JOURNAL OF MATERIALS SCIENCE 40(2005)5443-5450 Design of Si3N4-based ceramic laminates by the residual stresses N. ORLOVSKAYA Drexel University, Philadelphia, PA 19104, USA E-mail: orlovsk @ drexeledu J KUEBLER EMPA, Swiss Federal Laboratories for Materials Testing and Research, CH-8600 Duebendorf switzerland V. SUBBOTIN M. LUGOVY Institute for Problems of materials science, 03142 Kiev Ukraine Ceramic laminates with strong interfaces between layers are considered a very promising material for different engineering applications because of the potential for increasing fracture toughness by designing high residual compressive and low residual tensile in separate layers. In this work, Si3 Na/Si3N4-TiN ceramic laminates with strong interfaces were manufactured by rolling and hot pressing techniques. The investigation of their mechanical properties has shown that the increase in apparent fracture toughness can be achieved for the Si3N4/Si3 N4-20 wt %TiN composite, while further increase of Tin content in the layers with residual tensile stresses lead to a formation of multiple cracks, and as a result, a significant decrease in the mechanical performance of the composites Micro-Raman spectroscopy was used to measure the frequency shift across the Si3Na/Si3N4-20 wt %TiN laminate. These preliminary Raman results can be useful for further analysis of residual stress distribution in the laminate 2005 Springer Science Business Media, Inc 1. Introduction apparent fracture toughness of ceramics by creating a Ceramics have found their use in numerous crosscut- layer with compressive stresses on the surface. In such a ng industrial applications because of excellent hard- way, surface cracks will be arrested and achieve higher ness, wear, corrosion resistance, and ability to with- failure stresses [5]. The variable layer composition,as stand high temperatures. However, ceramics'reliabil- well as the systems geometry, allows the designer to y and ductility compared to metals are not very high. control the magnitude of the residual stresses in such a The best approach to increasing the fracture toughness way that compressive stresses in the outer layers near which enables the structural application of ceramics is the surface increase strength, flaw tolerance, fatigue through the development of ceramic composites. Fiber strength, resistance to oxidation, and stress corrosion reinforced composites demonstrate the highest fracture cracking. In the case of symmetrical laminates, thi oughness and damage tolerance. However, since these can be done by choosing layer compositions such that materials have a very high density of weak interfaces, the coefficient of thermal expansion(CTE) of the odd they are not very strong. In addition, their high cost layers is smaller than the CtE of the even ones. The limits their commercial applications. Particulate com- changes in compressive and tensile stresses depend on posites are less expensive to manufacture, but com- the mismatch of CTE's, Youngs moduli, as well as on pared to monolithic ceramics, their fracture toughness the thickness ratio of layers(even/odd). However, if the increases are insignificant. The several publications on compressive stresses exist only at or near the surface ceramics show that the use of layered materials is the of ceramics and are not placed inside the material, they most promising method for controlling cracks by de- will not effectively hinder internal cracks and flaws flection, microcracking, or internal stresses [1-3]. Lay- [6, 7] ered structures clearly offer the key to greater reliabilit It is clear that control over the mechanical behay at a moderate cost and new applications may result as ior and reliability of laminates can be obtained only more complex structures are tailored to specific appli- through design, control of residual stresses, and redis- cations [4] tribution of stresses during loading in laminate mate The way to achieve the highest possible mechanical rials. The sign and value of residual stresses can be properties is to control the level of residual stresses in established by theoretical prediction. There exists a individual layers. One can increase the strength and theoretical background that allows for the design of 005 Springer Science Business Media, Inc. DOI:10.1007/s10853-005-1918-7 5443
JOURNAL OF MATERIALS SCIENCE 4 0 (2 0 0 5 ) 5 4 4 3 –5 4 5 0 Design of Si3N4-based ceramic laminates by the residual stresses N. ORLOVSKAYA Drexel University, Philadelphia, PA 19104, USA E-mail: orlovsk@drexel.edu J. KUEBLER EMPA, Swiss Federal Laboratories for Materials Testing and Research, CH-8600, Duebendorf, Switzerland V. SUBBOTIN, M. LUGOVY Institute for Problems of Materials Science, 03142, Kiev, Ukraine Ceramic laminates with strong interfaces between layers are considered a very promising material for different engineering applications because of the potential for increasing fracture toughness by designing high residual compressive and low residual tensile stresses in separate layers. In this work, Si3N4/Si3N4-TiN ceramic laminates with strong interfaces were manufactured by rolling and hot pressing techniques. The investigation of their mechanical properties has shown that the increase in apparent fracture toughness can be achieved for the Si3N4/Si3N4-20 wt.%TiN composite, while further increase of TiN content in the layers with residual tensile stresses lead to a formation of multiple cracks, and as a result, a significant decrease in the mechanical performance of the composites. Micro-Raman spectroscopy was used to measure the frequency shift across the Si3N4/Si3N4-20 wt.%TiN laminate. These preliminary Raman results can be useful for further analysis of residual stress distribution in the laminate. C 2005 Springer Science + Business Media, Inc. 1. Introduction Ceramics have found their use in numerous crosscutting industrial applications because of excellent hardness, wear, corrosion resistance, and ability to withstand high temperatures. However, ceramics’ reliability and ductility compared to metals are not very high. The best approach to increasing the fracture toughness which enables the structural application of ceramics is through the development of ceramic composites. Fiber reinforced composites demonstrate the highest fracture toughness and damage tolerance. However, since these materials have a very high density of weak interfaces, they are not very strong. In addition, their high cost limits their commercial applications. Particulate composites are less expensive to manufacture, but compared to monolithic ceramics, their fracture toughness increases are insignificant. The several publications on ceramics show that the use of layered materials is the most promising method for controlling cracks by de- flection, microcracking, or internal stresses [1–3]. Layered structures clearly offer the key to greater reliability at a moderate cost and new applications may result as more complex structures are tailored to specific applications [4]. The way to achieve the highest possible mechanical properties is to control the level of residual stresses in individual layers. One can increase the