Science Direct E噩≈RS ELSEVIER Joumal of the European Ceramic Society 30(2010)1131-1137 r. com/loca Fabrication of cnt-SiC/Sic composites by electrophoretic deposition Katja Konig Sasa Novak a,, Aljaz Ivekovic a, Katja Rade Decheng Meng Aldo r Boccaccini, Spom Department for Nanostructured Materials, Jozef Stefan Institute, Ljubljana, Slovenia Available online 20 August 2009 Due to the outstanding mechanical and thermal properties of carbon nanotubes( CNTs), they are considered suitable reinforcement for structural materials. In this study, for the first time, electrophoretic deposition(EPD)was used to deposit(multi-walled) CNTs onto SiC fibres(SiCt) orm an effective CNT interphase layer for SiC:/SiC composites. This deposition was followed by electrophoretic infiltration of the CNT-coated Sic fibre mats with Sic powder to fabricate a new CNT-SiC-fibre-reinforced Sic-matrix(SiC/SiC)composite for fusion applications. In these EPD experiments, a commercial aqueous suspension of negatively charged CNTs and an optimized aqueous suspension of negatively charged Sic particles were used. The CNT-coatings on the SiC fibres were firm and homogenous, and uniformly distributed nanotubes were observed on the fibre surfaces. In a following step of EPD, a thick Sic layer was formed on the fibre mat when the CNT-coated SiC fibres were in contact with the positive electrode of the EPD cell; however, spaces between the fibres were not fully filled with SiC. Conversely, when CNT-coated SiC fibres were isolated from the electrode, the Sic particles were able to gradually fill the fibre mat resulting in relatively high infiltration, which leads to dense composites 2009 Elsevier Ltd. All rights reserved. Keywords: Electrophoretic deposition; SiC:/SiC composite; Carbon nanotubes(CNTs) Coatings: Interphase 1. Introduction an interphase layer, typically pyrolitic carbon of 100-500nm that is usually deposited by a chemical vapour deposition(CVD) Continuous SiC-fibre-reinforced Sic composites have been process, and has the main task to protect the material from catas- recognized as promising materials for use in the structural parts trophic fracture by deflecting the propagating crack and inducing of future fusion reactors due to the high thermal stability of Sic fibre pull-out, which is the most efficient toughening mechanism offering the possibility of high temperature operation of the reac- in this class of composites tor and hence more economic energy production. In addition, However, no state-of-the-art material meets all of the due to the low neutron activation of pure SiC, these composites highly demanding requirements for use in fusion reactors. The represent a way to minimise the amount of radioactive waste ceramic composites, for example, are associated with several generated during reactor operation in comparison with ferrous still unresolved issues, including gas permeability, insufficient materials. To ensure sufficient fracture toughness and reliabil- mechanical properties at high temperature, too high neutron ty of the material, Sic is proposed to be used in the form of activation, too low thermal conductivity, and instability of the continuous SiC-fibre-reinforced SiC ceramics, i.e., a( SiCr/Sic) fibres at high temperatures. The thermal conductivity and the composite. The composite is composed of SiC fibres woven mechanical properties of a composite are highly dependent on in a 2D or 3D architecture, filled with a matrix material that is the chemical composition of its three constituents(the matrix either pure Sic produced by chemical vapour infiltration or by material, the fibres, and the interphase layer), and on its porosity infiltration of a polymer precursor, or with a particulate com- Among the strategies to improve the overall performance of posite of a liquid phase sintered SiC. The fibres are coated with SiCrSiC composites, in particular to increase their mechani cal strength and thermal conductivity, incorporation of carbon nanotubes(CNTs)is highly promising, due to their outstanding Corresponding author at: Jamova cesta 39. SI-1000 Ljubljana. Sloveni mechanical and thermal properties. The CNTs are characterised Te.:+38614773271;fax:+38614773221l not only by a very high thermal conductivity(2000 W/mkp-7 E-mail address: sasa. novak@ijs si(S. Novak) and ability to increase the toughness of intrinsically brittle 0955-2219 front matter@ 2009 Elsevier Ltd. All rights reserved. doi: 10. 1016/j-jeurceramsoc. 2009.07.027
