April 2007 Constant Current EPID of Fiber-Reinforced Ceramic Composites Table I. Suspension and Electrophoretic Infiltration Deposi- For uniform Epid under constant current conditions. the tion Parameters in the Capillaries and in the Suspension deposit thickness is proportional to the time of EPlD, i.e 中a=C+kI(Rdep-Rsus-e甲p)t (18) capillaries suspension Suspension concentration cs(g/mL) 0.040.04 where k is a constant. Raep depends on the type of stabilizing Suspension conductivity os (uS/cm) agent in the suspension 0. If the deposit is uniform and the specific resistance remains Electrophoretic mobility HEp um.s (V/cm) constant, the applied voltage will increase linearly with time. At 0.3 mA, the observed voltage change is almost linear with time Sub-cell coefficient indicating that the deposit is uniform and its thickness increases Current density A(mA/cm linearly with time. This corresponds to the linear weight gain Electric field E(V/cm) with time(Fig. 7). The observed initial voltage drop may be re- Deposition/electrophoretic rate(mg/min) 23 lated to the preliminary suspension infiltration of the pores. For This increase is related to highly resistant entrapped air in the capillaries, but <l outside. Thus, the electric field differs inside deposit. The presence of the latter was confirmed by the max- d outside the capillaries. Under a constant current of 0.3 mA imum Pe board weight gain versus the deposition current. Max- the electric field in the capillaries is 44 V/cm. The suspension imum deposition(53 vol%)occurs at a current of 0.3 mA with resistance is calculated via the applied current and recorded an increase in linear voltage. When the current >0.6 mA, the voltage; hence the outside sub-cell coefficient can be determined deposited alumina only occupies 45% of the pore volume, ind between the same electrodes without intervening porous cating that air bubbles are trapped in the deposit. Clearly, the board in the cell, i.e., f=0.52. The electric field across the sus- deposition rate must be controlled to optimize the green density pension is 10 V/cm, and hence the EPID suspension paramete This is why constant current must be used for the EPID process. are as summarized in Table i According Eq.(ID). the deposition rate in V. Conclusions Conductive fibers are not suitable for fabrication of fiber-re- particle electrophoretic migration rate in an open suspension is inforced CMCs by electrophoretic deposition due to the electric 12 mg/min. The latter value is close to the measured deposition field shielding effect. The field penetration depth is only double he fiber filament opening distance and a central cavity inevit particle electrophoretic migration rate in the open suspension ably results when a large bundle of conductive fibers is used. (outside of the capillaries), i.e., all particles electrophoresing te A modified EPd cell can be used to fabricate non-conductive- the board outer surface enter the capillaries and infiltrate faster fiber-reinforced CMCs by EPlD. Particles are infiltrated into a there than in the suspension. This finding is not surprising as fiber preform to deposit on an electrode behind the preform and particles are""streamed"by the repulsion from the pore surfaces then backfill the preform. Thus, the front clogging encountered once they enter the porous board. This phenomenon also ex- in direct EPD is avoided. The repulsive interaction between the plains why no deleterious surface deposition(clogging) occurs fibers and the particles promotes particle streaming through until the porosity is fully filled. If the sub-cell coefficient outside the fiber preform, resulting in dense, uniform green composit ual the infiltration rate by EPID Constant-current EPID is modeled as capillary infiltration In an EPD cell with the cathode as the deposition electrode electrophoresis. Owing to the zero flow across a closed capillary the applied potentia s consumed in three steps:(a)a cross section, no effect of electro-osmotic flow is expected. The tential drop at each electrode(cathode and anode), (b)an ohmic infiltration deposition yield is proportional to the electric field tential drop over the suspension, and (c)a potential drop over inside the capillary; however, the total EPID yield is found to be the deposit on the cathode. In an EPID cell, on the other hand controlled by particle electrophoresis outside the capillaries, i. e, the potential drop over the suspension can be divided into two (I the potential drop across the open suspension outside the in the open suspension, due to the much lower electric field capillaries(between the anode and board outer surface), and (2) entrapment in the