Influence of additive system (Al2O3-RE2O3 , RE = Y, La, Nd, Dy, Yb) on microstructure and mechanical properties of silicon nitride-based ceramics

REGULAR ARTICLES

Influence of additive system (Al₂O₃-RE₂O₃ , RE = Y, La, Nd, Dy, Yb) on microstructure and mechanical properties of silicon nitride-based ceramics

Juliana Marchi^I,*; Cecilia Chaves Guedes e Silva^II; Bruno Batista Silva^III; José Carlos Bressiani^III; Ana Helena de Almeida Bressiani^III

^ICentro de Ciências Naturais e Humanas,Universidade Federal do ABC, Av. Santa Adélia, 166, Bangu, 09210-170 Santo André - SP, Brazil

^IICentro Tecnológico da Marinha em São Paulo - CTMSP, Av. Prof Lineu Prestes, 2468, Cidade Universitária, 05508-000 São Paulo - SP, Brazil

^IIIInstituto de Pesquisas Energéticas e Nucleares - IPEN/CNEN, Av. Prof. Lineu Prestes, 2242, Cidade Universitária, 05508-000 São Paulo - SP, Brazil

ABSTRACT

Silicon nitride based ceramics have been widely used as structural ceramics, due mainly to their thermo-mechanical properties such as high density, high thermal shock resistance, corrosion resistance and chemical stability. The aim of this study was to determine the influence of rare earth and aluminum oxide additions as sintering aids on densification, microstructure and mechanical properties of silicon nitride. Silicon nitride mixtures with 91 wt. (%) Si₃N₄ and 9% wt. (%) additives were prepared and sintered. The density, microstructure and mechanical properties of the sintered specimens of these mixtures were determined. In most specimens, scanning electron microscopic examination and X ray diffraction analysis revealed elongated grains of β-Si₃N₄ with aspect ratio of about 2.0 and dispersed in a glassy phase. The density of the sintered specimens was higher than 94% of the theoretical density (td) and specimens with La₂O₃ and Al₂O₃ additions exhibited the highest value. The results of this investigation indicate that the rare earth ion size influences densification of silicon nitride, but this correlation was not observed in specimens containing two different rare earth oxides. The hardness values varied in direct proportion to the density of the specimens and the fracture toughness values were influenced by the composition of the intergranular glassy phase.

Keywords: silicon nitride, sintering, rare earth additives, mechanical properties

1. Introduction

Silicon nitride based ceramics have been used as structural materials in a wide range of applications due to their high hardness, high heat resistance, high fracture toughness and excellent wear resistance. These applications include heat exchangers, turbine and automotive engine components, valves and cam roller followers for gasoline and diesel engines, insulators, electronic substrates, high Tc superconductors, tool bits, wear surfaces, etc^1,2,3. Even though silicon nitride ceramics have found a variety of applications, a number of issues still remain to be addressed for it to reach its full potential. This includes optimization of processing routes, choice of additives and sintering parameters.^4,5,6,7,8

Recent studies on silicon nitride include fabrication of near-net-shaped specimens⁹ and development of porous ceramics with several rare earth additives¹⁰. It has been suggested that porous ceramics with tailored microstructures are potential structural component materials¹¹. Silicon nitride is considered to be a promising biomaterial as it is non-cytotoxic¹². Moreover, total hip arthroplastic bearings have been successfully manufactured from silicon nitride powder¹³. Despite these studies, a suitable additive to increase the density of silicon nitride based ceramics is still being sought. It is well known that silicon nitride has two polymorphic states, α and β, with differences in the sequence of Si-N stacking layers. Besides the Si-N covalent bond and the low diffusion coefficient, additives are required to promote densification¹⁴. These additives are in general oxides, such as alumina, yttria, magnesia and rare earth oxides¹⁵ that react with the silica layer on the silicon nitride powder surfaces to form a liquid phase. Liquid phase sintering involves α-Si₃N₄ particle rearrangement, dissolution of the α-phase, diffusion of Si and N atoms and precipitation of the β-phase¹⁶, followed by grain growth¹⁷.

