Optimization of Process Conditions for Continuous Growth of Cnts on the Surface of Carbon Fibers

Article Optimization of Process Conditions for Continuous Growth of CNTs on the Surface of Carbon Fibers

Chengjuan Wang 1,2, Yanxiang Wang 1,2,* and Shunsheng Su 2

1 Key Laboratory for Liquid-Solid Structural Evolution and Processing of Materials (Ministry of Education), State Key Laboratory of Crystal Materials, Shandong University, Jinan 250061, China; [email protected] 2 Carbon Fiber Engineering Research Center, School of Materials Science and Engineering, Shandong University, Jinan 250061, China; [email protected] * Correspondence: [email protected]

Abstract: Grafting carbon nanotubes (CNTs) is one of the most commonly used methods for modify- ing carbon fiber surface, during which complex device is usually needed and the growth of CNTs is difficult to control. Herein, we provide an implementable and continuous chemical vapor deposition (CVD) process, by which the novel multiscale reinforcement of carbon nanotube (CNT)-grafted carbon fiber is prepared. After exploring the effects of the moving speed and growth atmosphere on the morphology and mechanical properties of carbon nanotubes/carbon fiber (CNTs/CF) reinforcement, the optimal CVD process conditions are determined. The results show that low moving speeds of carbon fibers passing through the reactor can prolong the growth time of CNTs, increasing the thickness and density of the CNTs layer. When the moving speed is 3 cm/min or 4 cm/min, the surface graphitization degree and tensile strength of CNTs/CF almost simultaneously reach the

highest value. It is also found that H2 in the growth atmosphere can inhibit the cracking of C2H2 and has a certain effect on prolonging the life of the catalyst. Meanwhile, the graphitization degree

is promoted gradually with the increase in H2 ﬂow rate from 0 to 0.9 L/min, which is beneﬁcial to CNTs/CF tensile properties.

Citation: Wang, C.; Wang, Y.; Su, S. Optimization of Process Conditions Keywords: CNTs/CF; CVD; graphitization degree; growth atmosphere for Continuous Growth of CNTs on the Surface of Carbon Fibers. J. Compos. Sci. 2021, 5, 111. https://doi.org/10.3390/jcs5040111 1. Introduction Carbon fiber has a series of excellent features such as high modulus, high strength, Academic Editor: Michela Simoncini low density, high temperature resistance, and remarkable electrical conductivity. Therefore, it is often chosen as a reinforcing material to composite with a resin matrix, widely used Received: 29 March 2021 in aerospace, transportation, new energy, building facilities, medical equipment, and Accepted: 14 April 2021 military fields [1–3]. As the reinforcement, the performances of carbon fibers and the Published: 17 April 2021 interfacial characteristics of the composites are important factors to determine whether the composites are excellent or not. However, during the carbonization treatment at high Publisher’s Note: MDPI stays neutral temperatures with inert gas, carbon fibers will meet the process of non-carbon element with regard to jurisdictional claims in escaping and carbon enrichment, which will reduce the surface activity and surface tension published maps and institutional affil- iations. of carbon fibers [4,5]. In addition, in order to ensure the tensile strength of carbon fibers, surface defects should be avoided as much as possible, so the surface of carbon fibers is smooth, and the specific surface area is not large. For these reasons, carbon fibers cannot be firmly compounded with the matrix material, which greatly reduces the properties of the composites. Thus, it is very important to modify the surface of carbon fiber to obtain Copyright: © 2021 by the authors. proper surface roughness and activity. Licensee MDPI, Basel, Switzerland. At present, a great number of traditional surface modification methods of carbon This article is an open access article fiber have been proposed to improve the wettability between carbon fibers and matrix distributed under the terms and materials, such as plasma modification, chemical oxidation, and electrochemical treat- conditions of the Creative Commons Attribution (CC BY) license (https:// ments [6–9]. Even though these methods do create the desired surface chemistry, the creativecommons.org/licenses/by/ effects on the base fiber strength are adverse, resulting in poor mechanical properties. By 4.0/).

J. Compos. Sci. 2021, 5, 111. https://doi.org/10.3390/jcs5040111 https://www.mdpi.com/journal/jcs J. Compos. Sci. 2021, 5, 111 2 of 12