Available online at www.sciencedirect.com Journal of the European Ceramic Society 30 (2010) 1131–1137 Fabrication of CNT-SiC/SiC composites by electrophoretic deposition Katja König a, Sasa Novak ˇ a,∗, Aljaz Ivekovi ˇ cˇ a, Katja Rade a, Decheng Meng b, Aldo R. Boccaccini b, Spomenka Kobe a a Department for Nanostructured Materials, Joˇzef Stefan Institute, Ljubljana, Slovenia b Department of Materials, Imperial College London, London, UK Available online 20 August 2009 Abstract Due to the outstanding mechanical and thermal properties of carbon nanotubes (CNTs), they are considered suitable reinforcement for structural materials. In this study, for the first time, electrophoretic deposition (EPD) was used to deposit (multi-walled) CNTs onto SiC fibres (SiCf) to form an effective CNT interphase layer for SiCf/SiC composites. This deposition was followed by electrophoretic infiltration of the CNT-coated SiC fibre mats with SiC powder to fabricate a new CNT-SiC-fibre-reinforced SiC-matrix (SiCf/SiC) composite for fusion applications. In these EPD experiments, a commercial aqueous suspension of negatively charged CNTs and an optimized aqueous suspension of negatively charged SiC particles were used. The CNT-coatings on the SiC fibres were firm and homogenous, and uniformly distributed nanotubes were observed on the fibre surfaces. In a following step of EPD, a thick SiC layer was formed on the fibre mat when the CNT-coated SiC fibres were in contact with the positive electrode of the EPD cell; however, spaces between the fibres were not fully filled with SiC. Conversely, when CNT-coated SiC fibres were isolated from the electrode, the SiC particles were able to gradually fill the fibre mat resulting in relatively high infiltration, which leads to dense composites. © 2009 Elsevier Ltd. All rights reserved. Keywords: Electrophoretic deposition; SiCf/SiC composite; Carbon nanotubes (CNTs); Coatings; Interphase 1. Introduction Continuous SiC-fibre-reinforced SiC composites have been recognized as promising materials for use in the structural parts of future fusion reactors due to the high thermal stability of SiC offering the possibility of high temperature operation of the reactor and hence more economic energy production.1 In addition, due to the low neutron activation of pure SiC, these composites represent a way to minimise the amount of radioactive waste generated during reactor operation in comparison with ferrous materials. To ensure sufficient fracture toughness and reliability of the material, SiC is proposed to be used in the form of continuous SiC-fibre-reinforced SiC ceramics, i.e., a (SiCf/SiC) composite.1–3 The composite is composed of SiC fibres woven in a 2D or 3D architecture, filled with a matrix material that is either pure SiC produced by chemical vapour infiltration or by infiltration of a polymer precursor, or with a particulate composite of a liquid phase sintered SiC. The fibres are coated with ∗ Corresponding author at: Jamova cesta 39, SI-1000 Ljubljana, Slovenia. Tel.: +386 1 477 3271; fax: +386 1 477 3221. E-mail address: sasa.novak@ijs.si (S. Novak). an interphase layer, typically pyrolitic carbon of 100–500 nm that is usually deposited by a chemical vapour deposition (CVD) process, and has the main task to protect the material from catastrophic fracture by deflecting the propagating crack and inducing fibre pull-out, which is the most efficient toughening mechanism in this class of composites.4 However, no state-of-the-art material meets all of the highly demanding requirements for use in fusion reactors. The ceramic composites, for example, are associated with several still unresolved issues, including gas permeability, insufficient mechanical properties at high temperature, too high neutron activation, too low thermal conductivity, and instability of the fibres at high temperatures. The thermal conductivity and the mechanical properties of a composite are highly dependent on the chemical composition of its three constituents (the matrix material, the fibres, and the interphase layer), and on its porosity. Among the strategies to improve the overall performance of SiCf/SiC composites, in particular to increase their mechanical strength and thermal conductivity, incorporation of carbon nanotubes (CNTs) is highly promising, due to their outstanding mechanical and thermal properties. The CNTs are characterised not only by a very high thermal conductivity (>2000 W/mK)5–7 and ability to increase the toughness of intrinsically brittle 0955-2219/$ – see front matter © 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.jeurceramsoc.2009.07.027
