Meltblown Solvated Mesophase Pitch-Based Carbon Fibers: Fiber Evolution and Characteristics

Journal of C Carbon Research

Article Meltblown Solvated Mesophase Pitch-Based Carbon Fibers: Fiber Evolution and Characteristics

Zhongren Yue * ID , Chang Liu and Ahmad Vakili * University of Tennessee Space Institute, 411 B.H. Goethert Parkway, Tullahoma, TN 37388, USA; [email protected] * Correspondence: [email protected] (Z.Y.); [email protected] (A.V.); Tel.: +1-931-393-7362 (Z.Y.); +1-931-393-7483 (A.V.)

Received: 12 July 2017; Accepted: 7 August 2017; Published: 8 August 2017

Abstract: Potentially low-cost continuous carbon fibers are produced from solvated mesophase pitch through a patented meltblowing process. The structural evolution and properties of the fibers are characterized by various analytical methods. The meltblown fibers are continuous fibers which are collected into a fibrous web form, and the diameter of the filaments is attenuated by the flow rate of air streams. The spun fibers can be rapidly stabilized in air due to the high melting mesogens and the removable solvent. The carbonized fibers show a high carbon yield of 75 wt % (or 86 wt % if the solvents are neglected) and a mean diameter of 8–22 µm with typical fiber diameter distribution and variation. The evolution of the fiber structure depends not only on the processing temperature but also on the fiber diameter. The processed carbon fibers retain the same form as the spun fibers and have a low packing density and reasonable mechanical properties.

Keywords: carbon fibers; mesophase pitch; thermal analysis; fiber conversion processes; microstructure

1. Introduction Carbon fibers (CFs) as a reinforcement in composite materials have been widely used in industry mainly due to their high specific strength and modulus. The large scale use of CFs in aircraft and aerospace is driven by maximum performance and fuel efficiency, while the cost factor and the production requirements are not critical. The use of CFs in general engineering and surface transportation is dominated by cost constraints, high production rate requirements, and generally less critical performance needs [1]. Therefore, the large-volume application of commercial CFs in the automotive industry has been hindered due to the high fiber costs and the lack of high-speed composite fabrication techniques [2]. It is believed that the use of low cost precursors, low cost spinning and thermal processes, and mass production is a critical step toward lower cost CFs and their composites for use in multiple industries [2–4]. Polyacrylonitrile (PAN) and conventional mesophase pitch (MP) are the two most important CF precursors for the manufacture of commercial CFs. Over the past two decades, many polymers, such as PE and lignin, have been evaluated as low-cost CF precursor materials. A significant amount of work has been done on engineering new precursors and processing technologies, relating fiber structure to properties, and translating the relationship into production for either reducing production cost [2,5,6] or increasing fiber properties [7–9]. Potentially low-cost CF has been produced from solvated MP pitch using a highly efficient meltblowing process [10]. Unlike conventional MP, solvated MP normally contains at least 70% by volume optical anisotropy, or MP in which at least 50 wt % is MP soluble quinoline insoluble (MSQI) pitch [11,12]. Due to its high MSQI content and adequate utilization of high melting mesogen materials, such an MP is of relatively lower cost. The added solvents make the pitch materials more spinnable at lower temperatures. When the solvent is removed, it has a high melting point, or may be unmeltable, which permits the formation of fibers that require

C 2017, 3, 26; doi:10.3390/c3030026 www.mdpi.com/journal/carbon C 2017, 3, 26 2 of 14 C 2017, 3, 26 2 of 14 removed, it has a high melting point, or may be unmeltable, which permits the formation of fibers that require little or no stabilization as spun [11–13]. It is well known that the meltblowing process is littlewidely or no used stabilization to produce as spuneconomical [11–13 ].nonwoven It is well knownfiber web. that Similarly the meltblowing to the spun-bond process is process, widely usedthe tomelt-blown produce economical process directly nonwoven transforms fiber the web. spun Similarly fibers into to a the nonwoven spun-bond fabric process, in a single, the melt-blownintegrated processprocess directly [14]. Therefore, transforms the the use spun of solvated fibers into MP aan nonwovend meltblow fabric technology in a single, would integrated help to prepare process low- [14]. Therefore,cost CFs. the use of solvated MP and meltblow technology would help to prepare low-cost CFs. MostMost of of the the previous previous studiesstudies of of pitch-based pitch-based CFs CFs have have been been conducted conducted by using by using conventional conventional MP MPand and melt-spinning melt-spinning processes processes [15–22]. [15–22 Very]. Very few few studies studies focused focused on on the the newly newly developed, developed, highly highly efficientefficient meltblowing meltblowing process process of of lower-cost lower-cost solvatedsolvated MP. ItIt was reported that that short short CF CF was was prepared prepared fromfrom naphthalene naphthalene MP MP through through the the meltblowing meltblowing processprocess [[23].23]. However However the the short short fibers fibers will will reduce reduce theirtheir reinforcement reinforcement in in the the composites composites and and limit limit their their application.application. InIn the the present present study, study, low-cost, low-cost, continuous continuous CFs wereCFs were prepared prepared from solvatedfrom solvated MP by MP using by a using patented a meltblownpatented fibermeltblown spinning fiber technique spinning [24 technique] and relatively [24] and fast relatively thermal stabilization fast thermal in stabilization air and carbonization in air and in carbonization in N2. The structure, evolution and properties of the fibers were then characterized N2. The structure, evolution and properties of the fibers were then characterized with various analyses and comparedwith various with commerciallyanalyses and availablecompared CFs. with The commercia results helpedlly available us to optimize CFs. The the CFresults preparation helped process,us to optimize the CF preparation process, to better understand the characteristics of the prepared CFs, to better understand the characteristics of the prepared CFs, and improve their uses in composite materials. and improve their uses in composite materials. 2. Results and Discussion 2. Results and Discussion 2.1. Characteristics of the Spun Fibers 2.1. Characteristics of the Spun Fibers The solvated MP fibers (Figure1) were continuously produced in the meltblown fiber spinning apparatus.The solvated The meltblown MP fibers fibers (Figure were 1) collected were continuously as shown inproduced Figure1 ina, andthe meltblown are continuous fiber longspinning fiber websapparatus. with a width The meltblown and a thickness fibers were of approximately collected as shown 10 cm in and Figure 1 cm, 1a, respectively. and are continuous In the web long of fiber fibers, webs with a width and a thickness of approximately 10 cm and 1 cm, respectively. In the web of individual filaments are continuous fibers and are loosely assembled together, their orientation is fibers, individual filaments are continuous fibers and are loosely assembled together, their controlled if needed, here only slightly toward the length of the web, as shown in Figure1b. The rate orientation is controlled if needed, here only slightly toward the length of the web, as shown in Figure of spinning, compared to the rate of collection of the fibers in a nonvowen web format, determines 1b. The rate of spinning, compared to the rate of collection of the fibers in a nonvowen web format, the orientation of the fibers in the web. The spun fibers are very brittle and need to be very carefully determines the orientation of the fibers in the web. The spun fibers are very brittle and need to be handled when transported. very carefully handled when transported.

Figure 1. Typical continuous pitch fiber web (a) collected from the meltblown fiber spinning Figure 1. Typical continuous pitch fiber web (a) collected from the meltblown fiber spinning apparatus, apparatus, and (b) showing the orientation of the individual long fibers. and (b) showing the orientation of the individual long fibers. Due to the use of a solvated MP precursor, the spun fibers contained solvent. A thermogravimetricDue to the use of analysis a solvated (TGA) MP precursor,running in the N2spun was fibersperformed contained to compare solvent. the A spun thermogravimetric fibers (spun analysiswith a (TGA)hot air flow running rate of in 40 N 2andwas 100 performed L per minute to compare(Lpm), respectively) the spun fibers with (spuntheir precursor with a hot (solvated air flow rateMP) of and 40 andthe conventional 100 L per minute isotropic (Lpm), and MP respectively) materials (both with solvent their precursor free). The (solvatedresults of the MP) TGA and are the conventionalshown in Figure isotropic 2. Without and MP solvents, materials both (bothisotropic solvent and MP free). materials The results showed of a therelatively TGA high are shown weight- in Figureloss starting2. Without point solvents, (335 °C). In both contrast, isotropic the solvated and MP MP materials and its showedfibers showed a relatively a relatively high low weight-loss weight- startingloss starting point(335 point◦C). (175 In °C) contrast, due to the the solvated solvent remo MP andval. itsBased fibers on showed the TGA a relativelyresults, the low solvated weight-loss MP startingand its point fiber contained (175 ◦C) due 12–16 to wt the % solvent solvent. removal. The solvent Based content on the of the TGA spun results, fibers the was solvated less than MP 1% andof itsits fiber precursor, contained indicating 12–16 wt that % solvent.only a small The solvent amount content of solvents of the was spun removed fibers was during less the than spinning 1% of its precursor, indicating that only a small amount of solvents was removed during the spinning process. Although the solvated MP and fibers contained a certain amount solvent, their carbon yields at 900 ◦C C 2017, 3, 26 3 of 14