deposit decrease green density due to air the suspension in the capillaries, i.e aa= Voel+ IRdepd+IRsus-cap(d-d)+IRsusd2 References +Voe (16) K.K. Chawla, Ceramic Matrix Composites. Chapman, New York, 1993 where Voe and voe are the potential drop at the anode and cathode, respectively, I is the current flowing through the cell, Oxide Matrix Composite Reinforced with Continuous Oxide Fibers "/Am. Ce defined as the resistivity divided by the surface area), Rsus the J.J. Haslam, K.E. Berroth, and F. F. Lange. "Processing and Properties of an specific resistance of the suspension over the board outer surface ance of the <a Oxide Composite with a Porous Matrix, "J. Eu. Ceram Soc. 20 15]607-18 in the capillaries, d the deposit thickness, di the board thickness, Fibre Mats with Sic Powder suspensions"s in , th conference of the efCrarean spe erammic Society, September 9-13, 2001. Edited by A.R. Boccaccini, Trans Tech Neglecting the observed small change of suspension conduct- 1. MacLaren. M. H. Lewis and C. B. Ponton,"Electro sition Infiltration of 2-D Woven Sic Fibl ivity(<+3%), Rsus and Rsus-cap are constant, and hence Mullite Comp J. Eur. Ceran. Soc., 17[13] 1545-50(1997) utler and C. B. Ponton. ""Novel Tech- da=C+I( Rdep-Rsus-cap)d (17) T J Illston, C. B. Ponton. P where C is a constant. If Rdep is constant during deposition, the eposition of Silica/Alumina Colloids for the Manufacture of Cme's":pp applied voltage will increase linearly with deposit growth into Ceramic Engineering and Science Proceedings, Proceedings of the 18th Ar Advanced Ceramic Materials-B. Part 2(of the capillaries. 9-14.1993,1994.15(5)capillaries, but o1 outside. Thus, the electric field differs inside and outside the capillaries. Under a constant current of 0.3 mA, the electric field in the capillaries is 44 V/cm. The suspension resistance is calculated via the applied current and recorded voltage; hence the outside sub-cell coefficient can be determined between the same electrodes without an intervening porous board in the cell, i.e., f 5 0.52. The electric field across the sus￾pension is 10 V/cm, and hence the EPID suspension parameters are as summarized in Table I. According Eq. (11), the deposition rate in the capillaries should be 23 mg/min, i.e., much higher than the measured value (10 mg/min). However, according to Hamaker’s equation, the particle electrophoretic migration rate in an open suspension is 12 mg/min. The latter value is close to the measured deposition rate, which indicates that the deposition is controlled by the particle electrophoretic migration rate in the open suspension (outside of the capillaries), i.e., all particles electrophoresing to the board outer surface enter the capillaries and infiltrate faster there than in the suspension. This finding is not surprising as particles are ‘‘streamed’’ by the repulsion from the pore surfaces once they enter the porous board. This phenomenon also ex￾plains why no deleterious surface deposition (clogging) occurs until the porosity is fully filled. If the sub-cell coefficient outside the capillaries is 1, then the electrophoretic migration rate must equal the infiltration rate. In an EPD cell with the cathode as the deposition electrode, the applied potential, fa, is consumed in three steps: (a) a po￾tential drop at each electrode (cathode and anode), (b) an ohmic potential drop over the suspension, and (c) a potential drop over the deposit on the cathode. In an EPID cell, on the other hand, the potential drop over the suspension can be divided into two: (1) the potential drop across the open suspension outside the capillaries (between the anode and board outer surface), and (2) the suspension in the capillaries, i.e. fa ¼ Hfc el þ IRdepd þ IRsuscapðd1 dÞ þ IRsusd2 þ Hfa el (16) where Vfel a and Vfel c are the potential drop at the anode and cathode, respectively, I is the current flowing through the cell, Rdep the specific resistance of the deposit (specific resistance is defined as the resistivity divided by the surface area), Rsus the specific resistance of the suspension over the board outer surface and the anode, Rsus–cap the specific resistance of the suspension in the capillaries, d the deposit thickness, d1 the board thickness, and d2 the distance between the board outer surface and the anode. Neglecting the observed small change of suspension conduct￾ivity (o73%), Rsus and Rsus–cap are constant, and hence fa ¼ C þ IðRdep RsuscapÞd (17) where C is a constant. If Rdep is constant during deposition, the applied voltage will increase linearly with deposit growth into the capillaries. For uniform EPID under constant current conditions, the deposit thickness is proportional