Thermal, electrical, and mechanical properties of pressureless sintered silicon carbide ceramics with yttria-scandia-aluminum nitride

Journal of the European Ceramic Society 36 (2016) 2659–2665

Contents lists available at www.sciencedirect.com

Journal of the European Ceramic Society journal homepage: www.elsevier.com/locate/jeurceramsoc

Thermal, electrical, and mechanical properties of pressureless sintered silicon carbide ceramics with yttria-scandia-aluminum nitride Tae-Young Cho a , Young-Wook Kim a,∗ , Kwang Joo Kim b a b

Functional Ceramics Laboratory, Department of Materials Science and Engineering, The University of Seoul, Seoul 02504, Republic of Korea Department of Physics, Konkuk University, Seoul 05029, Republic of Korea

a r t i c l e

i n f o

Article history: Received 22 January 2016 Received in revised form 1 April 2016 Accepted 11 April 2016 Available online 18 April 2016 Keywords: SiC Pressureless sintering Electrical properties Thermal properties Mechanical properties

a b s t r a c t The effects of the polytype of SiC starting powders on the thermal, electrical, and mechanical properties of pressureless sintered SiC ceramics with a new additive system (6.5 vol% Y2 O3 -Sc2 O3 -AlN) were investigated. Powder mixtures prepared from ␣- or ␤-SiC powders were sintered at 1950 ◦ C for 6 h in a nitrogen atmosphere without an applied pressure. We found that both specimens could be sintered to >96% of the theoretical density without an applied pressure. The SiC ceramic fabricated from ␤-SiC powders showed lower electrical resistivity, higher thermal conductivity, and better mechanical properties than that from ␣-SiC powders. The flexural strength, fracture toughness, hardness, electrical resistivity and thermal conductivity values of the SiC ceramics fabricated from ␤-SiC powders were 520 MPa, 5.1 MPa m1/2 , 25.0 GPa, 6.7 × 10−1  cm and 110 Wm−1 K−1 at room temperature, respectively. The new additive system achieved the highest thermal conductivity in pressureless liquid-phase sintered SiC ceramics. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction Silicon carbide (SiC) is an important engineering ceramic because of its good thermal conductivity, oxidation resistance, wear resistance, corrosion resistance, and high-temperature mechanical properties [1–8]. SiC is being used in many applications, especially in heaters, heater plates, susceptors, dummy wafers, focus rings for semiconductors and in light-emitting diode (LED) processing. These applications take advantage of its excellent thermal conductivity as well as other properties. Heat is carried by phonons in SiC ceramics because of the deficiency of free electrons. Although the thermal conductivity values of single crystalline SiC ceramics are very high (347 for 4HSiC and 490 Wm−1 K−1 for 6H-SiC) [9,10], those of polycrystalline SiC ceramics are quite low because of the scattering of phonons by point defects, phase boundaries, grain boundaries, and secondary phases [8,11,12]. Several strategies for selecting additives for improving thermal conductivity of polycrystalline liquid-phase sintered SiC (LPS-SiC) ceramics were suggested including the following [8,11–15]:

∗ Corresponding author. E-mail address: [email protected] (Y.-W. Kim). http://dx.doi.org/10.1016/j.jeurceramsoc.2016.04.014 0955-2219/© 2016 Elsevier Ltd. All rights reserved.

(1) Use of sintering additives that can pick up oxygen from the SiC lattice because the lattice oxygen can cause additional Si vacancies according to the following defect equation, SiO2 → SiSi + 2OC + VSi . These vacancies result in phonon scattering. (2) Use of sintering additives without Al and Al compounds because the Al can create additional Si vacancies according to the following defect equation, Al2 O3 → 2AlSi + 3OC + VSi . These vacancies can also cause phonon scattering. (3) Minimizing oxide additive content because the thermal conductivity of oxide phases is much lower than that of the SiC lattice, resulting in lower thermal conductivity. (4) Use of additives that can suppress the ␤ → ␣ phase transformation of SiC because the phase transformation can lead to increased phonon scattering at the 3C/4H and 3C/6H interfaces in SiC grains.

Examples of additive compositions which can satisfy the above requirements include Y2 O3 -La2 O3 , Y2 O3 -TiN, and Y2 O3 -Sc2 O3 systems. The maximum thermal conductivity values achieved in each system were 211 Wm−1 K−1 in SiC ceramics sintered with 5 vol% Y2 O3 -La2 O3 [11], 211 Wm−1 K−1 in SiC ceramics sintered with 3 vol% Y2 O3 -TiN [14] and 234 Wm−1 K−1 in SiC ceramics sintered with 1 vol% Y2 O3 -Sc2 O3 [8]. However, all of the above LPS-SiC ceramics were processed with an applied pressure, i.e., via a hot-pressing route. The thermal conductivity values of LPS-SiC