strength and apparent fracture toughness of ceramics by creating a layer with compressive stresses on the surface. In such a way, surface cracks will be arrested and achieve higher failure stresses [5]. The variable layer composition, as well as the system’s geometry, allows the designer to control the magnitude of the residual stresses in such a way that compressive stresses in the outer layers near the surface increase strength, flaw tolerance, fatigue strength, resistance to oxidation, and stress corrosion cracking. In the case of symmetrical laminates, this can be done by choosing layer compositions such that the coefficient of thermal expansion (CTE) of the odd layers is smaller than the CTE of the even ones. The changes in compressive and tensile stresses depend on the mismatch of CTE’s, Young’s moduli, as well as on the thickness ratio of layers (even/odd). However, if the compressive stresses exist only at or near the surface of ceramics and are not placed inside the material, they will not effectively hinder internal cracks and flaws [6, 7]. It is clear that control over the mechanical behavior and reliability of laminates can be obtained only through design, control of residual stresses, and redistribution of stresses during loading in laminate materials. The sign and value of residual stresses can be established by theoretical prediction. There exists a theoretical background that allows for the design of 0022-2461 C 2005 Springer Science + Business Media, Inc. DOI: 10.1007/s10853-005-1918-7 5443
laminated ceramics [8]. There have also been a number material are [19 of experimental studies of laminated ceramics that were conducted using these models, attempting to maximize f2(ar2-ar1)△T the mechanical properti CrI EIfi+ E2f2 Silicon nitride is the most promising and well- developed ceramics for structural application because and of its outstanding mechanical properties as well as its superior wear resistance [9]. The addition of Tin to E1fi(ar1-ar2)△T Si3N4 leads to an increase of Youngs modulus, electri Or2= ELfi+ E2f2 cal conductivity, and CTE of Si3N4 ceramics [10]. B varying the amount of TiN in silicon nitride ceramics we can increase the Cte/Youngs modulus mismatch where E/=E/(1-v),fi="*Dm, f2="2. and develop composites with compressive and tensile E and V: is the elastic modulus and Poissons ratio of j-th component respectively, II and l2 are the thickness stresses in alternative layers. This may further improve of layers for the first and second component, aTI and the mechanical properties of laminates [11-13]. B Si3N4 belongs to the space group C 6h(P63/m) and T2 are the thermal expansion coefficients(CTE)of the first and second components respectively, AT is the irreducible representation for the optical phonons the difference between the joining temperature and the en re ported[ 14-17] current temperature, and h is the total thickness of the specimen Topic= 4Ag+ 2Au 3Bg+ 4Bu+ 2Elg 5E2g The choice of composition for Si3N4-based ceramic laminates is dependent on the coefficient of thermal 4Elu+ 2E expansion and Youngs modulus of the compound Four compositions of composite layers were used: where Ag, Elg and E2g modes are Raman active and a 1. Si3 N4-5 wt %Y203-2 wt% AlO3 and Elu are infrared active Raman and infrared active bands are mutually excluded since the crystal structure 2. TIN. 3. Si3N4(5 wt%Y2O3-2 wt %Al,O3)-20 wt %TIN; has a center of symmetry. The goal of this work is to study the interrelation 4. Si3 N4(5 wt %Y2O3-2 wt% Al2O3)-50 wt %TiN. etween structure, residual stresses, mechanical prop- The residual stresses in each laver of the erties, and fracture behavior of complex particulate layered Si3N4/Si3N4-TiN based composites Si3 N4/Si3 N4-20%TiN, Si3 N4/Si3 N4-50%Tin, and Si3N4/Tin laminates, each sample having different numbers of layers and known layer thickness, were calculated using Equations I and 2 [19]. The joining temperature, used to determine the residual stresses. 2. Analysis of residual stresses was assumed to be 1200Crather than the hot pressing In this work, two-component brittle layered compos- temperature of 1800C. It was found that these ites with symmetric macrostructure are considered. The materials are sufficiently soft at the temperature above lyers consisting of different components alternate one 1200C to have a zero stress state due to ductile glassy after another, but the external layers consist of the same phases at the grain boundaries. Youngs moduli and omponent. Thus, the total number of layers N in such a CTE's of the components were calculated by the rule composite sample is odd. The layers of the first compo- of mixture and are presented in Table L Results of the nent including two extermal(top)layers are designated residual stress calculations are shown in Table II by index 10=1), and the layers of the second compo- nent(internal)are designated by index 2(=2). The number of layers designated by index I is(N+1)/2 and 3. Experimental the number of layers designated by index 2 is(N-1)/2. a-Si3N4(dso= l um) and Tin (dso=3 um)was used The layer of each component has some constant thick- formixture preparation Grinding of mixtures of certain ness,and the layers of same component have identical compositions was done in the ball mill for 5 h. After grinding, the plastification and rolling of thin tapes was There are effective residual stresses in the lay ayers of ach component in the layered ceramic composite dur- ing cooling the difference in deformation due to th TABLE I Youngs moduli and CTE of the components different thermal factors of the componen Composition CTE. 1/K is accommodated by creep as long as the temperature is high enough. Below a certain temperature, which Si3Na-5 wt% Y203-2 wt%320 3×10-6 called the"joining"temperature, the different com- A1,O 935×10-6 ponents become bonded together and internal stresses SigN4(5 wt%Y2 03 335.62 3.826×10-6 appear. In each layer, the total strain after sintering is 2 wt%Al203)-20 wt. the sum of an elastic component and a thermal com ponent [18]. In the case of a perfectly rigid bondin Si3N4(5wt%Y2O3-2w%364.93 5378×10-6 the residual stresses in the layers of a two-component AlO3)-50 wt%TIN 5444
