1.

3 Pitch Precursor-Based Carbon Fibers

Victor Bermudez, Sam Lukubira, and Amod A Ogale, Clemson University, Clemson, SC, United States
r 2018 Elsevier Ltd. All rights reserved.

1.3.1 Introduction 41
1.3.2 Pitch Precursors 42
[Link] Introduction 42
[Link] Pitch Production 43
[Link] Chemical Composition of Pitch Precursors 44
[Link].1 Coal tar pitch 44
[Link].2 Petroleum pitch 44
[Link].3 Synthetic pitch 45
[Link].4 Other pitches 45
1.3.3 Fiber Spinning and Processing 45
[Link] Rheology 45
[Link] Melt Spinning of Pitch 47
[Link].1 Extrusion and roll drawing 47
[Link].2 Melt blowing 47
[Link].3 Microstructure during spinning 47
[Link] Thermosetting/Stabilization 49
[Link] Carbonization and Graphitization 50
1.3.4 Microstructure and Properties of Pitch-Based CFs 50
[Link] Thermo-Physical Properties 51
[Link].1 Density 51
[Link].2 Coefficient of thermal expansion 51
[Link].3 Oxidation 51
[Link] Microstructure 51
[Link].1 X-ray diffraction 52
[Link].2 SEM and TEM 57
[Link].3 Raman spectroscopy 58
[Link] Mechanical Properties 58
[Link].1 Modulus 60
[Link].2 Strength 60
[Link] Transport Properties 62
1.3.5 Concluding Remarks 63
References 63
Further Reading 65
Relevant Websites 65

1.3.1 Introduction

Carbon fibers (CFs) are among the strongest and stiffest fiber reinforcements available commercially for composite materials.1 CFs
were developed in the 1960s for structural applications and their production and enhancement of properties took place over the
next three decades. Several commercial grades possess a tensile strength of over 5 GPa; coupled with their light weight
(densityE1800 kg m
3), CFs are almost five times stronger than steel, and are being used in ultrahigh-performance aerospace and
defense applications.2 Their outstanding strength-to-weight ratio makes CFs ideal reinforcements for military and civilian aircraft
applications. Components of F-16 and Stealth bombers have been made of carbon–epoxy composites for almost 25 years now.
Within the past decade, Boeing 787 “Dreamliner” commercial airplane has been designed to use about 50 wt% of CF composites
for its primary structure, including the fuselage and wings.3
In addition to strength and stiffness, other desirable CF characteristics are high dimensional stability at high/low temperatures
for aircraft brakes and low thermal expansion coefficient such as for outer space applications. Other properties and applications
include good vibration damping for audio equipment, high modulus and stiffness for robotic arms, and high electrical con-
ductivity for electromagnetic interference (EMI) and RF shielding. Because carbon does not melt or dissolve in any solvents, CFs
are ideal for chemical inertness, high corrosion resistance, and thermal resistance against melt-down for chemical and nuclear
industries.3,4

Comprehensive Composite Materials II, Volume 1 doi:10.1016/B978-0-12-803581-8.10312-1 41

42 Pitch Precursor-Based Carbon Fibers

Significant efforts have been devoted to improving CF production processes that have led to not only improvement of
properties but also a reduction in their cost since their initial commercialization 40 years ago. Thus, CFs are now being used in
industrial components, nuclear reactor components, rehabilitation of bridges and buildings,5 high performance sporting goods,
and packaging of electronics.6 However, CFs remain significantly more expensive relative to other reinforcing fibers, such as glass
fibers, when compared on performance/cost basis. In land-based applications, the high cost of CFs (US$10 to 1500 per pound),
somewhat limits its use to high-performance sporting goods (golf clubs and tennis rackets), and racing cars and high-performance
automobiles such as BMW electric vehicles “E-series”.7 The majority of CFs, around 75%, are used in the industrial market which
includes the automotive, pressure vessels and wind turbine sectors. The remainder is distributed between the aerospace market
(14%) and consumer market (11%). The global CF demand in 2016 was estimated at 101,600 metric tons, and it is expected to
grow over the next 9 years with a CAGR of around 10%, from 181,900 metric tons in 2021 to 262,400 in 2025. The global CF
supply in 2016 was estimated in 102,130 metric tons, with the expectation to reach 153,720 metric tons by 2021.8
All CFs are produced from hydrocarbon-type precursors because carbon in its elemental form does not melt or dissolve in any
solvent. The melt- or solution-processable precursor materials are first spun into thin precursor fibers (E10 mm), which are then
converted into CFs. Thus, the precursor material should be spinnable and have sufficient reactivity to be cross-linked (also referred
to as “stabilized”) without fibers sticking to each other. Finally, the precursor fibers should have adequate carbon yield when
subjected to thermal treatment at temperatures ranging from 1000 to 30001C to produce CFs. Because of these constraints, CFs
have been produced commercially only from three types of precursors: (i) rayon, (ii) polyacrylonitrile (PAN), and (iii) pitch.
Regenerated cellulose in its continuous fiber form, rayon, was used in the earliest CF development. In fact, Edison’s light bulb
filament was cellulose-based CF. However, tensile strength and modulus are quite low, as is the thermal conductivity due to the
significant imperfection in carbon structure thus formed. However, because of their low thermal conductivity and excellent heat
resistance, these CFs are used in missile/rocket exhaust nozzles where use temperature exceeds 10001C. The limited strength and
environmental issues of producing rayon have limited further development of rayon-based CFs.
PAN precursor has led to CFs that possess outstanding mechanical strength, and are used almost exclusively for high perfor-
mance structural composites.9 The inability of PAN to melt, but to dissolve in solvent, makes it a unique precursor. Thus, PAN
dissolved in a solvent (namely, dimethyl sulfoxide or dimethylformamide) is wet-spun into precursor fibers that can be thermo-
oxidatively stabilized at 200–3001C in air. An intractable ring structure is formed from carbon–nitrogen double bonds, often know
as a “ladder” structure.10–12 This is considered the most critical step during the production of CFs because it determines the final
structure and mechanical properties of the resulting CFs.12,13 Subsequently, the stabilized PAN fibers are carbonized at 1000-
–15001C to provide a carbon yield of about 45 wt%. Thus, a large fraction of precursor mass (over 55%) is lost during carbo-
nization as reaction products that include toxic gases such as hydrogen cyanide and carbon monoxide. Also, tension needs to be
applied on the fiber tows throughout the spinning, stabilization and carbonization steps to retain molecular/crystallite orientation
that provide PAN-based CFs outstanding tensile strength of up to 7 GPa.14 A detailed discussion of PAN-based CFs is outside the
scope of this chapter, and the reader is referred to numerous excellent studies have already been reported in the literature.15–17
This chapter focuses on CFs derived from pitch precursors. As noted briefly above, the polymer cost associated with PAN
precursor limits how low its cost can get, as over half of the precursor mass is lost during carbonization. However, a large fraction
of the increased demand for CFs is anticipated in the industrial sector that is cost-sensitive, such as automotive application.8,18 For
cost-effective products, there is a need for the development of novel precursors and appropriate processes to reduce the CF
production cost and significantly expand their use. A very inexpensive source of carbon precursor is pitch, formed from the
residues of petroleum and coal tar refining. By some estimates, pitch cost can be lowered below US$2/lb.19,20 The low precursor
cost coupled with a high carbon yield (over 75 wt%) makes it potentially attractive for low-cost, moderate-performance CFs.
Therefore, pitch-based CFs are the focus of this chapter, which is a revised version of the original chapter by Diefendorf21 that
readers should refer to for an excellent discussion of chemical and physical structure of graphite and the highly graphitic CFs. This
revision focusses on recently available pitches and processing studies reported in the literature. Readers are also referred to other
excellent publications for a more detailed discussion of specific topics related to such CFs.1,22–25

1.3.2 Pitch Precursors

[Link] Introduction
While studying the carbonization of polyvinyl chloride (PVC), Otani discovered that a carbonaceous pitch was formed as a
pyrolysis product at E4001C and that this pitch could be melted and processed into carbon filaments.26,27 In 1970, pitch-based
CFs were produced industrially by Kureha Chemical Industry. The pitch used for such CFs (KRECA) was isotropic and derived from
petroleum residues.28 Such general purpose CFs have axial tensile strengths up to 700 MPa, axial moduli up to 70 GPa, and strain-
to-failure of about 1%.15 Such CFs find use in CF felts for insulation in high temperature furnaces working under inert atmosphere,
such as sintering furnaces, heat treatment furnaces, and furnaces that manufacture multicrystalline silicon ingots, sapphire ingots
and optical fiber preforms.29 Isotropic pitch-based CFs are also chopped or milled for compounding with synthetic rubber and
various plastics that are utilized in disk brake pads, gaskets for gas sealing and seal rings.29
Spheres of mesophase or liquid crystal were first observed to form from an isotropic pitch upon thermal treatment by Brooks
and Taylor.30 Their finding was complemented by White et al.,31 who reported that the mesophase domains could be subjected to
 Pitch Precursor-Based Carbon Fibers 43

deformation by flow, which remained after high temperature treatment. In the early 1970s, Singer and co-workers at Union
Carbide Corporation made use of those and other studies to show that a mesophase pitch (MP) could be melt spun into fibers,
which in turn could be oxidized and carbonized to obtain CFs of high modulus and high thermal conductivity.32 These studies led
to the Union Carbide process for MP-based CF production. The mesophase-pitch derived CFs have axial tensile strengths up to 4
GPa, axial moduli up to 900 GPa, and axial thermal conductivity approaching 1000 W∙m
1∙K
1 depending on the maximum
graphitization temperature used.15 Some applications of such high performance MP-based CFs include aerospace and industrial
products such as structural materials for antennas of large radio telescopes, printing rollers, and carbon matrix composites, owing
to their low longitudinal CTE, light weight and high stiffness.33