The type and total amount of additives used can be adjusted to optimize the process and to increase the desired properties of the final silicon nitride-based ceramics, depending on the application. These modifications render controlled microstructures that are composed of elongated grains. Because most rare earth additives do not form a solid solution with Si₃N₄, the liquid phase, after cooling, forms at grain boundaries as an amorphous or partially crystalline phase¹⁸. In other words, the silicon nitride grains are surrounded by an intergranular glass or an intergranular film containing the rare earth element³. It has been mentioned that the glassy phase deteriorates high temperature creep and strength of the ceramics¹⁹.

Changes in microstructure are closely related to mechanical properties. In this context, increase in fracture toughness of such materials has been reported¹⁷. One way to improve mechanical properties is to control the formation of elongated grains through anisotropic grain growth during sintering¹⁷. This can be circumvented by using different sintering additives, especially rare earth elements²⁰.

According to Becher et al.²¹, the aspect ratio of silicon nitride grains is influenced by the size of the rare earth element used as an additive, and this effect is valid for a wide range of rare earths. However, studies so far have not produced a clear relationship. While some studies have shown that an increase in ionic radius of such elements causes an increase in the aspect ratio of silicon nitride grains²², others have concluded the opposite, i.e. additions of rare earth elements with larger ionic radii tend to reduce the liquid viscosity, which in turn decreases the aspect ratio of the silicon nitride grains²³. Moreover, since the rare earth ion acts as a network modifier in glasses, a decrease in its radius increases the bond strength between the rare earth ion and the surrounding oxygen²⁴.

Hence, the objective of this investigation was to evaluate the effect of alumina and rare earth oxide additions on the densification, microstructure and mechanical properties of silicon nitride ceramics.

2. Materials and Methods

The following starting materials were used: powders of Si₃N₄ (M11, Hermann C. Starck; with 92.7% α-Si₃N₄ and 1.14% wt oxygen), Y₂O₃ (Hermann C. Starck; purity > 99.9%) and other rare earth oxides (Sigma-Aldrich, purity > 99.9%) - La₂O₃, Nd₂O₃, Dy₂O₃ and Yb₂O₃.

Silicon nitride mixtures containing 9 wt. (%) additives (Table 1) were prepared using an attritor mill with silicon nitride spheres and isopropyl alcohol as the liquid media.

The milled and homogenized powder mixtures were dried at 90 °C and uniaxially compacted at 50 MPa followed by cold isostatic pressing at 200 MPa. The samples were then sintered at 1750 °C for 60 minutes in a furnace (Astro 1000, 4560, FP 20, Thermal -Technology Inc.) with graphite heating elements and a nitrogen atmosphere. A powder bed of silicon nitride mixed with the same oxides used as additives was prepared to avoid evaporation of the oxide and decomposition of the silicon nitride.

Density measurements were performed using the Archimedes method and the final densities of the samples were expressed in terms of the theoretical density for each mixture composition. The theoretical densities were determined using the rule of mixtures.²⁵

Sintered samples were analyzed using X ray powder diffraction analysis (XRD; Siemens^® D5000 X ray powder diffractometer, Cu K_α) to identify the crystalline phases. To observe the crystal shapes and grain sizes as well as their distribution, scanning electron microscopy (SEM, XL 30, Philips^®) was used. The sample surfaces were polished and plasma-etched using SF₆ and O₀ in a 3:1 ratio (Polotron, PT, 7160). The aspect ratio was estimated using the Quantikov software²⁶.

The mechanical properties of the different silicon nitride samples were determined from Vickers hardness tests. The fracture toughness was then obtained from the hardness indents (Zwick & Co., KG), using a 100 N load. Ten indents were made on the surfaces of samples corresponding to every silicon nitride composition. The hardness indents were observed in an optical microscope and lengths of the diagonals measured using an Image Tool software²⁷. Vickers hardness (Hv) was determined using Equation 1. The crack propagation behavior of silicon nitride based ceramics was also evaluated from the scanning electron micrographs of the indents.

The Poisson's ratio (ν) and Young's modulus (E) were determined using a 200 MHz ultrasonic pulser-receiver (Panametrics, USA, 5900 PR), 20 MHz longitudinal pulses, as described by Yoshimura et al.²⁸ The measured bulk density, longitudinal velocity and shear velocity were used as the parameters. Ten measurements at both velocities were made on polished samples with composition corresponding to the different mixtures.