laminated ceramics [8]. There have also been a number of experimental studies of laminated ceramics that were conducted using these models, attempting to maximize the mechanical properties. Silicon nitride is the most promising and welldeveloped ceramics for structural application because of its outstanding mechanical properties as well as its superior wear resistance [9]. The addition of TiN to Si3N4 leads to an increase of Young’s modulus, electrical conductivity, and CTE of Si3N4 ceramics [10]. By varying the amount of TiN in silicon nitride ceramics, we can increase the CTE/Young’s modulus mismatch and develop composites with compressive and tensile stresses in alternative layers. This may further improve the mechanical properties of laminates [11–13]. β- Si3N4 belongs to the space group C2 6h (P63/m) and the irreducible representation for the optical phonons has been reported [14–17] optic = 4Ag + 2Au + 3Bg + 4Bu + 2E1g + 5E2g + 4E1u + 2E2u where Ag, E1g and E2g modes are Raman active and Au and E1u are infrared active. Raman and infrared active bands are mutually excluded since the crystal structure has a center of symmetry. The goal of this work is to study the interrelation between structure, residual stresses, mechanical properties, and fracture behavior of complex particulatelayered Si3N4/Si3N4-TiN based composites. 2. Analysis of residual stresses In this work, two-component brittle layered composites with symmetric macrostructure are considered. The layers consisting of different components alternate one after another, but the external layers consist of the same component. Thus, the total number of layers N in such a composite sample is odd. The layers of the first component including two external (top) layers are designated by index 1 (j = 1), and the layers of the second component (internal) are designated by index 2 (j = 2). The number of layers designated by index 1 is (N+1)/2 and the number of layers designated by index 2 is (N−1)/2 . The layer of each component has some constant thickness, and the layers of same component have identical thickness. There are effective residual stresses in the layers of each component in the layered ceramic composite. During cooling, the difference in deformation due to the different thermal expansion factors of the components is accommodated by creep as long as the temperature is high enough. Below a certain temperature, which is called the “joining” temperature, the different components become bonded together and internal stresses appear. In each layer, the total strain after sintering is the sum of an elastic component and a thermal component [18]. In the case of a perfectly rigid bonding, the residual stresses in the layers of a two-component material are [19]: σr1 = E 1E 2 f2(αT 2 − αT 1)T E 1 f1 + E 2 f2 (1) and σr2 = E 2E 1 f1(αT 1 − αT 2)T E 1 f1 + E 2 f2 (2) where Ej = Ej/(1 − νj), f1 = (N+1)l1 2h , f2 = (N−1)l2 2h , Ej and Vj is the elastic modulus and Poisson’s ratio of j-th component respectively, l1 and l2 are the thickness of layers for the first and second component, αT1 and αT2 are the thermal expansion coefficients (CTE) of the first and second components respectively, T is the difference between the joining temperature and the current temperature, and h is the total thickness of the specimen. The choice of composition for Si3N4-based ceramic laminates is dependent on the coefficient of thermal expansion and Young’s modulus of the compounds. Four compositions of composite layers were used: 1. Si3N4-5 wt.% Y2O3-2 wt.% Al2O3; 2. TiN; 3. Si3N4 (5 wt.%Y2O3-2 wt.%Al2O3)-20 wt.%TiN; 4. Si3N4 (5 wt.% Y2O3-2 wt.% Al2O3)-50 wt.%TiN. The residual stresses in each layer of the Si3N4/Si3N4-20%TiN, Si3N4/Si3N4-50%TiN, and Si3N4/TiN laminates, each sample having different numbers of layers and known layer thickness, were calculated using Equations 1 and 2 [19]. The joining temperature, used to determine the residual stresses, was assumed to be 1200 ◦C rather than the hot pressing temperature of 1800 ◦C. It was found that these materials are sufficiently soft at the temperature above 1200 ◦C to have a zero stress state due to ductile glassy phases at the grain boundaries. Young’s moduli and CTE’s of the components were calculated by the rule of mixture and are presented in Table I. Results of the residual stress calculations are shown in Table II. 3. Experimental α-Si3N4 (d50 = 1 µm) and TiN (d50 = 3 µm) was used for mixture preparation. Grinding of mixtures of certain compositions was done in the ball mill for 5 h. After grinding, the plastification and rolling of thin tapes was T AB L E I Young’s moduli and CTE of the components Composition E, GPa CTE, 1/K Si3N4-5 wt.% Y2O3-2 wt.% Al2O3 320 3 × 10−6 TiN 440 9.35 × 10−6 Si3N4(5 wt.% Y2O3- 2 wt.%Al2O3)-20 wt.% TiN 335.62 3.826 × 10−6 Si3N4(5 wt.% Y2O3-2 wt.% Al2O3)-50 wt.%TiN 364.93 5.378 × 10−6 5444
TABLE II. Calculated residual stresses in Si3 Na based laminates Thickness of layers (um) Composition Si3N4 Si3N4 with Tin (MPa) (MPa) 4. Results and discussion Si3 Na/Sin 188246.5 4.1. Mechanical proper t o wt oTiN Mechanical properties such as the strength, Young Si3N4/2(Si3N4-245 279.5 151 modulus, and fracture toughness of the laminates are presented in Table Ill. The parameters of the tested 765515.5 50 wt oTiN laminates, such as composition and layer thickness are Si3 NaTiN 2467 1078 given in Table Il. Besides these four designs, one more design of Si3 N4/Si3 N4 laminate was used as a base for comparison. The laminates of this design were pre pared in the same way as the others, however, all layers were of the same composition. Therefore, no residual done. For rolling, a crude rubber(4 wt %)was added stresses can appear during cooling. It is worth noting to the mixture of powders as a plasticizer through a that both the Young's modulus and fracture toughness 39 solution in petrol. The powders were then dried up of these Sis N/Sis N4 laminates were measured to be on o a 2 wt% residual amount of petrol in the mixture. the same level as standard Si N4 ceramics prepared by After sieving powders with a 500 um sieve, granulated e standard powder route, which includes no rolling powders were dried up to the 0.5 wt. residual petrol. The strength of the Si3 N4/Si3 N4 laminate was less than A roll mill with 40 mm rolls was used for rolling. The that of the standard Si3 N4 ceramics with values of 507.6 velocity of rolling was 1.5 m/min. Working pressure +3.2 and 750+ 20.7 MPa, respectively. As one can varied from 10 MPa for a 64% relative density of tapes see from Table Ill, while the strength of Si3 N4/Si3N to 100 MPa for a 74% relative density. The thickness of 20wt. TiN laminates are approximately on the same tapes was either 0. 40.5 mm or 0.8-1.0 mm, the width level as the Sis N/Si N4 laminates, further increase of was 60-65 mm. Samples of ceramics were prepared by the TiN content to 50 and 100% results in a significant the hot pressing of tapes stacked together. Each layer decrease of both strength and Young's modulus. The contained one or a few tapes. Graphite dies were used measured fracture toughness of the Si3N4/TiN lami for hot pressing, and the hot pressing was performed nates also showed a decrease similar to strength and at a temperature of 1820C, with a dwelling time of Young's modulus values 20 min and a pressure of 30 MPa [20]. During hot The Si3 N4/Si3N4-20 wt TiN laminates showed an pressing,the shrinkage of layers occurred 3 times such increase in apparent fracture toughness. This increase that after rolling, the thickness of the individual tape can be explained by the introduction of the residual bulk was 450 um, while after hot pressing it decreased to 150 um. After hot pressing, the thickness of the Si3N compressive stresses in Si3N4 layers In the case where the thicknesses of the Si3 N4 and the si3 N4-20 wt TIN layers was in the range of 150-300 um, and the thick ness of the Si3N4 layers with TiN additive varied fr layers were similar, the calculated residual compressive stress was about 1 88 MPa and the residual tensile stress 200to500m. about 246.5 MPa The measured value of the apparent Fracture toughness was also measured by Single- fracture toughness was 7.41+ 1.79 MPa m/2.There Edge-V-Notched-Beam (SEVNB) method [21]. 