[Link] Pitch Production

As briefly noted above, pitch precursors can generally be classified as isotropic or anisotropic. The anisotropy of pitches emanates
from the disk-like architecture of a certain category of pitch molecules that results in the layering of such discotic molecules, which
leads to liquid crystalline behavior. Such pitches are also called “mesophase pitch” to identify its state as being somewhere
between a liquid and a crystalline solid because these materials possess orientational order, but not positional order. CFs are
produced from both groups of pitches with isotropic pitch leading to general performance CFs (GPCFs) whereas the latter group
produces high performance carbon fibers (HPCFs).
A schematic for the processing of pitch-precursor CFs is displayed in Fig. 1. HPCFs (derived from MP) have a higher production
cost due to the additional processing steps needed to produce the mesophase from isotropic pitch, which limits their use to
high-end applications where performance is the critical factor, such as aerospace. The lower production cost of GPCF (isotropic
pitch-based) makes them more suitable for applications where mechanical properties are not the most important factor, such as
concrete reinforcement,34 electrodes for energy storage systems,35 and activated CFs.36 Main producers of pitch-precursor CFs (sold
under trade names Dialead, Kreca, Thornel, Granoc) are Mitsubishi Rayon Co., Kureha Chemical Industries, Cytec Industries, Inc.
and Nippon Graphite Fiber Corporation. These four companies had a combined annual capacity of approximately 3000 t in
2010.37,38

Fig. 1 Manufacturing processes for pitch precursors and resulting carbon fibers (CFs). Adapted from Morgan, P., 2005. Carbon Fibers and Their
Composites. Boca Raton, FL: CRC Press.
44 Pitch Precursor-Based Carbon Fibers

[Link] Chemical Composition of Pitch Precursors

Overall, pitch is a mixture of polycyclic aromatic hydrocarbons (PAHs) with alkyl side-chains and heterocyclic compounds, and is
a black and shiny solid at ambient conditions. Although there are natural deposits of pitch in some regions of the world (Lake
Guanoco in Venezuela and Pitch Lake in Trinidad and Tobago), most pitch is produced from the refining of crude oil, known as
petroleum pitch, or from the pyrolysis of coal, known as coal tar pitch. In general, industrial pitches require some pretreatment to
make them suitable for CF production. Both isotropic and anisotropic pitches can be produced from petroleum pitch and coal tar
pitch.21
Typical pitches consist of four fractions: (1) aliphatic saturated components of low molecular weight; (2) naphthalene aromatic
compounds of low molecular weight and saturated ring structures; (3) polar aromatic compounds with heterocyclic structures; and
(4) asphaltenes of high molecular weight, high aromaticity, and more thermal stability. A high proportion of the latter fraction is
known to characterize the most suitable pitches for CF processing.17
Regardless of its raw material, a pitch precursor for either isotropic or anisotropic pitch-based CFs must provide a high carbon
yield, a low volatile content, a narrow molecular weight distribution, an adequate viscosity at the spinning temperature, and a low
concentration of inorganic impurities.39 Additionally, the softening point of the pitch must be appropriate to balance good
spinning performance (favored by a low softening point) versus a short stabilization step (favored by a high softening point).
Therefore, raw pitches, whether petroleum- or coal tar-based, require an adjustment by chemical reaction and/or separation of
components to achieve the desired properties. These industrial pitches typically contain some solid impurity and quinoline
insolubles (QI), which need to be removed in order to obtain either isotropic pitch or MP, otherwise coalescence of mesophase
spheres is impeded and the formation of a continuous, spinnable anisotropic domain is prevented, and at the same time a
homogenous isotropic pitch is neither formed.15

[Link].1 Coal tar pitch

Coal tar pitch is defined as the residue (nonvolatile at 4501C) from the distillation of coal tars produced by the high temperature
pyrolysis (410001C) of coal.40 The molecular structure of coal tar pitches usually consists of an aromatic core with only a few
methyl groups attached. In that sense, coal tar pitch is more aromatic and has a smaller degree of substitution than does petroleum
pitch. This makes coal tar pitches more thermally stable than natural pitches because bubbling, caused by the separation of alkyl
chains commonly present in natural pitches, is not extensive.21
Typical methods to modify coal tar pitch include thermal treatment, solvent extraction, or catalytic reactions. Mora et al.39
modified a coal tar pitch by thermal treatment at 4301C under nitrogen, followed by hot filtration to separate the isotropic phase
from the mesophase. The isotropic pitch was further extracted with solvents to adjust its softening point; the most suitable pitch
precursor for CFs was obtained with a mixture of 40% acetone–60% acetonitrile. Much larger fractions of polynuclear aromatics
make coal tar pitches less attractive than petroleum-derived pitches, since the increased thermal stability given by the larger
aromaticity results in inferior chemical reactivity which is needed in the stabilization step.

[Link].2 Petroleum pitch

Typical by-products resulting from petroleum distillation are catalytic cracker bottoms, pyrolysis fuel oil and catalytic slurry oil and
steam cracker tar, which can be used to produce pitch.17,41,42 The production of pitch from these materials can include one of the
following methods or a combination of several of them: (1) further heat treatment to increase molecular weight; (2) air blowing at
around 2501C; (3) steam stripping and vacuum to remove the more volatile components; and (4) distillation. Physical and
chemical properties of the resulting pitch are dependent on the process conditions used in any of the four methods above. Process
temperature and process time play an important role: higher temperatures and longer times typically lead to a more aromatic and
more anisotropic pitch.17
Li et al.43 turned catalytic slurry oil into a pitch precursor by thermal polymerization followed by centrifugal separation, solvent
extraction and vacuum distillation, and obtained CFs from it. The thermal polymerization time of the slurry oil proved to have an
important effect on CF microstructure and CF properties. Although average property values of CFs obtained through 7 and 12 h of
thermal polymerization met the criteria for GPCF, the latter showed a smoother microstructure and more uniform mechanical
properties.
The formation of isotropic pitch from pyrolysis fuel oil and slurry oil under the same processing conditions, a two-step
distillation without using additives, was investigated by Kim et al.,42 finding that pyrolysis fuel oil led to a more isotropic precursor
while the slurry oil-based precursor showed some mesophase development. This partial development mesophase is thought to
have had a negative effect on spinnability, considering that the spinning of the slurry oil-based precursor was difficult.
MP is produced from isotropic petroleum pitches that have a softening point of only about 1201C. The isotropic pitch is
thermally polymerized using process temperatures greater than 3801C. The reactants are agitated at high rates and stripped with an
inert gas until the softening point is in the desired range of 275–3251C and the desired level of mesophase content is reached.
Typical reaction times range from about 12 to 18 h. MP can be produced from isotropic petroleum pitches in batches of about
1000 kg using standard processing equipment and multiple reactors. The large batch processing can produce MP at about 25 t per
year, if operated commercially. Based on such a large scale production facility, it is estimated that MP can be produced for less than
US$5 per kg and moderate performance CFs can be manufactured for approximately US$10 per kg.44
 Pitch Precursor-Based Carbon Fibers 45

Fig. 2 Matrix-assisted laser desorption/ionization (MALDI) spectra of AR– synthetic mesophase pitch (MP) (left) and petroleum-based MP (right)
with intensity represented in arbitrary units. Reproduced with permission from Jeon, Y.-P., Zhang, M., Levan, C.D., Ogale, A.A., 2012. High thermal
conductivity carbon fibers from petroleum-based experimental mesophase pitch. In: Annual World Conference on Carbon 2012, Krakow, Poland.

[Link].3 Synthetic pitch

Given the source of the crude oil or coal, petroleum residues and coal tars have not only a wide range of chemical structures, but also
have a highly variable content of these species. This makes it difficult to control the structure of the isotropic or anisotropic pitch to
get the desired balance between low softening point and high stabilization reactivity. With the aim of obtaining a more controlled
precursor of higher purity, Mochida et al.45 developed a synthetic pitch from aromatic hydrocarbons, such as naphthalene and
anthracene using HF/BF3 as a catalyst. The process consisted of catalytic polymerization in the range 220–2601C under autogenous
pressure for 4 h. The catalyst, unreacted monomer and volatiles were distilled out of the reaction product by heating to 3401C under
nitrogen flow to obtain 100% anisotropic pitch. Synthetic pitch represents an attractive precursor for MP-based CFs, given their
adequate balance between softening point and reactivity, and their high carbon yield of over 75%. Mitsubishi Gas Chemical Co. has
commercialized a naphthalene-based, synthetic pitch under the AR brand, with a softening point in the 275–2951C range.33
MPs are very difficult to characterize because they do not dissolve in solvents. Nominal molecular weight information may be
obtained from matrix-assisted laser desorption/ionization (MALDI) spectroscopy. For the spectra displayed in Fig. 2, MP was
dispersed in a matrix consisting of 7,7,8,8-tetracyanoquinodimethane (TCNQ). Matrix-to-analyte mass ratio was 20 to 1, and all
signals below 210 Da were suppressed. It is interesting to note that synthetic pitches derived from thermal polymerization of
naphthalene display an MW distribution with a maximum at about 500 Da. In contrast, petroleum-derived MP displays bimodal
distribution with a large peak at about 500 Da, but a smaller one at 750 Da corresponding to nominally dimer, and trimer
homologs with the lowest molecular “monomer” response at 250 Da likely muted.46,47

[Link].4 Other pitches

Tars obtained from the condensation of volatiles evolved during slow pyrolysis of wood have been considered as a raw material for
pitch-precursor GPCF. For instance, eucalyptus tar was distilled to obtain a crude isotropic pitch.48 Yang et al. prepared an isotropic
pitch from a mixture of biotar and ethylene bottom oil by two separate methods: (1) conventional atmospheric distillation and
(2) bromination at 1101C followed by dehydrobromination at 3201C under nitrogen. The addition of ethylene bottom oil to
biotar was necessary to decrease the relatively high oxygen content of biotar and prevent the oxygen functional groups from
causing crosslinking. After removal of volatiles by thin layer evaporation at 3501C, a spinnable isotropic pitch with a softening
point of 2051C was obtained. When comparing CFs of the same diameter, 7 mm, it was found that CFs processed from isotropic
pitch obtained by bromination–dehydrobromination showed a higher tensile strength than CFs processed from isotropic pitch
obtained by conventional atmospheric distillation (1200 MPa vs. 800 MPa).49 To date, the mechanical properties of such
biopitch-based CFs are too low for primary structural reinforcement. Rather, such CFs have potential applications as activated CFs
and thermal insulating carbon felts.48

1.3.3 Fiber Spinning and Processing

[Link] Rheology
MP is a thermoplastic liquid crystalline material that softens and flows at elevated temperatures usually in the range of 250–3501C
and hence it is melt spinnable. A successful melt spinning depends on the rheological properties of molten pitch, which are a
strong function of temperature.50 Several rheological studies51–53 have correlated the development of texture in MP CFs to the
46 Pitch Precursor-Based Carbon Fibers