The fracture toughness was evaluated from Equation 2, using the method proposed by Anstis et al.²⁹

P = load applied, d = diagonal impression

ε = constant of deformation geometry, 0.016, c = crack dimension (from centre indents).

3. Results and Discussion

The densities of the different silicon nitride samples are summarized in Table 2. It can be seen that the theoretical density values are dependent on the additive used. The final densities were between 94 and 98.5% of the theoretical density (td), indicating the importance of silicon nitride as a structural material.

The results obtained with samples containing alumina and a single rare earth oxide (SN-Y, SN-La, SN-Nd, SN-Dy and SN-Yb) suggest that the density values tend to increase with increasing ionic radius of the rare earth (ionic radii of the rare earth elements are sequenced in Equation 3). As a result, the SN-La sample (with 98.5% td) had the highest density and the SN-Yb sample (with 94% td), the lowest.

R_La > R_Nd > R_Y > R_Dy > R_Yb (3)

The results obtained in this study are in agreement with Becher et al.³⁰. They showed that the viscosity of bulk Si-RE-Al oxynitride glasses with specific compositions and in the 800-1100 °C range, increased with decrease in rare earth ion radius. Thus rare earth cations with larger radii (La³⁺) contribute towards decrease in the eutectic temperature of the mixture. According to Murakami et al.³¹, the energy of formation of a solid phase (H) that coexists with the liquid is dependent on the ionic radius of the RE element. SN-La exhibited the lowest temperature for liquid formation, due to the large radius of La³⁺, and this implies a liquid with lower viscosity. This lower viscosity aids mass transport, which in turn increased densification of silicon nitride-based materials. On the other hand, SN-Yb revealed the highest temperature for liquid formation, due to the small radius of Yb³⁺. This implied a liquid with higher viscosity, hindered mass transport and consequent reduction in densification.

When two rare earth oxides are used simultaneously, the final density results require careful analysis as these mixtures (SN-Y/La, SN-Y/Yb and SN-Nd/Dy) contain a lesser amount of alumina compared to those containing a single rare earth oxide. This observation suggests that the sintering behavior of such materials cannot be directly related to the ionic radius of the rare earth elements used as additives. The higher amounts of rare earth oxides in these mixtures probably modify the viscosity and amount of the liquid at the sintering temperature³², resulting thereby in changes in the densification process and consequently, the microstructure of the silicon nitride-based materials. The density of SN-Y/La is in between those observed for SN-Y and SN-La. The density of SN-Y/Yb is close to that of SN-Y, even though SN-Yb exhibited a slightly lower density. The density of SN-Nd/Dy was higher than that observed for SN-Nd and SN-Dy.

The X ray diffraction patterns of the silicon nitride samples are shown in Figure 1. The β-Si₃N₄ phase can be observed in all the samples indicating dissolution of the Si₃N₄ phase and re-precipitation of the β-Si₃N₄ phase during sintering. This is a typical behavior in liquid phase sintering mechanisms³³. Secondary crystalline phases did not form during cooling, indicating that the secondary phase is amorphous in the grain boundary regions.

The mechanical properties determined from the indents on the different polished silicon nitride samples are summarized in Table 3. Vickers hardness values were estimated to be between 11.5 and 15 GPa (Table 3). In general, higher density samples reveal higher Vickers hardness values. As expected, SN-La had the highest Vickers hardness. No significant differences were found among SN-Nd, SN-Dy and SN-Yb, but these had lower hardness values than SN-La and SN-Y.

The Young's modulus values were between 298 and 264 GPa (Table 3), slightly lower than the typical value for silicon nitride of 300 GPa³⁴. The Young's modulus is closely related to the final density and Vickers hardness of silicon nitride. Moreover, a decrease in ionic radius implies a decrease in Young's modulus. The Poisson ratios varied between 0.26 and 0.31 (Table 3) without revealing any pattern, and in accordance with data in the literature³⁵.

The fracture toughness values were estimated to be between 4.17 and 4.83 MPa.m^1/2. The values are similar to each other when only one rare earth oxide is used as a sintering aid. The same behavior was also observed when two oxides were used simultaneously, but the fracture toughness values were higher, compared to that of samples with a single rare earth oxide.

The microstructures of polished and plasma-etched cross-sections of sintered silicon nitride with different additives are shown in Figure 2 and these reveal uniformly distributed elongated grains of β-Si₃N₄. No significant differences in features were noticed when different oxides were used as additives. The estimated aspect ratio was approximately 2 for all the samples (Table 3).

