《复合材料 Composites》课程教学资源（学习资料）第五章 陶瓷基复合材料_fiber153

Am.Ccrm.Soc,8814550(205 DO:10.111)1551-20162004007 urna Improved Filtration Performance of Continuous Alumina-fiber reinforced Mullite Composites for Hot-Gas Cleaning Satoshi Kitaoka and Naoki kawashima Japan Fine Ceramics Center. Nagoya 456-8587 Japan Yoshinobu Komatsubara Mitsui Mining Material Co.. Omuta 836-0062 Japan Akira ya hamaguchi Nagoya Institute of Technology, Nagoya 466-8555 Japan sao suzukI Shizuoka University, Hamamatsu 432-8561 Japan Ihe effect of filtration layer morphology on filtration perform- nificantly aflects the nature of the ash at high temperatures) ance at 673 K was investigated for continuous alumina-fiber- As the SiO KO phase diagram indicates, mullite is einforced mullite composite filters that capture fly ash on their one of the equilibrium phases that can form from ashes of outer surfaces. Two types of filtration layers were prepared on simple composition (e. g, Sio: AlO K,=71: 28: I wt% for the outer surfaces: a mono-laver consisting of mullite agglom Ash No. 10. distributed by the Association of Powder Proc erated particulates and a bi-layer, with mullite whiskers formed ess Industry and Engineering. Japan: SiO2: Al O: K 0= by a vapor-phase reaction and strongly adhered to the particu- 63.8: 358: 0. 4 wt% for the fly ash that was collected from a hop- lates exposed on the outer surfaces, For filters with mono-lil- per at the PFBC demonstration plant). The phase diagram tration layers, the filtration efficiency was improved slightly by indicates that Al-O, reacts with fly ash and therefore direct creasing the filtration laver thickness. However, the maximum contact should be avoided differential pressure increased during the filtration tests. Adhe- Because of the poor thermal conductivity of unreinforced ce- sion of whiskers to the outer surface decreased the maximum ramic filters. however, large temperature gradients form between differential pressure to about one-third less than that with th the inner and outer surfaces of the filters during excessive gas mono-filtration layer. This low differential pressure remained cleaning, or as the result of sudden ignition of char accumulated constant throughout the duration of the test, with corresponding on the filtration surfaces or from a combination of the two rea- increases in filtration efficiency sons, Thermal stresses induced by the temperature gradient lead to catastrophic failure of the filters, resulting in low reliability To be used successfully. these filters require precise control of 1. Introduction operating conditions and constant observation H OT-GAs particulate filters are key components for advanced Continuous-fiber- reinforced filters are expected to show ex- coal-based power generation systems such as pressurized cellent damage tolerance, durability, and reliability. Oxide ma fluidized-bed combustion(PFBC)and integrated coal gasifica terials are inherently stable in oxidizing environments such as tion combined cycle (IGCC). Filters should allow easy removal the PFBC system and have more corrosion resistance to alkali of dust cakes accumulated on the filter surfaces and excellent sinterability of mullite, if this material is used as the matrix of reversibility of the differential pressure during pulse cleaning. the fiber-reinforced filter, the composite must be sintered at high For this reason, adhesion of the dust cakes to the filters should be suppressed as far as possible. Adhesion is generally related to temperatures to increase the bonding strength among the mullite he formation of alkali-containing liquids in the cake and chem rains. Unfortunately, mullite fibers stable at high temperatures ical reactions between the filter element and the fly ash. Chem. have not yet been produced, so that only high-purity g-ALO ical reactions have a particularly strong eflect on dust adhesion fiber (which has the highest thermal stability. 1673 K, of an and make pulse cleaning diflicult. To simplify the study of the oxide fiber) is available. However, if the AlO, fibers are exposed interaction of fly ash with filters, we have considered a repre on the filtration surfaces. the fibers react with the ash cakes leading to an increased diferential pressure of the filters. There sentative ash containing SiO2 and Al- O,(the principal constit- fore. the filtration surfaces of the filters should be fully covered tents) and KO(one of the most reactive constituents, which with mullite We have developed cylindrical composite filters that capture Ralf Riedel contributing editer fly ash on their outer surfaces. The filters consist of a mullite filtration layer and a substrate layer of continuous AlO fibe einforced mullite matrix composite, The filters were fabricated by dipping a two-dimensional woven sheet of the tibers into a mentally Friendly Coal Combustion Technology), promoted by the New Energy and In- Ikoxide-derived mullite precursor solution. The precursor coat Cmm上mCE ed onto the large mullite particles formed very fine mullite ict from METL. as a part of regional Author to whom correspondence should he addresed. e-mail: kitaoka a jfc or ip

journal ./. Atii. Cerum. .Soe., K8 [I] 45 5t) (2t)O5) DOl- 10.1111 i.l55l-29l6.2l)(l4.l)O(i:7.i( Improved Filtration Performance of Continuous Alumina-fiber￾reinforced Mullite Composites for Hot-Gas Cleaning Satoshi Kitaoka^ and Naoki Kawashima .litpiin Kinc Ceramics Center. Nagoya 456-8587. Japan Yoshinobu Komatsubara Mitsui Mining Material Co.. Oniula S36-0062. Japan Akira Yamaguchi Nagoya Instittitc d'Technology. Nagoya 466-8555. Japan Hisao Suzuki ShizLioka University., Hamamalsu 432-8561, Japan The effect (>f filtration layer morphology on filtration perform￾ance at 673 K was hnesti^ated for continuous alumina-liber￾reinforced mulIJte composite filters that capture fly ash on their outer surfaces. I wo types of Nitration lavers were prepared on the outer surfaces: a mono-layer consisting of mullite af|;glom￾erated particulates and a bi-layer. with mullite whiskers formed hy a vapor-phase reaetion and strongh adhered to the