《复合材料 Composites》课程教学资源（学习资料）第二章 增强体_carbon fiber_Biopitch-based general purpose carbon fibers：Processing and properties

Availableonlineatwww.sciencedirect.com BCIENCE DRECT● CARBON ELSEVIER Carbon4302005)591-597 Biopitch-based general purpose carbon fibers Processing and properties M.J. Prauchner a, *v.M.D. Pasa b.s. otani c.C. otani c Instituto de quimica, Universidade de brasilia, C P. 4478, CEP: 70910 970, Bro Departamento de quimica, Universidade Federal de Minas gerais, At. Antonio Carlos, 6627, CEP: 31270 901, Belo Horizonte, Brazil Departamento de Fisica, Instituto de Tecnologia Aeronautica, Centro Tecnico Aeroespacial, CEP: 12228 901, Sao Jose dos Campos, Brazil Received 17 September 2003: accepted 17 October 2004 vailable online 24 November 2004 Abstract Eucalyptus tar pitches are generated on a large scale in Brazil as by-products of the charcoal manufacturing industry. They pres- ent a macromolecular structure constituted mainly of phenolic, guaiacyl, and siringyl units common to lignin. The low aromaticity (60-70%), high O/C atomic ratios (0.20-0.27%), and large molar mass distribution are peculiar features which make biopitches behave far differently from fossil pitches. In the present work, eucalyptus tar pitches are evaluated as precursors of general purpose carbon fibers(GPCF)through a four-step process: pitch pre-treatment and melt spinning, and fiber stabilization and carbonization Homogeneous isotropic fibers with a diameter of 27 um were obtained. The fibers had an apparent density of 1.84g/cm,, an electrical esistivity of 2 x 10Q2m, a tensile strength of 130 MPa, and a tensile modulus of 14 GPa. Although the tensile properties advise against using the produced fibers as structural reinforcement, other properties give rise to different potential applications, as for example in the manufacture of activated carbon fibers or felts for electrical insulation. C 2004 Elsevier Ltd. All rights reserved Keywords: A Carbon fibers, Pitch; B Heat treatment, Stabilization; D. Mechanical properties 1. Introduction from isotropic pitches, usually called general purpose carbon fibers(GPCF) or isotropic carbon fibers, have Fossil pitches(petroleum pitches and coal tar pitches) poor ordering and low or no preferred orientation have been largely investigated as carbon fiber precursors therefore having moderate or poor mechanical proper mainly because pitches are cheaper raw materials than es[46] the traditionally used polyacrylonitrile(PAN) and pres Due to the comparatively high production cost of ent a simpler and less expensive processing [1-6. Pitch- HPCFs, their applications are mainly restrained to the based carbon fibers can be classified into two types aerospace and sports goods industries, where the perfor- according to their properties: the first type, usually mance/weight quotient is by far the predominant factor called high performance carbon fibers(HPCF), are pro- [7]. In contrast, GPCFs are produced at a much lower duced from mesophase pitches and have high degrees of price, and therefore, they are employed in applications ordering and orientation, and consequently, improved where mechanical properties are not the most relevant mechanical properties [1-3]. Conversely, those obtained factor. For example, as concrete reinforcement [8], elec- trodes for energy storage systems [9, 10), and in the pro Corresponding author. Tel. +55 61 307 2167: fax: +55 61 273 duction of activated carbon fibers(ACF). In turn, ACFs are used as catalysts [ll], catalyst supports [12], mole- cular sieves for the separation and purification of gas 6223/. see front matter 2004 Elsevier Ltd. all rights reserved 0.1016 carbon2004.10.023

Biopitch-based general purpose carbon fibers: Processing and properties M.J. Prauchner a,*, V.M.D. Pasa b , S. Otani c , C. Otani c a Instituto de Quı´mica, Universidade de Brası´lia, C.P. 4478, CEP: 70910 970, Brası´lia, Brazil b Departamento de Quı´mica, Universidade Federal de Minas Gerais, Av. Antoˆ nio Carlos, 6627, CEP: 31270 901, Belo Horizonte, Brazil c Departamento de Fı´sica, Instituto de Tecnologia Aerona´utica, Centro Te´cnico Aeroespacial, CEP: 12228 901, Sa˜ o Jose´ dos Campos, Brazil Received 17 September 2003; accepted 17 October 2004 Available online 24 November 2004 Abstract Eucalyptus tar pitches are generated on a large scale in Brazil as by-products of the charcoal manufacturing industry. They pres￾ent a macromolecular structure constituted mainly of phenolic, guaiacyl, and siringyl units common to lignin. The low aromaticity (60–70%), high O/C atomic ratios (0.20–0.27%), and large molar mass distribution are peculiar features which make biopitches behave far differently from fossil pitches. In the present work, eucalyptus tar pitches are evaluated as precursors of general purpose carbon fibers (GPCF) through a four-step process: pitch pre-treatment and melt spinning, and fiber stabilization and carbonization. Homogeneous isotropic fibers with a diameter of 27lm were obtained. The fibers had an apparent density of 1.84 g/cm3 , an electrical resistivity of 2 · 10
4 Xm, a tensile strength of 130MPa, and a tensile modulus of 14GPa. Although the tensile properties advise against using the produced fibers as structural reinforcement, other properties give rise to different potential applications, as for example in the manufacture of activated carbon fibers or felts for electrical insulation. 2004 Elsevier Ltd. All rights reserved. Keywords: A. Carbon fibers, Pitch; B. Heat treatment, Stabilization; D. Mechanical properties 1. Introduction Fossil pitches (petroleum pitches and coal tar pitches) have been largely investigated as carbon fiber precursors mainly because pitches are cheaper raw materials than the traditionally used polyacrylonitrile (PAN) and pres￾ent a simpler and less expensive processing [1–6]. Pitch￾based carbon fibers can be classified into two types according to their properties: the first type, usually called high performance carbon fibers (HPCF), are pro￾duced from mesophase pitches and have high degrees of ordering and orientation, and consequently, improved mechanical properties [1–3]. Conversely, those obtained from isotropic pitches, usually called general purpose carbon fibers (GPCF) or isotropic carbon fibers, have poor ordering and low or no preferred orientation, therefore having moderate or poor mechanical proper￾ties [4–6]. Due to the comparatively high production cost of HPCFs, their applications are mainly restrained to the aerospace and sports goods industries, where the perfor￾mance/weight quotient is by far the predominant factor [7]. In contrast, GPCFs are produced at a much lower price, and therefore, they are employed in applications where mechanical properties are not the most relevant factor. For example, as concrete reinforcement [8], elec￾trodes for energy storage systems [9,10], and in the pro￾duction of activated carbon fibers (ACF). In turn, ACFs are used as catalysts [11], catalyst supports [12], mole￾cular sieves for the separation and purification of gas 0008-6223/$ - see front matter 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.carbon.2004.10.023 * Corresponding author. Tel.: +55 61 307 2167; fax: +55 61 273 4149. E-mail address: marcosjp@unb.br (M.J. Prauchner). Carbon 43 (2005) 591–597 www.elsevier.com/locate/carbon