Fig. 3 Steady viscosity of mesophase pitch (MP) from rate-sweep experiments at different temperatures for (a) increasing shear rates and (b)
decreasing shear rates. Dotted lines represent least-squares fit. Reproduced with permission from Kundu, S., Ogale, A.A., 2006. Rheostructural
studies on a synthetic mesophase pitch during transient shear flow. Carbon 44(11), 2224–2235.

shear rates and temperature during spinning. The rheological response of pitch depends on the chemical composition of the pitch.
For naphthalene derived pitches, a three region steady viscosity-shear behavior has been observed at 2751C with an initial shear-
thinning region at shear rates below 2000 s
1 followed by a Newtonian plateau region and shear thinning again at shear rates
above 7000 s
1.52
However, the rheological response is a strong function of the initial texture of the liquid crystalline melt, i.e., measured viscosity
values can vary greatly depending on the state of the mesophase texture. Steady shear studies on thermally stable, AR-HP synthetic
MP indicated a two-region viscosity-shear pattern for shear rates between 0.1 and 10 s
1 over a range of temperatures (290–3051C)
typically used for melt spinning. A shear-thinning viscosity response from 0.1 to 1 s
1 was followed by a Newtonian plateau
(Fig. 3(a)) for increasing shear rates. Interestingly, the viscosity values were lower when subsequent experiments were conducted
for decreasing shear rates (Fig. 3(b)), indicating a change in the texture of the melt. Studies at very high shear rates (up to
10,000 s
1) did not show an ultimate shear-thinning Region III as a function of shear rates measured from various capillaries
attached to a single-screw extruder, Fig. 4.54 In Fig. 5, the transient shear tests revealed that the shear stress displayed a local
maximum and a minimum before reaching steady state; this phenomenon was observed at all temperatures and shear rates tested
in this study. The maxima and minima appeared at smaller strain units with increasing temperature, indicating accelerated kinetics
of textural transformation. However, the extent of maxima and minima decreased with increasing temperature, due likely to the
weaker intermolecular forces at high temperature. The absence of maximum/minimum during flow reversal following steady flow,
Fig. 5(b), confirmed that the local maximum/minimum response during flow startup can be attributed to textural changes in the
discotic liquid crystalline fluid. A lack of Region III shear-thinning response is also not surprising given that MP consists of small,
disk-like molecules, unlike long-chain polymers that typically display shear thinning due to molecular chain extension. It is
believed that the breakdown of liquid crystalline texture results in Region I shear thinning.
 Pitch Precursor-Based Carbon Fibers 47

Fig. 4 Viscosities as a function of shear rates measured from a cone–plate rheometer and a single-screw extruder. In the extruder, measurement
was performed for capillaries with two different L/D ratios, 3 and 20, respectively. Reproduced with permission from Kundu, S., Ogale, A.A., 2010.
Rheostructural studies of a discotic mesophase pitch at processing flow conditions. Rheologica Acta 49, 845–854.

[Link] Melt Spinning of Pitch

The first step in pitch-based CF production is the formation of pitch-precursor fibers, which are then thermoset/stabilized,
carbonized and finally graphitized, as displayed in Fig. 6. All commercial pitch-precursor fibers are produced by melt spinning.
High quality pitch, free of particles and ash content well below 1000 ppm is required for melt spinning.55

[Link].1 Extrusion and roll drawing

The typical melt process utilizes a combined extruder and roll draw down set up. In the extruder, pitch chips are fed, melted and
metered through a spin pack consisting of a filter and a spinneret with a series of capillaries (0.2–0.05 mm in diameter) where
extrudates are formed. The molten extrudate is drawn down by taking up on rolls and quenched by surrounding air to form thin
pitch fibers approximately 12–30 mm in diameter.

[Link].2 Melt blowing

An alternative to roll drawing is the melt blowing of the extrudate using a specialized spinneret that has orifices for hot air. The
melt blowing spinneret is composed of a V-slot through which hot air is supplied encompassing one to several hundreds of orifices
from which the fibers originate (Fig. 7). As the molten extrudates exit the orifice, the drag force exerted by the air draws down the
melt extrudate into fibers, which are collected a few feet on a screen or moving belt.56 A 1999 US patent to Mishimura57 describes
the melt blowing process for MP. In their example, a MP (density ¼1250 kg m
3) was spun through a nozzle of a 3-mm-wide slit
comprising alternate lines of 500 spinning orifices each of 0.2 mm diameter at 3601C. Hot air injected through the slit at 120 kPa
was used to draw down the molten pitch into fibers of an average diameter of 13 mm.

[Link].3 Microstructure during spinning

Due to the high rate of fiber production, melt spinning is a preferred process over wet and dry spinning, but it is a complex
process due to the quasi-stable chemical nature of MP. The stability problems stem from the fact the MP viscosity is highly
temperature dependent, as displayed in Fig. 8. Further, as-spun MP fibers are very brittle and weak with a tensile strength of only
0.04 GPa.55,58
Mesophase melt spinning is also extremely sensitive to small changes in process conditions that can alter fiber microstructure.
Studies have shown that pitch CF microstructure is developed during spinning58 and the subsequent conversion steps only refine
the microstructure. Microstructure is largely influenced by spinneret geometry, pre-spinneret flow deformation and spinning
conditions (especially temperature). Therefore, numerous researchers have produced pitch fiber microstructures, each with certain
advantages by adjusting the above parameters. For example, Hamada et al.59 produced pitch fiber microstructures ranging from
random, onion, quasi-onion by disrupting the flow prior to extrusion through the capillary by stirring, whereas Edie et al.58 used
noncircular capillary cross-section configuration to produce line-centered ribbon microstructure. For many applications, circular
cross-section fibers are desirable, although for some other specialized applications other shapes like ribbon and star-shaped cross-
sections offer better performance. Ribbon cross-sections result in high degree of graphitization, which enhances lattice-dependent
properties of electrical and thermal conductivity along fiber axis.
As mentioned earlier, spinning temperature has been shown to play an important role in transverse microstructure evolution of
pitch-based CFs as illustrated in Fig. 8.55 Radial (r) texture forms at lower temperatures due to a higher viscosity that results in
higher shear stresses. The radial texture changes to random (m), or mixed onion-radial (m) texture at moderate temperature and
48 Pitch Precursor-Based Carbon Fibers

Fig. 5 Rheostructural evolution in mesophase pitch (MP) at a shear rate of 1 s
1 and 2971C: (a) Shear stress and micrographs are for vorticity-
flow section (Reproduced with permission from Kundu, S., Grecov, D., Ogale, A.A., Rey, A.D., 2009. Shear flow induced microstructure of a
synthetic mesophase pitch. Journal of Rheology 53(1), 85–113.). (b) Shear stress during flow reversal (shown in diamond symbols). Reproduced
with permission from Kundu, S., Ogale, A.A., 2006. Rheostructural studies on a synthetic mesophase pitch during transient shear flow. Carbon 44
(11), 2224–2235.

viscosity. Further increase in temperature significantly reduced the pitch viscosity resulting in flow instability in the spinneret that
produces the onion-like texture (on). In terms of tensile strength of MP-based CFs, the non-radial texture has been found to have
superior tensile and compressive strengths relative to the radial texture. However, increasing the spinning temperature 501C above
 Pitch Precursor-Based Carbon Fibers 49

Fig. 6 Schematic of pitch carbon fiber (CF) production. There are four main steps in conversion of pitch into CFs, namely, pitch fiber spinning
and drawdown, thermosetting/stabilization, carbonization, and finally graphitization. Reproduced with permission from Jeon, Y.-P., Alway-Cooper,
R., Morales, M., Ogale, A.A., 2013. Carbon fibers. In: Somiya, S. (Ed.), Handbook of Advanced Ceramics, second ed. Amsterdam: Elsevier,
pp. 143–154.

Fig. 7 Schematic of a melt blowing spinneret where the draw down occurs due to air flow.

the softening point approaches the decomposition temperature of the low molecular weight fraction of pitch, which introduces
gross defects such as voids and bubbles in the pitch fibers, and the resulting CFs can have poor mechanical properties.58

[Link] Thermosetting/Stabilization
Matsumoto and colleagues attempted to eliminate the thermal stabilization step using solvents to extract the soluble fractions
of the MP fibers.60 They hypothesized that the extraction step would remove the smaller disordered molecules and concentrate
the higher molecular weight components, which would render the fibers infusible or at least decrease the extents of thermal
stabilization required or raise the softening point and enable rapid stabilization at higher temperature. They found that thinner
fiber (10 mm) required a 30 min extraction time with benzene at room temperature and less oxidation for the complete
stabilization.60
50 Pitch Precursor-Based Carbon Fibers

Fig. 8 Evolution of mesophase texture with varying melt spinning temperature for two different types of pitch. Reprinted with permission from
Ō tani, S., Ō ya, A., 1986. Progress of pitch-based carbon fiber in Japan. In: Petroleum Derived Carbons. Washington, DC: ACS Publications,
pp. 323–334. Copyright 1986 American Chemical Society.

Advances to thermo-oxidative methods of pitch fiber stabilization have mostly concentrated on optimizing the heating rate in
air. Conventional thermosetting is by heating the precursor pitch fibers in air at temperatures of about 30–1001C below the pitch
softening temperature. During this process, oxygen diffuses into the pitch fibers and reacts with the aliphatic groups on the pitch
molecule to form ether or other bonds. Such reactions crosslink the pitch molecules, thereby rendering the fibers infusible and
thus suitable for the next high heat treatment step (carbonization).

[Link] Carbonization and Graphitization

After thermosetting, pitch fibers are heat treated at high temperatures (900–35001C) in an inert atmosphere, usually nitrogen or
argon (or helium in some cases) to form CFs. Heat treatment of 900–18001C is referred to as carbonization. During this step, most
of noncarbon elements are removed in the form of gases such as carbon monoxide, carbon dioxide, methane, hydrogen, and
hydrogen sulfide. Higher heat treatment of 2000–33001C results in evolution of mesophase domains into 3D crystalline structures
that approach those of graphite. This step is referred to as graphitization.21 Increasing the final heat treatment temperature
improves the degree of preferred orientation within the fibers and results in increased lattice-dependent properties of tensile
modulus and thermal conductivity of pitch-based CFs, as discussed in the next section.
The final diameter of the pitch-precursor CF depends on the diameter of the as-spun fiber and the temperature used during the
carbonization and graphitization steps. Starting from the stabilized state, MP-based CFs showed a reduction in diameter of around
15% for a heat treatment temperature (HTT) of 9001C and around 25% for a HTT of 24001C.61 In general, this decrease in CF
diameter is smaller for pitch-based than for PAN-based, for which values of B35% have been reported.62 Diameters of pitch-based
CFs are usually in the range of 10–12 mm. By contrast, diameters of PAN-based CFs are in the range of 5–7 mm, and those of rayon-
based CFs are in the range of 6–10 mm.
Finally, surface treatment and sizing is applied to the carbonized fibers using proprietary conditions. This step is critical for MP-
based CFs due to the inert surface of such fibers resulting from a highly graphitic structure, a consequence of thermodynamically
stable crystalline structure. Generally, this step of surface treatment consists of mild oxidation of the CFs by exposure to air or
plasma treatment, or more significant treatment by electro-chemical oxidation in a strong acid such as nitric acid.23,24 This
provides surface functionality in the form of acidic or hydroxyl groups that promote chemical bonding between the fibers and
matrix, producing composites with good mechanical properties.