1132 K. Konig et al. Journal of the European Ceramic Society 30(2010)1131-1137 materials, probably through"bridging"and"pull-out"tough The EPD experiments were performed with(1)aqueous sus- ening mechanisms. -l but also by a relatively high thermal pensions containing 0.0625-0.5 wt of carbon nanotubes at a stability in non-oxidising atmospheres. The potential benefit of constant dc voltage of 2. 8V for 10 or 20 min, and/or(2)sus- CNTs in ceramic matrix composites which arises from their pensions containing 60 wt %o of Sic powder at a constant dc tubular three-dimensional structure, has prompted significant voltage of 30V for 5 min. Electrodes made of steel, copper research in the field, with CNTs and CNT-reinforced mate-( both of dimensions: 20 mm x 10 mm x0. 2 mm), or graphite rials being among the most studied materials within the last 20mm x 5 x 2 mm) were vertically immersed in the sus- decade.2 The results from these studies strengthen the belief pension in the EPD cell. The distance between electrodes was that CNTs will enable a breakthrough in the development of 2 cm. In the CNT- and SiC-coating processes, the Sic-fibre high-performance structural materials. 3 mat(Tyrano SA, Ube Industries, Ltd, Japan) was placed in Carbon nanotubes have been incorporated into many front of the electrode, whereas for the Sic-infiltration, the Sic different inorganic matrices, ranging from glass and bioac- fibre mat was electrically insulated from the electrode with a tive glass-ceramicl4-18 to various polycrystalline engineering Teflon% spacer, as described elsewhere. 40 Before deposition, (e.g, Al2O3, BaTiO3, TiN, Si3N4, SiC). 9-2) There the electrodes were thoroughly cleaned in acetone using an however,due to a lack of quantitative data available no firm con- with compressed air. The SiC fabric was pre-treated wil o are also a few reports on CNTs grown on ceramic fibres 26-29 ultrasonic treatment, washed with distilled water, and drie clusion can be drawn about the property improvement achieved sodium dioctyl-sulfosuccinate(SDOSS) solution to improve by the in situ synthesised CNTs. Since the as-prepared CNTs wettability. After the deposition, the samples were dried in re intrinsically inert and uncharged, it is known that they are air at room temperature, vacuum-infiltrated by an aluminium ifficult to disperse in liquids. Acid treatment and use of specific dihydrogen phosphate(Al(H2 PO4)3)aqueous solution (TKI surfactants enable a relatively good dispersion based on a cer- Hrastnik, Slovenia), and sintered at 1300C in argon(Ar)for tain surface charge. This in turn enables processing in an electric 3 h. The microstructures of the green and sintered samples were field to form deposits on different substrates. - Recent papers observed using field emission gun scanning electron micro- also report on composite deposits fabricated by co-deposition or scopes(FEG-SEM, LEO 1525, and Zeiss SUPRA 35VP). The subsequent deposition of CNTs with different ceramic powders sintered samples were inspected by EDs (e.g. SiO2, Fe304, TiO2, and HA).34-38 To the best knowledge of the authors, the use of EPD to coat 3. Results and discussion SiC fibre mats with CNTs has however never been reported CNTs have been electrophoretically deposited on Ito glass, The scanning electron micrograph in Fig. I shows the mor- graphite and metallic substrates, as reviewed elsewhere 30 and phology of the carbon nanotubes used in this study. The carbon on carbon fibre paper.In this study, we report on the elec- nanotubes have a high aspect ratio and a random and curled trophoretic deposition of CNTs onto SiC fibre mats, thereby structure, which is typically associated with CNTs produced by forming a continuous interphase layer on the fibres, followed by thermal chemical vapour deposition( CVD). 