C 2017, 3, 26 3 of 14 process. Although the solvated MP and fibers contained a certain amount solvent, their carbon yields C 2017, 3, 26 3 of 14 at 900 C were still higher than that of the isotropic and MP (solvent free) materials. The apparent wereprocess.reason still is higherAlthough that the than thesolvated that solvated of MP the MP isotropiccontains and fibers MSQI and contai MP pitch (solventned and a certain adequately free) amount materials. utilizes solvent, The high apparenttheir melting carbon reason yieldsmesogen is that theatmaterials solvated900 C were[11,12]. MP still contains higher MSQI than pitchthat of and the adequately isotropic and utilizes MP (solvent high melting free) materials. mesogen The materials apparent [11 ,12]. reason is that the solvated MP contains MSQI pitch and adequately utilizes high melting mesogen 100 materials [11,12].

10090

9080 Weight (%) Weight

80 Isotropic (free solvent) 70 MF (free solvent)

Weight (%) Weight Solvated MF Solvated MF fiber (100 Lpm) Isotropic Solvated (free MF solvent) fiber (40 Lpm) 70 60 MF (free solvent) 0 Solvated 100 200 MF 300 400 500 600 700 800 900 Solvated MF fiber (100 Lpm) Solvated MF fiber Temperature(40 Lpm) (C) 60 0 100 200 300 400 500 600 700 800 900 Figure 2. Thermogravimetric analysis (TGA) in N2 comparing solvated mesophase pitch (MP) fibers Figure 2. Thermogravimetric analysis (TGA)Temperature in N2 comparing (C) solvated mesophase pitch (MP) fibers withwith their their precursor precursor and and isotropic isotropic and and MP MP materials. materials. Figure 2. Thermogravimetric analysis (TGA) in N2 comparing solvated mesophase pitch (MP) fibers withDue theirto the precursor use of theand meltblowing isotropic and MP process materials. involvin g attenuation via air streams, the diameter of the Duespun to fibers the use can of be the attenuated meltblowing by the process flow rate involving of air streams. attenuation During via airthe streams, meltblowing the diameter process, of the the Due to the use of the meltblowing process involving attenuation via air streams, the diameter of spunprimary fibers gas can is be used attenuated to keep by the the fiber flow centered rate of airand streams. reduce Duringthe diameter the meltblowing of the resulting process, filament the primary as it the spun fibers can be attenuated by the flow rate of air streams. During the meltblowing process, the gascools is used and tosolidifies keep the after fiber exiting centered the andspinneret. reduce The the diametersecondary of gas the is resulting used to filamentfurther entrain, as it cools apply and primary gas is used to keep the fiber centered and reduce the diameter of the resulting filament as it solidifiesand maintain after exiting tension the on spinneret. the filaments The as secondary the air flow gas conveys is used to the further filament entrain,s downward. apply and The maintain fibers cools and solidifies after exiting the spinneret. The secondary gas is used to further entrain, apply tensionspun from on the solvated filaments MP as have the aira very flow rapid conveys quenching the filaments rate and downward. solidify within The fibers a short spun distance from solvated of the and maintain tension on the filaments as the air flow conveys the filaments downward. The fibers MPspinneret have a veryexit [11,12]. rapidquenching Thus, for a rategiven and spinneret solidify he withinad, the a shortdiameter distance of the of spun the spinneretfibers mainly exit varies [11,12 ]. spun from solvated MP have a very rapid quenching rate and solidify within a short distance of the Thus,with for the a flow given rate spinneret of the primary head, the gas. diameter of the spun fibers mainly varies with the flow rate of the spinneret exit [11,12]. Thus, for a given spinneret head, the diameter of the spun fibers mainly varies primarySince gas. the spun fibers are very brittle and not feasible to handle mechanically for testing, the with the flow rate of the primary gas. stabilizedSince thefibers spun were fibers used to are test very and brittledetermine and the not relationship feasible to between handle the mechanically fiber diameter for and testing, the Since the spun fibers are very brittle and not feasible to handle mechanically for testing, the the stabilized fibers were used to test and determine the relationship between the fiber diameter stabilizedflow rate offibers the wereair streams, used to because test and itdetermine was found the that relationship the applied between thermal the processing fiber diameter does and not the seem andflowto thechange rate flow of the ratethe mean air of streams, the diameter air streams, because from becauseit the was spun found it wasfiber that found to the the applied that stabilized the thermal applied fibers. processing thermal Figure processing 3does shows not theseem does fiber not seemtodiameter change to change andthe itsmean the standard mean diameter diameter deviation from from the measured spun the spun fiber by fiber tooptical the to stabilized themicroscopy stabilized fibers. analysis fibers. Figure Figureon 3 the shows3 traverse shows the fiber thesection fiber diameterdiameterof fiber bundle-embedded and and its its standardstandard deviation samples. measured measuredWithin a givenby by optical optical flow microscopy rate microscopy range (40–100analysis analysis Lpm),on the on the thetraverse diameter traverse section sectionof the ofofstabilized fiber fiber bundle-embedded bundle-embedded fibers decreases samples. as the flowWithin Within rate a a givenof given primary flow flow rate rategas range increases. range (40–100 (40–100 Typically, Lpm), Lpm), the all thediameter fiber diameter diameters of the of the stabilizedstabilizedshow a fibershigh fibers standard decreases decreases deviation, as as the the flow flow indicating rate rate of of primary primary that gasthe gas increases.meltblown increases. Typically, Typically,fibers have all all fiber poorfiber diameters diametersdimensional show ashow highuniformity standarda high and standard deviation, need to deviation,be indicating improved. indicating that It theis believed meltblown that the that meltblown fibers this is have due fibers poor to local dimensionalhave turbulence poor dimensional uniformity and can andbe needuniformityimproved to be improved. byand higher need quality Itto is be believed improved. design that of Itthe this is flowbelieved is due stream to that local passages. this turbulence is due to andlocalcan turbulence be improved and can by be higher qualityimproved design by higher of the flowquality stream design passages. of the flow stream passages. 30 100 30 100 25 96 25 20 96 m)

 92 20 m)

 15 92 15 88 10 Diameter ( Diameter 88 yieldMass (%) 10 Diameter ( Diameter Mass yield 84 yieldMass (%) 5 Mass Diameter yield 84 5 Diameter 0 80 0 40 50 60 70 80 90 10080 40Flow 50 rate 60 of 70air streams 80 90 (Lpm) 100 Flow rate of air streams (Lpm) Figure 3. Diameter (measured by optical microscopy analysis of the fiber-bundle-embedded samples) FigureFigureand mass 3. 3.Diameter Diameter yield of stabilized (measured fi by bybers optical optical as a functionmicroscopy microscopy of theanalysis analysis flow ofrate the of ofthe fiber-bundle-embedded the fiber-bundle-embedded air streams. samples) samples) andand mass mass yield yield ofof stabilizedstabilized fi fibersbers as as a a function function of of the the flow flow rate rate of ofthe the air airstreams. streams. C 2017, 3, 26 4 of 14 C 2017, 3, 26 4 of 14

It waswas alsoalso foundfound that that the the high high flow flow rates rates resulted resulted in fibersin fibers having having a relatively a relatively small small diameter diameter and moreand more waves, waves, potentially potentially due due to higher to higher stream stream turbulence. turbulence. In contrast, In contrast, the lowthe low flow flow rate rate allowed allowed the fibersthe fibers to have to have a relatively a relatively large large diameter diameter but to but be straighter.to be straighter. Therefore, Therefore, the next the step next was step to explorewas to theexplore relationship the relationship between between waves and waves air and flow air rate flow and ra tote tryand to to eliminate try to eliminate the fibers’ the fibers’ wave. wave. This could This alsocould be also achieved be achieved in future in future studies studies by applying by applying tension tension to the fibersto the duringfibers during thermal thermal processes. processes.