particu￾lates exposed on the outer surfaces. F(tr filters with mono-fil￾tration layers, the filtration efficiency was improved slightly hy iiicreasin;; the filtration layer thickness. However, the maximum differential pressure increased durinj; the filtration tests. Adhe￾sion of whiskers to the outer surface decreased the maximum differential pressure to ahout one-third less than that with the mono-filtration layer. This low differential pressure remained constant throughout the duration ofthe test, with corresponding increases in filtration efficiency. I. Introduction H OT-Ci.AS particuialc filters are key cotnponenls for advanced eoal-based power generation systetns sttch as pressurized tlLtidi7cd-bed combiislioii (PFBC) and integrated eoal gLtsifica￾lion eombined eyclc |IGCC). Filters should allow easy removal of dusl cakes accumtiluted on ihe filter surfaces and excellent reversibility of ihe dilTerential pressure during pulse cleitning. For this reason, adhesion of the dust cakes to the filters should be suppressed as far as possible. Adhesion is generally related to the formation of alkali-containing liquids in ihe cake and chem￾ical reactions between the filter eletnent and the f1y ash.' Chetn￾ieal reaclions have a particularly strong effect on dust adhesion and make pulse eleanini: difficult. To simplify the study of the interaction of fly ash with filters, we have considered a repre￾sentative ash containing SiO2 and AliOi (the principal constit￾uents) and KiO (one of the most reactive constituents, which .liiirihutiiiji ciljior Miinustripl No. 10492. Rca.-ivcd August 2». 2()(B: iipprovcd July 20. :(MI4. This rewarcli wiis ^uppnned in pan by ihe Center for Co;il L'lili/atiiin. Jypan (t'CUJ). under the Neil-Generalion COLII LllilLZiilion Tei:hn<>ihould bu addressed. e-iULiil: kilaokaii jti:i:.i>r.|p significantly alTects the nautre ofthe ash at high temperatures). As the SiO: AUOi K^O phase diagram indicates." mullite is one of the equilibriittii phases that can tbrtn from ashes of simple cotnposition (e.g., SiO2;Al2O,;K2O - 71:28:1 wt% for Ash No. 10. distributed by the Association of Powder Proc￾ess Industry and Engineering. Japan; SiO2:Al2Oi:K2O- 63.8:35.8:0.4 wt% for the fly ash lhat was collected frotn a hop￾per at the PFBC dctnonstratlon plant'). The phase diagram indicates that A!: ^ reaets with fly ash and. therefore, direct contact shn 45

Journal of the merican Ceramic Societ- kitaoka et al Vol. 88. No. I particulates during heating. These particulates accelerated the consisting of 15 wt mullite particles and the precursor solution neck growth between the large particles, even at low tempera was sprayed onto the outer surface of sample A and the filter tures. Therefore, the precursor allowed the sintering of larg was sintered again at 1673 K. in air. for I h. This filter is referred mullite particles without Al-O, fiber damage, resulting inin- to as sample B creased flexural strength for the composites while inhibiting de- Another hilter was prepared by adhering mullite whiskers lamination of the woven fiber sheets onto the outer surface of sample a by a vapor-phase reaction Although the morphology of the filtration layers strongly af- This whisker- synthesizing method has been referred to in previ fects filtration behavior. there is limited information on the op- ous studies. First. TEOS and aluminum nitrate enneahydrate timal morphology for improving filtration performance. In Al(NO,)3.9H-O were dissolved in dehydrated ethanol. AIF articular. if a layer consisting of particles with large aspect ra tios. such as whiskers is formed on the filter surfaces, it can be expected to have a small open pore size and large open porosity Si molar ratio of the mixture was 3. Ana2 q fully mixed in.The amount of AIF, was 2 wt% of the Al(NO:).9H-O, and the Al resulting in a decreased differential pressure and more efficient added to the mixture and precipitates were produced. The pre dust removal. In this case, the particles need to be strongly fixed cipitates were dried at 573 K for 24 h. A slurry consisting of the to the filter surfaces to prevent them from falling out as a result dried powder (15 wt%), polyvinyl butyral (I wt %a), and dehy o For the present study. the effect of filtration layer morphol- the outer surface of the sample, and the coated sample then shock during gas cleaning drated ethanol was prepared. The slurry was applied liberally mullite composites was investigated. Two types of filtration lay placed in an airtight container and fired at 1473 K for 2 h. After ers were produced on the fiber-reinforced composites: a mono- firing, the surplus mullite whiskers that had formed on layer consisting of mullite particulates and a bi-layer with mul the outer surface were removed by ultrasonic washing in water. lite whiskers formed by vapor-phase reaction and strongly ad- The edges of the remaining whiskers were strongly adhered to hered to mullite particulates exposed on the outer surfaces. he mullite particles exposed on the outer surface, which was fully covered by the whiskers. This filter is referred to as sample C. When the adhesion of the whiskers to the top surface was lI Experimental Procedure achieved at higher temperatures and for longer times than those for sample C, the whiskers on the surface grew, and neighboring 71) Materials whiskers combined, resulting in the formation of plate-like The mullite precursor coated onto the large mullite particles ac- particles. celerated sintering of the large particles, resulting in a lower sin- tering temperature without degradation of the Al-O, fibers. The precursor solution was prepared by a polymeric sol gel (2) Measurement and Analysis process from aluminum and silicon alkoxides as follows: "First. The filtration surfaces and cross sections of the filters were ob- tetraethoxysilane (TEOS) was added to dehydrated ethanol (TEOS concentration of 10 mol m,). Distilled water and hydro- mullite whiskers on the surface were examined by transmission chloric acid were then dropped into the solution in molar ratio of 2 and 0. 