2660

T.-Y. Cho et al. / Journal of the European Ceramic Society 36 (2016) 2659–2665

ceramics processed without an applied pressure were in the range of 55–90 Wm−1 K−1 . Specifically, the maximum thermal conductivity values achieved in each system without an applied pressure were 70 Wm−1 K−1 for SiC ceramics sintered with 10 vol% Al2 O3 Y2 O3 [16], 85 Wm−1 K−1 for SiC ceramics sintered with 3 vol% Y3 Al5 O12 -AlN [17], 85 Wm−1 K−1 for SiC ceramics sintered with 7 wt% Al2 O3 -Er2 O3 [18], and 90 Wm−1 K−1 in SiC ceramics sintered with 7 wt% Al2 O3 -CeO2 [18]. Generally, solid-state sintered SiC (SSS-SiC) ceramics with boron (B) or B-containing additives showed higher thermal conductivity than LPS-SiC ceramics processed without an applied pressure. The thermal conductivity values were 100 Wm−1 K−1 for SSS-SiC ceramics with 2 vol% BN [19], 140 Wm−1 K−1 for SiC ceramics with 2.1 wt% B4 C-C [20], 124 Wm−1 K−1 for SiC ceramics with 2.6 wt% AlN-B4 C-C [21], and 192 Wm−1 K−1 for SiC ceramics with 0.8 wt% B-C additives [22]. However, the SSS-SiC ceramics have a few drawbacks compared to LPS-SiC, such as poor mechanical properties and the possibility of boron contamination during semiconductor and LED processing. Thus, LPS-SiC ceramics are preferred for applications in semiconductor and LED processing. Industrial demands for applications of LPS-SiC parts in semiconductor and LED processing, e.g., susceptors, trays, and wafer boats, require the development of SiC ceramics with a higher thermal conductivity and lower electrical resistivity than the current LPS-SiC ceramics processed without an applied pressure, i.e., via a pressureless sintering route. This is because industry requires more rapid heating and cooling schedules during production. Most previous researches on pressureless LPS-SiC ceramics have focused on the Al2 O3 -Y2 O3 additive system [23–25], although a few other additive compositions have also been investigated for pressureless LPS-SiC. Those additive compositions include Y3 Al5 O12 -AlN [17] and Al2 O3 -RE2 O3 (Re = Ce, Er, Lu) [18]. The addition of a third oxide, in addition to Al2 O3 -Y2 O3 , was also investigated for lowering the sintering temperature and/or for improving the fracture toughness of LPS-SiC ceramics. Those additive compositions include Al2 O3 -Y2 O3 -MgO [26], Al2 O3 -Y2 O3 -TiO2 [27], and Al2 O3 -Y2 O3 -CaO [28] systems. The flexural strength and fracture toughness of the SiC ceramics sintered with 10 wt% Al2 O3 Y2 O3 -MgO were 377–440 MPa and 4.8–5.2 MPa m1/2 , respectively [26]. Those values of the SiC ceramics sintered with 10 wt% Al2 O3 Y2 O3 -TiO2 were 516 MPa and 4.6 MPa m1/2 , respectively [27]. The fracture toughness and hardness of the SiC ceramics sintered with 10 wt% Al2 O3 -Y2 O3 -CaO were 5.7 MPa m1/2 and 23.2 GPa, respectively [28]. However, no thermal conductivity data were reported for the pressureless sintered SiC ceramics with the above ternary additive systems. Based on the above review, we suggest several strategies that may improve the thermal conductivity of pressureless LPS-SiC ceramics: (1) sintering in a nitrogen atmosphere is preferred for suppressing the ␤→␣ phase transformation of SiC; (2) the addition of Al-containing additives should be avoided or minimized for minimizing VSi concentration in SiC lattice; and (3) if an Al-containing additive is added for successful densification, AlN is preferred over Al2 O3 for minimizing oxygen contamination in SiC lattice. To obtain LPS-SiC ceramics with higher thermal conductivity than the reported values (55–90 Wm−1 K−1 ), a SiC ceramic with 5 vol% equimolar Y2 O3 -Sc2 O3 additives was fabricated by pressureless-sintering. However, the relative density of the pressureless-sintered SiC ceramic with Y2 O3 -Sc2 O3 was only 88%. In this study, to densify the SiC ceramic containing Y2 O3 -Sc2 O3 additives without an applied pressure, a small amount of AlN (1.5 vol%) was introduced as an additional sintering additive in addition to the 5 vol% Y2 O3 -Sc2 O3 . Previous works on LPS-SiC suggested that the crystallographic modification of the initial SiC powder has a strong influence on the microstructure and properties of the resulting ceramics [23,29]. The objective of this research

is to investigate the effects of polytype of SiC starting powders on the sintered density, crystalline phase, and microstructure of the pressureless sintered SiC ceramics with a new additive system (Y2 O3 -Sc2 O3 -AlN). The thermal, electrical, and mechanical properties of the ceramics were measured, and the results were correlated with the sintered density, crystalline phase, and microstructure of the ceramics. 2. Experimental procedure Commercially available ␣-SiC (∼0.5 ␮m, UF-15, H.C. Starck, Berlin, Germany), ␤-SiC (∼0.5 ␮m, BF-17, H.C. Starck, Berlin, Germany), Y2 O3 (0.4 ␮m, 99.99% pure, Kojundo Chemical Lab Co., Ltd., Sakado-shi, Japan), Sc2 O3 (∼9 ␮m, 99.99% pure, Kojundo Chemical Lab Co., Ltd., Sakado-shi, Japan), and AlN (∼1.5 ␮m, Grade F, Tokuyama Soda Co., Ltd., Sakado-shi, Japan) were used as the starting materials. Two batches of powder mixtures were mixed in ethanol for 24 h using SiC balls and a polypropylene jar (Table 1). The ethanol used contains 1 wt% poly(ethylene glycol) as a binder. The additive content was fixed at 6.5 vol% and consisted of 1.5 vol% AlN and 5 vol% equimolar Y2 O3 -Sc2 O3 . The mixtures were dried, sieved (60 mesh), and uniaxially pressed under an applied pressure of 25 MPa. The green bodies (40 × 40 × 5 mm) were subsequently cold isostatically pressed at 280 MPa. Burning out prior to pressureless sintering was done at 450 ◦ C for 30 min with a heating rate of 2 ◦ C/min in a nitrogen atmosphere. The pressureless sintering of the samples (three samples at each condition) was carried out in a graphite furnace at 1850–2000 ◦ C for 6 h in a nitrogen atmosphere. All the tested specimens in this investigation were cut from the as-sintered specimens. The relative densities of the pressureless-sintered specimens were determined using the Archimedes method. Theoretical densities of the specimens were calculated according to the rule of mixtures, as shown in Table 1. X-ray diffraction (XRD; D8 Discover, Bruker AXS Gmbh, Karlsruhe, Germany) using Cu K␣ radiation was performed on the ground powders. The XRD data were analyzed using the Rietveld refinement method for quantitative phase analysis of SiC polytypes. The sintered specimens were polished and etched with CF4 plasma containing 10% oxygen. The etched microstructure and fracture surface were observed by scanning electron microscopy (SEM, S4300, Hitachi Ltd., Hitachi, Japan). Mean grain sizes were measured by the line intercept method, where the multiplication constant was 1.56 based on a tetrakaidecahedral grain shape [30]. Seven hundred to nine hundred grains were measured for each specimen to obtain a mean grain size. The thermal diffusivity was measured using the laser flash method. Differential scanning calorimetry (DSC, Model Q200; TA instrument Inc., New Castle, DE, USA) and thermal diffusivity measurement equipment (Model LFA 447; NET-ZSCH GmbH, Selb, Germany) were used for measuring the heat capacity (Cp ) and thermal diffusivity (␣), respectively. Samples (2.84 mm × 2.84 mm × 1 mm for measuring the heat capacity and 10 mm × 10 mm × 4 mm for measuring the thermal diffusivity) were cut from the pressureless sintered specimens and were polished. The Cp and ␣ were measured five times each, and the average values were used to calculate the thermal conductivity at 25 ◦ C. The thermal conductivity (␬) was calculated according the following equation [31], ␬ = ␣␳Cp