1.3.4 Microstructure and Properties of Pitch-Based CFs

CFs derived from pitch exhibit a wide range of physical and mechanical properties due to the wide range of microstructure that
develops in such CFs. For MP-based CFs, the tensile modulus and electrical conductivity are exceedingly large, whereas transverse
properties and compressive strength are poor due to the high degree of molecular orientation present in the graphene planes along
the fiber axis. In contrast, a lack of crystalline structure and absence of any major molecular alignment leads to quite low
mechanical properties in isotropic pitch-based CFs.
 Pitch Precursor-Based Carbon Fibers 51

[Link] Thermo-Physical Properties

[Link].1 Density
For any given temperature of heat treatment, the density of MP-based CFs is larger than that of isotropic pitch-based CFs due to the
tight layer packing resulting from the orientation of the anisotropic pitch. Density of both isotropic pitch-based and MP-based CFs
increases with increasing HTT, although the dependence is much stronger for the latter group. The density of MP-based CFs
increases from about 1800 kg m
3 for a HTT of 10001C to around 2200 kg m
3 for a HTT of 30001C. As the graphene layers
become more graphitic with increasing temperature, the overall packing is improved and the density is increased. Such rearran-
gement of graphene layers takes place at a much smaller scale in isotropic pitch-based CFs, hence the slight increase in density
(from 1600 to 1700 kg m
3) with increasing temperature (1000 to 25001C).21

[Link].2 Coefficient of thermal expansion

While it is relatively easy to measure the coefficient of thermal expansion (CTE) in the axial direction, the measurement in the
transverse direction is more difficult because of the smaller dimensions. However, some measurements of CTE in the transverse
direction have been made on CF composites using SEM and laser diffraction. Fig. 9 presents the CTE along the fiber axis of two
pitch-precursor CFs and two PAN-based CFs as a function of temperature. The axial CTE of T300 PAN-based CF is positive for all
the temperature range, and up to 5 times greater than the axial CTE of the pitch-based CFs, particularly at the highest temperatures.
In contrast, for the pitch-based CFs, the axial CTE is negative for a large range of temperature, and it only becomes positive near
10001C. A negative CTE along the fiber axis, coupled with a high axial modulus, allows the preparation of composites with
virtually zero thermal expansion by controlling the fiber orientation in the composite. The axial shrinkage that CFs experiment
near room temperature is caused by the bending of the graphene layers. The CTE in the transverse direction is always positive with
temperature.

[Link].3 Oxidation
Although pitch-based CFs have remarkable high temperature properties, there is a limitation for long-term application because of
their oxidation resistance. The more perfect the graphitic structure, the slower the oxidation rate is. That makes the highly graphitic,
high temperature treated MP-based CFs remarkably resistant to high temperatures in nonoxidizing environments, because melting
does not occur. The oxidation rate is affected by trace amounts of impurities, both within and outside the CF; the latter case is
more difficult to control and therefore the fiber is susceptible to localized pitting caused by contact with metals. The highest
oxidizing temperature that a CF can withstand is in the 450–5001C range, hence applications at higher temperatures must be
short-term or require boron-based protective systems. Boron can be added to the carbon lattice to increase the maximum use
temperature by up to 1001C, and even several hundred degrees when boron nitride coatings are used. The downside of a boron-
based strategy to prevent oxidation is the hygroscopicity of the oxidation product of boron, which may limit applications.

[Link] Microstructure
As discussed above, MP precursor is a nematic liquid crystal in which disk-shaped molecules, mainly consisting of PAHs, are
present in a layered state. Melt spinning of MPes, more specifically the shear experienced during extrusion, forces those molecules
to align parallel to the fiber axis. Small changes are introduced by drawing, which slightly increases axial preferred orientation and
decreases the graphene layer stack height, and by spinning temperature which decreases the two aforementioned parameters.

Fig. 9 Thermal expansion as a function of temperature for several carbon fibers (CFs). Two of the fibers are from PAN precursors, T-300 and M-
40, while P-55 and P100 are from mesophase pitch (MP). Reproduced with permission from Diefendorf, R.J., 2000. Pitch precursor carbon fibers.
In: Kelly, A. (Ed.), Comprehensive Composite Materials. Oxford: Pergamon, pp. 35–83.
52 Pitch Precursor-Based Carbon Fibers

Figure 10 Crystal structure of graphite crystal (left) and turbostratic carbon (right). Reproduced with permission from Buckley, J.D., 1992.
Carbon–carbon overview. In: Buckley, J.D., Edie, D.D. (Eds.), Carbon–Carbon Materials and Composites. Washington, DC: National Aeronautics and
Space Administration.

The fundamental structural unit of CFs is often considered to be a stack of carbon layers, usually in turbostratic form, that create
coherent domains. They usually exist in turbostratic form where the interplanar distance is 0.34 nm and the layers are not planar,
although a graphitic structure with a characteristic interplanar spacing of 0.335 nm is present in some MP-based CFs, particularly
the ones heat treated at high temperatures (Fig. 10).
These fundamental structural units are originated in the fiber formation process; they can split, fold, twist and join other units
to form larger domains, which in turn can experience the same dislocations and interactions. Thus, the spinning process creates
microdomains which are elongated along the fiber axis, and are connected to other microdomains by regions of pores or
dislocations.63 Although they are present in all CF types, the microdomains can interact with each other to a higher level of
organization in MP-based CFs, creating meso and macrodomains, which form the different types of transverse microtextures seen
in MP-based CFs (Fig. 11). Such microtextures represent the existence of a very long range, statistical orientation of the macro-
domains. The various types of transverse microtextures depend on the precursor and the process conditions in melt spinning.
Ideally, three textures can be readily defined: radial, onion-skin, and completely random. In real CFs, some combinations of these
are found. As shown in the schematics (bottom two), the core has random texture arising from shear stresses that are negligible at
the centerline, whereas the periphery has radial texture because shear stresses are maximum at the die wall.
Although the micro-, meso- and macrodomains are created in the spinning step, the subsequent stabilization and particularly
the high temperature treatment refines the microstructure of the fibers, by removing heteroatoms, increasing density and bringing
domains closer, to the point that turbostratic carbon layers appear. Planar layers of crystalline carbon appear at extreme tem-
peratures (420001C). The graphitic structure relative to the dimensions of a CF is shown in Fig. 12. Analysis of longitudinal
sections of fiber by scanning tunneling microscopy has evidenced that the atomic structure of pitch-precursor CFs resembles the
hexagonal structure of graphite.64
Fig. 13 shows a schematic representation of a carbon crystallite within a CF. These crystallites are typically described by four
values: La8 (persistence length, parallel to the fiber axis), La> (persistence length, perpendicular to the fiber axis), Lc (graphene layer
stack height) and d002 (spacing between graphene layers). These dimensions can be measured by wide-angle X-ray diffraction
(WAXD), Raman spectroscopy and transmission electron microscopy (TEM), as will be discussed below.
A more recent model for high modulus MP-based CFs (axial modulus 4350 GPa) by Emmerich65 has proposed as a structural
unit a lamellar stack of elongated sheets of carbon layers that contain a series of defects, called defective graphene nanoribbons
(dGNRs). The dGNRs have an average length La8 and are arranged longitudinally in series, separated by defective regions of
approximate constant length δ (Fig. 14(a)). Defective regions consist of disordered grain boundaries that may include 5/7
membered rings, vacancies, distorted hexagon rings, and octagon rings (Fig. 14(b)).

[Link].1 X-ray diffraction

The crystalline structure of CFs derived from AR-synthetic mesophase and heat treated from 1200 to 28751C is discussed next. The
WAXD spectra, presented in Figs. 15–20, display the evolution of the microstructure as evidenced by the narrowing and increasing
intensity of (00,2), (11,0), and (11,2) peaks. The calculated values of d-spacing between the (00,2) planes and crystallite sizes,
displayed in Fig. 18, show that the graphene layer stack height, Lc, and the persistence length of the graphene sheets parallel to the
 Pitch Precursor-Based Carbon Fibers 53

Figure 11 Schematics of cross-sectional textures found in mesophase pitch (MP)-based carbon fibers (CFs).

Fig. 12 Crystal structure of graphite with its schematic relationship to a carbon fiber (CF). Reproduced with permission from Diefendorf, R.J.,
2000. Pitch precursor carbon fibers. In: Kelly, A. (Ed.), Comprehensive Composite Materials. Oxford: Pergamon, pp. 35–83.

fiber axis, La, both show an increase to a HTT above 20001C (where the resolution limit of the WAXD diffractometer was reached).
The d-spacing shows a gradual decrease as HTT increases from 0.348 nm for MP to an almost graphitic spacing of 0.337 nm at the
highest HTT. The presence of (10,1) and (11,2) peaks shows the existence of three-dimensional graphitic crystallinity at higher
HTT (as opposed to turbostratic structure).
The misorientation angle of the graphene planes with reference to the fiber axis (sometimes called degree of orientation) is
determined by the full width at half maximum (FWHM or Z) (misorientation angle of 0 degree represents perfect alignment of the
graphene planes in the fiber direction). Azimuthal scans of a single filament were used to measure the misorientation angle for
single filaments using monochromated synchrotron radiation source at a wavelength of 0.1366 nm at Brookhaven National Lab.61
As displayed in Fig. 19, data from single-filament as well as fiber bundles indicate that the graphene layer misalignment is fairly
54 Pitch Precursor-Based Carbon Fibers

Fig. 13 Schematic representation of a carbon crystallite inside a carbon fiber (CF). Redrawn from Frank, E., Steudle, L.M., Ingildeev, D., et al.,
2014. Carbon fibers: precursor systems, processing, structure, and properties. Angewandte Chemie International Edition 53(21), 5262–5298.