4 was a further increase in KIe(8.5+0.01 MPa m/)for point bending strength of the machined specimens was the laminates with 20 wt%tin when the relative thick determined using a jig with an inner span of 20 mm and ness of the SiN4- 20 wt %TiN layers was increased an outer span of 40 mm. The notch tip was located in a second Si3 N, layer in the case of layered composites. reason for this is a significant increase of the residual Strength and Youngs modulus were also calculated at room temperature by measuring the deflection of the residual stress in the Si3N4-20 wt TiN layers(Ta- samples during 4-point bending tests according to the ble ID). However, an increase of TiN content to 50 wt% standard. The bending strength calculation was based resulted in a significant increase of the residual tensile on a monolithic sample analysis. Optical and scanning stress in the laminates. The calculated tensile stress val electron microscopy was used for a microstructure in- vestigation equipped with a Leica microscope, an XYZ mapping different layer compositon proper A Renishaw 1000 Raman microspectrometer was TABLE III. Mechanical properties of SigNa based laminates with stage and 514.5 nm argon ion laser. The laser generated Composition ar(Mpa) E(GPa) KIc(MPa/2) 12.5 mW of power. a plasma filter was used to remove plasma lines from the spectra taken. The laser spot wa SiN4/Si3N4507.6±32 306.6 5.54±0.01 about 1-2 um for the 100 x objective lens used during S seIN the measurements. Autofocusing was used to collect SisN/(SigNa- 450.4+82.9 8.5±0.01 the Raman spectra because it maintains a good focus 20 wt %TIN) on the sample during line mapping experiments. The SiaNa/Si3N4- 157.9+ 14.9 297.7 system was set up to take spectra from all points along 50 wt%TiN a single line of interest on the surface. Before the si3N4 SinA/TiN 140.8±1091574 3.97±0.52 5445
T A B L E I I . Calculated residual stresses in Si3N4 based laminates Thickness of layers (µm) Composition Si3N4 Si3N4 with TiN σcom. (MPa) σtens. (MPa) Si3N4/Si3N4— 20 wt.%TiN 250 210 188 246.5 Si3N4/2(Si3N4— 20 wt.%TiN) 245 530 279.5 151 Si3N4/Si3N4— 50 wt.%TiN 200 330 765 515.5 Si3N4/TiN 200 400 2467 1078 done. For rolling, a crude rubber (4 wt.%) was added to the mixture of powders as a plasticizer through a 3% solution in petrol. The powders were then dried up to a 2 wt.% residual amount of petrol in the mixture. After sieving powders with a 500 µm sieve, granulated powders were dried up to the 0.5 wt.% residual petrol. A roll mill with 40 mm rolls was used for rolling. The velocity of rolling was 1.5 m/min. Working pressure varied from 10 MPa for a 64% relative density of tapes to 100 MPa for a 74% relative density. The thickness of tapes was either 0.4–0.5 mm or 0.8–1.0 mm, the width was 60–65 mm. Samples of ceramics were prepared by the hot pressing of tapes stacked together. Each layer contained one or a few tapes. Graphite dies were used for hot pressing, and the hot pressing was performed at a temperature of 1820 ◦C, with a dwelling time of 20 min and a pressure of 30 MPa [20]. During hot pressing, the shrinkage of layers occurred 3 times such that after rolling, the thickness of the individual tape was 450 µm, while after hot pressing it decreased to 150 µm. After hot pressing, the thickness of the Si3N4 layers was in the range of 150–300 µm, and the thickness of the Si3N4 layers with TiN additive varied from 200 to 500 µm. Fracture toughness was also measured by SingleEdge-V-Notched-Beam (SEVNB) method [21]. 4- point bending strength of the machined specimens was determined using a jig with an inner span of 20 mm and an outer span of 40 mm. The notch tip was located in a second Si3N4 layer in the case of layered composites. Strength and Young’s modulus were also calculated at room temperature by measuring the deflection of samples during 4-point bending tests according to the standard. The bending strength calculation was based on a monolithic sample analysis. Optical and scanning electron microscopy was used for a microstructure investigation. A Renishaw 1000 Raman microspectrometer was equipped with a Leica microscope, an XYZ mapping stage and 514.5 nm argon ion laser. The laser generated 12.5 mW of power. A plasma filter was used to remove plasma lines from the spectra taken. The laser spot was about 1–2 µm for the 100 × objective lens used during the measurements. Autofocusing was used to collect the Raman spectra because it maintains a good focus on the sample during line mapping experiments. The system was set up to take spectra from all points along a single line of interest on the surface. Before the Si3N4 measurements, the spectrometer was calibrated using a standard Si wafer band with position at 520.3 cm−1. 4. Results and discussion 4.1. Mechanical properties Mechanical properties such as the strength, Young’s modulus, and fracture toughness of the laminates are presented in Table III. The parameters of the tested laminates, such as composition and layer thickness are given in Table II. Besides these four designs, one more design of Si3N4/Si3N4 laminate was used as a base for comparison. The laminates of this design were prepared in the same way as the others, however, all layers were of the same composition. Therefore, no residual stresses can appear during cooling. It is worth noting that both the Young’s modulus and fracture toughness of these Si3N4/Si3N4 laminates were measured to be on the same level as standard Si3N4 ceramics prepared by the standard powder route, which includes no rolling. The strength of the Si3N4/Si3N4 laminate was less than that of the standard Si3N4 ceramics with values of 507.6 ± 3.2 and 750 ± 20.7 MPa, respectively. As one can see from Table III, while the strength of Si3N4/Si3N4- 20wt.% TiN laminates are approximately on the same level as the Si3N4/Si3N4 laminates, further increase of the TiN content to 50 and 100% results in a significant decrease of both strength and Young’s modulus. The measured fracture toughness of the Si3N4/TiN laminates also showed a decrease similar to strength and Young’s modulus values. The Si3N4/Si3N4-20 wt.% TiN laminates showed an increase in apparent fracture toughness. This increase can be explained by the introduction of the residual bulk compressive stresses in Si3N4 layers. In the case where the thicknesses of the Si3N4 and the Si3N4-20 wt.% TiN layers were similar, the calculated residual compressive stress was about 188 MPa and the residual tensile stress about 246.5 MPa. The measured value of the apparent fracture toughness was 7.41 ± 1.79 MPa m1/2. There was a further increase in K1c (8.5 ± 0.01 MPa m1/2) for the laminates with 20 wt.%TiN when the relative thickness of the Si3N4-20 wt.%TiN layers was increased compared to the thickness of pure Si3N4 layers. The reason for this is a significant increase of the residual compressive stress, and at the same time, a decrease of the residual stress in the Si3N4-20 wt.% TiN layers (Table II). However, an increase of TiN content to 50 wt.% resulted in a significant increase of the residual tensile stress in the laminates. The calculated tensile stress valT A B L E I I I . Mechanical properties of Si3N4 based laminates with different layer compositions Composition σf (Mpa) E (GPa) KIC (MPa. m1/2) Si3N4/Si3N4 507.6 ± 3.2 306.6 5.54 ± 0.01 Si3N4/Si3N4- 20 wt.%TiN 356.2 ± 76.4 312.9 7.41 ± 1.79 Si3N4/2(Si3N4- 20 wt.%TiN) 450.4 ± 82.9 – 8.5 ± 0.01 Si3N4/Si3N4- 50 wt.%TiN 157.9 ± 14.9 297.7 – Si3N4/TiN 140.8 ± 10.9 157.4 3.97 ± 0.52 5445