(1)

where ␳ is the density of the sample. The average phonon mean free path (␫) was calculated according the equation [15], ␫ = 3␬/Cp ␳␯

(2)

where ␯ is the average velocity of sound in the SiC ceramic. The velocity of sound in SiC is 11820 m/s at room temperature [32]. The

T.-Y. Cho et al. / Journal of the European Ceramic Society 36 (2016) 2659–2665

2661

Table 1 Batch composition and relative density of SiC specimens. Sample designation

Batch composition (wt%)

Theoretical density (g/cm3 )

Relative density (%)

SCA SCB

91.775% ␣-SiC + 4.180% Y2 O3 + 2.553% Sc2 O3 + 1.492% AlN 91.770% ␤-SiC + 4.183% Y2 O3 + 2.554% Sc2 O3 + 1.493% AlN

3.278 3.277

98.6 96.7

lattice oxygen content of the sintered samples was measured using a hot-gas extraction method. The bulk samples were pulverized using iron pulverizer and screened through a 500 mesh sieve. The powder was treated first with 50% HF at 60 ◦ C for 3 h and then, with 50% H2 SO4 at 120 ◦ C for 2 h, to remove the grain boundary phases. The slurry was repeatedly washed with distilled water till a pH value of 7, then dried at 110 ◦ C for 8 h and passed through a 100 mesh sieve. The oxygen content of the bulk specimens was measured quantitatively using an oxygen analyzer (TC600, LECO Co., St. Joseph, MI, USA). Hall-effect measurements were carried out at room temperature using the van der Pauw technique in order to determine the electrical resistivity, carrier density, and carrier mobility of the samples. An external magnetic field of 1 T was applied perpendicular to the square-shaped sample plane (10 mm × 10 mm) during the Hall measurements. In order to perform flexural strength measurements, bar-shaped samples were cut and polished to a size of 2 mm × 1.5 mm × 25 mm. The tensile surface of the bars was polished to a 1-␮m diamond finish, and the tensile edges were chamfered to avoid stress concentration and edge flaws caused by sectioning. Bending tests were performed at a crosshead speed of 0.2 mm/min using a four-point bending method with inner and outer spans of 10 and 20 mm, respectively. The fracture toughness was determined according to ASTM C1421-15 [33]. The dimensions of the test specimen were 3 mm × 4 mm × 25 mm. A Knoop indenter was used to make an indent in the middle of the polished surface of the test specimen. The indentation force was 49 N, and the full force dwell time was 15 s. The residual stress damage zone was removed by mild grinding before the fracture test. The hardness was measured using a Vickers indenter (Model AVK-C2, Akashi Corp., Yokohama, Japan) with a load of 1.96 N and a dwell time of 15 s.

3. Results and discussion (1) Microstructure

Table 2 Polytype contents in the starting SiC powders and sintered SiC specimens. The polytype content was determined by Rietveld method. Specimen designation

␣-SiC powder ␤-SiC powder SCA SCB

Polytype content (%) 3C

6H

4H

– 86.6 – 25.8

90.5 13.4 80.8 26.9

9.5 – 19.2 47.3

The SiC specimens fabricated from submicron ␣-SiC and ␤-SiC powders with 6.5 vol% Y2 O3 -Sc2 O3 -AlN additives were designated as SCA and SCB, respectively. The relative densities of the SCA and SCB sintered at various temperatures are shown in Fig. 1. The sintered density increased from 75.6% to 98.6% for SCA and from 75.0% to 96.7% for SCB, respectively, with increasing sintering temperature from 1850 ◦ C to 1950 ◦ C, and then, the density decreased to 92.8% for SCA and 86.0% for SCB, respectively, after sintering at 2000 ◦ C for 6 h in a nitrogen atmosphere. The decrease in sintered density with increasing sintering temperature from 1950 ◦ C to 2000 ◦ C was due to the partial evaporation of sintering additives. Thus, further characterization of the samples in this investigation was carried out on the samples sintered at 1950 ◦ C. The maximum densities obtained in the SCA and SCB were 98.6% and 96.7%, respectively, when sintered at 1950 ◦ C for 6 h in a nitrogen atmosphere. Y2 O3 -Sc2 O3 -AlN additives react with SiO2 , a native oxide film on SiC particles, forming a Y-Sc-Al-Si-ON melt during heating and, with increasing temperature, a Y-Sc-Al-Si-OCN melt due to the dissolution of SiC during sintering [34]. The liquid phase was responsible for the densification of specimens via liquid-phase sintering. The XRD analysis and quantitative phase analysis of SiC polytypes obtained by the Rietveld refinement method for the starting SiC powders and sintered specimens are shown in Fig. 2 and Table 2, respectively. The results showed that (i) the 6H → 4H phase transformation of SiC took place in SCA during sintering at 1950 ◦ C; (ii) both the 3C → 6H and the 6H → 4H phase transformation for SiC took place in SCB during sintering at 1950 ◦ C; (iii) the major phase

100

95

Relative density (%)

98.6

SCA SCB

96.7 92.8

90

85.8 85

86.0

80

75.6

80.2

75

75.0 70 1850

1900

1950

2000

o

Sintering temperature ( C) Fig. 1. Relative densities of SiC specimens sintered with 6.5 vol% Y2 O3 -Sc2 O3 -AlN additives as a function of sintering temperature: (a) SCA and (b) SCB (refer to Table 1).