Fig. 14 (a) Proposed model for structural unit of mesophase pitch (MP)-based carbon fibers (CFs) and (b) suggested type of defective region.
Reproduced with permission from Emmerich, F., 2014. Young’s modulus, thermal conductivity, electrical resistivity and coefficient of thermal
expansion of mesophase pitch-based carbon fibers. Carbon 79, 274–293.

large before heat treatment but decreases as HTT increases. A slight increase in misalignment is observed as the HTT increases from
300 to 9001C, followed by an eventual decrease of misalignment at the highest HTTs. As expected, the single-filament data show
less misorientation at the highest HTTs where any fiber misalignment within a bundle is larger than the graphene plane mis-
alignment in the fibers. Misalignment angles of less than 2 degrees were measured at the highest HTTs. These values are con-
siderably smaller than those obtained from fiber bundles, and indicate that fiber misorientation within a bundle limits the
measurement of small values of Z that are likely achieved after heat treatment at elevated temperatures approaching 30001C.
Data presented in Fig. 19 also indicate that at lower HTTs, the graphene layer misalignment angles showed a maximum at
600–9001C. The small increase in the misorientation angle from stabilized state to 9001C may be the result of relaxation of
molecular orientation as a result of increased mobility at elevated temperatures given that the carbonization process is not
completed until about 10001C. Another reason for the increase might be the disruption caused by evolution of heteroatoms
during the carbonization step. It is noted, however, that the larger values of Z for the single-filament specimens at lower HTTs was
contrary to expectation. Due to the necessary background corrections, XRD results from high HTTs display a variation of 1–2% but
at low HTTs the numbers are less reliable. For the high HTTs, the correction is not even noticeable but at the low HTTs the
background correction is of the order of the peak itself. Therefore, an absolute comparison of Z-values from the single-filament
 Pitch Precursor-Based Carbon Fibers 55

Fig. 15 Representative equatorial scans in the (00,2) region of fibers heat-treated to various temperatures showing the peak narrowing and peak
position shift toward more graphitic nature. Reproduced with permission from Ogale, A.A., Lin, C., Anderson, D.P., Kearns, K.M., 2002. Orientation
and dimensional changes in mesophase pitch-based carbon fibers. Carbon 40, 1309–1319.

Fig. 16 Representative meridonal scans in the (11,0) and (11,2) region of fibers heat-treated to various temperatures showing the narrowing of
the (11,0) peak as a result of cyrstallite size increases. Reproduced with permission from Ogale, A.A., Lin, C., Anderson, D.P., Kearns, K.M., 2002.
Orientation and dimensional changes in mesophase pitch-based carbon fibers. Carbon 40, 1309–1319.

and fiber-bundle specimens is not statistically significant. However, for a given type of specimens, because the relative corrections
are similar, the existence of a small maximum in the misorientation angle.
Z p=2
IðfÞsinfcos2 fdf
〈cos fhkl 〉 ¼
2 0
Z p=2
IðfÞsinfdf
0

This value can be calculated from the Z or FWHM value of the original X-ray intensity data. If one assumes that the average
projected length of the crystals along the fiber axis is equal to the average cosine of the (00,2) plane, the relative lengths of fiber
should increase 3.3% when the Z value decreases from 25 degrees at a HTT of 9001C to 4 degrees at a HTT of 24001C. Whereas the
predicted length increase is larger than the measured value, results from the present study support theoretical arguments that the
fiber length should increase as the graphene layer planes align closer to the fiber axis.
Fig. 20(a) and (b) display WAXD results for MP-based CFs. Fig. 20(a) indicates a two-theta peak associated with (002) planes at
values between 26 and 26.4 degrees. This corresponds to a large fraction of the (002) planes being stacked with an interplanar
spacing of 0.3354 nm and the registry of the planes being the abab sequence found in crystalline graphite. The highest two-theta
value of about 26.4 degrees for K1100 grade is consistent with the highest extent of interplanar packing and three-dimensional
56 Pitch Precursor-Based Carbon Fibers

Fig. 17 Representative off-axis scans (fiber at chi¼70 degrees) in the (11,0) and (11,2) region of fibers heat-treated to various temperatures
showing the evolution of the 3-D graphitic structure. Reproduced with permission from Ogale, A.A., Lin, C., Anderson, D.P., Kearns, K.M., 2002.
Orientation and dimensional changes in mesophase pitch-based carbon fibers. Carbon 40, 1309–1319.

Fig. 18 Graphitic crystalline parameters measured by WAXD for the AR-mesophase pitch (MP) fiber bundles heat treated to various temperatures.
Reproduced with permission from Ogale, A.A., Lin, C., Anderson, D.P., Kearns, K.M., 2002. Orientation and dimensional changes in mesophase
pitch-based carbon fibers. Carbon 40, 1309–1319.

Fig. 19 Degree of misorientation of graphene planes measured in AR-mesophase pitch (MP) fibers heat-treated to various temperatures.
Reproduced with permission from Ogale, A.A., Lin, C., Anderson, D.P., Kearns, K.M., 2002. Orientation and dimensional changes in mesophase
pitch-based carbon fibers. Carbon 40, 1309–1319.
 Pitch Precursor-Based Carbon Fibers 57

Fig. 20 WAXD diffractograms for P25 and K1100 MP pitch-based carbon fibers (CFs) as compared with that for T300 PAN-based fibers: (a)
integrated azimuthal distribution as a function of two-theta; and (b) azimuthal distribution of the (002) planes centered along the fiber axis.
Reproduced with permission from Jeon, Y.-P., Zhang, M., Levan, C.D., Ogale, A.A., 2012. High thermal conductivity carbon fibers from petroleum-
based experimental mesophase pitch. In: Annual World Conference on Carbon 2012, Krakow, Poland.

Fig. 21 SEM micrographs of unmodified AR-mesophase pitch (MP)-based carbon fibers (CFs) heat treated (HT) at: (a) 6001C and (b) 26001C.

crystalline order resulting from an extremely high HTT of 3000 to 33001C. For P25 grade MP-based CFs, the two-theta value is
about 26.11, consistent with other lattice-dominated properties that indicate a lower HTT.
The (002) peak is very sharp for K1100 fibers with a FWHM of about 0.4 degrees. For P25 fibers, the peak is less sharp with a
higher FWHM of about 0.8 degrees. These FWHM values of B1 degrees for MP-based fibers are significantly smaller than that of
about 4.6 degrees for PAN-based fibers. The stacking height, LC, for K1100, P25, and T300 are of the order of 200, 20, and 2 nm.
This clearly illustrates the larger three-dimensional stacking developed in CFs derived from disk-like MP precursors versus the
chain-like PAN precursors. From the azimuthal orientation distribution of the (002) graphene layer planes within the fibers,
displayed in Fig. 20(b), the FWHM values were about 4 degrees for K1100 and 14 degrees for P25. These FWHM values for MP-
based fibers are much smaller than those for PAN-based fibers (FWHM greater than 30 degrees for T300).

[Link].2 SEM and TEM

Fig. 21 shows the radial structure development that is typical of CFs derived from AR-synthetic MP. With increasing HTT, graphitic
layers become highly developed, as is evident in the series of SEM micrographs displayed in Fig. 21(a) and (b). As shown in
Fig. 21(b), high temperature treated MP-based CFs generally show a well-developed planar radial structure due to higher graphitic
crystallinity, rather than polar or random structure. Another typical phenomenon evident in Fig. 21(b) is the “Pac-man” splitting, a
typical behavior of the high temperature treated relatively thick anisotropic pitch-based CFs with radial alignment. The radially aligned
layer planes develop high graphitic crystallinity and pack tighter (volume shrinkage), which results in excessive hoop-direction stress.
Due to weak p–p bonding in the direction transverse to layer planes, the fiber cracks in the hoop direction as Pac-man splitting, usually
triggered by small defects in the fibers. The sharp v-notch at the center of the split fiber makes the fiber rather brittle. Typically, such CFs
have poor strength and a low strain-to-failure, which lower the handleability of such fiber during composite processing operations.
Detailed microstructure from TEM of these fibers is presented next in Fig. 22 that shows the development of graphitic layers due
to crystallinity improvement by increasing HTTs. At low HTT of 6001C, the CF matrix consists of short and randomly orientated
layer planes, which become longer and oriented with heat treatment. After an HTT of 26001C, very well oriented long graphitic
58 Pitch Precursor-Based Carbon Fibers

Fig. 22 TEM micrographs of carbon fibers (CFs) obtained from unmodified AR-MP heat treated at two temperatures: (a) 6001C and (c) 26001C.
Reproduced with permission from Jeon, Y.-P., Zhang, M., Levan, C.D., Ogale, A.A., 2012. High thermal conductivity carbon fibers from petroleum-
based experimental mesophase pitch. In: Annual World Conference on Carbon 2012, Krakow, Poland.

Fig. 23 SEM micrographs of carbon fibers (CFs) obtained from (a) unmodified AR-mesophase pitch (MP). (b) AR-MP modified with 0.3 wt%
carbon nanotubes, respectively. Reproduced with permission from Cho, T., Lee, Y.S., Rao, R., et al., 2003. Structure of carbon fiber obtained from
nanotube-reinforced mesophase pitch. Carbon 41, 1419–1424.

layered structure was found. As expected, fibers at higher HTTs show highly graphitic layers with spacing between 0.336 and 0.338
nm, depending upon the HTT.
It is hypothesized that the reduction of the high degree of orientation within these CFs can help reduce the anisotropic gap
between axial and radial properties. In prior studies, we have established that the nanotexture of CFs can be modified by
incorporating ultradiluted concentrations of multiwall carbon nanotubes (MWCNTs) that render the nanostructure random in-
plane therefore offering a cost-effective means of modifying CF.66 Only 0.3 wt% of MWCNTs led to a homogenization of texture
within the plane (Fig. 23(b)). The compressive strengths of the nanocomposite fibers (as measured by tensile-recoil compressive
tests) increased almost 50% from 0.7 to 1.1 GPa, equivalent to values of some commercially available PAN- and pure pitch-based
CFs. Although the compressive strength of the “nanocomposite” fibers increased, it was accompanied by a decrease in the tensile
strength from 2.6 to 1.8 GPa, due to the presence of MWCNTs that also act as flaws.

