Sie sind auf Seite 1von 4

Available online at www.sciencedirect.

com

Scripta Materialia 59 (2008) 638641 www.elsevier.com/locate/scriptamat

Spark plasma sintering and mechanical behaviour of ZrC-based composites


Diletta Sciti,a,* Stefano Guicciardia and Mats Nygrenb
b

CNR-ISTEC, Institute of Science and Technology for Ceramics, Via Granarolo 64, I-48018 Faenza, Italy Department of Inorganic Chemistry, Arrhenius Laboratory, Stockholm University, SE-10691, Stockholm, Sweden
Received 5 May 2008; accepted 17 May 2008 Available online 5 June 2008

ZrC materials containing MoSi2 as a sintering aid in the volumetric range 09 vol.% were fully consolidated by spark plasma sintering at temperatures between 1750 and 2100 C. Addition of MoSi2 resulted in a decrease in the sintering temperature, renement of the microstructure and improvement of the mechanical properties in comparison with the monolithic material. Hardness values were in the range 2528 GPa, Youngs modulus was in the range 420470 GPa and the exural strength was in the range 400600 MPa. 2008 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved.
Keywords: Ceramic; Composites; Spark plasma sintering; Microstructure; Mechanical properties

Zirconium carbide is one of the refractory transition metal carbides from Groups IV to VI of the periodic table. Its most important properties are its high melting point ($3420 C) [1], high hardness ($25.5 GPa) [1], high electrical conductivity (78 106 X cm) and high modulus of elasticity ($440 MPa) [1]. This compound is suitable for many applications such as eld emitters, coating of nuclear particle fuels and ultrahigh-temperature environments [25]. However, the use of zirconium carbide in engineering applications has been limited by the lack of a commercially viable sintering process. As an ultra-refractory compound, very high temperatures and pressure-assisted techniques are required to achieve dense bodies, due to its highly covalent bonding character and low self-diusion coecients [38]. One promising approach for improving the densication is the use of eld-assisted sintering techniques such as spark plasma sintering (SPS). This technique employs a pulsed DC current to activate and improve sintering kinetics [912]. Higher densities, rened microstructures, clean grain boundaries and elimination of surface impurities have been reported, which, in turn, result in an overall improvement in the materials performance [912]. The SPS process is hence particularly attractive for industrial applications of poorly sinterable materials.
* Corresponding author. Tel.: +39 0546 699748; fax: +39 0546 46381; e-mail: diletta.sciti@istec.cnr.it

This contribution is part of an extensive study including ZrB2, HfB2 and HfC, consolidated by SPS with MoSi2 as sintering additive [13,14]. The aim of this work is to exploit the potential of the SPS technique, attempting the densication with as little sintering aid as possible, i.e. 0, 1, 3 and 9 vol.% MoSi2. The microstructures and mechanical properties of dense ZrC composites were then studied as a function of the starting MoSi2 amount and compared to the monolithic material. The following commercial powders were used as raw materials: cubic ZrC Grade B (H.C. Starck, Germany), mean particle size 3.8 lm, impurities: Cfree 1.5%, O 0.6%, N 0.8%, Fe 0.05%, Hf: 2%; tetragonal MoSi2 < 2 lm (Aldrich, Milwaukee, USA), mean particle size 2.8 lm, O content $1 wt.%. The monolithic material was labelled Z0. The composites were prepared adding 1, 3, 9 vol.% MoSi2, and were labelled Z1, Z3 and Z9, respectively. The raw powders were ultrasonically treated in absolute ethanol, ball milled for 24 h using zirconia milling media, dried in a rotary evaporator and sieved to 60 mesh screen. For the sintering routes a double layer of graphite blanket was placed around the die in order to minimize the heat loss. The temperature was measured by an optical pyrometer focused on the graphite die. The heating rate was set at 300 C min1 up to 600 C, 200 C min1 up to 1400 C, 100 C min1 from 1400 C to the maximum temperature. The uniaxial pressure of 100 MPa (65 MPa for monolithic ZrC) was applied at 600 C.