M.. Prauchner et al. Carbon 43(2005)591-597 mixtures [13], gas storage adsorbents [14], and water are peculiar features which make biopitches behave far fying and filtering ele 5-17 differently from fossil pitches [22, 23] In Brazil, a different type of pitch, derived from bio- The isotropic character and low carbon yield (<50%) mass, is generated in large scale as a by-product of the of eucalyptus tar pitches allows them to develop easily charcoal-making industry. In charcoal production, the microporosity during carbonization. Even without any volatiles released during wood pyrolysis can be recov- additional activation process, eucalyptus tar pitch-based ered by condensation to give rise to an oily liquid called carbons obtained at 800C have BET microporous sur- wood tar. This is separated by decanting to give rise to face areas over 200m/g [24]. Although these values are an aqueous fraction(so-called pyroligneous acid)and still low if compared to those typical for activated car an organic fraction (insoluble tar), which corresponds bons, in the range 700-2000m1g [27, 28]. the pores to around 35% and 7% of the initial mass of wood, formed during carbonization will certainly facilitate respectively. The insoluble tar can be distilled to sepa he diffusion and attack of oxidizing agents during a rate fractions used as flavors, fragrances and sources subsequent activation process [29]. In this context, and of fine chemical products [18, 19]. A heavier fraction, taking into account the current interest in the filamen the so-called wood tar pitch, is then obtained as a distil- tary form of activated carbons, which presents higher lation residue(about 50% in mass). adsorption and desorption rates and can be easily pro- t Most part of the Brazilian charcoal production uses cessed into sheets, felts and cloths, the forms preferred ood from planted eucalyptus forests [20]. This is an for manufacturing filtering elements, eucalyptus tar important and environmentally friendly activity because pitches have been studied as precursors of GPCf. The biomass is a renewable energy source which, on the con- present paper reports on pitch processing and the char- trary that fossil fuels, gives rise to a positive balance of acterization of the fibers obtained CO2 fixation and O2 release and to a comparatively low level of emission of toxic gases such as SO2, NO2 and No [21]. In turn, developing applications for wood tar 2. Experimental fractions is important as a means to stimulate tar recov- ering in the industrys chimneys, therefore preventing 2. 1. Raw materials the release of a large volume of pollutants into the atmo phere, and to aggregate revenue to the charcoal-making The precursor pitch used in the present work, the so- industry. In this context, our research group has devel- called crude pitch, was obtained by vacuum distillation oped work aiming to characterize eucalyptus tar pitch of eucalyptus tar in a pilot batch plant. Tar was ob- and investigate its potential uses [22-26 tained by condensation of the volatile released during Previous studies demonstrated that eucalyptus tar the slow pyrolysis (500C, 12-14 C/h)of eucalyptus pitch presents a macromolecular structure constituted wood in industrial masonry ovens. The distillation cut mainly of phenolic, guaiacyl, and siringyl units(Fig. 1) temperature was 180C at 30-38 mmHg and pitch yield resulting from lignin degradation during wood pyroly- was about 50%(w/w) sis. Their low aromaticity(60-70%), high O/C atomic In order to increase the softening point of the crude ratios(0.20-0.27%), and large molar mass distribution pitch(76 1C), about 400g was heat-treated in a 1000- mL Kettle vessel connected to a vigreaux column. Pitch homogenization was achieved through a mechanical H/OCH stirrer. Different temperatures and treatment times, in HaCO/H the range of 190-250C and 1-8h, respectively, were used [22, 2 2.2. Carbon fiber preparation (CH2)n The pitches were converted into the filamentary form H3 CO/H OCHvH by melt spinning. The melted pitches were extruded der nitrogen pressure through a circular-shaped spin- ning nozzle (diameter =0. 5mm; length over diameter (L/D)=4.4)using a laboratory-scale monofilament apparatus. After being condensed by cooling, the fibers were wound onto a winding bobbin. This step involved adjusting the spinning temperature, pressure and rate The as-spun fibers, also called green fibers, were stabi- ig. 1. Model structure illustrating the main functional groups lized by oxidative thermal treatments. For that, the presents in eucalyptus tar pitches fibers were cut into M6-in. bundles weighing 1.5-2g

mixtures [13], gas storage adsorbents [14], and water purifying and filtering elements [15–17]. In Brazil, a different type of pitch, derived from bio￾mass, is generated in large scale as a by-product of the charcoal-making industry. In charcoal production, the volatiles released during wood pyrolysis can be recov￾ered by condensation to give rise to an oily liquid called wood tar. This is separated by decanting to give rise to an aqueous fraction (so-called pyroligneous acid) and an organic fraction (insoluble tar), which corresponds to around 35% and 7% of the initial mass of wood, respectively. The insoluble tar can be distilled to sepa￾rate fractions used as flavors, fragrances and sources of fine chemical products [18,19]. A heavier fraction, the so-called wood tar pitch, is then obtained as a distil￾lation residue (about 50% in mass). Most part of the Brazilian charcoal production uses wood from planted eucalyptus forests [20]. This is an important and environmentally friendly activity because biomass is a renewable energy source which, on the con￾trary that fossil fuels, gives rise to a positive balance of CO2 fixation and O2 release and to a comparatively low level of emission of toxic gases such as SO2, NO2 and NO [21]. In turn, developing applications for wood tar fractions is important as a means to stimulate tar recov￾ering in the industry