[Link].3 Raman spectroscopy

Studies by Yang et al.67 found that solid-phase annealing of synthetic MP spun fiber at temperature between Tg and softening
point, prior to graphitization, increased the stacking height of the graphitized fibers and subsequently the tensile modulus and
thermal conductivity. Fig 24 compares Raman spectra of annealed methylnaphthalene-derived carbon fibers to those derived
without annealing of the pitch fibers (mNPGF2). In both spectra, the four typical bands for graphitized carbon fibers are observed:
the G-band (1580 cm–1) associated with graphite sp2 bonds, a D-band (1360 cm–1) attributed to disordered carbon, a D’ band
(1620 cm–1) and a 2D (or G’) band (2720 cm–1). However, they found that annealing of the precursor fibers weakened the D band
relative to the G band, which indicates enhanced graphitic content for CFs derived from annealed fibers.

[Link] Mechanical Properties

Strength (both tensile and compressive) and modulus (axial and transversal) are the most relevant mechanical properties of pitch-
precursor CFs. Table 1 presents tensile strength and modulus values for various commercial pitch-based CFs. Although it is known
 Pitch Precursor-Based Carbon Fibers 59

Fig. 24 Raman spectra of methylnaphthalene pitch-based graphitized fiber, mNPGF2, annealed (top) and not annealed (bottom). Reproduced with
permission from Yang, H. Yoon, S.-H., Korai, Y., Mocida, I., Katou, O., 2003. Improving graphitization degree of mesophase pitch-derived carbon
fiber by solid-phase annealing of spun fiber. Carbon 41(3), 397–403.

Table 1 Tensile modulus and strength for various commercial pitch-based carbon
fibers (CFs)

Tensile modulus (GPa) Tensile strength (GPa) Density (g cm
3)

BP-Amoco
P-25 160 1.4 1.90
P-55S 380 1.9 2.00
P-75S 520 2.1 2.00
P-100 760 2.4 2.16
P-120 830 2.4 2.17
K-800x 930 2.9 2.18
K-1100 930 3.1 2.20
Mitsubishi Kasei
K133 440 2.4 2.08
K135 540 2.6 2.10
K137 640 2.7 2.11
K139 740 2.8 2.12
K321 180 2.0 1.90
Nippon graphite fiber
YS-50A 520 3.8 2.09
YS-60 590 3.5 2.12
YS-70A 720 3.6 2.14
YS-80 785 3.5 2.15
Isotropic Kureha pitch
T101F 33.0 0.80 1.81
T201F 33.0 0.70 1.57

Source: Diefendorf, R.J., 2000. Pitch precursor carbon fibers. In: Kelly, A. (Ed.), Comprehensive Composite
Materials. Oxford: Pergamon, pp. 35–83.

that the properties of CFs are determined by the structure of the fibers, these two mechanical properties are a function of different
structural characteristics of pitch-based CFs. While strength is controlled by the density of defects and the occurrence of micro-
buckling, modulus is a strong function of the degree of orientation of the graphene layers with respect to the fiber axis.
60 Pitch Precursor-Based Carbon Fibers

Fig. 25 Stress–strain response of AR-MP pitch-based carbon fibers (CFs) heat treated at various temperatures. Reproduced with permission from
Jeon, Y.-P., Zhang, M., Levan, C.D., Ogale, A.A., 2012. High thermal conductivity carbon fibers from petroleum-based experimental mesophase
pitch. In: Annual World Conference on Carbon 2012, Krakow, Poland.

Two methods used to determine the tensile properties of CFs are the single-filament tensile test and the composite strand
tensile test. In the composite strand method, a tow of fiber is impregnated with an epoxy resin, and the rigid test specimen is
tensile loaded to failure. This method requires the elongation-to-break of the cured epoxy resin to be larger than that of the fiber.
CF properties thus obtained are an average for the fiber tow. Although such values are of practical relevance (as CFs are always used
as tows, not single filaments), the single-filament test can provide a distribution of CF properties for a given sample.
The single-filament method consists of mounting a filament on a cardboard tab ensuring that the specimen is axially aligned
with the tensile direction. The tab is burned (or cut) and the fiber is pulled apart to failure at a constant strain rate. Since it is not
practical to attach an extensometer to the specimen to record the actual strain, system compliance is measured using different gage
lengths to obtain accurate modulus values. Typical stress–strain curves for CFs derived from synthetic AR-MP are displayed in
Fig. 25. It should be noted that the stress–strain response of CFs obtained at different HTTs is almost linear and does not show any
yielding phenomenon, typical of brittle materials.

[Link].1 Modulus
Axial modulus of pitch-precursor CFs is highly dependent on HTT. Fig. 26 shows axial modulus versus HTT for a CF based on
solvent-fractionated mesophase-pitch obtained from a catalytic cracker fraction. It is believed that upon heat treatment at 30001C,
CFs could have a modulus close to the single crystal value of 1000 GPa. CFs based on other MPes would show the same general
trend, although with some quantitative differences. For instance, the modulus curve would be shifted down for a less oriented and
less graphitizable pitch precursor. The increase in axial modulus with increasing temperature is a confirmation that the modulus
increases with the improvement in preferred orientation. Values from 200 to 900 GPa can be achieved for the axial modulus of
MP-based CFs.15
The variation of axial modulus with test temperature has not been widely studied for pitch-precursor CFs. However, structural
changes are expected to occur when the test temperature exceeds the HTT of the CF. Data for Thornel P-50 CF have revealed a
decrease in modulus to around 75% of its room-temperature value at 16001C, followed by a more rapid decrease at higher
temperatures.21 It is expected that the higher the room-temperature modulus of the CF, the smaller the decrease in modulus with
increasing temperature, because of the better alignment of the graphene layers.
Some measurements of the transverse modulus of single filaments of pitch-based CFs have been made. Data for Thornel P-
series CFs suggest an inversely proportional relationship between transverse modulus and axial modulus: for an axial modulus of
200, 400 and 800 GPa, the transverse modulus is around 10, 6 and 3 GPa, respectively. The reader can refer to Diefendorf21 for a
detailed plot.

[Link].2 Strength
For pitch-precursor CFs, the tensile strength is always greater than the compressive strength. The measured strength is a function of
the gage length used in testing. A gage length of 25 mm is typically used in the single-filament method, whereas a gage length ten
times greater is commonly used in the composite strand method. In order to get a more accurate estimate of the intrinsic strength,
and to determine flaw distribution, studies with different gage lengths are required.
Structural imperfections that influence the strengths of pitch-based CFs are surface flaws, internal particulates, internal voids,
and misoriented crystallite regions. Therefore, processing determines the strengths to a great extent, given that it is during the
manufacturing steps (melt spinning, stabilization and high temperature treatment) that the aforementioned defects arise. An
initial approach to minimize the emergence of imperfections is the thorough filtering of the precursor pitch to remove particles,
which can act as stress concentrators. The removal of metals contaminants can prevent surface pitting during stabilization. Fiber
 Pitch Precursor-Based Carbon Fibers 61

Fig. 26 Modulus and tensile strength of carbon fibers (CFs) as a function of heat treatment temperature (HTT). A solvent-fractionated catalytic
cracker mesophase was used as precursor. Reproduced with permission from Diefendorf, R.J., 2000. Pitch precursor carbon fibers. In: Kelly, A.
(Ed.), Comprehensive Composite Materials. Oxford: Pergamon, pp. 35–83.

sticking must also be prevented during the stabilization step, and general fiber abrasion during and after processing needs to be
minimized. The elimination of the external sources of defects listed above would leave the inherent internal microstructure of the
pitch-based CF as the only factor directly controlling the strengths. Strengths of the early commercial pitch-precursor CFs were
relatively low compared to those of the PAN-based CFs. Improvements allowed MP-based CFs to match the tensile strength values
of standard PAN-based CFs, although the values of high strength PAN-based CFs have not been achieved in pitch-based CFs.
The dependence of tensile strength with HTT shows a different behavior than for modulus, as seen earlier in Fig. 26. The tensile
strength of the petroleum pitch-based CF increases rapidly with HTT in the range from 1500 to 17001C, and it starts to plateau at
18001C. While the tensile strength remains rather constant at higher HTTs, the modulus continues to increase in the same
temperature range and in consequence the strain-to-failure decreases to relatively low values.
Models of the strength of CFs have assumed that fractures are induced by shear stresses in misoriented crystallites, which
indicates that crystallite size affects tensile strength. Besides the fact that there is a strong correlation between strain-to-failure and
crystallite size, it is believed that a small crystallite size favors a high strain-to-failure (for an explanatory figure see Ref. [21]).
Therefore, a small crystallite size Lc must be a characteristic of high strength MP-base CFs. The tensile strength of pitch-based CFs
increases with test temperature, a trait commonly observed in graphite, before decreasing steeply as the fiber creeps at higher
temperatures (see [21]).
The compressive strength of CFs is usually measured by a single fiber compression test, in which a fiber is cut after being loaded in
tension, and the subsequent compressive wave loads the fiber in compression. A composite-based method is also used although in
general the results show disagreement with the single-filament method. As discussed above, the compressive strength is always
smaller than the tensile strength, with the values for early CFs being as low as 1% of the tensile strength. Later developments have
raised the compressive strengths of pitch-based CFs to around 0.7 GPa for high modulus mesophase-pitch-based CFs; such values are
still low when compared to the compressive strength of around 2 GPa observed for PAN-based CFs. Measurements of compressive
strength as a function of axial modulus have been made for Thornel P-series and Granoc pitch-based CFs. For both brands,
compressive strength decreases with increasing axial modulus, a consequence of the higher degree of orientation along the fiber axis
in the high modulus CFs. Thornel P-series CFs with axial modulus of 200, 400 and 600 GPa have an approximate compressive
strength of 1.1, 0.8 and 0.6 GPa, respectively. The reader is referred to detailed figures provided in the original chapter.21
Fiber diameter has an effect on the properties of a pitch-based CF. In one of the first papers on pitch-precursor CFs, Otani26
reported an increase in tensile strength and initial modulus with a decrease in diameter from 30 to 8 mm and especially below 15
mm. Although it has been observed that preferred orientation increases with as-spun fiber diameter, it is hard to maintain the
orientation in the CF state because of the development of a skin–core structure, which results in decreased tensile strength, axial
modulus, and electrical resistivity.68
62 Pitch Precursor-Based Carbon Fibers