ues are higher than the tensile strength of the material, zones on the fracture surface. The first zone near the and therefore there is much cracking and a decrease in notch tip has a rough surface and corresponds to a slow all mechanical properties crack growth. The second zone has a rather smooth surface with distinct steps only at the interfaces be tween layers. This zone corresponds to a fast crack 4.2. Fracture surfaces growth(Fig. 2B). No crack bifurcation occurred and The typical fracture surfaces of pure Si3 N4 layer and two equal parts of the sample could be found after fail Si3N4-20 wt %TiN layer are shown in Fig. 1. The bi- ure. The Si3N4/2(Si3N4-20 wt %TiN) laminates failed modal grain size distribution exists with a number of after crack bifurcation. The part of the fracture surface elongated grains being surrounded by small rounded near the notch tip was the same as the ones shown in grains of Si3N4(Fig. 1a). The average grain size in the Fig. IA and B. At the moment when the crack bifur- 13n4 layer was 0. 4-0.5 um. The micrograph of the cated, an unusually smooth fracture surface was ob- Si3N4-20%TiN fracture surface revealed that a major- served(Fig. 2C). When the value of residual tensile ity of the grain sizes were in the range of 1-2 um, with stresses approaches the value of the tensile strength of some grains of a size less than I um(Fig. Ib). It was the layer, cracks in the layers are generated, as was the shown that the tin has a homogeneous distribution in case of the Si3 Na/Si3N4-50 wt %TiN and Si3N4/TiN the Si3N4 matrix and no solid solution was detected laminates. The cracks originated during the cooling between Si3n4 and Tin particles [22] stage after the hot pressing of the laminates and ap Fracture surfaces of Si3 N/Si3 N4, Si N4/Si3N4- peared due to mismatching of CTEs and elastic mod- 20 wt%TiN, Si3N4/2(Si3N4-20 wt %TiN) and Si3N4/ uli of two different layers. Channel cracks were ob TiN laminates are shown in Fig. 2. 2(Si3N4-20 served in the laminates with a difference in composi- wt%TiN)means that thickness of Si3N4-20 wt%TiN) tion between layers, starting with 50 wt %TiN content layer is twice than that of Si3 N4 layer. The fracture and higher. Si3 N4/TiN laminates demonstrate channel surface of the Si3 N4/Si3 N4 laminate, where no resid- cracking(Fig. 2D) similar to the cracks described in stresses were generated during cooling, is flat and [23]. These cracks are responsible for the dramatic de smooth(Fig. 2A). As layers of different composition crease in the mechanical properties of Si]N4 based lam- tre used. the fracture surface becomes rougher. For inates. To reduce or eliminate cracking, it is necessary the Si3N/Si3N4-20 wt %TIN laminate, there are two to make composites with characteristics more close between layers, especially the CtE and elastic moduli The extent of channel cracking was decreased in lam- inates with Si3N4-50 wt. TiN layers in comparison to composites where one of the layers was pu Channel cracking was fully eliminated for composites with a Si3N4-20 wt %TiN layer composition. An ab- sence of pre-existent cracks resulted in an increase of the strength and fracture toughness The fracture surface of Si3 N4/Si3 N4- 50 wt% TIN is shown in Fig. 3. As one can see, there is a high roughness of the surface, and bifurcation of the mov ing crack occurred when it was inside the Si3N4 layer with residual compressive stresses. There are fracture steps and channel cracks at the Si3N4-50 wt % TIN lay ers which are perpendicular to the interfaces of com- ite. The fracture step ed only at layers wit tensile stresses. Such fracture steps and other defects are responsible for a decrease in mechanical proper B ties. Multiple bifurcations occur for preexisting cracks inside the layers with residual compressive stresses, and in addition, the moving crack bifurcates during e sa ding. The schematic presentation and an optical image of the crack bifurcation during the failure of this laminate are shown in Fig 4 4.3. Raman shift measurements A measurement of residual stresses is an important he development of the laminates ber of works have been published which use Raman spectroscopy for the determination of residual stresses. from the residual stresses in Figure/ Micrographs of fracture surfaces of Si3N4 layer(A)and Al2O3/ZrO2 composites has been evaluated [24]. The Si3N4-20 wt %oTiN layer(B)in Si3 N4/Si3 N4-20 wt. TiN laminate magnitude of bridging stresses in Si3 N4 and AlO
ues are higher than the tensile strength of the material, and therefore there is much cracking and a decrease in all mechanical properties. 4.2. Fracture surfaces The typical fracture surfaces of pure Si3N4 layer and Si3N4-20 wt.%TiN layer are shown in Fig. 1. The bimodal grain size distribution exists with a number of elongated grains being surrounded by small rounded grains of Si3N4 (Fig. 1a). The average grain size in the Si3N4 layer was 0.4–0.5 µm. The micrograph of the Si3N4-20%TiN fracture surface revealed that a majority of the grain sizes were in the range of 1–2 µm, with some grains of a size less than 1 µm (Fig. 1b). It was shown that the TiN has a homogeneous distribution in the Si3N4 matrix and no solid solution was detected between Si3N4 and TiN particles [22]. Fracture surfaces of Si3N4/Si3N4, Si3N4/Si3N4- 20 wt.%TiN, Si3N4/2(Si3N4-20 wt.%TiN) and Si3N4/ TiN laminates are shown in Fig. 2. 2(Si3N4-20 wt.%TiN) means that thickness of (Si3N4-20 wt.%TiN) layer is twice than that of Si3N4 layer. The fracture surface of the Si3N4/Si3N4 laminate, where no residual stresses were generated during cooling, is flat and smooth (Fig. 2A). As layers of different composition are used, the fracture surface becomes rougher. For the Si3N4/Si3N4-20 wt.%TiN laminate, there are two Figure 1 Micrographs of fracture surfaces of Si3N4 layer (A) and Si3N4-20 wt.%TiN layer (B) in Si3N4/Si3N4-20 wt.%TiN laminate. zones on the fracture surface. The first zone near the notch tip has a rough surface and corresponds to a slow crack growth. The second zone has a rather smooth surface with distinct steps only at the interfaces between layers. This zone corresponds to a fast crack growth (Fig. 2B). No crack bifurcation occurred and two equal parts of the sample could be found after failure. The Si3N4/2(Si3N4-20 wt.%TiN) laminates failed after crack bifurcation. The part of the fracture surface near the notch tip was the same as the ones shown in Fig. 1A and B. At the moment when the crack bifurcated, an unusually smooth fracture surface was observed (Fig. 2C). When the value of residual tensile stresses approaches the value of the tensile strength of the layer, cracks in the layers are generated, as was the case of the Si3N4/Si3N4-50 wt.