s chimneys, therefore preventing the release of a large volume of pollutants into the atmo￾sphere, and to aggregate revenue to the charcoal-making industry. In this context, our research group has devel￾oped work aiming to characterize eucalyptus tar pitch and investigate its potential uses [22–26]. Previous studies demonstrated that eucalyptus tar pitch presents a macromolecular structure constituted mainly of phenolic, guaiacyl, and siringyl units (Fig. 1) resulting from lignin degradation during wood pyroly￾sis. Their low aromaticity (60–70%), high O/C atomic ratios (0.20–0.27%), and large molar mass distribution are peculiar features which make biopitches behave far differently from fossil pitches [22,23]. The isotropic character and low carbon yield (<50%) of eucalyptus tar pitches allows them to develop easily microporosity during carbonization. Even without any additional activation process, eucalyptus tar pitch-based carbons obtained at 800C have BET microporous sur￾face areas over 200m2 /g [24]. Although these values are still low if compared to those typical for activated car￾bons, in the range 700–2000m2 /g [27,28], the pores formed during carbonization will certainly facilitate the diffusion and attack of oxidizing agents during a subsequent activation process [29]. In this context, and taking into account the current interest in the filamen￾tary form of activated carbons, which presents higher adsorption and desorption rates and can be easily pro￾cessed into sheets, felts and cloths, the forms preferred for manufacturing filtering elements, eucalyptus tar pitches have been studied as precursors of GPCF. The present paper reports on pitch processing and the char￾acterization of the fibers obtained. 2. Experimental 2.1. Raw materials The precursor pitch used in the present work, the so￾called crude pitch, was obtained by vacuum distillation of eucalyptus tar in a pilot batch plant. Tar was ob￾tained by condensation of the volatile released during the slow pyrolysis (500C; 12–14C/h) of eucalyptus wood in industrial masonry ovens. The distillation cut temperature was 180C at 30–38mmHg and pitch yield was about 50% (w/w). In order to increase the softening point of the crude pitch (76.1C), about 400 g was heat-treated in a 1000- mL Kettle vessel connected to a vigreaux column. Pitch homogenization was achieved through a mechanical stirrer. Different temperatures and treatment times, in the range of 190–250C and 1–8 h, respectively, were used [22,23]. 2.2. Carbon fiber preparation The pitches were converted into the filamentary form by melt spinning. The melted pitches were extruded un￾der nitrogen pressure through a circular-shaped spin￾ning nozzle (diameter = 0.5mm; length over diameter (L/D) = 4.4) using a laboratory-scale monofilament apparatus. After being condensed by cooling, the fibers were wound onto a winding bobbin. This step involved adjusting the spinning temperature, pressure and rate. The as-spun fibers, also called green fibers, were stabi￾lized by oxidative thermal treatments. For that, the fibers were cut into 6-in. bundles weighing 1.5–2 g. OH H3CO/H O H/OCH3 (CH2)n (CH2)n (CH2)n O H3CO/H OCH3/H O H n = 0, 1, 2, 3 H (CH2)n Fig. 1. Model structure illustrating the main functional groups presents in eucalyptus tar pitches. 592 M.J. Prauchner et al. / Carbon 43 (2005) 591–597

M.J. Prauchner et al. Carbon 43(2005)591-597 The bundles, laid onto a piece of mesh screen in order to 2.8. Tensile properties assure oxygen access on all over the fiber surface, were then placed into an air forced-convection oven and trea- The tensile strength and tensile modulus of the fibers ted at different heating rates and final temperatures. were determined from the stress-elongation curves. The After stabilization, the fibers were carbonized at measurements were carried out according to ASTM 1000C(2.0C/min; I h) under nitrogen atmosphere in D3379M standards (single-fiber method). The curves a stainless steel tubular furnace were obtained in a universal test machine Lloyd Lr K. the gauge length was 30mm and the crosshead 23. Softening point (SP) speed was 0.6mm/min. The test was repeated with 30 40 single filaments to make it possible to obtain an aver Pitch SPs were determined using the"ring and ball age value. The system compliance was determined using method"following the ASTM D2398--73-Standard commercial PAN-based carbon fibers. The average Test. In accordance with the standard, the tests were diameters of the fibers were visually observed using carried out in duplicate. If the measurements resulted SEM and the sectional area was then calculated in a difference larger than 1C, the analyses were repeated 2.9. Electrical resistivity 2. 4. GPC analvses Taking into account that isotropic carbon fibers have relatively high electrical resistivity at room temperature Analyses of gel permeation chromatography (GPC) if compared to those of metals such as copper and silver of the tetrahydrofuran-soluble fraction of eucalyptus we used a two-point probe technique for the measure- tar pitches were carried out in a Shimadzu LC-10AD Li- ments. The fibers(50mm in length)were submitted to quid Chromatograph coupled with a Shimadzu UV-Vis a voltage of 2.0V in an Eco Chemie PGSTAT Auto SPD-10AV Detector at 254nm. Elution was carried out lab potentiostat. Approximately 30 single filaments were in THF at 30C with a flow rate of l mL/min using two measured to obtain an average value. Once more fila coupled columns of polystyrene-divinyl benzene gel ment areas were obtained from sem observations (Shim-pack GPC-8025 and Shim-pack 803, in the cited order). Injections of 20uL were made with the samples 2. 10. Apparent density dissolved in the eluent (2mg/mL) The apparent densities of the fibers at 25C we 2.5. Elemental analysis determined by the sink-float method. The fibers were sank into carbon tetrachloride (d=1594), and then The samples were analyzed in a Perkin-Elmer 2400 1, 2-dibromo-ethane (d=2. 179) was added, drop by lemental Analyzer to determine carbon, hydrogen, drop, until the fibers stood still at an intermediate height and nitrogen content. The oxygen-plus-ash content in the liquid column. Then, the density of the corre- was calculated by difference sponding solution was measured using a picnometer and the value was assumed to be the same as the fiber 2.6. Rheological studies sample dio. parent vIscosity was measured using a Brookfield digital rheometer model HADV-Ill and a Brookfield 3. Results and discussions Thermosel accessory. An SC4-21 spindle was used inside a cylindrical sample holder for all tests. Sample volume 3.1. Pitch processing was 8mL. Because of the strong effect of temperature on apparent viscosity [25]. both the pitch and the spindle key step in the production of pitch-based carbon were preheated for 30 minutes before the spindle was fibers is stabilization, which aims to render the thermo- lowered into the sample. In addition, 30 more minutes plastic as-spun filaments infusible in order to prevent were allowed for temperature equilibrium before the fibers from melting and/or deforming during the carbon spindle was rotated. ization step and, in the particular case of anisotropic carbon fibers to retain the molecular orientation ac 2. 7. Scanning electron microscopy(SEM) quired during spinning. Stabilization of isotropic pitch filaments is usually carried out by thermal treatment un The micrographs were obtained from a JEOL JSM- der air atmosphere at temperatures below the softening 840A electron microscope(20kV; 2 x 10A). The as- point(SP)of the fiber. This treatment introduces oxy pun and stabilized fibers(non-conducting) were sput- genated crosslinks among the molecules As crosslinking er-coated with pure gold prior to scanning akes place, the SP increases, and hence the temperature