Fig. 2. X-ray diffraction patterns of SiC specimens sintered with 6.5 vol% Y2 O3 Sc2 O3 -AlN additives: (a) SCA and (b) SCB (refer to Table 1).

2662

T.-Y. Cho et al. / Journal of the European Ceramic Society 36 (2016) 2659–2665

Fig. 3. Typical microstructures of SiC ceramics sintered with 6.5 vol% Y2 O3 -Sc2 O3 -AlN additives: (a) SCA and (b) SCB (refer to Table 1).

(2) Thermal properties

60

(b) 0.658

140 0.694

43.6 40 30 20 10 0

SCA

SCB

Heat Capacity (J/g/K)

51.1

50

0.6

0.4

0.2

0.0

SCA

SCB

Thermal Conductivity (W/m.K)

0.8

(a)

2

Thermal Diffusivity (mm /s)

70

The thermal properties of the pressureless sintered SiC ceramics including thermal diffusivity, heat capacity, thermal conductivity, and phonon mean free path are shown in Fig. 4. Thermal diffusivity and heat capacity of the SiC specimens were 43.6 mm2 /s and 0.658 J (g K)−1 for SCA and 51.1 mm2 /s and 0.694 J (g K)−1 for SCB, respectively. The thermal conductivities of SCA and SCB were 91.9 Wm−1 K−1 and 110.3 Wm−1 K−1 , respectively. The thermal conductivity values measured here were quite a bit lower than those of single crystalline SiC ceramics (347 Wm−1 K−1 for 4H-SiC and 490 Wm−1 K−1 for 6H-SiC) [9,10]. This was due to the phonon scattering by point defects, phase boundaries, grain boundaries, and secondary phases [8,11–15]. The thermal conductivity of SiC ceramics sintered with 1 vol% Y2 O3 -Sc2 O3 was 234 Wm−1 K−1 . Such a high thermal conductivity was achieved in the previous work [8] due to the following beneficial effects of the Y2 O3 -Sc2 O3 additive system. (1) The Y2 O3 -Sc2 O3 additives served as an oxygen sink and reduced the oxygen content in the SiC lattice by forming a (Sc,Y)2 Si2 O7 phase. (2) Y and Sc were not soluble in the SiC lattice because of large differences in the ionic sizes of the elements, as evidenced by EDS analysis on SiC grains. Finally, (3) the additives minimized grain boundary segregation as evidenced by clean or crystallized SiC–SiC boundaries. The thermal conductivities of the present ceramics (∼92 and ∼110 Wm−1 K−1 for SCA and SCB, respectively) were lower than that (234 Wm−1 K−1 ) of the ceramics hot-pressed with 1 vol% Y2 O3 Sc2 O3 (hereafter referred to as SC1). This was due to the following reasons: (1) the addition of 1.5 vol% AlN as well as 5 vol% Y2 O3 Sc2 O3 , which was necessary for the densification of the ceramics by pressureless sintering. The Al can create additional Si vacancies (Al2 O3 → 2AlSi + 3OC + VSi ), and the vacancies caused an increase in phonon scattering [11]. (2) The 6.5 vol% Y2 O3 -Sc2 O3 -AlN additives

120 100

15

(c) 110.3 91.9

80 60 40 20 0

SCA

SCB

Phonon Mean Free Path (nm)

in the sintered specimen was the 6H phase in SCA and the 4H phase in SCB. These results indicate that the phase transformation of SiC took place in the 3C → 6H → 4H sequence during sintering at 1950 ◦ C in a nitrogen atmosphere. The results also suggest that significant ␤ → ␣ phase transformation of SiC took place at 1950 ◦ C in SCB (∼70% 3C phase in the starting powders transformed to 4H and 6H phases in SCB). In contrast, only a small amount of the 6H phase (∼10% of 6H phase in the starting powders) transformed to the 4H phase in the SCA. A comparison with the previous data [8] for sintering with only Y2 O3 -Sc2 O3 additives suggests that the addition of AlN into the Y2 O3 -Sc2 O3 additive system accelerated the ␤ → ␣ phase transformation of SiC at 1950 ◦ C, even in a nitrogen atmosphere. This was due to Al incorporation into the SiC lattice during sintering that increases point defect concentrations in the SiC lattice and accelerates the ␤ → ␣ phase transformation of SiC [34,35]. The microstructures of the sintered specimens are shown in Fig. 3. The SiC grains in both SCA and SCB clearly show the core/shell structure, indicating the growth of SiC grains by a solution-precipitation mechanism. The microstructure of the SCA sample consisted of mostly equiaxed grains. In contrast, the SCB consisted of relatively large elongated grains (platelet grains in 3dimensions) and relatively small equiaxed grains. The grain sizes of SCA and SCB specimens were 1.4 and 1.9 ␮m, respectively. The aspect ratios of SiC grains in SCA and SCB were 1.2 and 3.2, respectively. The growth of elongated grains in SCB was a result of the ␤ → ␣ phase transformation of SiC. It is well documented that the ␤ → ␣ phase transformation of SiC leads to growth of elongated grains with high aspect ratios [23,25,36].

(d) 12.7 11.0

10

5

0

SCA

SCB

Fig. 4. (a) Thermal diffusivity, (b) heat capacity, (c) thermal conductivity, and (d) phonon mean free path of SiC specimens sintered with 6.5 vol% Y2 O3 -Sc2 O3 -AlN additives.