1, respectively. The TEOS was partially hydroly resolution of -10 nm. was performed b ray spectroscopy(EDS)and electron diffraction(ED). The po. called the silicon precursor. Next. aluminum isopropoxide rosity and distribution of the pores in the composite were meas. (Al(OCH(CH,)):)at a concentration of 10 mol m was dis- solved in 2-methyl-1-propanol at room temperature and ref The permeability. k. of the filters was determined from the luxed at 368 K for 24 h. The solution obtained is referred to relationship between the flow rate and the applied N? pressure, hereafter as the aluminum precursor. A mullite precursor solu according to darey's law tion was synthesized by mixing the silicon precursor with the aluminum precursor(Al Si molar ratio=3)and stirring at room emperature for 5 h. The chemical copolymerization of partially hydrolyzed species resulted in high compositional homogeneity of the precursor solution where AP is the differential pressure across a filter of thickness A slurry was prepared from the mullite precursor solution L. The superficial velocity, u, of the gas with viscosity. H is as- and mullite particles(KM1Ol, mean diameter 0. 75 um: KCM sumed to be the same throughout the filter Co.. Nagoya. Japan). The concentration of the mullite particles The filtration performance of the samples was evaluated us- was 25 wto, A two-dimensional woven sheet of -Al,O, fibers ng an apparatus(Hosokawa Micron Corp. Osaka, Japan) in ALMAX. Al-O purity of 99.5 wt%, mean diameter 10 um accordance with the German VDI DIN standard 3926. The Mitsui Mining Material Co., Tokyo, Japan) was vacuum infil cylindrical composites were cut into 100-mm lengths and both trated with the slurry, The infiltrated sheet was wound around bases of the cylindrical samples were sealed by the holders in the an Al O, cylindrical mold and then dried during rotation at 100 apparatus. Dust-containing air was sucked inward through the rpm to prevent gravity segregation of the slurry in the sheet. The samples, and the dust cakes that accumulated on the outer sur- centrifugal force induced by the rotation moved part of the in- faces were removed by cyclically blowing air outward. The dif filtrated slurry to the outer surface of the sheet, resulting in the ferential pressure across the samples was monitored during the buildup of a mullite layer on the outer surface. Next, the sheet filtration tests. A sample of fly ash, which was collected from a with the mold was sintered at 1673 K in air for 2 h. The cylin hopper at the PFBC demonstration plant, was used as the dust drical composite obtained (outer diameter 55 mm, thickness for the filtration tests. Figure I shows the particle size distribu 2.5 mm width 1 50 mm) is hereafter referred to as sample A. The tion of the dust, which has two peaks. The average diameter of thickness of the mullite layer on the outer surface of sample A the dust particles was 7.4 Hm. Figure 2 shows an SEM micro- tion ors- 10 um, which acted as a filtration layer. The fiber frac- graph of the dust. Spherical particles. which were probably In case of dust filtration at the exterior wall of the cylindrical ticles were observed. The filtration performance using this dust composite, a decrease in pore size from the interior of the cyl was evaluated at 673 K, the upper-limit temperature of the ap- inder to its exterior, as well as a uniform size distribution of the paratus. for a dust concentration of 5 g m, a filtration velocity xterior pores. was necessary for easy removal of the cake dur- of 0.097 m s a pulse blowing pressure of 0.2 MPa. a blowin ng back blowing. To achieve this. two types of filters were pro- interval of 150 s. and a total number of 204 blowing cycles duced using sample A as a substrate layer. For one filter. a slurry These tests were conducted as a preliminary assessment of the

46 .Joiiival of the .'\iiicriciiii Ccnmuc Socich Kiuioka cl al. Vol. 88. No. pLiriiculates during healing. These particiilales iiL-ccierated lhe neck growth between the large partieies. even at low Icmpcra￾tLires, Therefore, the prectirsor iillowed the sintering of kirgc niLillite particles without AI^O? liber damage, resulting in in￾creased tkwural strength for the composites while inhibiting de￾laniination of lhe woven fiber sheets. Although the morphology ofthe filtration layers strongly af￾fects lillration behavior, there is limited information on the op￾timal morphology foi" improving (iltration performance. In particular, if a layer consisting of particles with large aspect ra￾tios, sueh as whiskers, is formed on the filter surfaces, it can be expeeted to have a small open pore size and large open porosity, resulting in a decreased differential pressure and more effieient dust removal. In this case, the partieies need to be strongly fixed to the lilter surfaces to prevent them from falling out as a resuli of shock during gas cleaning. For the present study, the effect of filtration layer morphol￾ogy on the performanee of continuous AUOvfiber-reinforced mullite composites was investigated, Tuo types of filtration lay￾ers were produeed on the fiber-reinforced composites: a mono￾layer consisting of mullite partieulates and a bi-layer wilh mul￾lilc whiskers formed by vapor-phase reaction and strongly ad￾hered to mullite particulates exposed on the outer surfaces. II. Experimental Procedure (1} Materials The mullite precursor coated onto the large mullile particles ac￾celerated sintering ofthe large particles, resulting in a lower sin￾tering temperature without degradation of the AUOi fibers.' The precursor solution was prepared by a polymeric sol gel process from aluminum and silicon alkoxides as follows:' First, tetraethoxysilane (TFOS) was added to deliydrated ethanoi (TEOS concentration of 10 mo! m'). Distilled water and hydro￾chloric acid were then dropped into lhe solution in molar ratios of 2 and 0,1. respeetively. The TEOS was partially hydroly￾zed by reflux at 343 K for 50 h. producing what is hereafter called the silicon precursor. Next, aluminum isopropoxide (AHOCHiCH,)^).!) at a concentration of 10 moi m was dis￾solved in 2-mcthyl"l-propanol at room temperature and ref￾luxed at 368 K for 24 h. The solution obtained is referred to hereafter as the aluminum precursor, A mullite precursor soiu￾tion was synthesized by mixing the silicon precursor wiih the aluminum precursor (Al.Si molar ralio = 3) and stirring at room temperature for 5 h. The ehemical copolymerization of partially hydrolyzed species resulted in high compositional homogeneity ofthe precursor solution. A slurry was prepared from the mullite precursor solution and mullite partieies (KMIOL mean diameter 0.75 \mx KC'M Co., Nagoya. Japan), The concentration O'L the mullite particles was 25 wt%. A two-dimensional woven sheel ol' •y.