The bundles, laid onto a piece of mesh screen in order to assure oxygen access on all over the fiber surface, were then placed into an air forced-convection oven and trea￾ted at different heating rates and final temperatures. After stabilization, the fibers were carbonized at 1000C (2.0C/min; 1 h) under nitrogen atmosphere in a stainless steel tubular furnace. 2.3. Softening point (SP) Pitch SPs were determined using the ‘‘ring and ball method’’ following the ASTM D2398—73—Standard Test. In accordance with the standard, the tests were carried out in duplicate. If the measurements resulted in a difference larger than 1 C, the analyses were repeated. 2.4. GPC analyses Analyses of gel permeation chromatography (GPC) of the tetrahydrofuran-soluble fraction of eucalyptus tar pitches were carried out in a Shimadzu LC-10AD Li￾quid Chromatograph coupled with a Shimadzu UV–VIS SPD-10AV Detector at 254 nm. Elution was carried out in THF at 30C with a flow rate of 1mL/min using two coupled columns of polystyrene–divinyl benzene gel (Shim-pack GPC-8025 and Shim-pack 803, in the cited order). Injections of 20lL were made with the samples dissolved in the eluent (2mg/mL). 2.5. Elemental analysis The samples were analyzed in a Perkin-Elmer 2400 Elemental Analyzer to determine carbon, hydrogen, and nitrogen content. The oxygen-plus-ash content was calculated by difference. 2.6. Rheological studies Apparent viscosity was measured using a Brookfield digital rheometer model HADV-III and a Brookfield Thermosel accessory. An SC4-21 spindle was used inside a cylindrical sample holder for all tests. Sample volume was 8mL. Because of the strong effect of temperature on apparent viscosity [25], both the pitch and the spindle were preheated for 30 minutes before the spindle was lowered into the sample. In addition, 30 more minutes were allowed for temperature equilibrium before the spindle was rotated. 2.7. Scanning electron microscopy (SEM) The micrographs were obtained from a JEOL JSM- 840A electron microscope (20 kV; 2 · 10
9 A). The as￾spun and stabilized fibers (non-conducting) were sput￾ter-coated with pure gold prior to scanning. 2.8. Tensile properties The tensile strength and tensile modulus of the fibers were determined from the stress–elongation curves. The measurements were carried out according to ASTM D3379M standards (single-fiber method). The curves were obtained in a universal test machine Lloyd LR 5K. The gauge length was 30mm and the crosshead speed was 0.6mm/min. The test was repeated with 30– 40 single filaments to make it possible to obtain an aver￾age value. The system compliance was determined using commercial PAN-based carbon fibers. The average diameters of the fibers were visually observed using SEM and the sectional area was then calculated. 2.9. Electrical resistivity Taking into account that isotropic carbon fibers have relatively high electrical resistivity at room temperature if compared to those of metals such as copper and silver, we used a two-point probe technique for the measure￾ments. The fibers (50mm in length) were submitted to a voltage of 2.0 V in an Eco Chemie PGSTAT Auto lab potentiostat. Approximately 30 single filaments were measured to obtain an average value. Once more fila￾ment areas were obtained from SEM observations. 2.10. Apparent density The apparent densities of the fibers at 25C were determined by the sink-float method. The fibers were sank into carbon tetrachloride (d = 1.594), and then 1,2-dibromo-ethane (d = 2.179) was added, drop by drop, until the fibers stood still at an intermediate height in the liquid column. Then, the density of the corre￾sponding solution was measured using a picnometer and the value was assumed to be the same as the fiber sample. 3. Results and discussions 3.1. Pitch processing A key step in the production of pitch-based carbon fibers is stabilization, which aims to render the thermo￾plastic as-spun filaments infusible in order to prevent fibers from melting and/or deforming during the carbon￾ization step and, in the particular case of anisotropic carbon fibers, to retain the molecular orientation ac￾quired during spinning. Stabilization of isotropic pitch filaments is usually carried out by thermal treatment un￾der air atmosphere at temperatures below the softening point (SP) of the fiber. This treatment introduces oxy￾genated crosslinks among the molecules. As crosslinking takes place, the SP increases, and hence the temperature M.J. Prauchner et al. / Carbon 43 (2005) 591–597 593

M.. Prauchner et al. Carbon 43(2005)591-597 can be raised until the fibers become infusible [30, 31]. As ization and carbonization caused pronounced filament a higher initial SP permits stabilization to be started at a shrinkage, as is discussed later temperature, and crosslinking reactions ha In spite of the pre-treatment and consequent in- higher kinetics at higher temperatures, pitches with creases in SP, even the filaments produced from the higher SP can be stabilized using more elevated heating pitch pre-treated at 250C for 6h required very low rates without incurring problems of fiber deformation or heating rates to be stabilized without coalescence, fusing 0.08C/min, therefore implicating in a too time-consum y the other hand, the pitch spinnability is reduced ing process even though a final temperature of 180C with increasing pitch SP. This occurs because the higher was sufficient to make the fibers infusible. This heating the pitch SP, the higher the temperature necessary to rate is much lower than those usually employed to stabi spin it Since eucalyptus tar pitches have reactive oxygen lize fossil pitch filaments, 0.5-2C/min [30, 31]. There are containing functional groups and high volatile contents, two main reasons for this. The first one is that filaments high spinning temperatures give rise to bubble forma- with much higher SP can be used in the case of fossil tion, which damages the spinning process stability. In pitches, which is made possible because fossil pitches fact, if rheological studies are carried out for pitches present elevated thermal stability, and therefore, can with different SP, and the temperature is adjusted for be easily spun at temperatures sometimes even higher each sample in order to provide approximately the same than 350 C [32]. Having higher SP, the fossil pitch fila- viscosity(e. g, the spinning viscosity), viscosity instabil- ments can be stabilized at temperatures at which cross- ity increases with increasing pitch SP(Fig. 2)because linking reactions occur more rapidly. The another the corresponding temperature is higher. As pitch vis- main reason is the large amount of low molar mass mol- cosity is highly temperature-dependent [25], the pitch fil- ecules present even in the polymerized samples of euca- aments draw down and cool very quickly during lyptus tar pitches, which correspond to higher retention spinning. Since fiber tensile stress during spinning is time in the GPC curves of Fig 3. During oxidative ther nearly that required to break the filaments due to the mal treatment, these molecules can easily acquire energy brittle nature of the as-spun fibers, it is understandable sufficient to provoke the coalescence of the filaments hat oscillations during spinning can easily lead to fila- even at temperatures below the pitch SP. However,un ment failure der heating rates as low as 0.08 C/min, these molecules In practice, we had to find a compromise by increas- were slowly eliminated either by sublimation or ing pitch SP to a value as high as practicable