N₄ surface during heating up to 900 ºC also oxidized. In this way in the accomplished corrosion tests, the true contact interface was Al₂O₃/SiO₂. For the microstructural analysis (Fig. 1), three different phases are observed.

By the observed tonality differences (identification of atomic number of the phases), we propose that supposedly Al₂O₃ in the area of Al, due to the sample was exposed to atmosphere oxidizer, during the preparation of the samples (sanding and polish), the enough time to oxidize the aluminum surface. Comparing the tonality of the intermediary phase (interface) with the two majority regions, we identified, by difference of atomic number, which the presented phases: Al₂O₃ has Z = 50, while for Si₃N₄ Z = 70.

Starting from the microstructural analysis and the X-ray diffraction results, for the SiO₂ presence, (Z = 30), we can affirm that the interface presented in Figs. 1 and 2 is an Si₃N₄ oxidation film composed mainly of SiO₂ (Z_SiO2 <Z_Al2O3 <Z_Si3N4), which was formed in the contact region of Al/Si₃N₄, and that was formed due to the presence of oxygen in the system, motivated by the lack of existent contact of solid phases.

CONCLUSIONS

High corrosion resistant Si₃N₄ ceramics crucibles were produced using an alternative sintering additive, CTR₂O₃, produced at this laboratory. This corrosion behavior for liquid aluminum, at 1150 ºC, in air, is due to the oxidized surface layer of SiO₂, stable and adherent, that promotes the adherence of Al₂O₃ present on the aluminum surface. Starting from the scanning electronic microscopy characterizations, the chemical analysis for energy dispersive spectroscopy, and phase analysis by X-ray diffraction, the detected phases could be identified and the mechanisms present during the corrosion tests were defined.

ACKNOWLEDGEMENT

The authors would like to thank to FAPESP (Processes 99/08976-8 and 01/08682-6) for financial support.

Rec. 25/02/03, Ac. 18/07/03

References

• [1] K. Strecker, R. Gonzaga, S. Ribeiro, M. J. Hoffmann, Mater. Lett. 45 (2000) 39-42.
• [2] G. Ziegler, J. Heinrich, G. Wotting, J. Mater. Sci. 22 (1987) 3041-86.
• [3] K. D. Mörgenthaler, H. Bühl, in "Tailoring of Mechanical Properties of Si₃N₄ Ceramics", ed. M.J. Hoffmann et al. (1994) p. 429-441, Dordrecht, NATO ASI Series.
• [4] N. S. Jacobson, J. Am. Ceram. Soc. 76, 1 (1993) 3-28.
• [5] H. Schwartz, Structural Ceramics 8 (1992) 8.1-8.85.
• [6] S. Tanimoto, K. Toyoda, K. Murai. Jap. Kokai Tokkyo Koho 85-142190 A2 (1985) 8 pp. from C.A 104, 1986, no. 23093.
• [7] U. Schwabe, L. R. Wolff, F. J. J. V. Loo, G. Ziegler, J. Eur. Ceram. Soc. 9 (1992) 407-415.
• [8] L. Mouradoff, A. Lachau-Durand, J. Desmaison, J. C. Labbe, O. Grisot, R. Rezakhanlou, J. Eur. Ceram. Soc. 13 (1994) 323-28.
• [9] L. Mouradoff, P. Tristan, J. Desmaison, J. C. Labbe, M. Desmaison-Brut, R. Rezahanlou, Key Eng. Mater. 113 (1997) 177-186.
• [10] R. Sangiorgi, Corrosion of Advanced Ceramics, Kluwer Academic Publishers, Netherlands (1994) 261-284.
• [11] M. Naka, Ultramicroscopy 39 (1991) 128-134.
• [12] K. Keller, T. Mah, T. A. Parthasarathy, Ceram. Eng. Sci. Proc. 11, 7-8 (1990) 1122-33.
• [13] J. I. Goldstein, D. E. Newbury, D. C. Joy, Advanced Scanning Electron Microscopy and X-Ray Microanalysis, Plenum Press, USA (1997) 454 p.
• [14] W. H. Gitzen, Alumina as a ceramic material, USA (1985) 253p.
• [15] R. J. Fordham, High Temperature Corrosion of Technical Ceramics, Elsevier Appl. Sci. (1990) 229p.
• [16] H. Du, R. E. Tressler, K. E.Spear, J. Electrochem. Soc. 136, 11 (1989) 3210-15.
• [17] J. Echeberria, F. Castro, Mater. Sci. Technol. 6 (1990) 497-503.
• [18] M. K. Cinibulk, G. Thomas, J. Am. Ceram. Soc. 75, 8 (1992) 2044-49.

Publication Dates

• Publication in this collection
  01 Apr 2004
• Date of issue
  Dec 2003

History

• Accepted
  18 July 2003
• Received
  25 Feb 2003