-M^Oy fibers (ALMAX. AUOi purity of 99,5 wt^ . mean diameter 10 |.im; Mitsui Mining Material Co.. Tokyo. Japan) was vacuum infil￾trated with the slurry. The infiltrated sheet was wound around an AKOi cylindrical mold and then dried during rotation at 100 rpm to prevent gravity segregation ofthe slurry in the sheel. The centrifugal force induced by the rotation moved part ofthe in￾filtrated slurry to lhe outer surface of lhe sheet, resulting in the buildup ofa muilite layer on the outer surface. Next, the sheet with the mold was sintered at 1673 K in air for 2 h. The cylin￾drical eomposite obtained (outer diameter 55 mm, thickness 2.5 mm. width 150 mm) is hereafter referred to as sample A. The thickness of lhe mullite layer on the outer surface of sample A was -^ 5-10 jam. which acted as a filtration layer. The fiber frac￾tion ofthe supporling layer was ^30 vol%- In case of dust filtration at Ihe exterior wall ofthe cylindrical composite, a decrease in pore size from the interior ofthe cyl￾inder to its exterior, as well as a uniform size distribution of the exterior pores, was necessary \'or easy removal ofthe cake dur￾ing back blowing. To achieve this, two types of filters were pro￾duced using sample A as a substrate layer. For one filter, a sltirry consistinji of 15 wt",> mullite particles and the precursor solution was sprayed onio the outer surface of sample A- and the filter was sintered again at 1673 K. in air. for 1 h. This filler is referred to as sample B. Another fitter was prepared by adhering mullite whiskers onto the outer surface of sample A by a vapor-phase reaction. This whisker-synthesi/ing [nethod has been referred to in previ￾ous studies,'^'' First. TEOS and aluminum nitiate enneahydrate AKNO.Or'^^20 were dissolved in dehydrated ethanoi, AIF, powders then uere added to the solution and fully mixed in. The amount of AIT', was 2 wt"o ofthe AI(NO,0i • yH20. and the Ai/ Si molar ratio of the mixture was 3. An ammonia solution was added to the mixture and precipitates were produced. The pre￾cipitates were dried at 573 K for 24 h. A slurry consisting ofthe dried powder (15 wt"'!.). polyvinyl bulyral (I wa"o). and dehy￾drated elhanol was prepared. The slurr> was applied liberally lo the outer surface o\~ the sample, and the coaled sample then was dried al room temperature. The cylindrical composite was placed in an airtight container and fired at 1473 K for 2 h. After firing, the surplus mullite whiskers that had formed on the outer surface were removed by ultrasonic washing in water. The edges ofthe remaining whiskers were strongly adhered to the mutlite partieies exposed on the outer surface, which was fully covered by the whiskers. This filter is referred to as sample C. When the adhesion of the whiskers to the top surface was achieved at higher temperatures and for longer times than those for sample C. the whiskers on the surface grew, and neighboring whiskers combined, resulting in lhe formation o\' plate-like panicles, (2) Measurement and Analysis The filtration surfaces and cross sections ofthe filters were ob￾served by seanning electron microscopy (SEM). For sample C. mullile whiskers on the surface were examined by transmission electron microscopy (TEM), Chemical analysis, wilh a spatial resolulion of ^ 10 nm. was performed by energy-dispersive X￾ray spectroseopy (EDS) and electron diffraction (ED). The po￾rosity and distribution ofthe pores in the composite were meas￾ured by mercury intrusion porosimetry. The permeability, k. of the filters was determined from the relationship between the flow rate and the applied N^ pressure, according to Darcy's law. (1) where AP is the differential pressure across a filler of thickness L. The superlieial velocity. », of the gas with viscosity, (a. is as￾sumed to be the same throughout the filter. The filtration performance of the samples was evaluated us￾ing an apparatus (Hosokawa Micron Corp.. Osaka, Japan) in accordance with the German VDI DIN standard 3926.'" The cylindrical composites were eut into 100-mm lengths, and both bases of the cylindrical samples were sealed by the holders in the apparatus. Dust-containing air was sucked inward through the ,samples, and the dust cakes that accumulated on the outer sur￾faces were removed by cyclically blowing air outward. The dif￾ferenliiil pressure aeross the samples was monitored during the filtration tests, A sample of fiy ash. which was collected from a hopper at the PFBC demonstration plant, was used as the dust for the lillration tests. Figure I shows lhe particle size distribu￾tion ofthe dust, whieh has two peaks. The average diameter of the dust particles was 7,4 [.tm. Figure 2 shows an SEM miero￾graph of the dust- Spherical particles, which were probably formed by solidification of melted dusl. and agglomerated par￾tieies were observed. The filtration performance using this dust was evaluated at 673 K, the upper-limit temperature of the ap￾paratus, for a dust concentration of 5 g, m . a filtration veloeity of 0,097 m/s, a pulse blowing pressure of 0,2 MPa. a blowing interval of 150 s, and a total number of 204 blowing cycles. These tests were eonducled as a preliminary assessment of the