The slightly lower fracture toughness of the silicon nitride samples can be attributed to the low volume fraction of needle-like grains (Figure 2). Moreover, this property is also dependent on the grain-boundary phase, which is similar in all the samples. As reported by³⁶ Becker et al., increase in interfacial de-bonding between the grains should result in high fracture toughness.

To study the crack propagation behavior in silicon nitride based ceramics, the Vickers indent (a) and the crack path (b) (shown in the example in Figure 3) were considered. It can be seen from the intergranular crack path, in all samples, that crack deflection occurred at the grain boundaries, resulting in the observed Si₃N₄ grain pullout. Similar toughening mechanisms tend to operate in the materials that have been studied here, independent of the additives.

As summarized by Rosenflaz³⁷, not only the microsctructure, composed of bimodal silicon nitride grain diameter distribution³⁸, but characteristics of the glass/grain interface, as determined by bond matching, are important to have a material with high fracture toughness. They showed that the decrease in rare earth ion size leads to a decrease in crack bridging (in terms of debonding length), and consequent lower fracture resistance. We did not observe this behavior, due probably to the method used to estimate the fracture toughness values.

Our results demonstrate the influence of intergranular phase composition on fracture toughness of silicon nitride-based ceramics. Samples with lower alumina content, i.e. samples containing two rare earth oxides, present a brittle interface compared with those with higher alumina content. A brittle interface induces intergranular fracture due to enhanced detachment between grains and the intergranular phase, promoting higher fracture toughness values²².

The aspect ratio is an important parameter to discuss the mechanical properties of silicon nitride based ceramics³⁹. Amongst rare earths, as shown by Hoffman et al.⁴⁰, there is an increase in the aspect ratio of silicon nitride grains (that grow dispersed in a Si-Λl-RE oxynitride glass) with increase in the ionic radius of the rare earth used as the additive. However, additions of alumina have a limited effect on this relationship, because the opposite behavior is observed⁵. These observations explain the similarity in aspect ratios observed in this study, amongst samples with different rare earth additives (~2, Table 3). Similar microstructure features amongst samples with different oxide additives can also imply similar aspect ratio values.

4. Conclusions

It can be concluded that the pressureless sintering conditions, associated with the amount and composition of additives, used in this study were adequate to obtain high density silicon nitride-based ceramics with relatively good mechanical properties. The results are related to the morphological features of several samples and especially to the aspect ratio of b-Si₃N₄ grains.

Comparison of the effects produced by the different additives revealed that the highest densification and Vickers hardness results were obtained in samples containing Al₂O₃ and La₂O₃. The use of two rare earth oxides simultaneously as sintering additives in combination with a lower alumina content increased the fracture toughness of the silicon nitride-based ceramics. The influence of the intergranular phase composition on fracture toughness of silicon nitride-based ceramics was also evaluated.

This revealed that the size of the rare earth ion can affect densification and, consequently, the mechanical properties. However, when rare earth oxides are used together and in combination with a lower alumina content, other factors are involved which hinder a direct correlation.

ACKNOWLEDGEMENTS

The authors would like to thank FAPESP (Fundação de Amparo à Pesquisa do Estado de São Paulo) and PRONEX/CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico) for their financial support.

The experimental work performed at the X ray laboratory (Laboratório de Difração de Raios X do Departamento de Mineralogia e Geotectônica do Instituto de Geociências da Universidade de São Paulo) and the Mechanical Laboratories (Laboratório de Ensaios Mecânicos do Departamento de Metalurgia e Materiais da Escola Politécnica da Universidade de São Paulo and Laboratório de Metalurgia e Materiais Cerâmicos do Instituto de Pesquisas -Tecnológicas - LMMC-IPT), are greatly appreciated. The authors also thank Catia Frederici and Marcelo Mendes Pinto for their help with the Young's Modulus experiments

Received: September 11, 2008;