T.-Y. Cho et al. / Journal of the European Ceramic Society 36 (2016) 2659–2665 20

19

10

18

10

17

2.5 10-1 -1

10

Electrical Resistivity

Carrier Density (cm-3)

19

1.4 10

10

(b)

-2

6.8 10

-2

10

SCB

10

(c)

1

10

5.4

0

-1

10

6.7 10

-1

10

-2

10

-3

-3

SCA

cm)

19

3.7 10

10

2

0

10

(a)

Carrier Mobility (cm2/V/s)

10

2663

SCA

SCB

10

SCA

SCB

Fig. 5. (a) Carrier density, (b) carrier mobility, and (c) electrical resistivity of SiC specimens sintered with 6.5 vol% Y2 O3 -Sc2 O3 -AlN additives.

formed a larger amount of grain boundary phases (probably from the Y-Sc-Al-Si-OCN melt) than that of the SC1 (1 vol% Y2 O3 -Sc2 O3 ). Since the thermal conductivity of the OCN phase should be lower than that of SiC ceramic, a larger amount of additives led to a lower thermal conductivity. (3) The grain sizes of the specimens (1.4 ␮m and 1.9 ␮m for SCA and SCB, respectively) were smaller than that of SC1 (∼10 ␮m). The smaller grain size led to more grain boundaries per unit volume, resulting in increased phonon scattering at the grain boundaries. Likewise, since the grain size of SCA was smaller than that of SCB, the thermal conductivity of SCA (∼92 Wm−1 K−1 ) was lower than that of SCB (∼110 Wm−1 K−1 ). This was evidenced by a shorter phonon mean free path of SCA than SCB (Fig. 4(d)). (4) Although the SCB was sintered at a lower temperature (1950 ◦ C) than the SC1 (2050 ◦ C), the SCB showed a greater amount of phase transformation (∼70% of 3C phase transformed to 4H or 6H phases) than the SC1 (30% of 3C phase transformed to 4H or 6H phases). This was due to (i) the presence of Al, which makes Si vacancies in the SiC lattice and (ii) the lower viscosity of the Y-Sc-Al-Si-OCN melt than the Y-Sc-Si-OCN melt [37]. The phase transformation creates more 3C/4H and 3C/6H phase boundaries in SiC grains, leading to increased phonon scattering at the interfaces. The thermal conductivity values of LPS-SiC ceramics processed without an applied pressure were in the range of 50–90 Wm−1 K−1 , depending on the additive composition and heat-treatment conditions after sintering [16–18,38]. Thus, the thermal conductivity value of SCB (∼110 Wm−1 K−1 ) is the highest value observed for LPS-SiC ceramics processed without an applied pressure. The high thermal conductivity of SCB was attributed to the beneficial effect of Y2 O3 -Sc2 O3 additives [8] and the minimal addition of an Alcontaining additive (1.5 vol% AlN). The additives reduced lattice oxygen content in the SiC lattice by forming a Y-Sc-Al-Si-OCN melt

from 0.67 wt% in the starting ␤-SiC powders to 0.25 wt% in the pressureless sintered SiC ceramic (SCB). The lattice oxygen content was also reduced from 0.63 wt% in the starting ␣-SiC powders to 0.27 wt% in the pressureless sintered SiC ceramic (SCA). (3) Electrical properties Fig. 5 shows the carrier density, carrier mobility, and electrical resistivity of the SCA and SCB; all these values were obtained by Hall measurements at room temperature. Both specimens were found to be n-type semiconductors. The SCB had a higher carrier density and carrier mobility (3.7 × 1019 cm−3 and 2.5 × 10−1 cm2 /V/s) than SCA (1.4 × 1019 cm−3 and 6.8 × 10−2 cm2 /V/s). The electrical resistivities of SCA and SCB were 5.4  cm and 6.7 × 10−1  cm, respectively. The lower electrical resistivity of SCB than SCA can be understood by considering: (i) the difference in N-doping content between SCA and SCB [Fig. 5(a)] and (ii) the difference in polytypes in the SCA and SCB specimens (Table 2). These can be explained as follows. A Y-Sc-Al-Si-OCN glass forms due to the dissolution of nitrogen from the atmosphere during the present sintering process. Since nitrogen has some solubility in the SiC lattice [39,40], the precipitating composition during grain growth of SiC contains N as well as Si and C, resulting in N-doped SiC grains [41]. Since the SCB has a larger grain size than SCA, SCB should have a higher nitrogen content than SCA, as evidenced by a higher n-type carrier density in SCB than SCA [Fig. 5(a)]. The latter can be explained by the fact that the ionization energies of nitrogen donors in SiC are 56.5meV for 3C-SiC, 100–155 meV for 6H-SiC, and 66–124 meV for 4H-SiC [42]. According to the polytype content listed in Table 2, SCA consisted of ∼81% 6H and ∼19% 4H, whereas SCB consisted of ∼47% 4H, ∼27% 6H and ∼26% 3C. Thus, the effective donor ionization barrier should

Fig. 6. SEM images of the crack paths from a Vickers indentation in the (a) SCA and (b) SCB specimens.

2664

T.-Y. Cho et al. / Journal of the European Ceramic Society 36 (2016) 2659–2665

Fig. 7. Typical fracture surfaces of (a) SCA and (b) SCB specimens.