January 2005 Filtration Performance of Mullite Composites for Hot-Gas Cleaning lated on the Al-O fiber sheets. The thickness of the particulate layer was -5-10 um This layer consisted of agglomerated mul lite particles with many pores (less than a few micrometers) =74pm among the agglomerates and acted as the filtration layer. Al nough regions among the fiber bundles were filled with the mullite particulates. there were many pores in the bundles. A shown in Fig 3. large gaps were observed among the bundles and agglomerated particulates. For sample B. the mullite par- ticulate layer, which was 20-30-um thick, acted as the filtration layer. The filtration surface seemed less uneven. and the pores In sample C the whiskers were arranged on the particulate layer perpendicular to the outer surface. The thickness of the whisker layer was the same as the length (-4 um)of the whiskers, The aspect ratio (length width)of the whiskers was -15. The filtra tion surface of sample C was fully covered with whiskers. There were no large depressions, such as those of sample A, and the pores observed on the surface of sample C seemed to be smaller Diameter. um than those of the other samples. Incidentally, the microstructure Fig. 1. Particle size distribution of the fly ash of the substrate layer of sample C was the same as those of the other samples. The ED pattern of a single whisker corresponded to that of a single crystal of mullite. TEM-EDS analysIs performance of the filters, which in practice would be used at whiskers indicated an average Si/ Al molar ratio of 0.34. similar temperatures >1100K. The dust passing through the filter to the stoichiometric composition (0.33)of mullite. No deter- samples was captured with filter paper. The average dust con- orated layers were observed at the interfaces between the mullite centration in the clean gas after filtration was determined by di whiskers and the mullite particulate layers that had accumulated iding the mass of the captured dust by the total amount(5. 1 on the Al O, fiber sheets m)of clean gas that passed through the filter paper. The filtra- We propose a mechanism for the formation of whiskers per- tion efficiency, n. was calculated as follows: pendicular to the outer surface based on thermodynamic argu ments, Figure 4 shows the equilibrium partial pressures of the n(%)=(1-C/C)×100 gaseous species and stable regions of the condensed phases un- der a constant F(e partial pressure of 1.3 x 10 Pa, calculated om the decomposition reaction of AIFys(AlF3-ALn+ where C, and C are the dust concentrations before and after fil- 3Fig) as a function of O; partial pressures at 1473 K. for the tration. respectively AlO-F-Si system. As shown in Fig 4. when AlF, particu- lates exist on the outer surface of the cylindrical composites, the II1. Results and Discussion AlF se) reacts with oxygen to mainly produce AlOFye, as follows Figure 3 shows SEM micrographs of typical cross-sections and filtration layer surfaces of the samples, For sample A. the outer AlF3s+ 1/20 g)-:g+I surface was covered with mullite particulates that had accumu- Consequently, an oxygen potential gradient forms on the outer urface. Under O, partial pressures of 3.2 x 10 Pa, the partial pressure of the alF decreases, and the AlF condenses into AlO3. Under O2 partial pressures of 9.6x 10-19 Pa Sirap/ Condenses into SiO follows SiFe+ o SIO s+4F The Gibbs free energy for the formation of mullite from AlO3 nd SiO, at |473 K is-207 kJ/mol. Ther AlOs and SiO, coexist as stable phases at higher O, partial ressures, as shown in Fig. 4. mullite may be produced accord 6AJOF Fig. 2. SEM micrograph of the fly ash. →3A2

Januarv 2005 Fillration Pcrfonmiiicc of MiiU'itc Compo.siles for Hol-Ga.'i Cleaning 47 -• 10" 10' 10' Diameter. )ini Fig, I. P;irliclc si/e distrihtition ofthe fly ash. performance of the filters, which in practice would be used at temperatures >11OO K, The dusl passing through the lilter samples was captured with filter paper. The average dust con￾cetitration in the clean gas after filtration was determined by di￾viding the mass ofthe captured dust by the lotal amount (5.1 nv') of clean gas that passed through the filter paper. The filtra￾tion etlicieiicy. y\. was calculated as follows: 100 (2) where C\ and C are the dust concentraiions before and afler fil￾tration, respectively. III. Results and Discussion Figure 3 shows SEM micrographs of typical cross-sections and filtration layer surfaces ofthe samples. For sample A. the outer surfaee was covered with mullite partieulates that had accumu￾lated on the A1:O, liber sheets. The thickness ofthe paniculate layer was ^5 10 |.im. This layer eonsisted of agglomerated mul￾lite particles wilh many pores (less than a few micrometers) among the agglomerates and acted as the filtration layer. Al￾though regions among the fiber bundles were filled with the mullite particulates, there were many pores in the bundles. As shown in Fig, 3, large gaps were observed among the bundles and agglomerated particulates. For sample B, the mullite par￾tieulate layer, which was 2O-3O-|.im thick, acted as the filtration layer. The filtration surface seemed less uneven, and the pores smaller, among the agglomerates of sample B than in sample A. In sample C, the whiskers were arranged on the partieulate layer perpendicular to the outer surface. The thickness ofthe whisker layer was the same as the length (-^4 \\m) ofthe whiskers. The aspect ratio (length/width) ofthe whiskers was -- 15. The filtra￾tion surface of sample C was fully covered wilh whiskers. There were no large depressions, such as those of sample A. and the pores observed on the surface of sample C seemed to be smaller than those of the olher samples. Incidentally, the microslrticture of the substrate layer of sample C was the same as those o'i the other samples. The ED pattern ofa singie whisker corresponded to that of a single crystal of mullite. TEM EDS analysis o\' the whiskers indicated an average Si/Al molar ratio of 0.34, similar to the stoichiometric composition (0.33) of mullite. No deteri￾orated layers were obser\ed at the interfaces between the mullile whiskers and the mullite partieulate layers that had accumulated on the ANOi fiber sheets. We propose a mechanism for the formation of whiskers per￾pendieular to the outer surface based