without o "1 adm tton, Derbyshire et al. 4 reported that the running the risk of introducing insuperable problems for the fiber forming step. In this context, crude eucalyp- maximum effective heating rate during oxidative stabil- tus tar pitch(SP=76C)was submitted to a pre-treat- ization of pitch filaments decreases with increasing fiber ment aiming to increase its SP. Pre-treatment involved diameter due to the limited oxygen diffusion. Therefore, hermal polymerization at 250C [22, 23 the relatively large diameter of the as-spun filaments The pitch sample pre-treated at 250C for 6h was the produced in the present work is possibly contributing one with the highest SP(134 C)which presented good to make stabilization difficult spinnability. This pitch was spun at 175-180C, at a rate Despite oxygen incorporation(Table 1), eucalyptus of 48-50m/min and under a pressure of 2 bar. The fila- tar pitch filaments underwent pronounced weight loss ments had an average diameter of 46+ 2 um, which is and consequent shrinkage during stabilization table elatively large. However, the subsequent steps of stabil- 2), which can be attributed to the release of volatiles, 1E+06 1E+05 人思x 2h250° E+04 0.05.010015.020.025030.035040.0 Fig. 2. Viscosity as a function of shear time for crude pitch and pre-treated at 250C for 2 and 4h. Tests were carried out at 101, 118, and 156C, spectively, in order to provide similar viscosities for all samples

can be raised until the fibers become infusible [30,31]. As a higher initial SP permits stabilization to be started at a higher temperature, and crosslinking reactions have a higher kinetics at higher temperatures, pitches with higher SP can be stabilized using more elevated heating rates without incurring problems of fiber deformation or fusing. By the other hand, the pitch spinnability is reduced with increasing pitch SP. This occurs because the higher the pitch SP, the higher the temperature necessary to spin it. Since eucalyptus tar pitches have reactive oxygen containing functional groups and high volatile contents, high spinning temperatures give rise to bubble forma￾tion, which damages the spinning process stability. In fact, if rheological studies are carried out for pitches with different SP, and the temperature is adjusted for each sample in order to provide approximately the same viscosity (e.g., the spinning viscosity), viscosity instabil￾ity increases with increasing pitch SP (Fig. 2) because the corresponding temperature is higher. As pitch vis￾cosity is highly temperature-dependent [25], the pitch fil￾aments draw down and cool very quickly during spinning. Since fiber tensile stress during spinning is nearly that required to break the filaments due to the brittle nature of the as-spun fibers, it is understandable that oscillations during spinning can easily lead to fila￾ment failure. In practice, we had to find a compromise by increas￾ing pitch SP to a value as high as practicable without running the risk of introducing insuperable problems for the fiber forming step. In this context, crude eucalyp￾tus tar pitch (SP = 76C) was submitted to a pre-treat￾ment aiming to increase its SP. Pre-treatment involved thermal polymerization at 250C [22,23]. The pitch sample pre-treated at 250C for 6 h was the one with the highest SP (134C) which presented good spinnability. This pitch was spun at 175–180C, at a rate of 48–50m/min and under a pressure of 2 bar. The fila￾ments had an average diameter of 46 ± 2lm, which is relatively large. However, the subsequent steps of stabil￾ization and carbonization caused pronounced filament shrinkage, as is discussed later. In spite of the pre-treatment and consequent in￾creases in SP, even the filaments produced from the pitch pre-treated at 250C for 6 h required very low heating rates to be stabilized without coalescence, 0.08C/min, therefore implicating in a too time-consum￾ing process even though a final temperature of 180C was sufficient to make the fibers infusible. This heating rate is much lower than those usually employed to stabi￾lize fossil pitch filaments, 0.5–2 C/min [30,31]. There are two main reasons for this. The first one is that filaments with much higher SP can be used in the case of fossil pitches, which is made possible because fossil pitches present elevated thermal stability, and therefore, can be easily spun at temperatures sometimes even higher than 350C [32]. Having higher SP, the fossil pitch fila￾ments can be stabilized at temperatures at which cross￾linking reactions occur more rapidly. The another main reason is the large amount of low molar mass mol￾ecules present even in the polymerized samples of euca￾lyptus tar pitches, which correspond to higher retention time in the GPC curves of Fig. 3. During oxidative ther￾mal treatment, these molecules can easily acquire energy sufficient to provoke the coalescence of the filaments even at temperatures below the pitch SP. However, un￾der heating rates as low as 0.08C/min, these molecules were slowly eliminated either by sublimation or polymerization. In addition, Derbyshire et al. [4] reported that the maximum effective heating rate during oxidative stabil￾ization of pitch filaments decreases with increasing fiber diameter due to the limited oxygen diffusion. Therefore, the relatively large diameter of the as-spun filaments produced in the present work is possibly contributing to make stabilization difficult. Despite oxygen incorporation (Table 1), eucalyptus tar pitch filaments underwent pronounced weight loss and consequent shrinkage during stabilization (Table 2), which can be attributed to the release of volatiles, shear time (min) viscosity (cP) 1.E+04 1.E+05 1.E+06 0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 crude pitch 2 h/250 ºC 4 h/250 ºC Fig. 2. Viscosity as a function of shear time for crude pitch and pre-treated at 250C for 2 and 4 h. Tests were carried out at 101, 118, and 156C, respectively, in order to provide similar viscosities for all samples. 594 M.J. Prauchner et al. / Carbon 43 (2005) 591–597

M.J. Prauchner et al. Carbon 43(2005)591-597 3.2. Carbon fiber properties During carbonization, the filaments underwent add tional shrinkage and weight loss(Table 2), leading to a net weight yield from as-spun fibers to carbon fibers of approximately 43%. The carbon fibers produced had an average diameter of 27 2 um and an apparent den sity of 1.84 g/cm. SEM micrographs(Fig. 4)show that the fibers present a smooth and homogeneous surface As usual for filamentary materi strength (a)of the eucalyptus tar pitch-based carbon fi- bers presented a large dispersion degree, in the range 40-309 MPa. Therefore the results were treated by a sta tistical tool, the Weibull distribution [33], the most used method to describe the strength variability of materials retention time(min) with catastrophic failure such as carbon fibers [34] Fig 3. GPC curves of crude pitch and pre-treated at 250C for 2 and The curve obtained(Fig. 5)clearly demonstrates that here is a bimodal distribution of o values. Since the fail ure in fibrous materials is induced by flaws, this result uggests the existence of two classes of fiber defects Table I the