Table 3 Mechanical properties of pressureless sintered silicon carbide ceramics with Y2 O3 SC2 O3 -AlN. Specimen designation

Fracture toughness (MPa·m1/2 )

Flexural strength (MPa)

Hardness (GPa)

SCA SCB

4.1 ± 0.4 5.1 ± 0.5

509 ± 65 520 ± 56

27.2 ± 1.0 25.0 ± 1.1

be higher in SCA than SCB, resulting in a lower carrier mobility in SCA than SCB. The electrical resistivities of the present specimens were one or two-orders of magnitude higher than the previous SiC ceramics sintered with Y2 O3 [43], yttrium nitrate [39], and Y2 O3 -RE2 O3 (RE = Sm, Gd, Lu) [15] in a nitrogen atmosphere. This was due to the addition of AlN as well as Y2 O3 -Sc2 O3 . Since Al tends to substitute for Si sites in the SiC lattice [35,43], Al-doping is expected to create acceptors. The Al-derived acceptors are likely to compensate for N-derived donors in the band gap of SiC [44], resulting in higher electrical resistivity compared to those of SiC ceramics sintered without Al-containing additives [15,39,43]. (4) Mechanical properties The fracture toughness values of SCA and SCB are 4.1 and 5.1 MPa m1/2 , respectively (Table 3). Fracture toughness of LPS-SiC ceramics is strongly dependent on their microstructure as well as sintering additive composition [45–48]. Since the chemistry of the sintering additives of the two specimens is the same, the difference in fracture toughness should be caused by the difference in microstructure. The aspect ratios of SiC grains in SCA and SCB are 1.2 and 3.2, respectively. Thus, the higher fracture toughness of SCB was attributed to the enhanced crack deflection and bridging by elongated grains with a higher aspect ratio, as shown in Fig. 6. The fracture mode of SiC grains was mostly intergranular (Fig. 7), resulting from the weak interface created by the difference in coefficients of thermal expansion of the liquid and SiC grains on cooling after sintering. The fracture toughness values of pressureless LPS-SiC ceramics from the literature [23,47–51] were in the range of 3.1–8.0 MPa m1/2 and depended on the microstructure and the additive composition. Thus, the toughness values of the current SiC ceramics were moderate or lower than the reported values. The flexural strengths of the SCA and SCB samples were 509 and 520 MPa, respectively. The flexural strength values of pressureless LPS-SiC ceramics depend on the additive composition and the resultant microstructure [23,50,51] and were in the 320–565 MPa range at room temperature. The flexural strength values obtained in this study for the SCA and SCB samples (509 and 520 MPa, respectively) are close to the upper bound of the reported values for pressureless LPS-SiC ceramics. The excellent strength of the SCA and SCB samples may be caused by the higher fracture energy of the Y2 O3 -

Sc2 O3 -AlN additive system as compared to the other oxide additive systems. Since the fracture mode in liquid-phase sintered SiC is mostly intergranular fracture, cracks tend to propagate along grain boundaries and the fracture energy of the intergranular phase contributes to the flexural strength of the ceramics. The flexural strength and fracture toughness of the SCB specimen were 520 MPa and 5.1 MPa m1/2 , respectively. When the current data are compared with the reported values [26–28], the SCB showed well-balanced mechanical properties among the pressureless LPS-SiC ceramics. The hardness values of the SCA and SCB samples were 27.2 and 25.0 GPa, respectively. The hardness of pressureless LPS-SiC ceramics varied in the 22.0–29.2 GPa range, depending on their microstructure, grain boundary phase content, and residual porosity [52–55]. The hardness values obtained here were comparable to the reported values in pressureless LPS-SiC ceramics.

4. Conclusions Dense SiC ceramics from ␣- or ␤-SiC powders (SCA or SCB, respectively) were successfully fabricated using a new additive composition (Y2 O3 -Sc2 O3 -AlN) by pressureless sintering. The SiC ceramics fabricated from ␤-SiC powders and 6.5 vol% Y2 O3 Sc2 O3 -AlN additives showed the highest thermal conductivity value (∼110 Wm−1 K−1 ) among the pressureless LPS-SiC ceramics. The additives reduced oxygen content in the SiC lattice from 0.67 wt% in the starting ␤-SiC powders to 0.25 wt% in the pressureless sintered SiC ceramic (SCB). The thermal conductivity of SCA (∼92 Wm−1 K−1 ) was lower than that of SCB (∼110 Wm−1 K−1 ) because of its smaller grain size (1.4 ␮m) and, probably, higher lattice oxygen content (0.27 wt%) than those of SCB (1.9 ␮m and 0.25 wt%, respectively). The electrical resistivity values of the SiC ceramics fabricated from ␣- or ␤-SiC powders were 5.4 and 6.7 × 10−1  cm, respectively. The fracture toughness, flexural strength, and Vickers hardness values of the SCB were 5.1 MPa m1/2 , 520 MPa, and 25.0 GPa, respectively. In contrast, those values of SCA were 4.1 MPa m1/2 , 509 MPa, and 27.2 GPa, respectively. The new additive system (6.5 vol% Y2 O3 -Sc2 O3 -AlN) achieved the highest thermal conductivity in pressureless LPS-SiC ceramics while maintaining comparable mechanical properties to the reported values in pressureless LPS-SiC ceramics.

Acknowledgements This work (C0296400) was supported by a Business for Cooperative R&D between Industry, Academy and Research Institute grant funded by the Korea Small and Medium Business Administration in 2015.