on thermodynamic argu￾ments. Figure 4 shows the equilibrium partial pressures ofthe gaseous species and stable regions of the condensed phases un￾der a constant F,,,, partial pressure of 1,3 x 10 ^ Pa. calculated from the decomposition reaction of AlF^i^i (AlE'iis,-* A1|||+ 3E'"(g)) as a function of O^ partial pressures at 1473 K for the Al-O F"" Si system. As shown in Fig, 4. when AlFi,^, particu￾lates exist on the outer surface ofthe cylindrical composites, the AlF^^i reacts with oxygen to mainly produce AI0F2(gi as follows: + -f- Fle) Consequently, an oxygen potential gradient forms on the outer surface. Under O^ partial pressures ol' 3.2 x 10 ""* Pa, the partial pressure of the AIF decreases, and the AIF condenses into AIiO^. -H3F, (5) Under O2 partial pressures of 9.6xlO~'' ' Pa, SiF4|j,, condenses into SiOi as follows: SiF. o2(gl +4F (E) (6) The Gibbs free energy for the formation of mullite from AhO, and SiO^ at 1473 K is -20.7 kJ/mol, Therefore, when both AUO, and SiO; coexist as stable phases at higher O2 partial pressures, as shown in Fig, 4. mullite may be produced accord￾ing to . 2. SLM ofthe tlj ;i 3AI22OO3 • (7)

Journal of the merican Ceramie Societ Kitaoka et al. Vol. 88. No. I Particle filtration Fig 3. SEM micrographs of typical cross sections and filtration layer surfaces in each sample Because the oxygen partial pressure at convex points on the substrate layers. pore size distributions were measured for the mullite particulate layer on the outer surface of the composite is entire sample. The thickness ratios of the filtration layers to higher than in the troughs, the convex points may serve as nu- the samples were 100 um in Fig. 5 probably correspond to crevices be- alone are difficult to evaluate and separate from those of the tween the fiber bundles and or sheets. The pores of micrometer size are probably gaps in the fiber bundles. and those measuring <I um correspond to the pores among the agglomerated par- ticulates and or whiskers Table I lists the properties of the samples. The average 10 ALO SiO ize and porosity of each sample were -50 Hm and 50 vol% respectively, The sample permeabilities were 2x 10 m"to A AIRy AIF 0.4 0.3 AlOF 0.3 AlOF 0.2 15 0旨 -45-403530-25-20-15-10-5 og(Po ig.4. Equilibrium partial pressures of densed phases under a cons 1.3x 10 Pa, which is calculated from The du species and stable partial pr mposition reaction ol AlF,(AIF3+ALn+ 3Figs). as a function of O: partial pressures at 1473 Diameter. um K for the Al- O F-Si system. All pressures in the figure are in pascals ze distribution in sample C

Jdunuil of the .•iDicriciin Ccrmiiic Socii'lv Kiltidkti el al. Voi, 88. No. Kig. 3. SFM microgniphs of typical cross sections Lind lillration kiycr surlaccs in each sample. Because the o,\ygen partial pressure at convex points on the mullite partieulate layer on the outer surface ofthe composite is higher than in the troughs, the convex points may serve as nu￾cleation sites for the vapor-phase reaction, forming mullite is￾lands. Since the tip of the island is exposed to a higher oxygen partial pressure than the edges, the islands will grou perpendic￾ular lo the outer surfaee to form whisker-shaped mullite crystals. Because the pore si/e distributions of the filtration layers alone are difficult to evaluate and separate from those of the -10 ^ -15 -45 -40 ^15 -30 -2 Fig.4. Equilibrium parlial presstires of gaseous species and siahlc re￾gions of condensed phases under a constant F,,., parii;!l piessiire of t,3 X 10 ** Pa. wliieli is calculated from the decomposilion reaction of AlFi (AIF,->Al,]|+3F,^i), as a function of O; partial pressures at 1473 K for the Al O F-Si system. All pressures in llie figure are in pascals. substrate layers, pore size distributions were measured for the entire sample. The thickness ratios of the filtration layers to the samples were 100 |.im in Fig. 5 probably correspond to crevices be￾tween the fiber bundles and/or sheets. The pores of micrometer size are probably gaps in the fiber bundles, and those measuring |02 Diameter, ^m FiR. 5. Pore size dislrihulion in sample C

January 2005 iltration Performance of Mullite Composites /or Hol-Gas Cleaning T able I. Properties of Samples Table Il. Filtration Performance of Sample D120-3g1a-3a7A10-3 size(Hm) Open porosity 54 49 51 (vol%o 8m/iltration Filtration efliciency 99.973 99,976 999 Permeability (n)2.92×10-12.35×10-22.75×1012(%) 3x 10-m". Therefore, the permeation behavior of the samples at 3 kPa. The pressure difference under steady-state conditions was almost the same as for the clean gas without dust. During the filtration performance tests. because the gas flow containing air had been sucked inward through the sample and in the inward direction through the samples caused the dust to remained constant until the air was blown outward. This resul accumulate on the outer surfaces of the samples. the differential suggests that a large amount of residual ash particles had be pressure across the samples gradually increased. When the dust come trapped in the filtration layer when the air was blown cake on the outer surface was removed by pulse blowing, the upwards. thereby blocking the gas flow in the inward direction differential pressure drastically decreased. Figure 6 shows pro- The particles captured on the filtration layer surface probably files of the differential pressures for the samples during the tests, could easily enter the pores formed among the agglomerated The horizontal time axis in Fig. 6 is too long to show the pres- particles. so that they became completely surrounded by the sure loss signal during one filtration-blowing cycle of 150 s. pore walls. As shown in Fig 6(b). although the time deseailar to pressure is shown. To illustrate clearly the pressure loss signals that of sample A when the thickness of the filtration layer was after 6 h, where the maximum differential pressures for the sam- ncreased. the maximum differential pressure of sample B was ples were at steady state. magnified portions of each profile are larger than that of sample A. However. for sample C(Fig. 