2.2. Stabilization and CarbonizationCarbonization of the Spun Fibers Since the spun fibersfibers contain high melting mesogens, they they can can be be rapidly rapidly stabilized stabilized in in air. air. Stabilization in air is considered to be the most economical process in the commercial manufacturing of CFs [[25].25]. AA typicaltypical stabilizationstabilization condition condition used used in in this this study study allowed allowed the the fiber fiber to beto be heated heated to 350to 350◦C atC aat heating a heating rate rate of 20of 20◦C/min C/min in in air air while while maintaining maintaining the the fiber fiber form. form. The The heating heating rate rate of of 20 20◦ C/minC/min is muchmuch higherhigher thanthan thethe reportedreported valuesvalues ofof 0.1–20.1–2◦ C/minC/min in the literature [[15–20].15–20]. After stabilization at 350 ◦C for 60 min, the stabilized fibers fibers showed a fi fiberber mass yield of 96–97% as shown in in Figure Figure 33.. This 3–4% weight loss is much lower than 13 wt % solvent content in the spun fibers,fibers, indicating that stabilization of solvated MPMP fibersfibers isis unusual.unusual. To studystudy thethe thermalthermal behaviorbehavior ofof thethe spunspun fiberfiber duringduring stabilization,stabilization, thethe samesame stabilizationstabilization conditions described above were used for TGATGA analysisanalysis (Figure(Figure4 4).). AsAs thethe stabilizationstabilization temperaturetemperature increased, thethe spun spun fiber fiber began began to loseto lose weight weight at 175 at◦C 175 due °C to solventdue to removalsolvent andremoval reached and maximum reached weightmaximum loss weight at 300 ◦ lossC. After at 300 that, °C. theAfter fiber that, gained the fiber weight gained (this weight may start (this at 180may◦ C[start26 at]) due180 °C to oxidation[26]) due (oxygento oxidation is introduced (oxygen is into introduced the fiber into [26– the28]). fiber The [26–28]). maximum The weight maximum gain weight was observed gain was at 350observed◦C at 15at 350 min, °C and at 15 then min, the and weight then lossthe dominatedweight loss thedomi restnated process the rest again. process These again. results These suggest results that, suggest (I) the solvatedthat, (I) the MP solvated fiber loses MP its fiber weight loses (solvents) its weight and (solvents) meanwhile andgains meanwhile weight gains (oxygen) weight in the (oxygen) temperature in the rangetemperature between range 150 andbetween 350 ◦ C;150 and and (II) 350 at °C; lower and stabilization (II) at lower temperatures, stabilization temperatures, weight loss dominates weight loss the thermaldominates process the thermal as the rateprocess of oxygen as the uptakerate of oxygen is lower uptake than the is lower rate of than solvent the removal.rate of solvent Since removal. the spun fiberSince losesthe spun solvent fiber and loses also solvent gains oxygenand also during gains ox stabilization,ygen during the stabilization, stabilized fiberthe stabilized in TGA presents fiber in aTGA total presents weight lossa total of 4%weight after loss stabilization of 4% after at stabilization 350 ◦C for 60 at min. 350 C for 60 min.

100 500

99 400 C) o 98 300

n i

m /

Weight (%) Weight 97 200 C o

0 Temperature ( Temperature 2 96 100

95 0 0 20406080 Time (min) Figure 4. TGA analysis of the spun fibers in air showing weight loss (solvent removal) and weight Figure 4. TGA analysis of the spun fibers in air showing weight loss (solvent removal) and weight gain (oxidation). gain (oxidation). Four stabilized fibers meltblown with four different air flow rates (40, 60, 80, and 100 Lpm) were then carbonizedFour stabilized in N fibers2 under meltblown the same conditionswith four different in a tube air furnace. flow rates The mass(40, 60, yield 80, and 100 mean Lpm) diameter were ◦ ofthen fibers carbonized after carbonization in N2 under at the 1050 sameC conditions are shown in in a Figure tube fu5.rnace. Based The on the mass original yield and weight mean of thediameter spun fibers,of fibers carbonized after carbonization fibers showed at 1050 a 75 wt°C %are yield. shown Disregarding in Figure 5. the Based 13 wt on % solventsthe original in the weight spun of fibers, the thespun carbon fibers, yield carbonized of the resulting fibers showed CFs is a 86 75 wtwt %.% yiel Thed. high Disregarding carbon yield the will13 wt undoubtedly % solvents in benefit the spun the manufacturingfibers, the carbon of thisyield low-cost of the resulting CF. CFs is 86 wt %. The high carbon yield will undoubtedly benefit the manufacturingAs compared with of this the low-cost diameter CF. of the stabilized fibers in Figure3, the diameter of the carbonized fibersAs in Figurecompared5 is significantly with the diameter reduced. of CFsthe withstabilized a mean fibe diameterrs in Figure of 8–22 3, theµm diameter were obtained of the incarbonized this study. fibersThe in Figure fiber diameter 5 is significantly variation reduced. along the CFs fiber with length a mean within diameter 10 mm of and 8–22 diameter m were distribution obtained in were this examinedstudy. with a laser scan micrometer. Typical results are shown in Figure6 comparing four CFs preparedThe fiber from diameter four different variation pitch along fibers the spun fiber with length flow within rates at10 40, mm 60, and 80, diameter and 100Lpm, distribution respectively, were examined with a laser scan micrometer. Typical results are shown in Figure 6 comparing four CFs C 2017, 3, 26 5 of 14 C 2017C 2017, 3, 26, 3, 26 5 of5 of14 14 prepared from four different pitch fibers spun with flow rates at 40, 60, 80, and 100 Lpm, respectively, prepared from four different pitch fibers spun with flow rates at 40, 60, 80, and 100 Lpm, respectively, and two commercially available products, pitch-based P25 and PAN-based CFs. All processed CFs andand two two commercially commercially available available products, products, pitch-based pitch-based P25P25 andand PAN-based CFs. CFs. All All processed processed CFs CFs have a certain degree of fiber diameter distribution and variation along the fiber length. The fibers havehave a certain a certain degree degree of of fiber fiber diameter diameter distribution distributionand and variationvariation along the fiber fiber length. length. The The fibers fibers processed at the low flow rate of 40 Lpm appeared to vary less in diameter along the fiber axis when processedprocessed at theat the low low flow flow rate rate of of 40 40 Lpm Lpm appeared appeared to tovary vary lessless in diameter along along the the fiber fiber axis axis when when compared to the other CFs spun at higher flow rates. comparedcompared to theto the other other CFs CFs spun spun at at higher higher flow flow rates. rates.

30 78 30 78 25 75 25 75 72 20 m) 72

 20 m)

 69 15 69 15 66 66 10 Diameter ( 10 63 yield (%) Mass Diameter ( Mass yield 63 yield (%) Mass 5 Mass yield 5 Diameter 60 Diameter 60 0 57 0 40 50 60 70 80 90 100 57 40 50 60 70 80 90 100 Flow rate of air streams (Lpm) Flow rate of air streams (Lpm) Figure 5. Diameter (measured with a laser scan micrometer) and mass yield of carbonized fibers as a FigureFigure 5. Diameter 5. Diameter (measured (measured with with a a laser laser scanscan micrometer)micrometer) and and mass mass yield yield of ofcarbonized carbonized fibers ﬁbers as a as function of the flow rate of air streams. a functionfunction of of the the ﬂow flow rate rate ofof airair streams.streams.