%TiN and Si3N4/TiN laminates. The cracks originated during the cooling stage after the hot pressing of the laminates and appeared due to mismatching of CTEs and elastic moduli of two different layers. Channel cracks were observed in the laminates with a difference in composition between layers, starting with 50 wt.%TiN content and higher. Si3N4/TiN laminates demonstrate channel cracking (Fig. 2D) similar to the cracks described in [23]. These cracks are responsible for the dramatic decrease in the mechanical properties of Si3N4 based laminates. To reduce or eliminate cracking, it is necessary to make composites with characteristics more close between layers, especially the CTE and elastic moduli. The extent of channel cracking was decreased in laminates with Si3N4-50 wt.%TiN layers in comparison to composites where one of the layers was pure TiN. Channel cracking was fully eliminated for composites with a Si3N4-20 wt.%TiN layer composition. An absence of pre-existent cracks resulted in an increase of the strength and fracture toughness. The fracture surface of Si3N4/Si3N4-50 wt.% TiN is shown in Fig. 3. As one can see, there is a high roughness of the surface, and bifurcation of the moving crack occurred when it was inside the Si3N4 layer with residual compressive stresses. There are fracture steps and channel cracks at the Si3N4-50 wt.%TiN layers which are perpendicular to the interfaces of composite. The fracture steps appeared only at layers with tensile stresses. Such fracture steps and other defects are responsible for a decrease in mechanical properties. Multiple bifurcations occur for preexisting cracks inside the layers with residual compressive stresses, and, in addition, the moving crack bifurcates during the sample loading. The schematic presentation and an optical image of the crack bifurcation during the failure of this laminate are shown in Fig. 4. 4.3. Raman shift measurements A measurement of residual stresses is an important issue in the development of the laminates. A number of works have been published which use Raman spectroscopy for the determination of residual stresses. Strengthening arising from the residual stresses in Al2O3/ZrO2 composites has been evaluated [24]. The magnitude of bridging stresses in Si3N4 and Al2O3 5446
A B C D Figure 2 Fracture surface of laminate composite(A)Si3N4/Si3N4 laminates;(B)Si3 Na/Si3N4-20 wt %TiN laminates;(C) Si3 Na/2(Si3N4 20 wt %TIN) laminates and(D) Si3 N4/TiN laminates. during crack propagation has also been estimated [25, however. The spectrum in Fig 5b was taken from the 26]. Some attempts to use Raman spectroscopy to es- center of a Vickers indentation(20 kg load) placed in timate the residual stresses around indentation in sil- the center of a thin Si3N4 layer from the same face icon nitride have been done [27, 28], but the results of Si3 N4/2(Si3 N4-20%TiN) laminate. The first three were contradictive and further clarification is needed. bands remain intact, but the other bands shifted to the The determination of residual stress in laminates is a higher wave numbers, which indicates the existence omplicated problem. Here we report the preliminary of a residual compressive stress in the center of the results of the Raman shift measurements that can be Vickers impression induced by the indentation. These further used to estimate the residual stresses in a lami- results are similar to published results [27, 28 nar composite One-dimensional maps of band shift, band inten- Two typical Si3 N4 Raman spectra are shown in Fig ity,FWHM, and other band parameters can be pro The spectrum in Fig 5a was taken at the center of a thin duced using a line scan technique [29]. Line map- Si, N4 layer from the side face of the Si3 N4/2(Si3 N4-20 ping of the 862 cm-I Raman band of silicon ni- wt%TiN) laminate. The thickness of Si3N4 layer was tride was performed across a thin Si3N4 layer from about 250 um and the thickness of Si3N4-20 wt%tin the Si3N4/2(Si3N4-20%TiN) laminate, starting at the layer was about 500 um. A first indication of existing Si3N4-20%TiN layer, crossing the interfaces, and end tensile mean stress came from the shift of the Si band to ing in the next Si3N4-20%TiN layer(Fig. 6). Maps of 518cm- because the spectrometer was calibrated w intensity(Fig 6A), FWHM(Fig. 6B), and peak shift a Si band being at 520.3 cm- at the beginning of exper-(Fig. 6C) were generated. As one can see, there is a iment. Free Si can sometimes be detected in Si3 N4 as shift in peak position from 862. 54 cm-in the Si N4 a result of desublimation of Si3N4. The band positions 20%TiN layer to 861.05 cm-I in the pure Si3N4 layer of unstressed Si3N4 were determined as 181, 203, 224,(Fig. 6C). Similar results have been published in [30] 446, 615, 728, 862, 926, 936, 1044 cm-. Also, Si3N4 The shift exists because of different surface stress states bands 862 cm,1044 cm, and others are shifted in layers with different composition [31-33]. There to lower wavenumbers in the center of a thin Si3N4 is tensile mean stress on the surface of the si3N4 layer. Three strong bands(181, 203, 224 cm)did not layers, since a down shift of the peak position was change their positions relative to the unstressed Si3N4, found. At the same time, a compressive mean stress
Figure 2 Fracture surface of laminate composite. (A) Si3N4/Si3N4 laminates; (B) Si3N4/Si3N4-20 wt.%TiN laminates; (C) Si3N4/2(Si3N4- 20 wt.%TiN) laminates and (D) Si3N4/TiN laminates. during crack propagation has also been estimated [25, 26]. Some attempts to use Raman spectroscopy to estimate the residual stresses around indentation in silicon nitride have been done [27, 28], but the results were contradictive and further clarification is needed. The determination of residual stress in laminates is a complicated problem. Here we report the preliminary results of the Raman shift measurements that can be further used to estimate the residual stresses in a laminar composite. Two typical Si3N4 Raman spectra are shown in Fig. 5. The spectrum in Fig. 5a was taken at the center of a thin Si3N4 layer from the side face of the Si3N4/2(Si3N4-20 wt%TiN) laminate. The thickness of Si3N4 layer was about 250 µm and the thickness of Si3N4-20 wt%TiN layer was about 500 µm. A first indication of existing tensile mean stress came from the shift of the Si band to 518 cm−1 because the spectrometer was calibrated with a Si band being at 520.3 cm−1 at the beginning of experiment. Free Si can sometimes be detected in Si3N4 as a result of desublimation of Si3N4. The band positions of unstressed Si3N4 were determined as 181, 203, 224, 446, 615, 728, 862, 926, 936, 1044 cm−1. Also, Si3N4 bands 862 cm−1, 1044 cm−1, and others are shifted to lower wavenumbers in the center of a thin Si3N4 layer. Three strong bands (181, 203, 224 cm−1) did not change their positions relative to the unstressed Si3N4, however. The spectrum in Fig. 5b was taken from the center of a Vickers indentation (20 kg load) placed in the center of a thin Si3N4 layer from the same face of Si3N4/2(Si3N4-20%TiN) laminate. The first three bands remain intact, but the other bands shifted to the higher wave numbers, which indicates the existence of a residual compressive stress in the center of the Vickers impression induced by the indentation. These results are similar to published results [27, 28]. One-dimensional maps of band shift, band intensity, FWHM, and other band parameters can be produced using a line scan technique [29]. Line mapping of the 862 cm−1 Raman band of silicon nitride was performed across a thin Si3N4 layer from the Si3N4/2(Si3N4-20%TiN) laminate, starting at the Si3N4-20% TiN layer, crossing the interfaces, and ending in the next Si3N4-20%TiN layer (Fig. 6). Maps of intensity (Fig. 6A), FWHM (Fig. 6B), and peak shift (Fig. 6C) were generated. As one can see, there is a shift in peak position from 862.54 cm−1 in the Si3N4- 20%TiN layer to 861.05 cm−1 in the pure Si3N4 layer (Fig. 6C). Similar results have been published in [30]. The shift exists because of different surface stress states in layers with different composition [31–33]. There is tensile mean stress on the surface of the Si3N4 layers, since a down shift of the peak position was found. At the same time, a compressive mean stress 5447