9 an infiltration of the coated-fibre mat with Sic powder, which forms a new CNT-SiC /SiC grade of fusion-relevant composite 3.1. Electrokinetic properties of SiC fibres and material characteristics of suspensions 2. Experimental Before performing the EPD experiments, we analysed the electrokinetic properties of the Sic powder and the CNTs sus- An aqueous suspension of multi-walled carbon nanotubes pensions(with solids contents of 25 and 1 wt %, respectively) (Aquacyl), with average diameter 9.5 nm and average length .5 um, was provided by Nanocyl S.A. ( Sambreville, Belgium) The solids content in the as-received suspension was l wt % In this study, we used a Sic powder with an average grain size of 0.5 um(SiC BF12, H Starck, Germany). The zeta-potential(ZP) of the SiC powder and of the CNTs was measured in undiluted suspensions using a ZetaProbe analyser( Colloidal Dynamics, USA). A ZetaPals instrument(Brookhaven, USA)was used to analyse the ZP of the crushed SiC fibres in diluted aqueous suspension The pH levels were adjusted using I M sodium droxide(NaOH) and 1 M hydrochloric acid (HCI). The sur- face charge of the Sic particles for EPD was modified with tetramethyl-ammonium hydroxide (TMAH, Sigma-Aldrich ), whereas the commercial CNT-suspension already contained anionic surfactant. The suspensions were homogenised using a 100 strong probe-type ultrasound(Hielscher Ultrasonics, Germany), before the ZP measurement and epd Fig 1. SEM micrograph of the used CNTs
1132 K. König et al. / Journal of the European Ceramic Society 30 (2010) 1131–1137 materials, probably through “bridging” and “pull-out” toughening mechanisms,8–11 but also by a relatively high thermal stability in non-oxidising atmospheres. The potential benefit of CNTs in ceramic matrix composites which arises from their tubular three-dimensional structure, has prompted significant research in the field, with CNTs and CNT-reinforced materials being among the most studied materials within the last decade.12 The results from these studies strengthen the belief that CNTs will enable a breakthrough in the development of high-performance structural materials.13 Carbon nanotubes have been incorporated into many different inorganic matrices, ranging from glass and bioactive glass–ceramic14–18 to various polycrystalline engineering ceramics (e.g., Al2O3, BaTiO3, TiN, Si3N4, SiC).19–25 There are also a few reports on CNTs grown on ceramic fibres;26–29 however, due to a lack of quantitative data available no firm conclusion can be drawn about the property improvement achieved by the in situ synthesised CNTs. Since the as-prepared CNTs are intrinsically inert and uncharged, it is known that they are difficult to disperse in liquids. Acid treatment and use of specific surfactants enable a relatively good dispersion based on a certain surface charge. This in turn enables processing in an electric field to form deposits on different substrates.31–33 Recent papers also report on composite deposits fabricated by co-deposition or subsequent deposition of CNTs with different ceramic powders (e.g., SiO2, Fe3O4, TiO2, and HA).34–38 To the best knowledge of the authors, the use of EPD to coat SiC fibre mats with CNTs has however never been reported; CNTs have been electrophoretically deposited on ITO glass, graphite and metallic substrates, as reviewed elsewhere30 and on carbon fibre paper.39 In this study, we report on the electrophoretic deposition of CNTs onto SiC fibre mats, thereby forming a continuous interphase layer on the fibres, followed by an infiltration of the coated-fibre mat with SiC powder, which forms a new CNT-SiCf/SiC grade of fusion-relevant composite material. 2. Experimental An aqueous suspension of multi-walled carbon nanotubes (Aquacyl), with average diameter 9.5 nm and average length 1.5m, was provided by Nanocyl S.A. (Sambreville, Belgium). The solids content in the as-received suspension was 1 wt.%. In this study, we used a SiC powder with an average grain size of 0.5m (SiC BF12, H. Starck, Germany). The zeta-potential (ZP) of the SiC powder and of the CNTs was measured in undiluted suspensions using a ZetaProbe analyser (Colloidal Dynamics, USA). A ZetaPals instrument (Brookhaven, USA) was used to analyse the ZP of the crushed SiC fibres in diluted aqueous suspensions. The pH levels were adjusted using 1 M sodium hydroxide (NaOH) and 1 M hydrochloric acid (HCl). The surface charge of the SiC particles for EPD was modified with tetramethyl-ammonium hydroxide (TMAH, Sigma–Aldrich), whereas the commercial CNT-suspension already contained anionic surfactant. The suspensions were homogenised using a strong probe-type ultrasound (Hielscher Ultrasonics, Germany), before the ZP measurement and EPD. The EPD experiments were performed with (1) aqueous suspensions containing 0.0625–0.5 wt.% of carbon nanotubes at a constant dc voltage of 2.8 V for 10 or 20 min, and/or (2) suspensions containing 60 wt.