Revised: April 7, 2009

• 1. Xikun L, Jing L, Like Q, Tong C, Guanming Q, Yanbin S. Composition, characteristics and development of advanced ceramic cutting tools. Journal of Rare Earths 2007; 25(2):287-294.
• 2. Danforth SC, Symons W, Nilsen KJ, Riman RE. Processing of silicon nitride powders. In: Binner JGP (Ed.). Advanced ceramic processing and technology Noyes: Willian Andrew Publishing; 1990. p. 39-71.
• 3. Berroth K, Prescher T. Development and industrial application of silicon nitride based ceramics. Key Engineering Materials. 2005; 287:3-9.
• 4. Mandal H, Thompson DP. New heat treatment methods for glass removal from silicon nitride and sialon ceramics. Journal of Materials Science. 2000; 35(24):6285-6292.
• 5. Chen WW, Sun WY, Wang PL, Li YW, Chen YB, Yan DS. Control of Microstructures in SiAlON Ceramics. Journal of the American Ceramic Society 2002; 85(1):276-278.
• 6. Ye F, Hoffmann MJ, Holzer S, Zhou Y, Iwasa M. Effect of the amount of additives and post-heat treatment on the microstructure and mechanical properties of yttrium-sialon ceramics. Journal of the American Ceramic Society 2003; 86(12):2136-2142.
• 7. Chockalingam S, Earl DA. Mechanical properties of 2.45 GHz microwave sintered Si₃N₄-Y₂O₃ -MgO-ZrO₂ system. Journal of the European Ceramic Society. 2009. In Press.
• 8. Jones MI, Valecillos MC, Hirao K, Yamauchi Y. Grain growth in microwave sintered Si₃N₄ ceramics sintered from different starting powders. Journal of the European Ceramic Society 2002; 22(16):2981-2988.
• 9. Bocanegra-Berna MH, Matovic B. Dense and near-net-shape fabrication of Si₃N₄ ceramics. Materials Science and Engineering: A. 2009; -500(-1-2):130-149.
• 10. Plucknett KP, Quinlan M, Garrido L, Genova L. Microstructural development in porous-Si₃N₄ ceramics prepared with low volume RE₂O₃-MgO-(CaO) additions (RE = La, Nd, Y, Yb). Materials Science and Engineering: A 2008; 489(1-2):337-350.
• 11. Ohji T. Microstructural design and mechanical properties of porous silicon nitride Ceramics. Materials Science and Engineering: A. 2008; 498(1-2): 5-11.
• 12. Guedes e Silva CC, Higa OZ, Bressiani, JC. Cytotoxic evaluation of silicon nitride-based ceramics. Materials Science and Engineering: C.  2004; 24(5):643-646.
• 13. Bal BS, Khandkar A, Lakshminarayanan R, Clarke I, Hoffman AA, Rahaman MN. Fabrication and Testing of Silicon Nitride Bearings in Total Hip Arthroplasty. The Journal of Arthroplasty. 2009; 24(1):110-116.
• 14. Ziegler G, Heinrich J, Wötting G. Relationships between processing, microstructure and properties of dense and reaction-bonded silicon nitride. Journal of Materials Science 1987; 22(9):3041-3085.
• 15. Guanming Q, Xikun L, Tai Q, Haitao Z, Honghao Y, Ruiting M. Application of rare earths in advanced ceramic materials. Journal of Rare Earths 2007; 25(2):281-286.
• 16. Sanders WA, Mieskowski DM. Strength and microstructure of sintered Si₃N₄ with rare-earth-oxide additions. American Ceramic Society Bulletin. 1985; 64(2):304-309.
• 17. Peillon FC, Thevenot F. Microstructural designing of silicon nitride related to toughness. Journal of the European Ceramic Society 2002; 22(3):271-278.
• 18. Chihara K, Hiratsuka D, Shinoda Y, Akatsu T, Wakai F, Tatami J. et al. High-temperature compressive deformation of SiAlON polycrystals containing minimum amount of intergranular glass phase. Materials Science and Engineering: B 2008, 148(1-3):203-206.
• 19. Meléndez-Martínez JJ, Domínguez-Rodríguez A. Creep of silicon nitride. Progress in Materials Science. 2004, 49(1):19-107.
• 20. Faber KT, Evans AG. Crack deflection processes - 1: Theory. Acta Metallurgica. 1983; 31(4):565-576.