6(c) shown in their sets. If the pressure difference suddenly increased adhesion of the whiskers to the outer surface dramatically de and quickly reached a constant value. the channels in the filtra- creased the maximum differential pressure to about one-third tion layer immediately filled with the dust, blocking the flow of lower than those of the other two samples. and this value re- gas into the filter, On the other hand. if the pressure difference mained constant during the test. The pressure difference at increased during filtration. the dust cake apparently grew with steady state gradually increased as the dust-containing air was out blocking the filtration layer. breathed, but it did not become saturated. This finding suggests The maximum differential pressure of sample A(Fig. 6(a)) that dust accumulated on the whisker surfaces could be removed ncreased with time for -I h after which it remained constant easily by pulse cleani Table II lists the filtration properties of the samples. The fil tration efficiencies of the samples were all very high and very similar, The amount of captured dust for the samples differed by 0.58 mg m'. a small but significant amount. For sample A the dust concentration in the clean gas after filtration was 1.37x 10' gm, and the filtration efliciency was 99,973% The filtration efficiency of sample B was slightly better than that of sample A. but the increased filtration-layer thickness led to an increase in the maximum differential pressure during the filtration tests. The filtration efliciency of sample C was the highest among all the samples (Table Il) The low maximum differential pressure and high filtration 4 efficiency were achieved simultaneously by the addition of he whisker layer for the following reasons: First, the large po osity and small pore diameters of the whisker layer maintained 6 high permeability. Second, because the whiskers were perpen dicular to the outer surlace. the captured dust was in point con tact with the edges of the whiskers and, thus, easy to remove during pulse cleaning. The whiskers rarely broke away or failed during the tests. retaining the same morphology as befor 2 IV. Conclusions The filtration performance of alumina-fiber-reinforced mullite omposite filters with respect to fly ash was evaluated at 673 K For filters with mullite particulate filtration layers. although the filtration efficiency was improved slightly by increasing the fil- ration layer thickness, the maximum differential pressure du ing the filtration tests also increased. Adhesion of whiskers to he top surface, with a thickness of only one grain. dramatically decreased the maximum differential pressure over that achieved with particulate filtration. A constant maximum differential pre sure was maintained throughout the testing. The filtration efhi Time ciency of the whisker-containing filter was the highest of all three Fig.6. Characteristic pressure loss signals for (a) sample A(b) sample filters. The whiskers remained intact and did not fail during B, and (e) sample C

January 2005 filiralioii Pcrfonmmcc oj MiiUilc Composites fur Ilcl-Gas Clctiiiing 1 able I. Properties of Saniplvs 1 able II. Filtration Performance of Samples Sumpk ( Open pore 52 si/c (j,mi) Open porosity 54 54 54 49 51 Permeability (nr) 2.92 x 10" '- 2,35 x 10"'- 2,75 x 10 '" Dust concentration 1,37 x 10 ' I.ISx 10 ' 0.784 x 10 ' after filtration (g/m') Filtration efficiency 99,973 99.976 99.984 3x1 0 " m". Therefore, the permeation behavior ofthe samples was almost the same as for the eleun gas without dust. During the filtration performance tests, because the gas Ilow in the inward direction through the samples caused the dust to accumulate on the outer surfaces ofthe samples, the differenlial pressure across the samples gradually inereased. When the dust cake on the oiiier surface was removed by pulse blowing, the differential pressure drastically decreased. Figure 6 shows pro￾liles ofthe differential pressures for the samples during the tests. The horizontal time axis in Fig. 6 is too long to show the pres￾sure loss signal during one iiltration-blowing cycle of 150 s. However, the time dependence of the maximum ditTerential pressure is shown. To illustrate clearly lhe pressure loss signals after fi h, where the ma,\imum differential pressures lor the sam￾ples were at steady state, magnified portions of each prolile are shown in their sets. If the pressure difference suddenly increased and quickly reached a constant value, the channels in the filtra￾tion layer immediately filled with the dust, blocking the flow of gas into the filter. On the other hand, if the pressure difference increased during filtration, the dust cake apparently grew with￾out blocking the liltralion layer. The maximum differential pressure of sample A (Fig, 6(a)) increased with time for — I h. after which it remained constant 10 Time, h Fig.6 . Chiiracteristi c pressur e los s signal s fo r (a ) siimplL- A . (h ) B. ;iiui (L'l siimplL" ('. at 3 kPa, The pressure difference under steady-state conditions suddenly reached a constant value immediately after the dust￾containing air had been sucked inward throngh the sample and remained constant until the air was blown outward. This resnlt suggests that a large amount of residual ash particles had be￾come trapped in the filtration layer when the air was blown outwards, thereby blocking the gas tlow in the inward direction. The particles captured on the (iltration layer surface probably couid easily enter the pores formed among the agglomerated particles, so that they became completely surrounded by the pore wails. As shown in Fig 6(b), although the time dependence ofthe maximum differential pressure of sample B was simiiar to that of sample A when the thickness ofthe filtration layer was increased, the maximum dilTerential pressure of sample B was larger than that of sample A, Howe\er. for sample C (Fig, 6(c)), adhesion ofthe whiskers to the outer surface dramalica!l\ de￾creased the maximum differential pressure to about one-third lower than those of the other tuo samples, and