to the time of EPID, i.e. fa ¼ C þ kIðRdep RsuscapÞt (18) where k is a constant. Rdep depends on the type of stabilizing agent in the suspension.34 If the deposit is uniform and the specific resistance remains constant, the applied voltage will increase linearly with time. At 0.3 mA, the observed voltage change is almost linear with time, indicating that the deposit is uniform and its thickness increases linearly with time. This corresponds to the linear weight gain with time (Fig. 7). The observed initial voltage drop may be re￾lated to the preliminary suspension infiltration of the pores. For a current 0.6 mA, the voltage jumps to a maximum (1000 V). This increase is related to highly resistant entrapped air in the deposit. The presence of the latter was confirmed by the max￾imum PE board weight gain versus the deposition current. Max￾imum deposition (53 vol%) occurs at a current of 0.3 mA with an increase in linear voltage. When the current 0.6 mA, the deposited alumina only occupies 45% of the pore volume, indi￾cating that air bubbles are trapped in the deposit. Clearly, the deposition rate must be controlled to optimize the green density. This is why constant current must be used for the EPID process. V. Conclusions Conductive fibers are not suitable for fabrication of fiber-re￾inforced CMCs by electrophoretic deposition due to the electric field shielding effect. The field penetration depth is only double the fiber filament opening distance and a central cavity inevit￾ably results when a large bundle of conductive fibers is used. A modified EPD cell can be used to fabricate non-conductive- fiber-reinforced CMCs by EPID. Particles are infiltrated into a fiber preform to deposit on an electrode behind the preform and then backfill the preform. Thus, the front clogging encountered in direct EPD is avoided. The repulsive interaction between the fibers and the particles promotes particle streaming through the fiber preform, resulting in dense, uniform green composites by EPID. Constant-current EPID is modeled as capillary infiltration electrophoresis. Owing to the zero flow across a closed capillary cross section, no effect of electro-osmotic flow is expected. The infiltration deposition yield is proportional to the electric field inside the capillary; however, the total EPID yield is found to be controlled by particle electrophoresis outside the capillaries, i.e., in the open suspension, due to the much lower electric field therein. High deposition rates decrease green density due to air entrapment in the deposit. References 1 K. K. Chawla, Ceramic Matrix Composites. Chapman, New York, 1993. 2 F. F. Lange and K. T. Miller, ‘‘Pressure Filtraion: Consolidation Kinetics and Mechanics,’’ Am. Ceram. Soc. Bull., 66, 1498–504 (1987). 3 M. G. Holmquist and F. F. Lange, ‘‘Processing and Properties of a Porous Oxide Matrix Composite Reinforced with Continuous Oxide Fibers,’’ J. Am. Cer￾am. Soc., 86 [10] 1733–40 (2003). 4 J. J. Haslam, K. E. Berroth, and F. F. Lange, ‘‘Processing and Properties of an All-Oxide Composite with a Porous Matrix,’’ J. Eu. Ceram. Soc., 20 [5] 607–18 (2000). 5 K. Moritz and E. Muller, ‘‘Electrophoretic Infiltration of Woven Carbon Fibre Mats with Sic Powder Suspensions’’; in 7th Conference of the European Ceramic Society, September 9—13, 2001. Edited by A. R. Boccaccini, Trans Tech Publications Ltd., Brugge, 2001. 6 A. R. Boccaccini, I. MacLaren, M. H. Lewis, and C. B. Ponton, ‘‘Electropho￾retic Deposition Infiltration of 2-D Woven Sic Fibre Mats with Mixed Sols of Mullite Composition,’’ J. Eur. Ceram. Soc., 17 [13] 1545–50 (1997). 7 P. A. Trusty, A. R. Boccaccini, E. G. Butler, and C. B. Ponton, ‘‘Novel Tech￾niques for Manufacturing Woven Fiber Reinforced Ceramic Matrix Composites. I. Preform Fabrication,’’ Mater. Manuf. Process., 10 [6] 1215–26 (1995). 8 T. J. Illston, C. B. Ponton, P. M. Marquis, and E. G. Butler, ‘‘Electrophoretic Deposition of Silica/Alumina Colloids for the Manufacture of Cmc’s’’; pp. 1052–9 Ceramic Engineering and Science Proceedings, Proceedings of the 18th Annual Con￾ference on Composites and Advanced Ceramic Materials—B. Part 2 (of 2), January 9–14, 1993, 1994, 15(5). Table I. Suspension and Electrophoretic Infiltration Deposi￾tion Parameters in the Capillaries and in the Suspension In capillaries In suspension Suspension concentration cs (g/mL) 0.04 0.04 Suspension conductivity ss (mS/cm) 2.5 2.5 Electrophoretic mobility mEP mm s (V/cm) 0.8 0.8 Sub-cell coefficient f 1.0 0.52 Current density L (mA/cm2 ) 0.109 0.048 Electric field E (V/cm) 44 10 Deposition/electrophoretic rate (mg/min) 23 12 April 2007 Constant Current EPID of Fiber-Reinforced Ceramic Composites 1069