[Link] Transport Properties

MP-based CFs can have an extremely high thermal conductivity, since their highly graphitic, axially aligned structure closely
recreates the basal plane of graphite, the material with the highest room-temperature thermal conductivity. Direct measurements
of thermal conductivity of single filaments are difficult to perform. Since there is a strong correlation between thermal and
electrical conductivities in CFs, electrical resistivity (reciprocal of electrical conductivity) measurements of single CF filaments are
made instead, and the thermal conductivity is estimated with the Issi–Lavin correlation69:

κ ¼ 440000=ðr þ 258Þ
295

where the electrical resistivity r is expressed in mO cm and the thermal conductivity κ is expressed in W m
1 K
1. The data and
correlation curve are shown in Fig. 27.
The electrical conductivity of CFs is a function of the HTT. In general, the electrical conductivity improves as the HTT increases,
and the same is true for thermal conductivity (Table 2). However, the exact behavior of this relationship depends on the type of
precursor. Rayon-based and isotropic pitch-based CFs show a relatively small increase in electrical conductivity in the range from
1000 to 30001C. Electrical conductivity of PAN-based CFs increases very rapidly between 1000 and 15001C, remains virtually
constant between 1500 and 25001C, and increases again between 2500 and 30001C. MP-based CFs experience a rapid and
continuous increase in electrical conductivity in the range from 1000 to 25001C, with little variation upon heating at higher
temperatures. For a given HTT, the electrical conductivity of MP-based CFs is greater than that of PAN-based CFs. MP-based CFs
have a greater electrical conductivity than isotropic pitch-based CFs and rayon-based CFs, in some cases by one order of mag-
nitude. For a descriptive figure of the electrical resistivity of CFs, the reader is referred to the original chapter.21
The change in electrical conductivity with test temperature also depends on the HTT. MP-based CFs heat treated at 1700 and
20001C show no major change in electrical conductivity in the test temperature range from 100 to 300K, while those heat treated

Fig. 27 Correlation of thermal conductivity and electrical resistivity in mesophase pitch (MP)-based carbon fibers (CFs) at 300K. Reproduced with
permission from Lavin, J.G., Boyington, D.R. Lahijani, J., Issi, J-P., 1993. The correlation of thermal conductivity with electrical resistivity in
mesophase pitch-based carbon fiber. Carbon, 31(6), 1001–1002.

Table 2 Electrical resistivity and thermal conductivity of a pitch-based carbon fiber (CF) for
different heat treatment temperatures (HTTs)

HTT D Electrical resistivity Thermal conductivity predicted from Issi–Lavin correlation (W

(1C) (mm) (mOm) m
1 K
1)

1500 9.4 7.0 165.3

2100 9.4 6.3 202.9
2400 9.3 5.0 286.9
2500 9.4 4.6 477.0

Source: Jeon, Y.-P., Zhang, M., Levan, C.D., Ogale, A.A., 2012. High thermal conductivity carbon fibers from petroleum-based
experimental mesophase pitch. In: Annual World Conference on Carbon 2012, Krakow, Poland.
 Pitch Precursor-Based Carbon Fibers 63

at higher temperatures show an increase in electrical conductivity up to room temperature. The observed piezoresistance could
make MP-based CFs useful as structural sensors.
Other factors influencing the transport properties of pitch-precursor CFs are the cross-sectional shape and cross-sectional
microstructure of the fiber. MP-based CFs with ribbon-shaped cross-sections achieve a more graphitic structure and in consequence
a higher thermal conductivity than circular fibers, for a given HTT. Ribbon-shaped MP-based CFs heat treated at 24001C have a
thermal conductivity equivalent to the one of circular fibers heat treated at 30001C. The reason behind the superior ability to
graphitize of ribbon-shaped fibers is their more aligned as-spun molecular orientation, parallel to the fiber axis. Thus, for a given
HTT, results have also shown that radial cross-sectional microstructures are more graphitizable than a random one, yielding CF
with higher thermal and electrical conductivities.
MP produced from a low-cost petroleum pitch has been used in recent studies to obtain CFs after high temperature treatment
in the range 1200–24001C.44 These treatments resulted in a control of the CF microstructure/graphitic crystallinity and mechanical
properties and thermal conductivity, with the highest tensile modulus measured at approximately 550 GPa, the highest tensile
strength at 2.2 GPa, and axial thermal conductivity of 500 W m
1 K
1 (estimated from Issi–Lavin correlation). These MP fibers
were produced from a standard circular-hole spinneret and continuous roll winding with resulting CFs ranging from about 10 to
18 mm in diameter. High HTTs enhance lattice-controlled properties such as modulus and conductivity, whereas low-to-moderate
temperatures favor strength by reducing textural defects. These CF properties, developed from a low-cost petroleum-derived MP
precursor, are comparable to those of commercial CFs derived from MP such as the P-series from Cytec Carbon Fibers.70 For
instance, P-series CFs display a modulus as high as 400 GPa and tensile strength of about 2 GPa. In earlier developments (1980s),
DuPont developed the E-series (now discontinued) with E75 grade CFs displaying modulus and strength as high as 500 and 2.5
GPa, respectively.
After a break of almost two decades, the topic of MP-based CFs has seen a resurgence arising from the increased demand for CFs
that possess good properties and yet are cost-competitive. Such commercial CFs have not yet been introduced, but recent
developments on the improvement of technology19 can possibly produce low-cost MP from petroleum residues at extremely low-
cost (under US$5 per kg at 2016 prices). Such developments present potential opportunities for pitch-based CFs in the expanding
automotive and industrial sectors.

1.3.5 Concluding Remarks

As discussed above, CFs are amongst the strongest reinforcing fibers available to reinforce polymeric matrices. CFs derived from
MP display a moderate tensile strength of about 3 GPa, but possess a very high tensile modulus of over 500 GPa. Such lattice-
controlled properties arise from the strong discotic alignment during melt spinning of precursor fibers that leads to the formation
of graphitic layer planes with d002 spacing approaching 0.336 nm (close to that of perfect graphite of 0.3354 nm) after heat
treatment above 2700K. Such graphitic CFs also possess outstanding thermal conductivity approaching 1000 W m
1 K
1 due to
enhancement of phonon transport in the crystalline lattice. Electron transport is also enhanced for fibers heat treated to over
3000K with electrical conductivity exceeding 5000 S cm
1, and are well suited for electrostatic dissipation and EMI shielding
applications. Although a vast majority of commercial CFs for high performance applications are currently derived from PAN
precursors because they possess very high strength (approaching 7 GPa), recent developments toward reducing costs of pitch
precursor/melt-processing indicate the potential for such pitches to lead to high volume, cost-competitive CFs suitable for light-
weight transportation and industrial applications.

References

1. Morgan, P., 2005. Carbon Fibers and Their Composites. Boca Raton, FL: CRC Press.
2. Edie, D.D., McHugh, J.J., 1999. High performance carbon fibers. In: Burchell, T.D. (Ed.), Carbon Materials for Advanced Technologies. Amsterdam: Elsevier, pp. 119–138.
3. Griffiths, B., 2005. High performance composites, Paris Air Show Report 2005, pp. 24–28.
4. Buckley, J.D., 1992. Carbon–carbon overview. In: Buckley, J.D., Edie, D.D. (Eds.), Carbon–Carbon Materials and Composites. Washington, DC: National Aeronautics and
Space Administration.
5. Castro, A.S., Kim, R.Y., Mistretta, J.P., 2001. The application of composites for the rehabilitation of concrete bridge infrastructure. Advanced Composite Materials 10,
147–157.
6. Chung, D.D.L., 2001. Electromagnetic interference shielding effectiveness of carbon materials. Carbon 39, 279–285.
7. Kurylko, D.T., 2010. BMW electric car using carbon fiber. Plastics News, 12 April.
8. Sloan, J., 2017. Carbon Fiber 2016 Report. CW Composites World 3 (1), 24–25.
9. National Research Council, 2005. High-performance fiber technology. In: High-Performance Structural Fibers for Advanced Polymer Matrix Composites. Washington, DC:
The National Academies Press, pp. 6–17.
10. Fitzer, E., Müller, D.J., 1975. The influence of oxygen on the chemical reactions during stabilization of pan as carbon fiber precursor. Carbon 13 (1), 63–69.
11. Edie, D.D., Diefendorf, R.J., 1992. Carbon fiber manufacturing. In: Buckley, J., Edie, D.D. (Eds.), Carbon–Carbon Materials and Composites. Washington, DC: National
Aeronautics and Space Administration, pp. 19–39.
12. Rahaman, M.S.A., Ismail, A.F., Mustafa, A., 2007. A review of heat treatment on polyacrylonitrile fiber. Polymer Degradation and Stability 92 (8), 1421–1432.
13. Fitzer, E., Frohs, W., Heine, M., 1986. Optimization of stabilization and carbonization treatment of PAN fibres and structural characterization of the resulting carbon fibres.
Carbon 24 (4), 387–395.
14. TORAY. T1000G Data Sheet, Available at: [Link]
64 Pitch Precursor-Based Carbon Fibers