1359-6462/$ - see front matter 2008 Acta Materialia Inc. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.scriptamat.2008.05.026

D. Sciti et al. / Scripta Materialia 59 (2008) 638641

639

The holding time was 3 min. The vacuum level ranged from 5 Pa at low temperatures to 10 Pa at high temperatures. The sintered pellets were 12 mm in diameter and 5 mm in height. Bulk densities were measured by the Archimedes method. Crystalline phases were identied by X-ray diraction (Siemens D500, Germany). The microstructures were analyzed by scanning electron microscopy (SEM, Cambridge S360, UK) and energy dispersive spectroscopy (EDS, INCA Energy 300, Oxford Instruments, UK). The nanoindentation tests were performed using a commercial nanoindenter (Nano Indenter XP, MTS Systems Corporation, Oak Ridge, TN, USA) with a diamond Berkovich indenter and a peak load of 500 mN. Hardness (H) and Youngs modulus (E) were calculated by the data software of the nanoindenter (TestWorksTM v. 4.06A) based on the model of Oliver and Pharr [15]. In each specimen, at least three runs of 10 indents each were made in dierent areas. Vickers microhardness (HV1.0) was measured with a load of 9.81 N, using a Zwick 3212 tester (Zwick, Germany). Eight indentations were made in each sample. Fracture toughness (KIc) was evaluated by the direct crack measurement (DCM) method, using the equation of Anstis et al. [16]. Five indentations were performed in each sample. Finally, the exural strength (r) was measured on 1.0 0.8 10 mm bars. Such tiny bars were a necessity due to the small dimensions of the starting pellets. The exural tests were performed on a three-point jig with a lower span of 8 mm on a universal testing machine (Instron 6025, Instron Ltd., UK). The nal relative densities are plotted against the sintering temperatures in Figure 1a. The relative densities were determined as the ratio between experimental bulk densities and theoretical ones calculated from the rule of mixtures on the basis of the starting compositions. As previously mentioned, the sintering temperatures were recorded on the external wall of the graphite die. Extensive studies on the SPS process have demonstrated that for electrically conductive materials, such as ZrC and MoSi2, the actual temperature inside the specimens is signicantly higher than on the external wall of the die where the pyrometer is focused [12]. The temperature mismatch depends on many factors such as the size of the die, the quality of the vacuum, the level of insulation of the die, the thermal and electrical conductivity of the

die, etc. Under the present conditions the temperature mismatch was estimated to increase from 150 to 250 C in the range 17502100 C [1214]. From data reported in Figure 1a, it is apparent that increasing the amount of MoSi2 lowered the maximum temperature required to obtain dense materials. Undoped ZrC (Z0) was densied at 2100 C (actual temperature >2300 C), reaching a relative density in excess of 99%. Densities in excess of 9798% were obtained at temperatures between 1900 and 1950 C for Z1 and Z3, respectively. For Z9, densities in excess of 99.5% were obtained at 1750 C (i.e. T > 1900 C). Correspondingly, increasing the MoSi2 content from 0 to 9%, reduced the mean grain size of the sintered materials from 13 to 3 lm (Fig. 1b). Typical relative density curves and shrinkage rate curves are shown in Figure 2a and b. The addition of MoSi2 caused an increase of the shrinkage rate, which can be ascribed to a change in the sintering mechanism from a solid-state one for undoped ZrC to a liquid phase one for the composites. The maximum in the shrinkage curve is believed to correspond to a change in the densication mechanism from being mainly surface diusion-controlled to being grain boundary diusion-controlled [10]. This maximum is shifted towards lower relative densities for higher MoSi2 content. This suggests that liquid phase formation promotes grain boundary diusion at lower densities, thus hindering the grain growth caused by surface diusion. Both the decrease in sintering temperature and the earlier activation of grain boundary diusion are believed to contribute to the microstructure renement. The X-ray diraction patterns of Z0, Z1 and Z3 contained only ZrC reections. The pattern of Z9 contained a few reections of low intensity originating from MoSi2 and SiC. The polished sections of dense monolith and composites are shown in Figure 3ae. Z0 contained about 2 vol.% of carbon which derives from carbon impurities present in the starting ZrC powder. Beside residual carbon, reaction phases recognized as SiC and ZrSi phases were identied in Z1, sintered at 1950 C, (Fig. 3b and c). Formation of SiC and ZrSi was also observed in Z3 sintered at 1900 C (Fig. 3d), while Z9 contained residual MoSi2 and newly formed SiC (Fig. 3e). The amount of SiC increased with increasing MoSi2 content, namely from $2% in Z1 to $5% in Z9 (Table 1).

1.00

b
Mean grain size, mm

14 13 12 11 10 9 8 7 6 5 4 3

0.98

Z0 Z1 Z3 Z9

Relative density

0.96

0.94 Z0 Z1 Z3 Z9 1700 1800 1900 2000 2100 2200

0.92

0.90
1700 1800 1900 2000 2100

Sintering temperature, C

Sintering Temperature, C

Figure 1. (a) Relative density and (b) mean grain size vs. sintering temperature.