32 32 32 32 28 28 28 28 24 24 m) m) 24 24   m) m)  20 20

20 20

16 16 16 16 Diameter ( Diameter ( Diameter

12 ( Diameter 12 ( Diameter 12 12 8 8 8 40 Lpm 8 60 Lpm 40 Lpm 60 Lpm 4 4 012345678910114 012345678910114 01234567891011 01234567891011 Position (mm) Position (mm) 32 Position (mm) 32 Position (mm) 32 32 80 Lpm 100 Lpm 28 28 100 Lpm 28 80 Lpm 28 24 24 m) 24 m) 24   m) m)  20 20

20 20

16 16 16 16 Diameter ( Diameter Diameter ( Diameter Diameter ( Diameter 12 12 ( Diameter 12 12 8 8 8 8 4 4 012345678910114 012345678910114 01234567891011 01234567891011 Position (mm) Position (mm) 32 Position (mm) 32 Position (mm) 32 32 P25 28 P25 28 PAN-CF 28 28 PAN-CF 24 24

24 m)

m) 24   m) m)

20 

 20

20 20 16 16 16 16 Diameter ( Diameter Diameter (

12 ( Diameter

Diameter ( 12 12 12 8 8 8 8 4 4 012345678910114 012345678910114 01234567891011 01234567891011 Position (mm) Position (mm) Position (mm) Position (mm)

FigureFigure 6. 6.Diameter Diametervariation variation alongalong the fiber fiber axis axis within within 10 10 mm mm and and diameter diameter distribution distribution of individual of individual Figure 6. Diameter variation along the fiber axis within 10 mm and diameter distribution of individual filamentsfilaments comparing comparing for fordifferent different carboncarbon fibersfibers (CFs). filaments comparing for different carbon fibers (CFs). For meltblown fibers, these changes in diameter compared to the data in the literature [29] are normal and are typical of the fiber diameter distribution and variation. However, when compared C 2017, 3, 26 6 of 14 C 2017, 3, 26 6 of 14 to two commerciallyFor meltblown availablefibers, these CFs, changes P25 and in diameter PAN-CF, co thempared CFs processedto the data byin the literature meltblown [29] method are normal and are typical of the fiber diameter distribution and variation. However, when compared to have a poorer fiber diameter uniformity in terms of the fiber diameter distribution and the diameter two commercially available CFs, P25 and PAN-CF, the CFs processed by the meltblown method have variation along the fiber length. In view of this, improving the uniformity of the spun fiber diameter a poorer fiber diameter uniformity in terms of the fiber diameter distribution and the diameter will be one of the main goals of future research studies. variation along the fiber length. In view of this, improving the uniformity of the spun fiber diameter will be one of the main goals of future research studies. 2.3. Structure Evolution 2.3.The Structure morphological Evolution structure of the spun fibers, stabilized fibers and carbonized fibers at 600 ◦C ◦ and 1050TheC morphological were examined structure using of an the optical spun microscope.fibers, stabilized Typical fibers images and carbonized of the fibers fibers are at 600 shown C in Figureand7 1050. Figure C were7a is examined a cross-sectional using an structure optical microscope. of the spun Typical fibers. images After stabilization,of the fibers are a sheath-coreshown in structureFigure is7. observedFigure 7a is in a largercross-sectional diameter structure (40 Lpm) of the fibers spun as fibers. shown After in Figure stabilization,7b. However, a sheath-core no such a structurestructure is isfound observed in smallerin larger diameter diameter (100(40 Lpm) Lpm) fibers fibers. as shown The sheath in Figure is believed 7b. However, to be composedno such a of well-stabilized,structure is found cross-linked, in smaller and diameter relatively (100 hard Lpm) material fibers. whileThe sheath the core is believed is composed to be ofcomposed less-stabilized of andwell-stabilized, relatively soft materialcross-linked, [30]. and The relatively reason why hard the material stabilized while fibers the exhibit core theis composed sheath–core of structureless- canstabilized be interpreted and relatively as being soft that material when [30]. the The fiber re embeddedason why the in stabilized epoxy resin fibers was exhibit polished, the sheath– the fiber cross-sectioncore structure could can bebe ainterpreted concave surfaceas being due that towh theen the different fiber embedded wear properties in epoxy of resin the was sheath polished, and core materials.the fiber Thecross-section concave could center be(core) a concave surrounded surface du bye to the the edge different (sheath) wear isproperties the explanation of the sheath for the and core materials. The concave center (core) surrounded by the edge (sheath) is the explanation for two-phase structure observed under an optical microscope. This result suggests that the stabilization the two-phase structure observed under an optical microscope. This result suggests that the of the spun fiber is controlled by the diffusion of oxygen from the outside layer to the internal core [31]. stabilization of the spun fiber is controlled by the diffusion of oxygen from the outside layer to the Thus, the smaller diameter fiber is easier to stabilize uniformly, but the larger diameter fiber may internal core [31]. Thus, the smaller diameter fiber is easier to stabilize uniformly, but the larger stabilizediameter gradually fiber may resulting stabilize in gradually a sheath–core resulting structure. in a sheath–core structure.

Figure 7. Optical microscope images showing the cross-sectional structures of (a) spun fibers Figure(40 Lpm); 7. Optical (b) stabilized microscope fibers; images (c) 600 showing C-carbonized the cross-sectional fibers; and (d structures) 1050 C-carbonized of (a) spun fibers. fibers (40 Lpm); (b) stabilized fibers; (c) 600 ◦C-carbonized fibers; and (d) 1050 ◦C-carbonized fibers. A radial crack structure or a “pac-man structure” [32] (fiber with wedge) as shown in Figure 7c,d isA formed radial on crack the structurecarbonized or fibers a “pac-man duringstructure” carbonization. [32] (fiberIt was with found wedge) that the as pac-man shown in structure Figure7 c,d is formedappears on at the600 carbonized°C and become fibers distinctly during carbonization.noticeable at 700 It was°C. From found 800 that °C the onwards pac-man pac-man structure ◦ ◦ ◦ appearsstructures at 600 canC be and observed become in distinctly almost all noticeablefibers with atthe 700 angleC. of From the pac-man 800 C onwards mouth being pac-man the greatest structures canat be 1050 observed °C. in almost all fibers with the angle of the pac-man mouth being the greatest at 1050 ◦C. TheThe crack crack formations formations in MP-based in MP-based CFs CFs have have been been studied studied [33 –[33–38]38] on melt-spunon melt-spun fibers. fibers. Previous Previous papers havepapers shown have that theshown spinning that conditions,the spinning such conditio as thens, spinning such as temperatures, the spinning the temperatures, Length/Depth the of the nozzleLength/Depth dimension, of the the diameter nozzle dimension, and wall material the diameter of the spinningand wall nozzle,material and of the the spinning type of MP nozzle, influence and the crackthe formation type of MP [31 –influence37]. The alignmentthe crack offormation the discotic [31–37]. mesophase The alignment molecules of in the the discotic radial direction mesophase results in the material being fairly weak in the hoop direction, and the hoop stresses generated during the heat C 2017C ,20173, 26, 3, 26 7 of 714 of 14

C 2017, 3, 26 7 of 14 molecules in the radial direction results in the material being fairly weak in the hoop direction, and treatment are relieved by the Pac-man split [18]. The strong radial orientation causes weak interplanar themolecules hoop stresses in the radial generated direction during results the inheat the treatm materialent beingare relieved fairly weak by the in Pac-manthe hoop splitdirection, [18]. Theand bonding leading to “pac-man” splitting during heat treatment as the fiber densifies [38]. However, for the strongthe hoop radial stresses orientation generated causes during weak theinterplanar heat treatm bondingent are leading relieved to ‘‘pac-man’’ by the Pac-man splitting split during [18]. heatThe meltblowntreatmentstrong radial solvated as the orientation fiber MP densifies fibers causes involved [38]. weak However, interplanar in this study, for bonding the it meltblown was leading found solvated to that ‘‘pac-man’’ the MP larger fibers splitting diameter involved during of in fibers heatthis or thestudy, sheath–coretreatment it was as structurethe found fiber that densifies formed the larger [38]. in the However,diameter stabilized offor fibersfibers the meltblown couldor the also sheath–core solvated lead to MP the structure fibers formation involved formed of a in pac-man thisthe structurestabilizedstudy, in it the wasfibers resulting found could that CFs.also the lead larger to the diameter formation of of fibers a pac-man or the structure sheath–core in the structure resulting formed CFs. in the stabilizedForFor smaller smaller fibers diameter coulddiameter also fibers, fibers, lead theto the the sheath–core sheath–core formation of structurestru a cturepac-man was was structure not not observed observed in the on resulting on the the stabilized stabilized CFs. fibers. fibers. As aAs result, a result,For thesmaller the pac-man pac-man diameter structure structure fibers, was thewas sheath–core not not observed observed stru incture the resulting resulting was not CFs.observed CFs. A Atypical typical on the optical opticalstabilized microscopy microscopy fibers. imageimageAs ofa result, theof the CFs the CFs madepac-man made from from structure small-diameter small-diameter was not observed pitchpitch fiberfibe in ther (100 resulting Lpm) Lpm) isCFs. is shown shown A typical in in Figure Figureoptical 8.8 microscopyThere. There is no is no pac-manpac-manimage structure of structurethe CFs in mostmade in most of from the of smallersmall-diameter the smaller diameter diameter pitch CFs, fibe butCFs,r such(100 but aLpm)such structure isa structureshown still existsin stillFigure in exists several 8. There in individualseveral is no individual larger diameter fibers. When viewed with polarized light, optically anisotropic graphitic- largerpac-man diameter structure fibers. Whenin most viewed of the withsmaller polarized diameter light, CFs, optically but such anisotropic a structure graphitic-crystallites still exists in several with crystallites with different orientations on a single fiber cross-section were detected by rotating the differentindividual orientations larger diameter on a single fibers. fiber When cross-section viewed with were polarized detected ligh byt, rotating optically the anisotropic polarized graphitic- lenses. polarizedcrystallites lenses. with different orientations on a single fiber cross-section were detected by rotating the polarized lenses.