% of SiC powder at a constant dc voltage of 30 V for 5 min. Electrodes made of steel, copper (both of dimensions: 20 mm × 10 mm × 0.2 mm), or graphite (20 mm × 5 mm × 2 mm) were vertically immersed in the suspension in the EPD cell. The distance between electrodes was 2 cm. In the CNT- and SiC-coating processes, the SiC-fibre mat (Tyrano SA, Ube Industries, Ltd., Japan) was placed in front of the electrode, whereas for the SiC-infiltration, the SiC fibre mat was electrically insulated from the electrode with a Teflon® spacer, as described elsewhere.40 Before deposition, the electrodes were thoroughly cleaned in acetone using an ultrasonic treatment, washed with distilled water, and dried with compressed air. The SiC fabric was pre-treated with a sodium dioctyl-sulfosuccinate (SDOSS) solution to improve wettability.41 After the deposition, the samples were dried in air at room temperature, vacuum-infiltrated by an aluminium dihydrogen phosphate (Al(H2PO4)3) aqueous solution (TKI Hrastnik, Slovenia), and sintered at 1300 ◦C in argon (Ar) for 3 h. The microstructures of the green and sintered samples were observed using field emission gun scanning electron microscopes (FEG-SEM, LEO 1525, and Zeiss SUPRA 35VP). The sintered samples were inspected by EDS. 3. Results and discussion The scanning electron micrograph in Fig. 1 shows the morphology of the carbon nanotubes used in this study. The carbon nanotubes have a high aspect ratio and a random and curled structure, which is typically associated with CNTs produced by thermal chemical vapour deposition (CVD).9 3.1. Electrokinetic properties of SiC fibres and characteristics of suspensions Before performing the EPD experiments, we analysed the electrokinetic properties of the SiC powder and the CNTs suspensions (with solids contents of 25 and 1 wt.%, respectively). Fig. 1. SEM micrograph of the used CNTs.
K. Konig et al. /Journal of the European Ceramic Society 30 (2010)1131-1137 30 20 SDOSS-treated SiC fibres 60 Fig 3. SEM image of a CNT-coating made on a graphite electrode by EPD Fig. 2. Zeta-pote (28V.5min) Since an appropriate net surface charge of fibres is very impor- tant for efficient electrophoretic infiltration, 42 we also analysed In addition to the appropriate surface charge of the fibres, the ZP of crushed SiC fibres. The ZP vs. pH relationship is good wettability by the suspension is also an important charac presented in Fig. 2. The as-received CNT-suspension is alkaline teristic for the infiltration of the fibre mats. Consequently, we 11), and due to the presence of anionic surfactant, the CNTs decreased the natural wetting angle of the alkaline aqueous sus- are negatively charged (ZP=-45 mV). The natural pH(pH4)of pension at the SiC fibre bundle(80)by pre-treating the fibres the as-received Sic-powder suspension nearly matches the iso- with the surfactant SDoSS, to approximately 300. As show electric point, resulting in a very unstable suspension. Titration in Fig. 2, the zeta-potential of the SDOSS-treated SiC fibres with hydrochloric acid is not efficient at increasing the zeta- negative over a wide range of alkaline pH values. Hence, as pre potential(described elsewhere); conversely, titration of the sented, the pre-treated SiC fibres, the SiC powder, and the CNts Sic suspension with an appropriate amount of NaOH or TMAH are characterised with a relatively high negative zeta-potential increases the negative ZP up to-50 mV at a pH value of 10 in the alkaline pH region Fig. 4. SEM micrographs of(a) non-coated, (b-d) coated SiC fibres with carbon nanotubes at different magnifications
K. König et al. / Journal of the European Ceramic Society 30 (2010) 1131–1137 1133 Fig. 2. Zeta-potential vs. pH of the materials investigated. Since an appropriate net surface charge of fibres is very important for efficient electrophoretic infiltration,42 we also analysed the ZP of crushed SiC fibres. The ZP vs. pH relationship is presented in Fig. 2. The as-received CNT-suspension is alkaline (pH 11), and due to the presence of anionic surfactant, the CNTs are negatively charged (ZP = −45 mV). The natural pH (pH 4) of the as-received SiC-powder suspension nearly matches the isoelectric point, resulting in a very unstable suspension. Titration with hydrochloric acid is not efficient at increasing the zetapotential (described elsewhere42); conversely, titration of the SiC suspension with an appropriate amount of NaOH or TMAH increases the negative ZP up to −50 mV at a pH value of 10. Fig. 3. SEM image of a CNT-coating made on a graphite electrode by EPD (2.8 V, 5 min). In addition to the appropriate surface charge of the fibres, good wettability by the suspension is also an important characteristic for the infiltration of the fibre mats. Consequently, we decreased the natural wetting angle of the alkaline aqueous suspension at the SiC fibre bundle (∼80◦) by pre-treating the fibres with the surfactant SDOSS, to approximately 30◦. As shown in Fig. 2, the zeta-potential of the SDOSS-treated SiC fibres is negative over a wide range of alkaline pH values. Hence, as presented, the pre-treated SiC fibres, the SiC powder, and the CNTs are characterised with a relatively high negative zeta-potential in the alkaline pH region. Fig. 4. SEM micrographs of (a) non-coated, (b–d) coated SiC fibres with carbon nanotubes at different magnifications