• 21. Becher PF, Painter GS, Shibata N, Satet RL, Hoffmann MJ, Pennycook SJ. Influence of additives on anisotropic grain growth in silicon nitride ceramics. Materials Science and Engineering: A. 2006; 422(1-2): 85-91.
• 22. Becher PF, Sun EY, Hsueh CH, Alexander KB, Hwang SL, Waters SB, Westmouland CG. Debonding of interfaces between beta-silicon nitride whiskers and Si Al Y oxynitride glasses. Acta Materialia. 1996; 44(10):3881.
• 23. Satet RL, Hoffmann MJ. Grain growth anisotropy of beta-silicon nitride in rare-earth doped oxynitride glasses. Journal of the European Ceramic Society. 2004; 24(12):3437-3445.
• 24. Tanabe S, Hirao K, Soga N. Elastic properties and molar volume of rare-earth aluminosilicate glasses. Journal of the American Ceramic Society 1992; 75(3):503-509.
• 25. Chiang YM, Birnie DP, Kingery WD. Physical ceramics: principles for ceramic science and engineering. New York: Wiley; 1997.
• 26. Pinto LCM. Quantikov: a microestrutural analyser [tese]. São Paulo: Universidade de São Paulo; 1996.
• 27. Wilcox D, Dove B, McDavid D, Greer D. Image Tool for Windows, Version 3.0. San Antonio: University of Texas Health Science Center in San Antonio; 1995-2002.
• 28. Yoshimura HN, Molisani AL, Narita NE, Cesar PF, Goldenstein H. Porosity dependence of elastic constants in aluminum nitride ceramics. Materials Research 2007; 10(2):127-133.
• 29. Anstis GR, Chantikul P, Lawn BR, Marshall DB. A critical evaluation of indentation techniques for measuring fracture toughness: I. Direct crack measurements. Journal of the American Ceramic Society. 1981; 64(9):533-542.
• 30. Becher PF, Ferber MK. The temperature dependent viscosity of SiREAl-based glasses as a function of N:O and RE:Al ratios where RE = La, Gd, Y and Lu. Journal of the American Ceramic Society. 2004; 87(7):1274-1279.
• 31. Murakami Y, Yamamoto H. Phase equilibria and properties of glasses in the Al₂O₃-Yb₂O₃ -SiO₂ system. Journal of the Ceramic Society Japan. 1993; 101(1316):1101-1106.
• 32. Shelby JE, Kohli JT. Rare-Earth Aluminosilicate Glasses. Journal of the American Ceramic Society. 2005; 73(1):39-42.
• 33. Ziegler G, Heinrich J, Wötting G. Review-relationshios between processing, microstructure and properties of dense and reaction-bonded silicon nitride. Journal of Materials Science. 1987; 22(22):3041-3085.
• 34. Sigulinski F, Boskovic S. Phase composition and fracture toughness of Si₃N₄-ZrO₂ with CeO₂ additions. Ceramics International 1999; 25(1):41-47.
• 35. Hampshire S. Engineering properties of nitrides. In: Schneider SJ. (Ed.). Ceramis and glasses: engineered materials handbook, 4. [S.L]: ASM International; 1991. p. 813-820.
• 36. Becher PF, Sun EY, Plucknett KP, Alexander KB, Hsueh CH, Lin HT. et al. Microstructural design of silicon nitride with improved fracture toughness. I. Effects of grain shape and size. Journal of the American Ceramic Society 1998; 81(11):2821-2830.
• 37. Rosen anz A. Silicon nitride and sialon ceramics. Current Opinion in Solid State and Materials Science. 1999; 4(5):453-459.
• 38. Ohji T, Hirao K, Kanzaki S. Fracture resistance behavior of highly anisotropic silicon nitride. Journal of the American Ceramic Society. 1995; 78(11):3125-3128.
• 39. Peillon FC, Thevenot F. Microstructural designing of silicon nitride related toughness. Journal of the European Ceramic Society. 2002; 22(3):271-278.
• 40. Hoffmann MJ. Microstructure: development, characterization and influence on mechanical properties. In: Hoffmann M, Petzow G. (Eds.). Tailoring of mechanical properties of Si₃N₄ ceramics The Netherlands: Kluwer Academic Publishers; 1994. p. 59-72.

Publication Dates

• Publication in this collection
  24 July 2009
• Date of issue
  June 2009

History

• Accepted
  07 Apr 2009
• Received
  11 Sept 2008