Figure 8. Optical microscopy image from rotation of polarized light showing cross-sectional Figure 8. Optical microscopy image from rotation of polarized light showing cross-sectional structures structuresFigure 8. ofOptical the CF microscopymade from smimageall diameter from rotation (100 Lpm) of pitchpolarized fibers. light showing cross-sectional of the CF made from small diameter (100 Lpm) pitch ﬁbers. structures of the CF made from small diameter (100 Lpm) pitch fibers. Typical images of pac-man structures under polarized light microscopy are shown in Figure 9. Typical images of pac-man structures under polarized light microscopy are shown in Figure9. The processedTypical images CF has of pac-mana heterogeneous structures microstructure under polarized which light contains microscopy a pac-man are shown crack, in Figurea radial 9. The processed CF has a heterogeneous microstructure which contains a pac-man crack, a radial transverseThe processed microstructure CF has a heterogeneousof the core, and microstructurea thin layer of skin.which contains a pac-man crack, a radial transversetransverse microstructure microstructure of of the the core, core, and and a a thin thin layer layer ofof skin.skin.

Figure 9. Optical microscope images showing a pac- man structure in larger diameter CFs. Figure 9. Optical microscope images showing a pac-man structure in larger diameter CFs. ScanningFigure 9. electronOptical microscopemicroscopy images (SEM) showing images adisplays pac-man the structure pac-man in larger cracks diameter which CFs.developed alongScanning the fiber electronaxis as shownmicroscopy in Figure (SEM) 10a,b. images The imagesdisplays also the confirmpac-man the cracks appearance which ofdeveloped a radial transversealongScanning the fibermicrostructure electron axis microscopyas shown on the in (SEM)fiberFigure cross-section. images10a,b. The displays imagesIt is evident the also pac-man confirmthat less cracks thepac-man appearance which cracks developed offormed a radial alongin thesmaller fibertransverse axis diameter as microstructure shown CFs, in as Figure shown on the 10 in a,b.fiber Figure The cross-section. 10c,d. images also It is confirm evident the that appearance less pac-man of cracks a radial formed transverse in microstructuresmaller diameter on the CFs, fiber as shown cross-section. in Figure It10c,d. is evident that less pac-man cracks formed in smaller diameter CFs, as shown in Figure 10c,d. C 2017, 3, 26 8 of 14

CC 2017 2017, ,3 3, ,26 26 8 8of of 14 14 The fiber microcrystalline structures were also characterized with an X-ray diffractometer as TheThe fiberfiber microcrystallinemicrocrystalline structures were alsoalso characterizedcharacterized withwith anan X-ray X-ray diffractometer diffractometer as as shown in Figure 11. The spun fiber and the resulting CF display similar diffraction patterns, but the shownshown inin FigureFigure 11.11. TheThe spun fiber and the resultinresultingg CFCF displaydisplay similar similar diffraction diffraction patterns, patterns, but but the the CF has a relatively sharp (002) main peak which poses a higher 2θ angle position and a narrower full CFCF has has a a relativelyrelatively sharpsharp (002) main peak which posesposes aa higherhigher 2 2θθ angle angle position position and and a a narrower narrower full full widthwidthwidth at at halfat halfhalf maximum maximummaximum (FWHM). (FWHM). According According tototo Bragg’sBragg’sBragg’s law law and and Scherer’s Scherer’s equation, equation, equation, used used used in in incarbon carbon carbon researchresearchresearch [39 [39,40], [39,40],,40], the thethe microcrystalline microcrystallinemicrocrystalline size size (Lc) (Lc) of ofof thethe fiber fiberfiber increases increases and and and the the the d-spacing d-spacing d-spacing (d002) (d002) (d002) decreases decreases decreases asas theas the the MP MPMP fiber fiberfiber is is convertedis convertedconverted into into CF. CF.

Figure 10. Scanning electron microscopy (SEM) images showing (a) and (b) larger diameter CFs with Figure 10. Scanning electron microscopy (SEM) images showing (a) and (b) larger diameter CFs with Figurea pac-man 10. Scanning crack along electron the fiber microscopy axis; (c) and (SEM) (d) smallerimages diametershowing (CFs.a) and (b) larger diameter CFs with a pac-mana pac-man crack crack along along the the ﬁber fiber axis; axis; ( c()c) and and ( d(d)) smaller smaller diameterdiameter CFs.

240 320 240 Spun Fiber 320 Carbon Fiber Spun Fiber 280 Carbon Fiber 200 Position FWHM Area % GL 280 25.48 2.510 793.79 100 200 Position FWHM Area % GL 240 Position FWHM Area % GL 24.97 4.813 771.423 41 25.48 2.510 793.79 100 160 Position FWHM Area % GL 240 24.97 4.813 771.423 41 200 160 200

CPS 120 CPS 160

CPS 120 CPS 160 120 80 120 80 80 40 80 40 40 40 0 0 10 20 30 40 50 60 10 20 30 40 50 60 0 0 10 20 30 40 50 60 10 20 30 40 50 60   Figure 11. X-ray diffraction patterns of solvated MP fibers and the resulting CFs. FigureFigure 11. 11.X-ray X-ray diffraction diffractionpatterns patterns ofof solvatedsolvated MP fibers fibers and and the the resulting resulting CFs. CFs. C 20172017,, 33,, 2626 9 of 14 C 2017, 3, 26 9 of 14 2.4. Characteristics of Carbonized Fibers 2.4.2.4. Characteristics ofof CarbonizedCarbonized FibersFibers The processed CFs retain the same form of fiber web as the spun fibers. The typical CF form looksTheThe like processedprocessed a non-woven CFs CFs retain retainweb theas the sameshown same form inform Figure of fiberof fiber 12a. web web The as the asfiber spunthe websspun fibers. arefibers. The continuously typicalThe typical CF formlong. CF looksform The likefilamentslooks a non-wovenlike are a non-wovencontinuous web as shownfibersweb as and in shown Figure somewhat in12 a.Figure The curved fiber 12a. (Figure webs The arefiber 12b). continuously webs As compared are continuously long. with The commercially filaments long. The are continuousavailablefilaments toware fibers continuousfibers, and this somewhat fiber fibers form and curved has somewhat a (Figuremuch curvedlower 12b). packingAs (Figure compared density12b). withAs at compared 0.012 commercially g/cm with3. It iscommercially available noteworthy tow 3 fibers, this fiber form has a much lower packing density at 0.012 g/cm . It is noteworthy3 to mention toavailable mention tow that fibers, the processed this fiber formCFs with has athis much fiber lower form packing are effectively density treatedat 0.012 with g/cm O.3 It gas is noteworthyto improve thattheirto mention the surface processed that chemistry the CFs processed with and this can CFs fiber be with directly form this are fiberused effectively form to fabricate are treated effectively CF with composites Otreated3 gas to with without improve O3 gas weaving their to improve surface and chemistryknittingtheir surface [41]. and chemistry can be directly and can used be to directly fabricate used CF compositesto fabricate withoutCF composites weaving without and knitting weaving [41]. and knitting [41]. (a) (b) (a) (b)