T.-Y. Cho et al. / Journal of the European Ceramic Society 36 (2016) 2659–2665

References [1] K. Biswas, G. Rixecker, F. Aldinger, Improved high temperature properties of SiC-ceramics sintered with Lu2 O3 -containing additives, J. Eur. Ceram. Soc. 23 (2003) 1099–1104. [2] M. Balog, P. Sajgalik, M. Hnatko, Z. Lences, F. Monteverde, J. Keckes, J.L. Huang, Nano- versus macro-hardness of liquid phase sintered SiC, J. Eur. Ceram. Soc. 25 (2005) 529–534. [3] K. Strecker, M.J. Hoffmann, Effect of AlN-content on the microstructure and fracture toughness of hot-pressed and heat-treated LPS-SiC ceramics, J. Eur. Ceram. Soc. 25 (2005) 801–807. [4] F. Rodriguez-Rojas, A.L. Ortiz, F. Guiberteau, M. Nygren, Anomalous oxidation behaviour of pressureless liquid-phase-sintered SiC, J. Eur. Ceram. Soc. 31 (2011) 2393–2400. [5] K.Y. Lim, Y.W. Kim, T. Nishimura, W.S. Seo, High temperature strength of silicon carbide sintered with 1 wt.% aluminum nitride and lutetium oxide, J. Eur. Ceram. Soc. 33 (2013) 345–350. [6] M. Herrmann, K. Sempf, M. Schneider, U. Sydow, K. Kremmer, A. Michaelis, Electrochemical corrosion of silicon carbide ceramics in H2 SO4 , J. Eur. Ceram. Soc. 34 (2014) 229–235. [7] S. Kaur, R. Riedel, E. Ionescu, Pressureless fabrication of dense monolithic SiC ceramics from a polycarbosilane, J. Eur. Ceram. Soc. 34 (2014) 3571–3578. [8] Y.W. Kim, K.Y. Lim, W.S. Seo, Microstructure and thermal conductivity of silicon carbide with yttria and scandia, J. Am. Ceram. Soc. 97 (2014) 923–928. [9] R. Wei, S. Song, K. Yang, Y. Cui, Y. Peng, X. Chen, X. Hu, X. Xu, Thermal conductivity of 4H-SiC single crystals, J. Appl. Phys. 113 (2013) 053503-1-4. [10] G.A. Slack, Nonmetallic crystals with high thermal conductivity, J. Phys. Chem. Solids 34 (1973) 321–335. [11] Y. Zhou, K. Hirao, Y. Yamauchi, S. Kanzaki, Effects of rare-earth oxide and alumina additives on thermal conductivity of liquid-phase-sintered silicon carbide, J. Mater. Res. 18 (2003) 1854–1862. [12] K.M. Kim, S.H. Jang, Y.W. Kim, W.S. Seo, Thermal and mechanical properties of SiC-TiC0.5 N0.5 composites, J. Am. Ceram. Soc. 98 (2015) 616–623. [13] Y. Zhou, K. Hirao, K. Watari, Y. Yamauchi, S. Kanzaki, Thermal conductivity of silicon carbide densified with rare-earth oxide additives, J. Eur. Ceram. Soc. 24 (2004) 265–270. [14] K.J. Kim, K.Y. Lim, Y.W. Kim, Electrical and thermal properties of SiC ceramics sintered with yttria and nitrides, J. Am. Ceram. Soc. 97 (2014) 2943–2949. [15] S.H. Jang, Y.W. Kim, K.J. Kim, S.J. Lee, K.Y. Lim, Effects of Y2 O3 -RE2 O3 (RE = Sm, Gd Lu) additives on electrical and thermal properties of silicon carbide ceramics, J. Am. Ceram. Soc. 99 (2016) 265–272. [16] E. Volz, A. Roosen, W. Hartung, A. Winnacker, Electrical and thermal conductivity of liquid phase sintered SiC, J. Eur. Ceram. Soc. 21 (2001) 2089–2093. [17] L.S. Sigl, Thermal conductivity of liquid phase sintered silicon carbide, J. Eur. Ceram. Soc. 23 (2003) 1115–1122. [18] H. Liang, X. Yao, J. Zhang, X. Liu, Z. Huang, The effect of rare earth oxides on the pressureless liquid phase sintering of ␣-SiC, J. Eur. Ceram. Soc. 34 (2014) 2865–2874. [19] K.J. Kim, T.Y. Cho, Y.W. Kim, T. Nishimura, E. Narimatsu, Electrical and thermal properties of silicon carbide–boron nitride composites prepared without sintering additives, J. Eur. Ceram. Soc. 35 (2015) 4423–4429. [20] Y. Li, H. Wu, J. Yin, P. Lu, Y. Yan, X. Liu, Z. Huang, D. Jiang, High electrical resistivity of pressureless sintered in situ SiC–BN composites, Scr. Mater. 69 (2013) 740–743. [21] C. Zhang, X. Yao, Y. Li, H. Liang, J. Chen, J. Zhang, J. Yang, X. Li, T. Qiu, Z. Chen, X. Liu, Z. Huang, Effect of AlN addition on the thermal conductivity of pressureless sintered SiC ceramics, Ceram. Int. 41 (2015) 9107–9114. [22] Y. Li, J. Yin, H. Wu, P. Lu, Y. Yan, X. Liu, Z. Huang, D. Jiang, High thermal conductivity in pressureless densified SiC ceramics with ultra-low contents of additives derived from novel boron–carbon sources, J. Eur. Ceram. Soc. 34 (2014) 2591–2595. [23] Y.W. Kim, M. Mitomo, H. Emoto, J.G. Lee, Effect of initial ␣-phase content on microstructure and mechanical properties of sintered silicon carbide, J. Am. Ceram. Soc. 81 (1998) 3136–3140. [24] C.W. Jang, J. Kim, S.J.L. Kang, Effect of sintering atmosphere on grain shape and grain growth in liquid-phase-sintered silicon carbide, J. Am. Ceram. Soc. 85 (2002) 1281–1284. [25] O. Borrero-Lopez, A.L. Ortiz, F. Guiberteau, N.P. Padture, Sliding-wear-resistant liquid-phase-sintered SiC processed using ␣-SiC starting powders, J. Am. Ceram. Soc. 90 (2007) 541–545. [26] A. Gubernat, L. Stobierski, P. Labaj, Microstructure and mechanical properties of silicon carbide pressureless sintered with oxide additives, J. Eur. Ceram. Soc. 27 (2007) 781–789. [27] H. Liang, X. Yao, H. Zhang, X. Liu, Z. Huang, In situ toughening of pressureless liquid phase sintered ␣-SiC by using TiO2 , Ceram. Int. 40 (2014) 10699–10704. [28] M.F. Zawrah, L. Shaw, Liquid-phase sintering of SiC in presence of CaO, Ceram. Int. 30 (2014) 721–725.