this value re￾mained constant during the test. The pressure difference at steady state gradually increased as the dust-containing air was breathed, but it did not become saturated. This finding suggests that dust accumulated on the w hisker surfaces could be removed easily by pulse cleaning. Table il lists the liltration properties of the samples. The fil￾tration efficiencies of the samples were ali \ery high and very simitar. The amount of captured dust for the sampies differed by 0.58 mg/nr\ a smail but signillcant amount. For sample A. the dust concentration in the clean gas after filtration was 1,37 X 10 -' g,m\ and the filtration efficiency was 99,973%, The filtration efficiency of sample B was slightly better than that of sample A, but the increased filtration-layer thickness led to an increase in the maximum differential pressure dvn'ing the filtration tests. The filtration efficiency of sample C was the highest among all the samples (Table II), The low maximum differential pressure and high filtration eiliciency were achieved simultaneously by the addition uf the whisker layer for the foliowing reasons: First, the large po￾rosity and small pore diameters ofthe whisker layer maintained high permeability. Second, because the whiskers were perpen￾dicular to the outer surface, the captured dust was in point con￾tact with the edges of lhe whiskers and. thus, easy to remove during pulse cleaning. The whiskers rarely broke away or failed during the tests, retaining the same morphology as before lhe testinc. IV. Conclusions The filtration performance of alumina-liber-reinforced mullite composite filters with respect to fly ash was evaluated at 673 K, For filters with mullite partieulate filtration layers, although the filtration efficiency was improved slightly by increasing the fil￾tration layer thickness, the maximum differential pressure dur￾ing the filtration tests aiso increased. Adhesion of whiskers to the top surface, with a thickness of only one grain, dramatically decreased the maximum differential pressure over that achieved w'itii partieulate liltralion, A constant maximum differential pres￾sure was maintained throughout tlie testing. The filtration efli￾ciency ofthe whisker-containing filter was the highest of all (hree filters. The whiskers remained intaci and did not fail during the tests

Journal of the American Ceramic Soxietr-Kitaoka et al Vol. 88. No. I R eferences pp 363-74 in High Temperature Gas Cleaning. Vol 2 Edited by A Ditter G P. Hancock."Cracking Spalling and Failure of Ash Deposits in "N. S. Jacobson. ]. L. Smialek, and d. 5. Fox,Molten salt 'c G Hemmer, and G Kasper. G. Braunems GmbH. Karlsruhe, Germany of Ceramics": pp. 205-22 in Corrasion of Advanced Ceres: Measurement and Modeling. Edited by K. G. Nickel. Kluwer Academie Publishers, Dordrecht -E F Osborn and A. Muan. "Phase Equilibrium Diagrams of Oxide Plate 3, The Amencan Ceramic Soctety and the Edward Orton, J H. Suzuki. M. Shimizu, H. Kamiya, M Takahashi, and T. Ota, "Preparation of Fine Mullite Powders with High Surface Area by Agglomeration Control S. Kitaoka, N, Kawashima. H. Muto. H. Suzuki. A. Yamaguchi, Y. Ta of Alkoxide- Derived Precursor Sol. "J, Sor. Ponder Technol, Jupon.34 Mullite Composite Filter. Cer. Eg. Sci. Proc. 22 H279-844(200T) K. Okada and N. Otsuka. ""Synthesis of Mullite Whiskers by Vapour-Phase Lippert, and J, E. Lane."Assessment of Po J Mater. Sci. Let 8, 1052(1989 Ceramie Materials for Hot Gas Filtration Applications, Cera. Bull. 70 19 K Okada and N Otsuka, Synthesis of Mullite Whiskers and Their Applic M. Davidson, X. Guan, H. HendrIx, and B. Shirley,""Power Systems Devel- opment Facility: Filter Element Evaluation During Combustion Testing Operational Conditions, Nv. Filtration Scpurutwon Technol. 13A. 10412(1999)

50 Journal of ihe Anieriain Ceramic Socielv—Kilaoka el al. Vol, 88. No. I References 'P, Haiiccx:k, "Cracking, Spelling and hailiirc of Ash Dcposils in Filtralion Syslem.s": pp. 3- 13 in Ili^li Tcnipi'miiirf GusCkiining. Vol Z. Edited hy A Ditllcr. G Hcmmer. and G Kasper,, (i, Braimems GmbH. Karlsruhe. Germimy, 1W9, "E, F, Osborn and A Miutn, "Phase Equilihrium Diaerauis of Oxidf Systems," Plate 5. The Ameriain Cciamic Sodety and ihc tidward Orion. ,Ir,. CiTamic Foiiniiation, l%{l, 'S, Kilaoka, N, Kawashima. H, Miiio, H, Suzuki, A. Yumaguchi. Y, T;i￾dakijma, unii S. ikeda, "f-abricalion ola Continuous Alumina Fibor-ReinlVirccd Mulliti: Composite Filler," Cer. Eng. Sci. Pro,.. 22 |4| 279 S4 (2(M)I), •"M, A, Alvin, T, H. Lipperl, and J, E, Lane, ""Assessment of Porous Ceramic Materials Ibr Hoi Gas Filtration Applicaiions," Ceram. Bul!.. 70 [9] "M. Davidson, X, Guan. H, HendriK, and B. Shirley. ""Power Systems Devel￾opment Facility: Filter Eletnetil Evaluation During Combustion Testing": pp. 363-74 ill Hi^h Ttwpt-itiitirc Gii.s Cleaiiiiig. \o\. 2. Ediled by A Dittler. G Hemnicr. and G Kasper. G. Braunems GmbH. Karlsruhe, Gcrmaiiy, 1999, ''N, S, Jacobson, J, L, Smiaiek, and D, S, Fox, '"Molten Suit Corrosion of Ceramics": pp. 205 22 in C"rrii.siim oJ Ailrtiiuvi/ dreiDiiiy: Mt'iisuremi'iit uml Moiklini;. F.dited by K. ti. Nickel, Kluwer Academic Publishers, Dordrecht. 19'M, H, Su/iiki, M, Shimi/u, II. Kamiva. M. Takahashi. and T, Ota. "Preparation of Fine Mullite Powders wilh High Surface Area by Agglomeration Control of Alkoxide-Dcrived Precursor Sol," ./. S«c. P-nnkr Ttrhuii. Jupun. 34. 170 75 (1997). "K, Okada and N Otsuka, ""Synihesis of Mulliti; Whiskers by Vapour-Phiise Reaction."" 7, Mala. .Sd. t<'/,, 8. "lO52-] (1989), ''K, Okwda and N, Otsuka. "Synthesis of Muliite Whiskers and Their Applica￾tion in Composites." ,/, Am. Cirwu. Sue. 74. 2414 8 (19911, '"D , -I, P, G'^'ng, "-Test Method for Cieanable Filters under Laboratory and Operalional Condilinns," Ailv. Filirmion Scpaidiinii Tiiliii"!.. 13A, UM I2(!9')9). •

Copyright of journal of the American Ceramic Society is the property of American Ceramic Society and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holders express written permission However, users may print, download, or email articles for individual use