10 cm 10 cm

Figure 12. Typical non-woven CF webs. (a) CF web in a plastic container; (b) fiber orientation under Figure 12. Typical non-woven CF webs. (a) CF web in a plastic container; (b) fiber orientation under opticalFigure microscope.12. Typical non-woven CF webs. (a) CF web in a plastic container; (b) fiber orientation under opticaloptical microscope.microscope. The tensile properties of the CFs processed at a carbonization temperature of 1050 C are shown ◦ in FigureTheThe tensile tensile13 as a propertiesproperties function of ofof the thethe flCFsCFsow processed processrate of edair at atstreams. aa carbonizationcarbonization As seen temperatureintemperature Figure 13, ofofthe 10501050 average CC areare strength shownshown inandin FigureFigure modulus 13 13 as as of a a functionthe function fiber of(8–22 of the the  flow m)flow are rate rate 800–1200 of of air air streams. streams.MPa and As As seen130–160 seen in Figurein GPa, Figure respectively.13, the13, averagethe average The strength maximum strength and modulusvaluesand modulus can of reach the of fiber the 1800 (8–22fiber MPa (8–22µm) and are  m)200 800–1200 are GPa, 800–1200 respectively. MPa andMPa 130–160 and All 130–160 test GPa, values respectively. GPa, for respectively. the Thesamples maximum The have maximum valuesa high candeviation.values reach can 1800 Thereach MPasmaller 1800 and MPadiameter 200 and GPa, 200CFs respectively. GPa, associated respectively. All with test valuesthe All highertest for values the flow samples forrates the haveof samples air a streams high have deviation. exhibita high Thehigherdeviation. smaller average The diameter tensilesmaller CFsstrength diameter associated and CFs modulus. with associated the But higher thewith presence flow the rates higher of of highair flow streamsfiber rates waves exhibitof aircaused streams higher by the average exhibit high tensileflowhigher rate strengthaverage could andtensilereduce modulus. strength the tensile But and theproperties modulus. presence of But ofthe highthe test presence fiber results. waves of high caused fiber by waves the high caused flow by rate the could high reduceflow rate the could tensile reduce properties the tensile of the properties test results. of the test results.

2000 220 2000 (a) 220 (b) (a) 200 (b) 1600 200 1600 180 1200 180 1200 160 160 800 140

800 Modulus (GPa) Strength (MPa) Strength 140 Modulus (GPa) Strength (MPa) Strength 400 Maximum 120 Maximum 400 AverageMaximum 120 AverageMaximum 0 Average 100 Average 0 40 50 60 70 80 90 100 100 40 50 60 70 80 90 100 40Flow 50 rate 60 of primary 70 gas 80 (Lpm) 90 100 40Flow 50 rate 60 of primary 70 gas 80 (Lpm) 90 100 Flow rate of primary gas (Lpm) Flow rate of primary gas (Lpm) Figure 13. Tensile strength (a) and modulus (b) of the processed CFs as a function of the flow rate of FigureairFigure streams. 13.13. TensileTensile strengthstrength ((aa)) andand modulusmodulus ((bb)) ofof thethe processedprocessed CFsCFs asas aa functionfunction ofof thethe flowflow raterate ofof airair streams.streams. The tensile properties of CFs could depend on many factors including precursor materials and processes,TheThe tensiletensile such propertiesasproperties fiber spinning, ofof CFsCFs couldcouldstabilization, dependdepend an onond many manycarbonization factorsfactors includingincluding conditions. precursorprecursor It is well-known materialsmaterials andthatand processes,bothprocesses, the suchtensilesuch asas fiberstrengthfiber spinning,spinning, and modulus stabilization,stabilization, of andpitch-basedand carbonizationcarbonization CFs are conditions.conditions. strongly ItIt dependent isis well-knownwell-known on thatthatthe carbonizationboth the tensile temperature strength [42].and For modulus the CFs preparedof pitch-based in this study,CFs are as thestrongly carbonization dependent temperature on the carbonization temperature [42]. For the CFs prepared in this study, as the carbonization temperature C 2017, 3, 26 10 of 14 bothC 2017 the, 3, tensile26 strength and modulus of pitch-based CFs are strongly dependent on the carbonization10 of 14 temperature [42]. For the CFs prepared in this study, as the carbonization temperature went to 1500 ◦C, thewent average to 1500 strength C, the and average modulus strength could beand up modulus to 2500 MPa could and be 200 up GPa, to 2500 respectively. MPa and A previous200 GPa, studyrespectively. has shown A previous that by study using has separate shown drying that by and using oxidation separate processes, drying and the oxidation solvents processes, in the spun the solvents in the spun fibers are first completely removed in N2, and then the dried fibers are oxidized fibers are first completely removed in N2, and then the dried fibers are oxidized in air. As a result, thein air. tensile As propertiesa result, the of thetensile resulting properties CFs are of remarkablythe resulting improved CFs are [remarkably13]. Efforts areimproved continuing [13]. toward Efforts optimizingare continuing the fabrication toward optimizing process to the reduce fabrication defects pr inocess the fibers to reduce and improvedefects in the the tensile fibers propertiesand improve of suchthe tensile low-cost properties CFs. of such low-cost CFs.

3.3. MaterialsMaterials and and Methods Methods SolvatedSolvated MPMP fromfrom ConocoPhillips ConocoPhillips Co.Co. USAUSA waswas employedemployed asas aa startingstarting materialmaterial forfor spinning spinning pitchpitch ﬁbers. fibers. InIn addition,addition, conventionalconventional isotropicisotropic andand MPMP materialsmaterials (both(both solventsolvent free)free) werewere usedused toto comparecompare with with the the solvated solvated MP. MP. As As a comparison, a comparison, commercially commercially available available pitch-based pitch-based CF Thornel CF Thornel® P25,® USAP25, andUSA PAN-based and PAN-based CF fabric CF (twill fabric weave, (twill ordered weave, fromordered Albany from Engineered Albany Engineered Composites, Composites, Rochester, NH,Rochester, USA) were NH, usedUSA) for were diameter used for measurement. diameter measurement. SolvatedSolvated MP MP was was spun spun into into fibers fibers in in a a meltblown meltblown device. device. The The spinning spinning system system comprised comprised a a pitch pitch feedingfeeding device, device, extruder, extruder, vacuum vacuum vent, vent, ballast ballast pump, pu spinningmp, spinning pump, pump, filter, spinneret filter, spinneret head, fiber head, collector, fiber aircollector, flow jet air regulator, flow jet andregulator, temperature and temperature and pressure, and and pressure, vacuum and control vacuum system, control etc. system, as shown etc. inas Figureshown 14 in. TheFigure fibers 14. were The formedfibers were aerodynamically formed aerodynamically at speeds of several at speeds hundred of several meters hundred per second meters and thenper second slowed and down then to aslowed few meters down per to seconda few meters at the collectionper second zone. at the The collection meltblown zone. device The comprisedmeltblown adevice spinneret comprised head where a spinneret exit orifices head form where filaments, exit aorifices venturi form adjacent filaments, to the spinneretsa venturi thatadjacent receives to thethe filamentsspinnerets from that the receives spinneret the head, filaments a diffuser from near the the spinneret venture that head, receives a diffuser the filaments near fromthe venture the venture, that onereceives or more the air filaments exhaust from ports the that venture, creates in one the or diffuser more air an airflowexhaust with ports a directionthat creates against in the the diffuser direction an ofairflow the flow with of thea direction filaments, against and a the fiber direction collection of bedthe thatflow receives of the filaments, the filaments and ina fiber the form collection of a finite bed widththat receives continuous the non-wovenfilaments in web the form [24]. of a finite width continuous non-woven web [24].

Figure 14. A schematic diagram of a meltblown fiber spinning apparatus and thermal processes for Figure 14. A schematic diagram of a meltblown fiber spinning apparatus and thermal processes manufacturing pitch-based CFs. 1-pitch feeding; 2-extruder; 3-exhaust vent; 4-ballast pump; 5- for manufacturing pitch-based CFs. 1-pitch feeding; 2-extruder; 3-exhaust vent; 4-ballast pump; 5spinning-spinning pump; pump; 66-filter;-filter; 77-spinneret-spinneret head; head; 8-primary-primary air stream; 99-secondary-secondary airair stream;stream; 1010-diffuser;-diffuser; 1111-air-air exhaust; exhaust;12 12-spun-spun fiber fiber web; web;13 13-moving-moving belt belt collector; collector;14 14-stabilization-stabilization oven; oven;15 15-carbonization-carbonization furnace;furnace;16 16-CF-CF web. web.