640

D. Sciti et al. / Scripta Materialia 59 (2008) 638641

a
Relative density

1.00

b
3.0x10
-3

0.95 2.5x10

-3

Z0 Z1 Z3 Z9

(l/Lo)/dt, s

0.90

-1

2.0x10 1.5x10 1.0x10 5.0x10

-3

0.85 Z0 Z1 Z3 Z9

-3

-3

0.80

-4

0.75 1600 1800 2000

0.0 2200 0.7 0.8 0.9 1.0

Sintering temperature, C

Relative density

Figure 2. (a) Typical relative density curves and (b) shrinkage rate curves for compositions Z0, Z1, Z3, Z9.

c
C

20 m

10 m

3 m 3 3 1
C Si
KeV

d
2

1 1 1

1 1

2 3
C

Si

Zr

KeV

3 3

10 m

5 m

Figure 3. Polished sections of ZrC materials and EDS spectra: (a) Z0, (b) and (c) Z1, (d) Z3 and (e) Z9 (BSE imaging). 1, SiC; 2, ZrSi; 3, MoSi2.

Table 1. Relevant parameters for selected high-density materials Sample label Z0 Z1 Z3 Z9 Initial composition (vol.%) ZrC ZrC + 1MoSi2 ZrC + 3MoSi2 ZrC + 9MoSi2 Sintering Mean Density Residual SiC H parametersa (g/cm3) (%) MoSi2 grain (vol.%) (GPa) (C/min/MPa) size (lm) (vol.%) 2100/3/65 1950/3/100 1900/3/100 1700/3/100 13 1 7.0 0.7 5.8 0.6 3.5 0.4 10.1 12.3 12.4 11.9 98.0 98.0 99.7 99.0 Traces Traces 67 0 2 4 5 25.2 1.4 24.9 1.0 25.1 2.3 26.8 1.3 E (GPa) 464 22 420 10 444 17 467 22 HV1.0 (GPa) 17.9 0.6 18.8 0.3 18.4 0.7 20.0 0.5 r (MPa) KIc (MPa m1/2) 2.1 0.2 3.2 0.4 3.3 0.4 407 38 495 61 591 48

H, nanoindentation hardness; E, nanoindentation Youngs modulus; HV1.0, Vickers hardness; KIc, fracture toughness; r, three-point exural strength. a Temperature recorded on the external wall of the graphite die.

The addition of MoSi2 as sintering aid enhanced the densication of ZrC even in amounts as low as 1 vol.%. Carbon impurities in the staring ZrC powder could have aided the densication of Z0, by cleaning the particles surface of oxides which hinder densication, as previously seen for other carbides such as SiC [17]. However, the sudden increase in the densication

rate observed during our sintering cycles as well as the high wetting morphology of MoSi2 suggest the formation of a liquid phase originating from this silicide. According to the calculated CMoSi2 pseudo-binary phase diagram [18], above 1700 C a liquid phase containing liquid phase + MoSi2 + SiC is predicted to form at any C/MoSi2 concentration. During cooling, SiC

D. Sciti et al. / Scripta Materialia 59 (2008) 638641

641

particles are predicted to precipitate from the liquid, which is in agreement with SEM observations. In Z3 and Z9, the starting MoSi2 amount was high enough to convert all C into SiC. In Z1, unreacted C was found in the microstructure. For compositions treated at temperatures > 1900 C, such as Z1 and Z3, the formation of ZrSi was also favoured. SiC and ZrSi phases can form according to the following reactions, which have a negative Gibbs free energy: 5MoSi2 7C 7SiC Mo5 Si3 ZrC 5MoSi2 5C ZrSi 6SiC Mo5 Si3 1 2