B Steps of Fracture Crack bifurcation N4-50wt° oTIN layer Figure 3 Fracture surface of Si3 Na/Si3 N4-50% wt %TiN composite (A)and(C)SEI image: (B)and(D)backscattered image. Surface under tension Surface under tension B Figure 4(A)A schematic drawing of a crack bifurcation during the fracture in Si3 N4/Si3 N4-TiN laminates:(B)An optical photograph of the Si3 N4/Si3 N4-50 wt %TiN laminate bar after fracture. 5448
Figure 3 Fracture surface of Si3N4/Si3N4-50% wt.%TiN composite. (A) and (C) SEI image; (B) and (D) backscattered image. Figure 4 (A) A schematic drawing of a crack bifurcation during the fracture in Si3N4/Si3N4-TiN laminates; (B) An optical photograph of the Si3N4/Si3N4-50 wt.%TiN laminate bar after fracture. 5448
(a) A。2E 1000 Counts/Raman Shift(cm-1) Figure 5 Positions of Raman bands of Si3 N4 in the center of (a)thin Si3 N4 layer with surface tensile stresses and(b) Vickers impression. Load is influences the intensity of the Raman signal, and par- tially because Si3 N4 is under less uniform compressive microstresses in the Si3N4-20 wt %TiN layer, as com- pared to pure Si3N4. The intensity scan of the 862 cm band reveals a strong maximum in the center of Si3n4 layer(Fig 6a). These preliminary results can be useful B for further analysis of residual stress distribution in the laminate 5. Conclusions SinA based multilayered ceramics with layers of dif- ferent thickness and compositions were manufactured C by rolling pressing techniques. The composi- tions and thickness of layers varied to design resid 862-scmSiNc20wt"TIN ual compressive and tensile stresses which affected the mechanical behavior of the composite. The increase of apparent fracture toughness(8.5+0.01 MPam/2)was achieved when the residual compressive stress in pure s61,05 cm SN, Layer Si3N4 layers was equal to 280 MPa, but at the same time the residual tensile stress in Si3 N4-20 wt%TIN 00200300 layers was 150 MPa. When the amount of Tin was in- Distance, um creased to 50wt %o or 100%0, multiple cracks appeared Figure 6 862 cm-1 band line mapping across SigNa Layer(A)A peak in the layers with a residual tensile stress, which lead to intensity map; (B)A FWHM map;(C)A peak shift map the degradation of the mechanical properties. Numer- ous crack bifurcations were observed after the failure of the Si3N4/Si3 N4- 50 wt %TiN laminates. A bifur exists on the Si3 N4-20 wt% TiN surface. Since the cation of the moving crack has also occurred during mansion coefficient of the Si3N4 is lower than that of the failure of Si3N4/2(Si3N4-20 wt %TiN) laminates have bulk Micro-Raman spectroscopy was used for Raman shift residual compressive stress in the Si3, layer and bulk measurements, and preliminary results have revealed residual tensile stress in the Si3 N4-TiN layer. However, that tensile mean stress exists on the surface of Si,N4 edge effects appear on a side face of the layered sample ayers, which have residual compressive stress in the ce of the bulk. At the same time, compressive mean stress layer and edge compressive stress appears on a side on the surface of Si]- 20 wt %nN layers, which have surface of the SiaN4-TiN layer. Raman shift depends residual tensile stress in the bulk. Further work is re- on the sum of bulk and edge components of stress with quired to estimate quantitatively the magnitude of the edge components dominating on a side face. In such surface residual stresses identified by the micro-Raman a way Raman shift indicates the presence of a tensile surface stress in Si3 N4 layer and compressive surface stress in Si3N4-TiN layer Acknowledgements The FWHM map reveals a large scatter of the This work was supported by the European Commis- 862 cm peak in Si3N4-20 wt %TIN layer(ig 6b). sion, the project"Silicon nitride based laminar and This is partially because of the TiN second phase, which functionally gradient ceramics for engineering applica- 5449
Figure 5 Positions of Raman bands of Si3N4 in the center of (a) thin Si3N4 layer with surface tensile stresses and (b) Vickers impression. Load is 20 kg. Figure 6 862 cm−1 band line mapping across Si3N4 Layer. (A) A peak intensity map; (B) A FWHM map; (C) A peak shift map. exists on the Si3N4-20 wt.% TiN surface. Since the expansion coefficient of the Si3N4 is lower than that of the Si3N4-TiN, therefore, after cooling we have bulk residual compressive stress in the Si3N4 layer and bulk residual tensile stress in the Si3N4-TiN layer. However, edge effects appear on a side face of the layered sample. Edge tensile stress exists on side surface of the Si3N4 layer and edge compressive stress appears on a side surface of the Si3N4-TiN layer. Raman shift depends on the sum of bulk and edge components of stress with edge components dominating on a side face. In such a way Raman shift indicates the presence of a tensile surface stress in Si3N4 layer and compressive surface stress in Si3N4-TiN layer. The FWHM map reveals a large scatter of the 862 cm−1 peak in Si3N4-20 wt.%TiN layer (Fig. 6b). This is partially because of the TiN second phase, which influences the intensity of the Raman signal, and partially because Si3N4 is under less uniform compressive microstresses in the Si3N4-20 wt.%TiN layer, as compared to pure Si3N4. The intensity scan of the 862 cm−1 band reveals a strong maximum in the center of Si3N4 layer (Fig. 6a). These preliminary results can be useful for further analysis of residual stress distribution in the laminate. 5. Conclusions Si3N4 based multilayered ceramics with layers of different thickness and compositions were manufactured by rolling and hot pressing techniques. The compositions and thickness of layers varied to design residual compressive and tensile stresses which affected the mechanical behavior of the composite. The increase of apparent fracture toughness (8.5 ± 0.01 MPa m1/2) was achieved when the residual compressive stress in pure Si3N4 layers was equal to 280 MPa, but at the same time the residual tensile stress in Si3N4-20 wt.%TiN layers was 150 MPa. When the amount of TiN was increased to 50wt.% or 100%, multiple cracks appeared in the layers with a residual tensile stress, which lead to the degradation of the mechanical properties. Numerous crack bifurcations were observed after the failure of the Si3N4/Si3N4-50 wt.%TiN laminates. A bifurcation of the moving crack has also occurred during the failure of Si3N4/2(Si3N4-20 wt.%TiN) laminates. Micro-Raman spectroscopy was used for Raman shift measurements, and preliminary results have revealed that tensile mean stress exists on the surface of Si3N4 layers, which have residual compressive stress in the bulk. At the same time, compressive mean stress exists on the surface of Si3N4-20 wt.%TiN layers, which have residual tensile stress in the bulk. Further work is required to estimate quantitatively the magnitude of the surface residual stresses identified by the micro-Raman spectroscopy. Acknowledgements This work was supported by the European Commission, the project “Silicon nitride based laminar and functionally gradient ceramics for engineering applica- 5449