In this study, the temperature of the spinneret head and primary air was set to 355 C. The number of filaments produced was 10 (adjustable from 1 to 1000). Fibers were collected on a moving belt in a web format. Four pitch fibers with four different nominal diameters corresponding to air flow rates of 40, 60, 80, and 100 Lpm, respectively, were investigated. C 2017, 3, 26 11 of 14

In this study, the temperature of the spinneret head and primary air was set to 355 ◦C. The number of filaments produced was 10 (adjustable from 1 to 1000). Fibers were collected on a moving belt in a web format. Four pitch fibers with four different nominal diameters corresponding to air flow rates of 40, 60, 80, and 100 Lpm, respectively, were investigated. The spun pitch fibers were stabilized in the flowing air by heating them at 20 ◦C/min in an oven from room temperature to 350 ◦C, and holding the temperature constant for 60 min. The stabilized ◦ fibers were then carbonized in a furnace purged with flowing N2, by heating them at 20 C/min to a given temperature, and holding the selected temperature constant for 5 min. No tension was applied on fibers in the above thermal processes. The fiber mass yield (%) was calculated from the measurements in weight changes prior to and after thermal treatments. Weight of the fiber after thermal treatment Fiber mass yield (%) = × 100% (1) Weight of spun fiber

Thermogravimetric analysis, SDT Q600, TGA/DSC combo (TA Instruments, New Castle, DE, USA) was used to measure the weight changes in the pitch and its fiber during the stabilization and carbonization. The tensile properties of CFs and fiber diameter variations were measured with a Diastron Limited FDAS 765 fiber analyzer (Dia-stron Limited, Andover, UK) that consisted of a high resolution tensile tester (LEX 810) with a force resolution of 0.005 gram and a Mitutoyo LSM-500 laser scan micrometer (Dia-stron Limited, Andover, UK) with a positional repeatability of 0.1 µm [43]. The filament was mounted on two plastic tabs with a distance of 10 mm by welding with jeweler’s wax, and placed on a tray. The tray held 15 filaments for tests. Usually 2–4 trays (30–60 filaments) were chosen from each sample for analysis. The laser micrometer secured the fiber in place for measurement of the cross-sectional area using a mechanized holder. The holder rotated and moved horizontally so that measurements could be taken along the length of fiber as well as at different points around the circumference of the fiber. Readings were taken along the length of the fiber at points referred to as “slices.” At each slice, the measurements of the cross-sectional area were taken around the fiber as it was rotated at 20 azimuthal angles, incrementally to 180◦. The cross-sectional areas for each angle of each slice were then averaged to give the mean cross-sectional area of the fiber. In this study, 10 slices were scanned for each fiber with a length of 10 mm. An optical microscope was used to examine the cross-sectional structure of the fibers. Fibers must be carefully prepared before they can be observed under the optical microscope. First, relatively straight portions of fibers were selected from a batch of fibers. Then this portion was cut out and wrapped with a short piece of clear vinyl tubing. Epoxy resin made with a 5:1 ratio of Epoxy 105 and Hardener 206 (West System, Bay City, MI, USA) was then injected into the tubing via a syringe. After the epoxy resin curing, the tubing was then peeled open to release the fiber/epoxy rod sample where fibers were embedded in the epoxy resin with high fiber volume and alignment. The prepared rod samples were then arranged in an orderly fashion and placed into an open cylinder tray where the axes of the samples were aligned with the axis of the cylinder. The same ratio of epoxy/hardener as the above was then also poured into the open cylinder to hold all the different samples in place. After fully curing, the clear epoxy piece was then taken out of the open cylinder and underwent polishing. Once the epoxy piece was polished then it was placed under an optical microscope for observation and imaging of the individual cross-sections of fiber bundles. The optical microscope Olympus® BX60M (Olympus Corporation, Tokyo, Japan) was used in conjunction with the camera Infinity 1 and software Infinity Analyze version 5.0.2 (Lumenera® Corporation, Nepean, Ottawa, ON, Canada) to capture images of the CF cross-sections [43]. The diameter of the stabilized fibers was analyzed with a freeware called ImageJ offered by the National Institute of Health, USA (https://imagej.nih.gov/ij/). Size selection and analysis were set as: size (pixel2) = 4-infinity, circularity = 0.85–1.00, exclude edges, and include holes. The number of C 2017, 3, 26 12 of 14

filaments obtained from each image was about 20–200 depending on the fiber diameter, fiber volume, and fiber alignment in the embedded samples. Usually 2–4 images were analyzed for each sample. The pixels obtained from 130–400 individual fibers were then read for different cross-sectional areas and the average diameter and standard deviation of the fibers were calculated. The morphological structure of CFs was examined with an SEM machine, SUPER-III A, International Scientific Instruments, Pleasanton, CA, USA. The image recorded was processed by the EDS 2004 version 1.3 Rev P (IXRF system, Austin, TX, USA). A thin layer of gold was coated on the fibers with a sputter coater to provide higher resolution images. The microcrystalline structures of pitch fibers and CFs were characterized with an X-ray diffractometer, Philips X’Pert–MPD system, and Philips Analytical X-ray PW 1700 (Philips, Almelo, The Netherlands). The fiber sample was ground into a fine powder and then impregnated with vacuum grease to adhere it to a mold. The wavelengths of the X-rays emitted were 1.54 Angstroms. All samples were set to be scanned from 15 to 65 degrees with a scan rate at 0.01 degrees per two seconds.

4. Conclusions With a patented meltblown high throughput fiber-spinning technique, solvated MP was spun into fibers at four different nominal diameters corresponding to air flow rates of 40, 60, 80, and 100 Lpm, respectively, and the continuous fibers were collected in a nonwoven fibrous web format. The meltblown fibers contained 13 wt % of solvents which was able to be removed by heating. Due to the high content of melting mesogens, the meltblown fiber was able to be rapidly (with a heating rate of 20 ◦C/min) stabilized in air. TGA results show that the meltblown fibers start to lose weight at 175 ◦C due to the solvent removal, and then gain weight at 300 ◦C due to the oxidation of pitch materials. As a result, the stabilized fibers only show 4 wt % weight loss. The carbonized fibers show a high carbon yield of 75 wt % (86 wt % if disregarding the solvents) after carbonization at 1050 ◦C. The diameter of the meltblown fibers was primarily attenuated by the air flow rates in fiber meltblowing process and decreased as the air flow rates increased. The diameter did not change significantly after stabilization but reduced after carbonization. Thus, the processed CFs had an average diameter of 8 to 22 µm depending on the flow rate of air streams. High air flow rates formed a relatively small diameter and more waves. In contrast, low flow rates made the fibers have a relatively large diameter, but be straighter. Typically, the processed CFs had a certain degree of fiber diameter distribution and diameter variation along the fiber length. A sheath–core structure was observed in the stabilized fibers of larger diameter. Such a structure was found to readily result in the formation of pac-man cracks in the resulting CFs. The pac-man cracks developed at 600◦C and the cracks opened larger as the carbonization temperature increased. X-ray diffraction shows that the size of crystallites increased and d-spacing decreased as the MP was converted to CFs. The fiber form was different to that of the commercial CFs. It was a continuous non-woven web with a very low packing density of 0.012 g/cm3, and its filament was continuously long but somewhat curved. The fiber tensile properties were strongly influenced by the carbonization temperatures. At 1050 ◦C, the average tensile strength and modulus of the processed fibers (8–22 µm) were 800–1200 MPa and 130–160 GPa, respectively. At 1500 ◦C, they increased up to 2500 MPa and 200 GPa, respectively.

Acknowledgments: Sincere thanks to Advanced Research Labs and Technical Support at The University of Tennessee Space Institute: Jim Goodman, Joel Davenport, Kate Lansford, Douglas Warnberg, and Gary Payne for their timely technical and machining support and fabrication including numerous helpful discussions. In particular many thanks and appreciation is due to W. Mark Southard and Daniel F. Rossillon for their various tangible and intangible support and advice for systems developments, integration and expert suggestions and advice. The authors would like to thank Federal Transit Administration (FTA), U.S. Department of Transportation (DoT) for their support (contract No. TN-26-7029-00). Author Contributions: Ahmad Vakili conceived the overall spinning idea and process concept, and carried out the fiber spinning process. Zhongren Yue designed the experiments and manuscript. Chang Liu and Zhongren Yue C 2017, 3, 26 13 of 14 carried out the thermal processes and characterization experiments. All the authors have contributed to the writing of this article. Conflicts of Interest: The authors declare no conflict of interest.