15. Bahl, O.P., Shen, Z., Lavin, J.G., Ross, R.A., 1998. Manufacture of carbon fibers. In: Donnet, J., Wang, T.K., Rebouillat, S., Peng, J.C.M. (Eds.), Carbon Fibers, third ed.
New York, NY: Marcel Dekker, pp. 1–83.
16. Jeon, Y.-P., Alway-Cooper, R., Morales, M., Ogale, A.A., 2013. Carbon fibers. In: Somiya, S. (Ed.), Handbook of Advanced Ceramics, second ed. Amsterdam: Elsevier,
pp. 143–154.
17. Park, S.-J., Heo, G.-Y., 2015. Precursors and manufacturing of carbon fibers. In: Carbon Fibers. Berlin: Springer, pp. 31–66.
18. Reisch, M.S., 2011. Getting the steel out. Chemical & Engineering News 89 (39), 10–14.
19. Malone, D.P., Lee, D.M., 2016. Turbulent mesophase pitch process and products. Patent No. US9376626 B1.
20. Sloan, J., 2016. Coming to carbon fiber: Low-cost mesophase pitch precursor. Available at: [Link]
mesophase-pitch-precursor
21. Diefendorf, R.J., 2000. Pitch precursor carbon fibers. In: Kelly, A. (Ed.), Comprehensive Composite Materials. Oxford: Pergamon, pp. 35–83.
22. Figueiredo, J.L., Bernardo, C., Baker, R.T.K., Hüttinger, K.J. (Eds.), 1990. Carbon Fibers Filaments and Composites. Berlin: Springer Science & Business Media.
23. Chung, D.D., 1994. Processing of carbon fibers. In: Carbon Fiber Composites. Boston, MA: Butterworth-Heinemann, pp. 13–53. ISBN 0-7506-9169-7 ed.
24. Chung, D.D.L., 1994. Carbon Fiber Composites. Newton, MA: Butterworth-Heinemann.
25. Donnet, J.-B., Wang, T.K., Rebouillat, S., Peng, J.C.M. (Eds.), 1998. Carbon Fibers, third ed. New York, NY: Marcel Dekker.
26. Otani, S., 1965. On the carbon fiber from the molten pyrolysis products. Carbon 3, 31–38.
27. White, J.L., 2000. Pioneer of carbonaceous pitch processing Sugio Otani. International Polymer Processing XV (4), 336.
28. Arai, Y., 1993. Structure and properties of pitch-based carbon fibers. Nippon Steel Technical Report, Issue 59, pp. 65–70.
29. Kureha Chemical Industries. KRECA carbon fiber. Available at: [Link]
30. Brooks, J.D., Taylor, G.H., 1965. The formation of graphitizing carbons from the liquid phase. Carbon 3, 185–193.
31. White, J.L., Guthrie, G.L., Gardner, J.O., 1697. Mesophase microstructures in carbonized coal tar pitch. Carbon 5, 517–518.
32. Lewis, I.C., McHenry, E.R., Singer, L.S., 1976. Process for producing carbon fibers from mesophase pitch, Patent No. US3976729 A.
33. Mitsubishi Gas Chemical Company, Inc. AR Resin. Available at: [Link]
34. Garcés, P., Fraile, J., Vilaplana Ortego, E., Andión, L.G., 2005. Effect of carbon fibres on the mechanical properties and corrosion levels of reinforced Portland cement
mortars. Cement and Concrete Research 35 (2), 324–331.
35. Lee, H.-M., Kwac, L.-K., An, K.-H., Kim, B.-J., 2016. Electrochemical behavior of pitch-based activated carbon fibers for electrochemical capacitors. Energy Conversion and
Management 125, 347–352.
36. Vilaplana-Ortego, E., Lillo-Ródenas, M.A., Alcañiz-Monge, J., Cazorla-Amorós, D., Linares-Solano, A., et al., 2009. Isotropic petroleum pitch as a carbon precursor for the
preparation of activated carbons by KOH activation. Carbon 47 (8), 2141–2142.
37. Kureha Corporation, 2008. Expansion of carbon fiber business. Available at: [Link]
38. Grégr, J., 2010. The complete list of commercial carbon fibers. Liberec, Czech Republic.
39. Mora, E., Blanco, C., Prada, V., Menendéz, R., 2002. A study of pitch-based precursors for general purpose carbon fibres. Carbon 40, 2719–2725.
40. Kandiyoti, R., Herod, A., Bartle, K., 2006. Solid Fuels and Heavy Hydrocarbon Liquids – Thermal Characterization and Analysis, first ed. Oxford: Elsevier.
41. Dickakian, G., 1986. Aromatic pitches from the distillate fraction of catalytic cracker bottoms and residue fractions. In: Bacha, J.D., Newman, J.W., White, J.L. (Eds.),
Petroleum-Derived Carbons. Washington, DC: ACS, pp. 126–133.
42. Kim, J., Im, U.-S., Lee, B., Bartle, K., 2016. Pitch-based carbon fibers from coal tar or petroleum residue under the same processing condition. Carbon Letters 19, 72–78.
43. Li, P.-P., Xiong, J.-M., Ge, M.-L., Song, Y.Y., 2015. Preparation of pitch-based general purpose carbon fibers from catalytic slurry oil. Fuel Processing Technology 140,
231–235.
44. Jeon, Y.-P., Zhang, M., Levan, C.D., Ogale, A.A., 2012. High thermal conductivity carbon fibers from petroleum-based experimental mesophase pitch. In: Annual World
Conference on Carbon 2012, Krakow, Poland.
45. Mochida, I., Shimizu, K., Korai, Y., et al., 1992. Mesophase pitch catalytically prepared from anthracene with HF/BF3. Carbon 30 (1), 55–61.
46. Esguerra, D.F., Hoffman, W.P., Thies, M.C., 2014. Liquid crystallinity in trimer oligomers isolated from petroleum and pyrene pitches. Carbon 79, 265–273.
47. Burgess, W.A., Thies, M.C., 2011. Molecular structures for the oligomeric constituents of petroleum pitch. Carbon 49 (2), 636–651.
48. Prauchner, M.J., Pasa, V.M.D., Otani, S., Otani, C., 2005. Biopitch-based general purpose carbon fibers: Processing and properties. Carbon 43, 591–597.
49. Yang, J., Nakabayashi, K., Miyawaki, J., Yoon, S.-H., 2017. Preparation of isotropic spinnable pitch and carbon fiber by the bromination–dehydrobromination of biotar and
ethylene bottom oil mixture. Journal of Materials Science 52, 1165–1171.
50. Tekinalp, H., 2011. Pitch-Based Activated Carbon Fibers: The Effect of Precursor Composition on Pore Structure. Clemson, SC: Clemson University.
51. Cato, A.D., Edie, D.D., 2003. Flow behavior of mesophase pitch. Carbon 41 (7), 1411–1417.
52. Mochida, I., Yoon, S.-H., Korai, Y., 2002. Mesoscopic structure and properties of liquid crystalline mesophase pitch and its transformation into carbon fiber. The Chemical
Record 2 (2), 81–101.
53. Kundu, S., Ogale, A.A., 2006. Rheostructural studies on a synthetic mesophase pitch during transient shear flow. Carbon 44 (11), 2224–2235.
54. Kundu, S., Ogale, A.A., 2010. Rheostructural studies of a discotic mesophase pitch at processing flow conditions. Rheologica Acta 49, 845–854.
55. Frank, E., Steudle, L.M., Ingildeev, D., et al., 2014. Carbon fibers: precursor systems, processing, structure, and properties. Angewandte Chemie International Edition 53
(21), 5262–5298.
56. Ellison, C.J., Phatak, A., Giles, D.W., Macosko, C.W., Bates, F.S., 2007. Melt blown nanofibers: fiber diameter distributions and onset of fiber breakup. Polymer 48 (11),
3306–3316.
57. Yoshiyuki Mishimura, K.-M., 1999. Mesophase pitch-based carbon fiber for use in negative electrode of secondary battery and process for producing the same, Patent No.
US005951959A, Japan.
58. Edie, D.D., 1998. The effect of processing on the structure and properties of carbon fibers. Carbon 36 (4), 345–362.
59. Hamada, T., Nishida, T., Sajiki, Y., Matsumoto, M., 1987. Structures and physical properties of carbon fibers from coal tar mesophase pitch. Journal of Materials Research
2 (6), 850–857.
60. Park, Y.D., Mochida, I., Matsumoto, T., 1988. Extractive stabilization of mesophase pitch fiber. Carbon 26 (3), 375–380.
61. Ogale, A.A., Lin, C., Anderson, D.P., Kearns, K.M., 2002. Orientation and dimensional changes in mesophase pitch-based carbon fibers. Carbon 40, 1309–1319.
62. Kong, L., Liu, H., Cao, W., Xu, L., 2014. PAN fiber diameter effect on the structure of PAN-based carbon fibers. Fibers and Polymers 15 (12), 2480–2488.
63. Peebles, L.H., 1995. Carbon Fibers. Formation, Structure and Properties. Boca Raton, FL: CRC Press.
64. de Bont, P.W., Scholte, P.M.L.O., Hottenhuis, T., et al., 1994. Atomic structure of longitudinal sections of a pitch-based carbon fiber studied by STM. Applied Surface
Science 74, 73–80.
65. Emmerich, F., 2014. Young’s modulus, thermal conductivity, electrical resistivity and coefficient of thermal expansion of mesophase pitch-based carbon fibers. Carbon 79,
274–293.
66. Cho, T., Lee, Y.S., Rao, R., et al., 2003. Structure of carbon fiber obtained from nanotube-reinforced mesophase pitch. Carbon 41, 1419–1424.
67. Yang, H., Yoon, S.-H., Korai, Y., Mocida, I., Katou, O., 2003. Improving graphitization degree of mesophase pitch-derived carbon fiber by solid-phase annealing of spun
fiber. Carbon 41 (3), 397–403.
68. Lu, S., Blanco, C., Rand, B., 2002. Large diameter carbon fibres from mesophase pitch. Carbon 40, 2109–2116.
 Pitch Precursor-Based Carbon Fibers 65

69. Lavin, J.G., Boyington, D.R., Lahijani, J., Issi, J-P., 1993. The correlation of thermal conductivity with electrical resistivity in mesophase pitch-based carbon fiber. Carbon
31 (6), 1001–1002.
70. Cytec Engineered Materials, Thornel P-55 pitch-based fiber. Technical Data Sheet. Available at: [Link]
pdf

Further Reading
Alway-Cooper, R.M., 2014. Structure–Property Relationship of Nanomodified Mesophase Pitch-Based Carbon Fibers. Clemson, SC.
ASTM International, , 1989. D 3379-75: Standard Test Method for Tensile Strength and Young’s Modulus for High-Modulus Single-Filament Materials. West Conshohocken,
PA: ASTM International.
Clarke, A.J., Bailey, J.E., 1973. oxidation of acrylic fibres for carbon fibre formation. Nature 243, 146–150.
Diefendorf, R.J., 1987. Engineered Materials Handbook, vol. 1: Composites. Materials Park, OH: ASM International.
Mitsubishi Rayon Co., Ltd., Pitch-based carbon fiber (CF) DIALEAD. Available at: [Link]
Morales, M.S., Ogale, A.A., 2015. UV-based dual mechanism for crosslinking and stabilization of PAN-based carbon fibers precursors. In: Tiwari, A., Polykarpov, A. (Eds.),
Photocured Materials. London: The Royal Society of Chemistry, pp. 256–289.

Relevant Websites

[Link]
American Chemical Society.
[Link]
Cytec Engineered Materials.
[Link]
Kureha Corporation.
[Link]
Mitsubishi Chemical Carbon Fiber and Composites.
[Link]
Mitsubishi Chemical Corporation.
[Link]
Nippon Graphite Fiber Corporation.