tion", the program Copernicus-2. Raman spectroscopy 18. T. CHARTIER, D. MERLE and J.L. BESSON, J Europ measurements were supported by AFOSR, the project Ceran.Soc.16(1995)101 "Measurement of stress in ceramic laminates with 19. M. LUGOVY, N. ORLOVSKAYA, K. BERROTH and J. micro- Raman”∵ EMPA work was funded by BBw the KUEBLER. Comp. Sci. Technol. 59(1999)1429 ORLOVSKAYA Swiss Federal office for education and Science under EINARSRUD and v contract no. 99.0785 NATO ARW Multilayered and Fibre-Reinforced Com Problems and Prospects", edited by Y. M. Haddad References J. KUEBLER, ASTM STP 1409, J. A Salem, edited by M. G 1. W. J. CLEGG, Science 286(1999)1097. Jenkins and G. D. Quinn, ASTM, West Conshohocken, PA, USA 2. W.J. CLEGG. K. KENDALL. N. ALFORD. T W. BUT ISBN0-8031-2880-0,(2002)p.93 TON, and J. D. BIRCHALL, Nature 347(1990)455 22. N. ORLOVSKAYA. V. SUBBOTIN and M. LUGOVY.un- 3. M. P RAO, A.J. SANCHEZ-HERENCIA, G. E. BELTZ R. M. MCMEEKING and F. F. LANGE, Science 286(1999) 23. N. BAl, D. D. POLLARD and H. GAO, Nature 403( 2000) 4. H. CHAN, Ann. Rev. Mater. Sci. 27(1997)249 24 G. PEZZOTTI, V. SERGO,O. SBAIZERO, N. MURAKI, 5. P. HONEYMAN-CLOVIN and F. F. LANGE.. Am. Cera S. MERIANI and T. NISHIDA, J Erop. Cera. Soc. 19(1999 Soc.79(1996)1810. 47 6. M. OECHSNER, C. HILLMAN and F. F. LANGE. ibid. 79 25 G. PEZZoTTI. N. MURAkI N. MAEDA. K. SATOU and (1996)1834 T. NISHIDA, J Am. Ceram Soc. 82(1999)1249 7. A. SANCHEZ-HERENCIA, C. PASCUAL, J. HE and F. 26. G. PEZZoTTI. H. SUENOBU. T. NISHIDA and o LANGE,iid.82(1999)1512 SBAIZERo, ibid. 82(1999)1257 8. C. A. FOLSOM, F. W. ZOK and F. F. LANGE, ibid. 77 27. N. MURAKI, G. KATAGIRI, V. SERGO, G. PEZZOTTI (1994)689 A.J. Mater: Sci. 32(1997) 9.F.L. RILEY,iid.83(2000)245 28 V. SERGO, G. PEZZOTTI, G. KATAGIRI, N. MURAKI 10. Y. GOGOTSI, J Mater. Sci. 29(1994)2541 11. P. SAJGALIK, Z. LENCES and J. DUSZA, ibid. 31(1996) 29. M. BOWDEN, Appl. Spectr. 51(1997)1405 4837 30. N. ORLOVSKAYA and Y. GOGOTSI Euro Ceramics VI. PT ORLOVSKAYA, 1-3, Key Eng. Mater..206(2,(2002)1025 EINARSRUD. K. BERROTH and V. KOVYLAYEV 31. V. SERGO, D. M. LIPKIN, G. DE PORTU and D. R Key eng. Mater.132-136(1997)2017 CLARKE, J Am Ceram. Soc. 80(1997)1633 13. T. OHI, L. SHIEGEGAKl, T. MIYAJIMA and S.KAN- ZAKI, J Am Ceram Soc. 80(1997)991 32. S. HO, C. HILLMAN, F. F. LANGE and Z. SUO,JAm. Ceran.Soc.78(1995)2353. 14. N. WADA, S A. SOLIN, J. WONG andS PROCHAZKA, 33 N. L. HARRISON and W. J. HARRISON, /. Adhes. 3(1972) J Non-Cryst. Solids 43(1981)7. 15 K. HONDA, S. YOKOYAMA and S. TANAKA, J. App Physics85(1999)7380. 16. J. DONG and O. SANKEY, ibid. 87(2000)958. Received &January 2004 17. A. TAKASE and E. TANI, J Mater. Sci. Lett. 6(1987)607 and accepted 16 February 2005 5450
tion”, the program Copernicus-2. Raman spectroscopy measurements were supported by AFOSR, the project “Measurement of stress in ceramic laminates with micro-Raman”. EMPA work was funded by BBW the Swiss Federal Office for Education and Science under contract no. 99.0785 References 1. W. J. CLEGG, Science 286 (1999) 1097. 2. W. J. CLEGG, K. KENDALL, N. ALFORD, T. W. BUT- TON, and J. D. BIRCHALL, Nature 347 (1990) 455. 3. M. P. RAO, A. J. SANCHEZ-HERENCIA, G. E. B E LT Z, R. M. MCMEEKING and F. F. LANGE, Science 286 (1999) 102. 4. H. CHAN, Ann. Rev. Mater. Sci. 27 (1997) 249. 5. P. HONEYMAN-CLOVIN and F. F. LANGE, J. Am. Ceram. Soc. 79 (1996) 1810. 6. M. OECHSNER, C. HILLMAN and F. F. LANGE, ibid. 79 (1996) 1834. 7. A. SANCHEZ-HERENCIA, C. PA S C UA L, J. H E and F. LANGE, ibid. 82 (1999) 1512. 8. C. A. FOLSOM, F. W. ZOK and F. F. LANGE, ibid. 77 (1994) 689. 9. F. L. RILEY, ibid. 83 (2000) 245. 10. Y. GOGOTSI, J. Mater. Sci. 29 (1994) 2541. 11. P. SAJGALIK, Z. LENCES and J. DUSZA, ibid. 31 (1996) 4837. 12. V. YA RO S H E N KO, N. O R L OV S K AYA, M.-A. EINARSRUD, K. BERROTH and V. KOV Y L AY E V, Key Eng. Mater. 132–136 (1997) 2017. 13. T. OHIJ, L. SHIEGEGAKI, T. M I YA J I M A and S . KANZAKI, J. Am. Ceram. Soc. 80 (1997) 991. 14. N. WA DA, S . A. SOLIN, J. WONG and S . PROCHAZKA, J. Non-Cryst. Solids 43 (1981) 7. 15. K. HONDA, S . YO KOYA M A and S . TA NA K A, J. Appl. Physics 85 (1999) 7380. 16. J. DONG and O. SANKEY, ibid. 87 (2000) 958. 17. A. TA K A S E and E. TA N I, J. Mater. Sci. Lett. 6 (1987) 607. 18. T. CHARTIER, D. MERLE and J. L. BESSON, J. Europ. Ceram. Soc. 16 (1995) 101. 19. M. LUGOVY, N. O R L OV S K AYA, K. BERROTH and J. KUEBLER, Comp. Sci. Technol. 59 (1999) 1429. 20. V. YA RO S H E N KO, N. ORLOVSKAYA, M.-A. EINARSRUD and V. KOVYLAYEV, in Proceedings of NATO ARW “ Multilayered and Fibre-Reinforced Composites: Problems and Prospects”, edited by Y. M. Haddad (Kluwer, Dordrech, 1998) p. 285. 21. J. KUEBLER, ASTM STP 1409, J. A. Salem, edited by M. G. Jenkins and G. D. Quinn, ASTM, West Conshohocken, PA, USA, ISBN 0-8031-2880-0, (2002) p. 93. 22. N. O R L OV S K AYA, V. SUBBOTIN and M. LUGOVY, unpublished work. 23. N. BAI, D. D. POLLARD and H. GAO, Nature 403 (2000) 753. 24. G. PEZZOTTI, V. SERGO, O. SBAIZERO, N. MURAKI, S . MERIANI and T. NISHIDA, J. Erop. Ceram. Soc. 19 (1999) 247. 25. G. PEZZOTTI, N. MURAKI, N. MAEDA, K. S ATO U and T. NISHIDA, J. Am. Ceram. Soc. 82 (1999) 1249. 26. G. PEZZOTTI, H. SUENOBU, T. NISHIDA and O. SBAIZERO, ibid. 82 (1999) 1257. 27. N. MURAKI, G. K ATAG I R I, V. SERGO, G. PEZZOTTI and T. NISHIDA, J. Mater. Sci. 32 (1997) 929. 28. V. SERGO, G. PEZZOTTI, G. K ATAG I R I, N. MURAKI and T. NISHIDA, J. Am. Ceram. Soc. 79 (1996) 5419. 29. M. BOWDEN, Appl. Spectr. 51 (1997) 1405. 30. N. O R L OV S K AYA and Y. GOGOTSI, Euro Ceramics VII, PT 1-3, Key Eng. Mater. 206(2), (2002) 1025. 31. V. SERGO, D. M. LIPKIN, G. DE PORTU and D. R. CLARKE, J. Am. Ceram. Soc. 80 (1997) 1633. 32. S . H O, C. HILLMAN, F. F. LANGE and Z. SUO, J. Am. Ceram. Soc. 78 (1995) 2353. 33. N. L. HARRISON and W. J. HARRISON, J. Adhes. 3 (1972) 195. Received 8 January 2004 and accepted 16 February 2005 5450
Copyright of Journal of Materials Science is the property of Springer Science Business Media B V. The copyright in an individual article may be maintained by the author in certain cases. Content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder,'s express written permission However, users may print, download, or email articles for individual use