K. Konig et al. Joumal of the European Ceramic Sociery 30(2010)1131-1137 Fig. 5.(a and b) SEM micrograph of the fracture surface of a CNT-coated SiC fibre after SiC deposition by electrophoresis at two different magnifications 3.2. Electrophoretic deposition of carbon nanotubes on per A solid CNT-coating was also formed on a graphite anode metallic electrodes and SiC-fibres at 2.8 V in 5 min From the higher magnification image in Fig. 3. it is evident that the CNTs form a fairly uniform coating and that The EPD experiments were first performed using a CNt- they are evenly distributed on the graphite electrode suspension with a solids content of 0.5 wt %, on steel, copper, In further EPD experiments, the Sic fibre mats were attached or graphite electrodes. Due to the negative ZPof CNTs in suspen- to the front side of the anode and were coated with a layer of sion, the deposits were always formed on the anode. Initially we CNTs at an applied voltage of 2.8 V for 10 min. In these exper performed an EPD experiment on a steel electrode At voltages iments, the SiC fabric was in contact with the anode so that it gher than 2. 8 V, the electrolysis of water apparently limited the acted itself as the electrode. a similar result could be obtained deposition, whereas at lower voltages, no deposit of carbon nan- by using the fibre mat directly as an electrode, as shown else- otubes was observed after 10 min. With a copper electrode, the where on the SiC/mullite system, however, due to the lower bubble formation was suppressed due to the oxidation of cop- conductivity of the SiC fibres compared with a metal electrode, Sic fabric anode teflon cover Fig. 6. Schematic representation of the EPD cell used for infiltration of SiC fibre fabric with SiC particles:(a)SiC-fabric is in contact with the electrode and(b) SiC-fabric is not in contact with the electr
1134 K. König et al. / Journal of the European Ceramic Society 30 (2010) 1131–1137 Fig. 5. (a and b) SEM micrograph of the fracture surface of a CNT-coated SiC fibre after SiC deposition by electrophoresis at two different magnifications. 3.2. Electrophoretic deposition of carbon nanotubes on metallic electrodes and SiC-fibres The EPD experiments were first performed using a CNTsuspension with a solids content of 0.5 wt.%, on steel, copper, or graphite electrodes. Due to the negative ZP of CNTs in suspension, the deposits were always formed on the anode. Initially we performed an EPD experiment on a steel electrode. At voltages higher than 2.8 V, the electrolysis of water apparently limited the deposition, whereas at lower voltages, no deposit of carbon nanotubes was observed after 10 min. With a copper electrode, the bubble formation was suppressed due to the oxidation of copper. A solid CNT-coating was also formed on a graphite anode at 2.8 V in 5 min. From the higher magnification image in Fig. 3, it is evident that the CNTs form a fairly uniform coating and that they are evenly distributed on the graphite electrode. In further EPD experiments, the SiC fibre mats were attached to the front side of the anode and were coated with a layer of CNTs at an applied voltage of 2.8 V for 10 min. In these experiments, the SiC fabric was in contact with the anode so that it acted itself as the electrode. A similar result could be obtained by using the fibre mat directly as an electrode, as shown elsewhere on the SiC/mullite system,43 however, due to the lower conductivity of the SiC fibres compared with a metal electrode, Fig. 6. Schematic representation of the EPD cell used for infiltration of SiC fibre fabric with SiC particles: (a) SiC-fabric is in contact with the electrode and (b) SiC-fabric is not in contact with the electrode