References

1. Chand, S. Review Carbon fibers for composites. J. Mater. Sci. 2000, 35, 1303–1313. [CrossRef] 2. Huang, X. Fabrication and properties of carbon fibers. Materials 2009, 2, 2369–2403. [CrossRef] 3. Oak Ridge National Laboratory. Transportation Technology Program. Annual Report; 2008. Available online: https://energy.gov/sites/prod/files/2014/03/f8/ornl_ttp_report_fy08.pdf (accessed on 12 February 2017). 4. Sullivan, R.A. Automotive carbon fiber: Opportunities and challenges. JOM 2006, 11, 77–79. [CrossRef] 5. Hosseinaei, O.; Harper, D.P.; Bozell, J.J.; Rials, T.G. Role of physicochemical structure of organosolv hardwood and herbaceous lignins on carbon fiber performance. ACS Sustain. Chem. Eng. 2016, 4, 5785–5798. [CrossRef] 6. Baker, D.A.; Rials, T.G. Recent advances in low-cost carbon fiber manufacture from lignin. J. Appl. Polym. Sci. 2013, 130, 713–728. [CrossRef] 7. Khayyam, H.; Fakhrhoseini, S.M.; Church, J.S.; Milani, A.S.; Bab-Hadiashar, A.; Jazar, R.; Naebe, M. Predictive Modelling and Optimization of Carbon Fiber Mechanical Properties through High Temperature Furnace. Appl. Therm. Eng. 2017. Available online: http://dx.doi.org/10.1016/j.applthermaleng.2017.06.071 (accessed on 1 July 2017). 8. Liu, J.; Yue, Z.; Fong, H. Continuous nanoscale carbon fibers with superior mechanical strength. Small 2009, 5, 536–542. [CrossRef][PubMed] 9. Chae, H.G.; Newcomb, B.A.; Gulgunje, P.V.; Liu, Y.; Gupta, K.K.; Kamath, M.G.; Lyons, K.M.; Ghoshal, S.; Pramanik, C.; Giannuzzi, L. High strength and high modulus carbon fibers. Carbon 2015, 93, 81–87. [CrossRef] 10. Vakili, A.; Yue, Z.; Fei, Y.; Cochran, H.; Allen, L.; Duran, M. Pitch based carbon fiber processing and composites. In Proceedings of the SAMPE Fall Technical Conference (39th ISTC), Cincinnati, OH, USA, 29 October–1 November 2007. 11. Walter, M.; Kalback, H.; Romine, E.; Bourrat, X.M. Solvated Mesophase Pitches. U.S. Patent 5,259,947, 9 November 1993. 12. Southard, W.M.; Romine, H.E.; Nanni, E.J.; Carel, M.W. Process for Making Solvated Mesophase Pitch. U.S. Patent 5,437,780, 1 August 1995. 13. Yue, Z.; Liu, C.; Vakili, A. Solvated mesophase pitch-based carbon fibers: Thermal-oxidative stabilization of the spun fiber. J. Mater. Sci. 2017, 52, 8176–8187. [CrossRef] 14. Silva, E. The Melt-Blowing Process. Available online: https://www.academia.edu/440299/The_melt- blowing_process_-_Nonwovens (accessed on 21 July 2017). 15. Mochida, I.; Ling, L.; Korai, Y. Some factors for the high performances of mesophase pitch based carbon fibre. J. Mater. Sci. 1994, 29, 3050–3056. [CrossRef]

16. Tomioka, T.; Arai, Y.; Hamada, T. Pitch-based carbon fibres derived from thermoset fibres oxidized with Cl2 containing air. J. Mater. Sci. 1995, 30, 1570–1576. [CrossRef] 17. Yao, Y.; Chen, J.; Liu, L.; Dong, Y.; Liu, A. Mesophase pitch-based carbon fiber spinning through a filter assembly and the microstructure evolution mechanism. J. Mater. Sci. 2014, 49, 191–198. [CrossRef] 18. Ogale, A.A.; Lin, C.; Anderson, D.P.; Kearns, K.M. Orientation and dimensional changes in mesophase pitch-based carbon fibers. Carbon 2002, 40, 1309–1319. [CrossRef] 19. Wazir, A.H.; Kakakhel, L. Preparation and characterization of pitch-based carbon fibers. New Carbon Mater. 2009, 24, 83–88. [CrossRef] 20. Lu, S.; Blanco, C.; Rand, B. Large diameter carbon fibres from mesophase pitch. Carbon 2002, 40, 2109–2116. [CrossRef] 21. Mora, E.; Blanco, C.; Santamaría, R.; Granda, M. A novel method to obtain a petroleum-derived mesophase pitch suitable as carbon fibre precursor. Carbon 2003, 41, 445–452. [CrossRef] 22. MInus, M.; Kumar, S. The processing, properties, and structure of carbon fibers. JOM 2005, 57, 52–58. [CrossRef] 23. Watanabe, F.; Korai, Y.; Mochida, I.; Nishimura, Y. Structure of melt-blown mesophase pitch-based carbon fiber. Carbon 2000, 38, 741–747. [CrossRef] C 2017, 3, 26 14 of 14

24. Vakili, A.; Rossillon, D. Process for Collection of Continuous Fibers as a Uniform Batt. U.S. Patent 8,206,640, 26 June 2012. 25. Morgan, P. Carbon Fibers and Their Composites; CRC Press: Boca Raton, FL, USA, 2005. 26. Singer, L.S.; Mitchell, S. Diffusion of oxygen into pitch. Carbon 1997, 35, 599–604. [CrossRef] 27. Matsumoto, T.; Mochida, I. A structural study on oxidative stabilization of MP pitch fibers derived from coal tar. Carbon 1992, 30, 1041–1046. [CrossRef] 28. Lavin, J.G. Chemical reactions in the stabilization of mesophase pitch-based fibers. Carbon 1992, 30, 351–357. [CrossRef] 29. Bresee, R.R.; Qureshi, U.A. Influence of process conditions on melt blown web structure. Part IV—Fiber diameter. J. Eng. Fibers Fabr. 2006, 1, 32–46. 30. Otani, S. Mechanism of the carbonization of mesophase carbon fiber at the low temperature range. Carbon 1967, 5, 219–220. [CrossRef] 31. Mochida, I.; Toshima, H.; Korai, Y.; Hino, T. Oxygen distribution in the mesophase pitch fiber after oxidative stabilization. J. Mater. Sci. 1989, 24, 389–394. [CrossRef] 32. Gerald, J.D.F.; Pennock, G.M.; Taylor, G.H. Domain structure in MP (mesophase pitch)-based fibres. Carbon 1991, 29, 139–164. [CrossRef] 33. Donnet, J.B.; Qin, R.Y. Study of carbon fiber surfaces by scanning tunnelling microscopy, part I. Carbon fibers from different precursors and after various heat treatment temperatures. Carbon 1992, 30, 787–796. [CrossRef] 34. Yoon, S.H.; Takano, N.; Korai, Y.; Mochida, I. Crack formation in mesophase pitch based carbon fibers: Part I Some influential factors for crack formation. J. Mater. Sci. 1997, 32, 2753–2758. [CrossRef] 35. Yoon, S.H.; Korai, Y.; Mochida, I. Crack formation in mesophase pitch based carbon fibres: Part II Detailed structure of pitch based carbon fibres with some types of open cracks. J. Mater. Sci. 1997, 32, 2759–2769. [CrossRef] 36. Matsumoto, M.; Iwashita, T.; Arai, Y.; Tomioka, T. Effect of spinning conditions on structures of pitch-based carbon fiber. Carbon 1993, 31, 715–720. [CrossRef] 37. Matsumoto, T. Mesophase pitch and its carbon fibers. Pure Appl. Chem. 1985, 57, 1553–1562. [CrossRef] 38. Alway-Cooper, R.M.; Anderson, D.P.; Ogale, A.A. Carbon black modification of mesophase pitch-based carbon fibers. Carbon 2013, 59, 40–48. [CrossRef] 39. Wang, P.H.; Liu, J.; Yue, Z.R.; Li, R.Y. Thermal oxidative stabilization of polyacrylonitrile precursor fiber—Progression of morphological structure and mechanical properties. Carbon 1992, 30, 113–120. [CrossRef] 40. Shirvanimoghaddam, K.; Abolhasani, M.M.; Li, Q.; Khayyam, H.; Naebe, M. Cheetah skin structure: A new approach for carbon-nano-patterning of carbon nanotubes. Compos. Part A 2017, 95, 304–314. [CrossRef] 41. Yue, Z.; Vakili, A.; Duran, M.P. Surface treatments of solvated mesophase pitch-based carbon fibers. J. Mater. Sci. 2017, 52, 10250–10260. [CrossRef] 42. Bright, A.A.; Singer, L.S. The electronic and structural characteristics of carbon fibers from mesophase pitch. Carbon 1979, 17, 59–69. [CrossRef] 43. Liu, C. Mesophase Pitch-Based Carbon Fiber and Its Composites: Preparation and Characterization. Master’s Thesis, The University of Tennessee, Knoxville, TN, USA, 2011.

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).