first class would correspond to minor defects which Elemental composition of as-spun and stabilized fibers (0.08C/min; are inherent to the process and the precursor. These de- fects include pores and structure imperfections. Being C()H() N()0+ ashes"(%) H/C O/C more numerous, they would be uniformly distributed As-spun fiber 72.5 5.9 0.3 along all filaments, giving rise to a relatively high Wei- 0.700.37 bull modulus (o) for the region of fibers with higher Ash content is around 1% strength(the narrower the dispersion of the values, the higher the o value). The other class of defects would Table 2 correspond to major ones, which are attributed to the shrinkage during stabilization (0.08C/ formation of cracks in the fibers due to excessive weight min;l8o°C)and ion(20°/min;1000°C loss during the stabilization and carbonization steps Process step Linear These defects would give rise to filaments with tensile shrinkage strengths as low as 40 MPa. Furthermore, being present in a smaller number, they would be less uniformly dis- Stabilization tributed, thus leading to a very low o value in the region Carbonization corresponding to the fibers with low strength Total The average tensile strength was 129 MPa. This value From as-spun fiber to the end carbon fiber. is low if compared to those reported by Hayashi et al. [35] for coal tar pitch-based isotropic carbon fibers, as well as to the degradation of the abundant aliphatic about 600 MPa. Such difference can be attributed to side chains. After stabilization, the filaments were then two main reasons: the first one is the lower carbon yield carbonized to be converted into carbon fiber of eucalyptus tar pitches, which gives rise to extensive Fig. 4. SEM of eucalyptus tar pitch-based carbon fibers

as well as to the degradation of the abundant aliphatic side chains. After stabilization, the filaments were then carbonized to be converted into carbon fibers. 3.2. Carbon fiber properties During carbonization, the filaments underwent addi￾tional shrinkage and weight loss (Table 2), leading to a net weight yield from as-spun fibers to carbon fibers of approximately 43%. The carbon fibers produced had an average diameter of 27 ± 2lm and an apparent den￾sity of 1.84 g/cm3 . SEM micrographs (Fig. 4) show that the fibers present a smooth and homogeneous surface. As usual for filamentary materials, the tensile strength (r) of the eucalyptus tar pitch-based carbon fi- bers presented a large dispersion degree, in the range of 40–309MPa. Therefore, the results were treated by a sta￾tistical tool, the Weibull distribution [33], the most used method to describe the strength variability of materials with catastrophic failure such as carbon fibers [34]. The curve obtained (Fig. 5) clearly demonstrates that there is a bimodal distribution of r values. Since the fail￾ure in fibrous materials is induced by flaws, this result suggests the existence of two classes of fiber defects: the first class would correspond to minor defects which are inherent to the process and the precursor. These de￾fects include pores and structure imperfections. Being more numerous, they would be uniformly distributed along all filaments, giving rise to a relatively high Wei￾bull modulus (x) for the region of fibers with higher strength (the narrower the dispersion of the values, the higher the x value). The other class of defects would correspond to major ones, which are attributed to the formation of cracks in the fibers due to excessive weight loss during the stabilization and carbonization steps. These defects would give rise to filaments with tensile strengths as low as 40MPa. Furthermore, being present in a smaller number, they would be less uniformly dis￾tributed, thus leading to a very low x value in the region corresponding to the fibers with low strength. The average tensile strength was 129MPa. This value is low if compared to those reported by Hayashi et al. [35] for coal tar pitch-based isotropic carbon fibers, about 600MPa. Such difference can be attributed to two main reasons: the first one is the lower carbon yield of eucalyptus tar pitches, which gives rise to extensive Fig. 3. GPC curves of crude pitch and pre-treated at 250C for 2 and 4 h. Table 1 Elemental composition of as-spun and stabilized fibers (0.08C/min; 180C) Sample C (%) H (%) N (%) O + ashesa (%) H/C O/C As-spun fiber 72.5 5.9 0.3 20.1 0.98 0.21 Stabilized 63.3 3.7 0.2 31.6 0.70 0.37 a Ash content is around 1%. Table 2 Weight loss and dimensional shrinkage during stabilization (0.08C/ min; 180C) and carbonization (2.0C/min; 1000C) Process step Weight loss (%) Diameter reduction (%) Linear shrinkage (%) Stabilization 13 20 18 Carbonization 51 27 17 Totala 57 42 32 a From as-spun fiber to the end carbon fiber. Fig. 4. SEM of eucalyptus tar pitch-based carbon fibers. M.J. Prauchner et al. / Carbon 43 (2005) 591–597 595

M.. Prauchner et al. Carbon 43(2005)591-597 y=19+33 matter is important because it allows the use of the whole tar generated as a by-product during charcoal production, therefore stimulating volatile recovery in industrial chimneys and aggregating revenue to the charcoal-making industry, which is particularly interest ing because planted biomass is an environmentally friendly and renewable energy source Although the tensile properties make clear that the produced fibers are not useful as structural reinforce- ment, other properties give rise to different potential 31y=61+0.63X applications. For example, the reduced carbon yield of the precursor and the high isotropic and disordered structure of the fibers suggest a great potential for the production of activated carbon fibers, besides using them in the manufacture of thermal insulation carbon In o(MPa) felts. Efforts have been made to improve the process Fig. 5. Weibull distribution for tensile strength measurements of so far employed mainly aiming to reduce the period of eucalyptus tar pitch-based carbon fibers. time necessary to stabilize the fibers porosity in carbonized material and even to crack for- mation, as already discussed. In that respect, Derbyshire Acknowledgments et al. [4] demonstrated that fiber tensile strength in creases with increasing precursor carbon yield. The sec The authors thank Fundacao de amparo a Pesquisa ond reason is that the relatively large diameters of the do estado de minas gerais(FAPEMIG), Empresa Bra- fibers produced in the present work is certainly contrib- sileira de Pesquisa Agropecuaria(EMBRAPA) and uting to reduce the tensile strength because thicker fibers Fundacao de empreendimentos cientificos e Tecnolog contain more flaws per unit length compared to smaller icos(FINATEC)for financial support diameter fibers. Fitzer and KOnkele [36] have reported that larger diameters imply lower tensile strengths. In turn, Mora et al. [5] found that the tensile strength of References coal tar pitch-based isotropic carbon fibers varies from 143 MPa to 414 MPa when the fiber diameter vari Edie DD, Diefendorf rj. Carbon fiber manufacturing. In: from 35 um to 16um Buckley JD, Edie DD, editors. Carbon-carbon materials and The tensile modulus of the carbon fibers obtained was composites. Park Ridge, NJ: Noyes: 1993. p 18-30 [2]Savage G. In: Carbon-carbon composites. London: Chapman 14+3 GPa. This value is lower than the usual for Hal;1993.p.53-7 GPCF. For example, the values reported by Mora et 3 Edie DD. The effect of pr structure and proper al. [S] in the previously quoted paper fall in the range of carbon fibers. Carbon 1998: 36: 345-62 29-36GPa. Besides the flaws. this result reflects the mis- [4] Derbyshire F, Andrews R, Jacques D, Jagtoyen M, Kimber G oriented structure of the fibers derived from the three- Antell T. Synthesis of isotrop arbon fibers from pitch precursors. Fuel 2001; 80: 345-56 dimensional structure of eucalyptus tar pitch molecules 5 Mora E, Blanco C, Prada V, Santamaria R. Granda M As expected, eucalyptus tar pitch-based carbon fiber Menendez r. A study of pitch-based precursors for general oresented a moderately low electrical resistivity, rpose carbon fibres. Carbon 2002: 40: 2719-25 2 x 10-Q2m. However, this value is by far higher than [6 Alcaniz-monge J, Cazorla-Amoros D, Linares-solano A, Oya A usual for HPCF produced from mesophase pitch, typi Sakamoto A, Hoshi K. Preparation of general purpose carbon fibers from coal tar pitches with low softening point. Carbon cally in the order of 10Q2m, and even higher than that 1997:35:1079-87 reported by Fu et al. [8] for carbon fiber produced from [7] Donnet JB, Bansal RC. In: Carbon fibers. New York: Marcel an isotropic petroleum pitch, 3 x 10 Q2m. Once more, Dekker: 1990. p. 367-446 [ chapter 7) the result reflects the disordered. misoriented and flawed [8 Fu x, Lu w, Chung DDL. Ozone treatment of carbon fiber for structure of the fibers, which partially hinders electro conduction along the filaments [9] Roh YB Jeong KM, Cho HG, Kang HY, Lee YS. Ryu SK, et al. Unique charge/discharge properties of carbon materials with different structures. J Power Sources 1997: 68: 271-6 [10 Egashira M, Takatsuji H, Okada S, Yamaki J. Properties of 4. Conclusions activated carbon fiber fo electrode in lithium batteries. J Power Sources 2002- 107- 56-60 (1 Mochida l, Kisamori S, Hironaka M, Kawano S, Matsumura Y, The present work shows that eucalyptus tar pitches Yoshikawa M. Oxidation of no into no, over active carbon are potential precursors of isotropic carbon fibers. This fibers. Energy Fuels 1994: 8: 1341-4

porosity in carbonized material and even to crack for￾mation, as already discussed. In that respect, Derbyshire et al. [4] demonstrated that fiber tensile strength in￾creases with increasing precursor carbon yield. The sec￾ond reason is that the relatively large diameters of the fibers produced in the present work is certainly contrib￾uting to reduce the tensile strength because thicker fibers contain more flaws per unit length compared to smaller diameter fibers. Fitzer and Ko¨nkele [36] have reported that larger diameters imply lower tensile strengths. In turn, Mora et al. [5] found that the tensile strength of coal tar pitch-based isotropic carbon fibers varies from 143MPa to 414MPa when the fiber diameter varies from 35lm to 16lm. The tensile modulus of the carbon fibers obtained was 14 ± 3GPa. This value is lower than the usual for GPCF. For example, the values reported by Mora et al. [5] in the previously quoted paper fall in the range 29–36GPa. Besides the flaws, this result reflects the mis￾oriented structure of the fibers derived from the three￾dimensional structure of eucalyptus tar pitch molecules. As expected, eucalyptus tar pitch-based carbon fibers presented a moderately low electrical resistivity, 2 · 10
4 Xm. However, this value is by far higher than usual for HPCF produced from mesophase pitch, typi￾cally in the order of 10
6 Xm, and even higher than that reported by Fu et al. [8] for carbon fiber produced from an isotropic petroleum pitch, 3 · 10
5 Xm. Once more, the result reflects the disordered, misoriented and flawed structure of the fibers, which partially hinders electron conduction along the filaments. 4. Conclusions The present work shows that eucalyptus tar pitches are potential precursors of isotropic carbon fibers. This matter is important because it allows the use of the whole tar generated as a by-product during charcoal production, therefore stimulating volatile recovery in industrial chimneys and aggregating revenue to the charcoal-making industry, which is particularly interest￾ing because planted biomass is an environmentally friendly and renewable energy source. Although the tensile properties make clear that the produced fibers are not useful as structural reinforce￾ment, other properties give rise to different potential applications. For example, the reduced carbon yield of the precursor and the high isotropic and disordered structure of the fibers suggest a great potential for the production of activated carbon fibers, besides using them in the manufacture of thermal insulation carbon felts. Efforts have been made to improve the process so far employed mainly aiming to reduce the period of time necessary to stabilize the fibers. Acknowledgments The authors thank Fundac¸a˜o de Amparo a Pesquisa do Estado de Minas Gerais (FAPEMIG), Empresa Bra￾sileira de Pesquisa Agropecua´ria (EMBRAPA) and Fundac¸a˜o de Empreendimentos Cientı´ficos e Tecnolo´g￾icos (FINATEC) for financial support. References [1] Edie DD, Diefendorf RJ. Carbon fiber manufacturing. In: Buckley JD, Edie DD, editors. Carbon–carbon materials and composites. Park Ridge, NJ: Noyes; 1993. p. 18–39. [2] Savage G. In: Carbon–carbon composites. London: Chapman & Hall; 1993. p. 53–7. [3] Edie DD. The effect of processing on the structure and properties of carbon fibers. Carbon 1998;36:345–62. [4] Derbyshire F, Andrews R, Jacques D, Jagtoyen M, Kimber G, Rantell T. Synthesis of isotropic carbon fibers and activated carbon fibers from pitch precursors. Fuel 2001;80:345–56. [5] Mora E, Blanco C, Prada V, Santamarı´a R, Granda M, Mene´ndez R. A study of pitch-based precursors for general purpose carbon fibres. Carbon 2002;40:2719–25. [6] Alcan˜iz-monge J, Cazorla-Amoro´s D, Linares-solano A, Oya A, Sakamoto A, Hoshi K. Preparation of general purpose carbon fibers from coal tar pitches with low softening point. Carbon 1997;35:1079–87. [7] Donnet JB, Bansal RC. In: Carbon fibers. New York: Marcel Dekker; 1990. p. 367–446. [chapter 7]. [8] Fu X, Lu W, Chung DDL. Ozone treatment of carbon fiber for reinforcing cement. Carbon 1998;36:1337–45. [9] Roh YB, Jeong KM, Cho HG, Kang HY, Lee YS, Ryu SK, et al.. Unique charge/discharge properties of carbon materials with different structures. J Power Sources 1997;68:271–6. [10] Egashira M, Takatsuji H, Okada S, Yamaki J. Properties of containing Sn nanoparticles activated carbon fiber for a negative electrode in lithium batteries. J Power Sources 2002;107:56–60. [11] Mochida I, Kisamori S, Hironaka M, Kawano S, Matsumura Y, Yoshikawa M. Oxidation of NO into NO2 over active carbon fibers. Energy Fuels 1994;8:1341–4. 3.5 4.0 4.5 5.0 5.5 -4 -3 -2 -1 0 y = -19 + 3.3 x y = -6.1 + 0.63 x ln{ln[1/(1-P)]} ln [σ (MPa)] Fig. 5. Weibull distribution for tensile strength measurements of eucalyptus tar pitch-based carbon fibers. 596 M.J. Prauchner et al. / Carbon 43 (2005) 591–597