K. Konig et al. /Journal of the European Ceramic Society 30 (2010)1131-1137 Fig. 7.(a) Macroscopic view of a green SiC fibre fabric, which was first coated with CNTs and then infiltrated with a SiC particle suspension by electrophoresis; (b, c, e)and(f) SEM micrographs of the sintered part(1300C in Ar) at different magnifications and different zones in the central part of the composite;(d) EDS spectrum of the matrix between the CNT-coated SiC fibres the applied voltage needs to be higher to achieve the same effect. 3.3. Electrophoretic infiltration of CNT-coated Sic fibre Another reason for using the solid rectangular electrode in this mats with Sic particle suspension study is that it results in a more homogenous electric field. SEM micrographs in Fig. 4a and b show the Sic fibres before and In the next step, we electrophoretically infiltrated the Cnt- after deposition with the CNTs, respectively. From Fig 4b it is coated fibre mat with a suspension of Sic powder. As reported evident that the fibres are completely covered with a thin layer elsewhere, 0-4244 Sic powder can be deposited either from of CNTs. The high-magnification SEM micrographs in Fig. 4c ethanol or water. However, higher ZP values can be obtained and d show a fairly uniform layer of CNTs(estimated thickness in aqueous suspensions, resulting in deposits with higher parti- up to 400 nm)on the SiC fibre. Due to their very small size and cking density. In the present investigation, we deposited well-dispersed state, the carbon nanotubes were able to penetrate SiC particles onto the SiC fibres from a suspension containing into the spaces between the fibres in the 2D fabric. By observing 60 wt %o solids and including TMAH, at 30 V for 5 min In the the cross-section of the sample we confirmed that the SiC-fibres first EPD trials, the CNT-coated fibre mat was placed in front in the central region of the fibre mat were also efficiently coated of the anode so that it was in direct contact with the electrode by the CNTs. This can be clearly seen in Fig. 7e and f, which This caused deposition of the negative SiC particles onto the Sic show CNT-coated fibres inside the central region of the sintered fibres, which acted as the anode. As expected, the SiC particles CNT-SiC/SiC composite. formed a firm deposit on the CNT-coated SiC fibres. Fig 5a and b
K. König et al. / Journal of the European Ceramic Society 30 (2010) 1131–1137 1135 Fig. 7. (a) Macroscopic view of a green SiC fibre fabric, which was first coated with CNTs and then infiltrated with a SiC particle suspension by electrophoresis; (b, c, e) and (f) SEM micrographs of the sintered part (1300 ◦C in Ar) at different magnifications and different zones in the central part of the composite; (d) EDS spectrum of the matrix between the CNT-coated SiC fibres. the applied voltage needs to be higher to achieve the same effect. Another reason for using the solid rectangular electrode in this study is that it results in a more homogenous electric field. SEM micrographs in Fig. 4a and b show the SiC fibres before and after deposition with the CNTs, respectively. From Fig. 4b it is evident that the fibres are completely covered with a thin layer of CNTs. The high-magnification SEM micrographs in Fig. 4c and d show a fairly uniform layer of CNTs (estimated thickness up to 400 nm) on the SiC fibre. Due to their very small size and well-dispersed state, the carbon nanotubes were able to penetrate into the spaces between the fibres in the 2D fabric. By observing the cross-section of the sample we confirmed that the SiC-fibres in the central region of the fibre mat were also efficiently coated by the CNTs. This can be clearly seen in Fig. 7e and f, which show CNT-coated fibres inside the central region of the sintered CNT-SiC/SiC composite. 3.3. Electrophoretic infiltration of CNT-coated SiC fibre mats with SiC particle suspension In the next step, we electrophoretically infiltrated the CNTcoated fibre mat with a suspension of SiC powder. As reported elsewhere,40–42,44 SiC powder can be deposited either from ethanol or water. However, higher ZP values can be obtained in aqueous suspensions, resulting in deposits with higher particle packing density. In the present investigation, we deposited SiC particles onto the SiC fibres from a suspension containing 60 wt.% solids and including TMAH, at 30 V for 5 min. In the first EPD trials, the CNT-coated fibre mat was placed in front of the anode so that it was in direct contact with the electrode. This caused deposition of the negative SiC particles onto the SiC fibres, which acted as the anode. As expected, the SiC particles formed a firm deposit on the CNT-coated SiC fibres. Fig. 5a and b