2665

[29] M. Nader, F. Aldinger, M.J. Hoffmann, Influence of the ␣/␤-SiC phase transformation on microstructural development and mechanical properties of liquid phase sintered silicon carbide, J. Mater. Sci. 34 (1999) 1197–1204. [30] M.I. Mendelson, Average grain size in polycrystalline ceramics, J. Am. Ceram. Soc. 52 (1969) 443–446. [31] W.J. Parker, R.J. Jenkins, C.P. Butler, G.L. Abbott, Flash method of determining thermal diffusivity, heat capacity, and thermal conductivity, J. Appl. Phys. 32 (1961) 1679–1684. [32] NIST Property Data Summaries for Sintered Silicon Carbide (SiC), http://www. ceramics.nist.gov/srd/summary/scdscs.htm. (accessed on 20.04.15.). [33] ASTM C1421-15, Standard test methods for determination of fracture toughness of advanced ceramics at ambient temperature Annual Book of Standards, vol. 15.01, ASTM Int., West Conshohocken, PA, 2015, pp. 1–33. [34] Y.W. Kim, S.H. Lee, T. Nishimura, M. Mitomo, Heat-resistant silicon carbide with aluminum nitride and scandium oxide, Acta Mater. 53 (2005) 4701–4708. [35] Y. Tajima, W.D. Kingery, Solid solubility of aluminum and boron in silicon carbide, J. Am. Ceram. Soc. 65 (1982) C-27–C-29. [36] J.H. Eom, Y.W. Kim, I.H. Song, Effects of the initial ␣-SiC content on the microstructure mechanical properties, and permeability of macroporous silicon carbide ceramics, J. Eur. Ceram. Soc. 32 (2012) 1283–1290. [37] S. Hampshire, M.J. Pomeroy, Effect of composition on viscosities of rare earth oxynitride glasses, J. Non-Cryst. Solids 344 (2004) 1–7. [38] H. Liang, X. Yao, H. Deng, H. Zhang, X. Liu, Z. Huang, High electrical resistivity of spark plasma sintered SiC ceramics with Al2 O3 and Er2 O3 as sintering additives, J. Eur. Ceram. Soc. 35 (2015) 399–403. [39] Y.W. Kim, K.J. Kim, H.C. Kim, N.H. Cho, K.Y. Lim, Electrodischarge machinable silicon carbide ceramics sintered with yttrium nitrate, J. Am. Ceram. Soc. 94 (2011) 991–993. [40] S.H. Jang, Y.W. Kim, K.J. Kim, Effects of Y2 O3 –RE2 O3 (RE = Sm, Gd, Lu) additives on electrical and mechanical properties of SiC ceramics containing Ti2 CN, J. Eur. Ceram. Soc. (2015), http://dx.doi.org/10.1016/j.jeurceramsoc.2015.09. 042. [41] Y.W. Kim, T.Y. Cho, K.J. Kim, Effect of grain growth on electrical properties of silicon carbide ceramics sintered with gadolinia and yttria, J. Eur. Ceram. Soc. 35 (2015) 4137–4142. [42] M. Ikeda, H. Matsunami, T. Tanaka, Site effect on the impurity levels in 4H 6H, and 15R SiC, Phys. Rev. B: Solid State. 22 (1980) 2842–2854. [43] K.J. Kim, K.Y. Lim, Y.W. Kim, Influence of Y2 O3 addition on electrical properties of ␤-SiC ceramics sintered in nitrogen atmosphere, J. Eur. Ceram. Soc. 32 (2012) 4401–4406. [44] K.J. Kim, K.Y. Lim, Y.W. Kim, Control of electrical resistivity in silicon carbide ceramics sintered with aluminum nitride and yttria, J. Am. Ceram. Soc. 96 (2013) 3463–3469. [45] J.Y. Kim, Y.W. Kim, M. Mitomo, G.D. Zhan, J.G. Lee, Microstructure and mechanical properties of ␣-silicon carbide sintered with yttrium–aluminum garnet and silica, J. Am. Ceram. Soc. 82 (1999) 441–444. [46] Y. Zhou, K. Hirao, M. Toriyama, Y. Yamauchi, S. Kanzaki, Effects of intergranular phase chemistry on the microstructure and mechanical properties of silicon carbide ceramics densified with rare-earth oxide and alumina additions, J. Am. Ceram. Soc. 84 (2001) 1642–1644. [47] Y.W. Kim, Y.S. Chun, T. Nishimura, M. Mitomo, Y.H. Lee, High-temperature strength of silicon carbide ceramics sintered with rare-earth oxide and aluminum nitride, Acta Mater. 55 (2007) 727–736. [48] E. Ciudad, O. Borrero-Lopez, F. Rodriguez-Rojas, A.L. Ortiz, F. Guiberteau, Effect of intergranular phase chemistry on the sliding-wear resistance of pressureless liquid-phase-sintered ␣-SiC, J. Eur. Ceram. Soc. 32 (2012) 511–516. [49] N.P. Padture, In situ-toughened silicon carbide, J. Am. Ceram. Soc. 77 (1994) 519–523. [50] J.H. Eom, Y.K. Seo, Y.W. Kim, S.J. Lee, Effect of additive composition on mechanical properties of pressureless sintered silicon carbide ceramics sintered with alumina, aluminum nitride and yttria, Met. Mater. Int. 21 (2015) 525–530. [51] A. Gubernat, L. Stobierski, P. Labaj, Microstructure and mechanical properties of silicon carbide pressureless sintered with oxide additives, J. Eur. Ceram. Soc. 27 (2007) 781–789. [52] S.H. Lee, H. Tanaka, Y. Kagawa, Spark plasma sintering and pressureless sintering of SiC using aluminum borocarbide additives, J. Eur. Ceram. Soc. 29 (2009) 2087–2095. [53] A.L. Ortiz, O. Borrero-Lopez, M.Z. Quadir, F. Guiberteau, A route for the pressureless liquid-phase sintering of SiC with low additive content for improved sliding-wear resistance, J. Eur. Ceram. Soc. 32 (2012) 965–973. [54] R. Vassen, D. Stover, Processing and properties of nanograin silicon carbide, J. Am. Ceram. Soc. 82 (1999) 2585–2593. [55] O. Borrero-Lopez, A. Pajares, A.L. Ortiz, F. Guiberteau, Hardness degradation in liquid-phase-sintered SiC with prolonged sintering, J. Eur. Ceram. Soc. 27 (2007) 3359–3364.