crease of the initial MoSi2 content (from 407 to 591 MPa). The increased strength is due to the slight increase of the fracture toughness and the renement of the microstructure. The measured values are higher than those reported in the literature for other ZrC-based composites, ranging between 220 and 320 MPa [5]; however, the absolute values of exural strength could be articially high due to the small dimensions of the samples. The authors gratefully acknowledge G. Celotti for XRD analysis and C. Melandri for mechanical tests.
[1] H.O. Pierson, Handbook of Refractory Carbides and Nitrides, William Andrew Publishing/Noyes, Westwood, NJ, 1996, p. 68. [2] T. Ogawa, K. Ikawa, J. Nucl. Mater. 105 (1982) 331. [3] P. Barnier, C. Brodhag, F. Thevenot, J. Mater. Sci. 21 (1986) 2547. [4] H.J. Ryu, Y.W. Lee, S.I. Cha, S.H. Hong, J. Nucl. Mater. 352 (2006) 341. [5] E. Min-Haga, W.D. Scott, J. Mater. Sci. 23 (1988) 2865. [6] K.H. Kim, K.B. Shim, Mater. Charact. 50 (2003) 31. [7] T. Tsuchida, S. Yamamoto, Solid State Ionics 172 (2004) 215. [8] W.B. Johnson, A.S. Nagelberg, E. Breval, J. Am. Soc. 74 (1991) 2093. [9] M. Nygren, Z. Shen, Sil. Ind. 69 (2004) 211. [10] Z. Shen, M. Nygren, The Chem. Rec. 5 (2005) 173184. [11] J.R. Groza, M. Garcia, J.A. Schneider, J. Mater. Res. 16 (2001) 286. [12] U. Anselmi-Tamburini, Y. Kodera, M. Gasch, C. Unuvar, Z.A. Munir, M. Ohyanagi, S.M. Johnson, J. Mater. Sci. 41 (2006) 30973104. [13] D. Sciti, L. Silvestroni, M. Nygren, J. Eur. Ceram. Soc. 28 (2008) 1287. [14] D. Sciti, S. Guicciardi, M. Nygren, J. Am. Ceram. Soc. 91 (2008) 1433. [15] W.C. Oliver, G.M. Pharr, J. Mater. Res. 7 (1992) 1564. [16] G.R. Anstis, P. Chantikul, B.R. Lawn, D.B. Marshall, J. Am. Ceram. Soc. 64 (1988) 533. [17] K. Motzfeld, in: Proceedings of the International Conference on Engineering Ceramics92, Retroprint, Bratislava, 1993. [18] X. Fan, K. Kack, T. Ishigawi, Mater. Sci. Eng. A 278 (2000) 46. [19] S.E. Landwehr, G.E. Hilmas, W. Fahrenholtz, I.G. Talmy, J. Am. Ceram. Soc. 90 (2007) 1998. [20] A.C. Fischer-Cripps, Nanoindentation, second ed., Springer-Verlag, New York, 2004. [21] I.J. McColm, Ceramic Hardness, Plenum Press, New York, 1990. [22] M. Taya, S. Hayashi, A.S. Kobayashi, H.S. Yoon, J. Am. Ceram. Soc. 73 (1990) 1382. [23] J.F. Shackelford, W. Alexander, CRC Materials Science and Engineering Handbook, CRC Press, Boca Raton, FL, 2001.

Reactions (1) and (2), however, imply the formation of Mo-based species, such as Mo5Si3, which was not detected in any of the compositions. It is possible that Mo enters the lattice of ZrC, forming a solid solution, instead of forming new Mo5Si3 phases. Recent TEM studies, which will be object of future publications, seem to suggest this possibility. Furthermore, it was reported that Mo has a signicant solubility in ZrC between 1800 and 2100 C [19]. The measured mechanical properties are displayed in Table 1. The nanoindentation hardness values are in the range 2527 GPa, with little variation with respect to the starting composition and the formation of new phases such as SiC or ZrSi. These values are in agreement with those reported in the literature for monolithic ZrC (25 GPa [1]). The slight increase observed for the composition with 9% could be related to the renement of the microstructure. Similar conclusions can be drawn for the microhardness values. The lower values of the Vickers microhardness with respect to the nanoindentation hardness can be explained by considering the dierent contact areas which are used in the calculation of the Vickers and Berkovich hardness [20] and the indentation size eect (ISE) [21]. The Youngs modulus values (420 470 GPa, Table 1) of the monolith and the composites are generally higher than values reported in the literature for monolithic ZrC (350440 GPa [1]). Apart from ZC1, all the values were not sensitive to the formation of new phases. Excessive spalling at the indentation sites prevented the measurement of fracture toughness by the DCM method in Z0. For doped compositions, there was a modest increase in toughness with increase in the initial MoSi2 content (Table 1). These features suggest that when the MoSi2 phase is retained in the microstructure a toughening mechanism can be activated, as for example that based on residual stresses [22]. The thermal expansion coecient of ZrC (6.7 106 C1 [1]), is lower than that of MoSi2 (89 106 C1 [23]). This thermal expansion mismatch could generate residual compressive stresses in the matrix which promote an increase in fracture toughness. The exural strength (Table 1) showed a monotonic increase with in-

Das könnte Ihnen auch gefallen