[12]Jin H, Park S-E, Lee JM, Ryu SK. The shape selectivity of [24] Prauchner MJ, Pasa VMD, Otani C, Otani S, Pardini LC. activated carbon fibers as a palladium catalyst support Carbon lohallen NDS Structural evolution of eucalyptus tar pitch-based 96:34:429-3 carbons during carbonization. Biomass Bioenergy 2005: 28: 53-61 [3]de la Casa-Lillo MA, Alcaniz-Monge J, Raymundo- Pinero E, [25 Prauchner MJ, Pasa VMD, Otani C, Otani S. Rheological study Cazorla-Amoros D, Linares-Solano A. Molecular sieve properties of eucalyptus tar pitches. J Appl Polym Sci 2002: 84: 900-8 carbon fibres. Carbon 1998: 36: 1353--60 26 Prauchner MJ. Sintese e cara arbonosos [14 Lozano-Castello D, Alcaniz- Monge J, de la Casa-Lillo MA, avanca a partir de piche de alcatrao de Cazorla-Amoros D, Linares-Solano A. Advances in the study of Thesis. Federal University of Minas Gerais, Belo Horizonte methane storage in porous carbonaceous materials. Fuel 2001.p.826. 281:1777-803. 27 Alcaniz-Monge J, Cazorla-Amoros D, Linares-Solano A. Effect of [15]Li CY, Wan YZ, Wang J, Wang YL, Jiang XQ, Han LM. Antibacterial pitch-based activated carbon fiber supporting silver based carbon fibres. Carbon 1994: 32- 1277-83 Carbon l998:36:61-5 28 Thwaites MW, Stewart ML, McNeese BE, Sumner MB. Synthesis [16 Ryu sk, Kim SY, Li z. Jaroniec M. Characterization of silver- and characterization of activated pitch-based carbon fibers. Fue ontaining pitch-based activated carbon fibers. J Colloid Interf Sci Process Technol 1993: 34: 137-45 [29] Bansal RC, Donnet J-B, Stoeckli F. In arbon. New [17 Navarri P, Marchal D, Ginestet A. Activated carbon fiber York: Marcel Dekker, 1988. p. 1-26.[cl materials for VOC removal. Filtrat Separat 2001: 38: 33-40 30] Matsumoto T, Mochida I. A structur on oxidative [18 Carazza F, Okuma AA. Alcatrao de Eucalyptus fonte de aromas e stabilization of mesophase pitch fibers sabores. Extended abstracts XI Seminario Carboquimicos. San Carbon1992;30:1041-6 tos. brazil: asao Brasileira deMetalurgia: 2000. P. 271-9 31] Matsumoto T, Mochida I. Oxygen distribution in oxidatively [19 Chen CA, Pakdel H, Roy C. Separation of phenols from stabilized mesophase pitch fiber Carbon 1993: 31: 143-7. Eucalyptus wood tar Biomass Bioenergy 1997: 13: 25-37 32] Edie DD, Dunham MG. Melt spinning pitch-based carbon fibers. o Statistical Yearbook. Associacao Brasileira de Carvao Vegetal, Carbon198927:647-55 Belo horizonte. brasil. 2000. 33] Weibull w. A statistical distribution function of wide applicabil- 21 Sampaio RS. Producao de metais com biomassa plantada. In: ty. J Appl Mech 1951: 18: 293-7 Mello MG, editor. Biomassa: energia dos tropicos em Minas 34] Stoner EG, Edie DD, Durham SD. An end-effect model for the Gerais. Belo Horizonte- Brasil: Editora UFMG: 2001. p. 163-78 single-filament tensile test. J Mater Sci 1994: 29: 6561-74 [22] Prauchner MJ. Pasa VMD, Otani C, Otani S. Characterization [35] Hayashi J-l, Nakashima M, Kusakabe K, Morooka S, Mitsuda s of eucalyptus tar pitches Rapid stabilization of pitch fiber precursor by multi-step thermal oxidation. Carbon 1995: 33: 1567-71 23 Prauchner MJ, Pasa VMD, Otani C, Otani S, Menezes SMC. [36 Fietzer E, KOnkele F. Today's carbon fibres-a new energy-saving Eucalyptus tar pitch pre-treatment for carbon material processing. and environment-friendly all-round material (a review). High J Appl Polym Sci 2004: 91: 1604-lI Temp High Press 1999: 22: 239-66

[12] Jin H, Park S-E, Lee JM, Ryu SK. The shape selectivity of activated carbon fibers as a palladium catalyst support. Carbon 1996;34:429–31. [13] de la Casa-Lillo MA, Alcan˜iz-Monge J, Raymundo-Pin˜ero E, Cazorla-Amoro´s D, Linares-Solano A. Molecular sieve properties of general-purpose carbon fibres. Carbon 1998;36:1353–60. [14] Lozano-Castello´ D, Alcan˜iz-Monge J, de la Casa-Lillo MA, Cazorla-Amoro´s D, Linares-Solano A. Advances in the study of methane storage in porous carbonaceous materials. Fuel 2002;81:1777–803. [15] Li CY, Wan YZ, Wang J, Wang YL, Jiang XQ, Han LM. Antibacterial pitch-based activated carbon fiber supporting silver. Carbon 1998;36:61–5. [16] Ryu SK, Kim SY, Li ZJ, Jaroniec M. Characterization of silver￾containing pitch-based activated carbon fibers. J Colloid Interf Sci 1999;220:157–62. [17] Navarri P, Marchal D, Ginestet A. Activated carbon fiber materials for VOC removal. Filtrat Separat 2001;38:33–40. [18] Carazza F, Okuma AA. Alcatra˜o de Eucalyptus fonte de aromas e sabores. Extended abstracts XI Semina´rio Carboquı´micos. San￾tos, Brazil: Associac¸a˜o Brasileira deMetalurgia; 2000. p. 271–9. [19] Chen CA, Pakdel H, Roy C. Separation of phenols from Eucalyptus wood tar. Biomass Bioenergy 1997;13:25–37. [20] Statistical Yearbook. Associac¸a˜o Brasileira de Carva˜o Vegetal, Belo Horizonte, Brasil, 2000. [21] Sampaio RS. Produc¸a˜o de metais com biomassa plantada. In: Mello MG, editor. Biomassa: energia dos tro´picos em Minas Gerais. Belo Horizonte-Brasil: Editora UFMG; 2001. p. 163–78. [22] Prauchner MJ, Pasa VMD, Otani C, Otani S. Characterization and thermal polymerization of eucalyptus tar pitches. Energy Fuels 2001;15:449–54. [23] Prauchner MJ, Pasa VMD, Otani C, Otani S, Menezes SMC. Eucalyptus tar pitch pre-treatment for carbon material processing. J Appl Polym Sci 2004;91:1604–11. [24] Prauchner MJ, Pasa VMD, Otani C, Otani S, Pardini LC, Mohallen NDS. Structural evolution of eucalyptus tar pitch-based carbons during carbonization. Biomass Bioenergy 2005;28:53–61. [25] Prauchner MJ, Pasa VMD, Otani C, Otani S. Rheological study of eucalyptus tar pitches. J Appl Polym Sci 2002;84:900–8. [26] Prauchner MJ. Sı´ntese e caracterizac¸a˜o de materiais carbonosos avanc¸ados a partir de piches de alcatra˜o de eucalyptus. PhD Thesis. Federal University of Minas Gerais, Belo Horizonte Brazil; 2001. p. 82–6. [27] Alcan˜iz-Monge J, Cazorla-Amoro´s D, Linares-Solano A. Effect of the activating gas on tensile strength and pore structure of pitch￾based carbon fibres. Carbon 1994;32:1277–83. [28] Thwaites MW, Stewart ML, McNeese BE, Sumner MB. Synthesis and characterization of activated pitch-based carbon fibers. Fuel Process Technol 1993;34:137–45. [29] Bansal RC, Donnet J-B, Stoeckli F. In: Active carbon. New York: Marcel Dekker; 1988. p. 1–26. [chapter 1]. [30] Matsumoto T, Mochida I. A structural study on oxidative stabilization of mesophase pitch fibers derived from coal tar. Carbon 1992;30:1041–6. [31] Matsumoto T, Mochida I. Oxygen distribution in oxidatively stabilized mesophase pitch fiber. Carbon 1993;31:143–7. [32] Edie DD, Dunham MG. Melt spinning pitch-based carbon fibers. Carbon 1989;27:647–55. [33] Weibull W. A statistical distribution function of wide applicabil￾ity. J Appl Mech 1951;18:293–7. [34] Stoner EG, Edie DD, Durham SD. An end-effect model for the single-filament tensile test. J Mater Sci 1994;29:6561–74. [35] Hayashi J-I, Nakashima M, Kusakabe K, Morooka S, Mitsuda S. Rapid stabilization of pitch fiber precursor by multi-step thermal oxidation. Carbon 1995;33:1567–71. [36] Fietzer E, Ku¨nkele F. Today
s carbon fibres—a new energy-saving and environment-friendly all-round material (a review). High Temp High Press 1999;22:239–66. M.J. Prauchner et al. / Carbon 43 (2005) 591–597 597