contrast, grafting carbon nanotubes (CNTs) is a more effective and moderate way [10,11]. The introduction and uniform distribution of CNTs on the surface of carbon fibers can significantly improve the mechanical properties of CNTs-grafted carbon fibers reinforced composites, and endow them with excellent electrical and thermal properties, expanding their application fields. Chemical vapor deposition (CVD) is usually used to prepare carbon nanotubes/carbon fiber (CNTs/CF) multiscale reinforcement [12–14], which can improve the adhesion between carbon fibers and the matrix [15–17], inhibit the crack growth and reduce the interfacial stress concentration [18–20]. An et al. [21] used a catalytic chemical vapor deposition (CCVD) method to grow vertically aligned CNTs arrays onto the carbon fiber fabric and found a nearly 110% increase in interfacial shear strength for the micro-droplet composite. Aerosol-assisted chemical vapor deposition was also developed to prepare carbon nanotube-hybridized carbon fiber [22]. Zhao et al. [23] reported an automatic fabrication of CNT fiber fabrics, which was achieved through knitting yarns composed of CNT fibers that are produced from a floating catalyst chemical vapor deposition. This kind of fabrics was proved to be flexible, lightweight, mechanically strong, and electrically and thermally conductive. Luo et al. [24] demonstrated the potential of carbon fibers modified by the vertical carbon nanotubes/polypyrrole composites for improving the electricity generation performances of the microbial fuel cells. However, ordinary CVD techniques are not able to achieve continuous industrial production of CNTs/CF. Su et al. [25] provided a continuous method for grafting CNTs on the surface of carbon fibers by CVD, where the control of the morphology of CNTs/CF could be achieved by adding thiourea to the cobalt nitrate precursor. Yao et al. [26] developed a bimetallic catalyst for the growth of CNTs on the surface of carbon fibers and found that the catalytic efficiency is the highest when the atomic fraction of Fe is 1:1. Furthermore, the continuous method mentioned above would offer a promising technique for CNTs-grown carbon fibers online manufacture [27]. The growth of CNTs could also be achieved at ultra-low temperatures using the one-step method [28]. In the CVD process, in addition to the growth temperature [29] and catalyst system [30], the growth time and growth atmosphere [10] of CNTs have a great influence on the surface morphology and mechanical properties of CNTs/CF. In this paper, this innovative continuous CVD process was used to grow CNTs on the surface of carbon fibers, with the mixture of C2H2 and H2 as the growth atmosphere. By changing the moving speeds of carbon fibers and the volume ratio of H2 to C2H2 in the mixed gas to adjust the growth time and atmosphere, a series of CNTs/CF multiscale reinforcements were synthesized under different process conditions. And the effects of process parameters such as carbon fiber moving speeds and mixed gas compositions on CNTs/CF multiscale reinforcements were explored and verified. J. Compos. Sci. 2021, 5, x FOR PEER REVIEW 3 of 12

2. Materials and Methods 2.1. Experiment Method

be carriedThe loading out synchronously, of the catalyst which precursor was and called the the continuous one-step growth method, of CNTswhile werethe N realized2 valve wasby using opened the from equipment the beginning designed of in the the experiment. laboratory, as shown in Figure1.

Figure 1. SchematicSchematic diagram diagram of catalyst loading and carbon nanotubes (CNTs) growth equipment.equipment.

Carbon fibers were driven by pulleys running at a directional and uniform speed. The moving speed of carbon fibers was adjusted by tuning the running speed of pulleys to control CNTs growth time, and the surface morphology and mechanical properties of CNTs/CF were studied. In addition, a variety of CNTs/CF was prepared by changing the flow rate of H2 and C2H2 in the growth atmosphere. The flow rate of C2H2 remained 0.3 L/min, while H2 flow rate increased as manifested in Table 1, influencing the volume ratio of H2 to C2H2 and gas composition in the reactor.

Table 1. Different gas compositions in the growth atmosphere (The unit of flow rate is L/min).

Gas Flow Rate 1 Flow Rate 2 Flow Rate 3 Flow Rate 4 Flow Rate 5 H2 0 0.15 0.3 0.6 0.9 C2H2 0.3 0.3 0.3 0.3 0.3

2.2. Characterization The surface morphology of carbon fibers and multiscale reinforcements and the microstructure of CNTs was observed by field emission scanning electron microscope (SEM, SU-70) and high-resolution transmission electron microscope (HRTEM, JEM-2100). The degree of graphitization on the surface of multiscale reinforcements was characterized using the LabRAM-HR800 Raman spectrometer. The test condition was He-Ne light source, the wavelength was 633 nm, the resolution was 1 cm−1, and the test range was 500– 2000 cm−1. The change of graphite microcrystals on the surface of carbon fiber was analyzed by X-ray diffraction (XRD, Rigaku D/max-RC). The single-filament tensile strength of carbon fiber was tested according to ASTM D3822-07 standard. Samples were made according to Figure 2, and a single carbon fiber of appropriate length was fixed on the paper frame cut by a piece of mark paper with strong glue. After the glue was cured, two ends of the sample were clamped on the chuck of the XQ-1C fiber strength tester, and the paper pieces on both sides of the carbon fiber were cut with scissors to start the test. Stretching rate was 2 mm/min during the test. At least 40 effective samples were prepared for each condition, and the average value was taken. The single-filament tensile strength was calculated according to the following For- mula. 4F δ = (1) π 2 d In the Formula, δ is carbon fiber single-filament tensile strength (Pa), F means carbon fiber single-filament fracture load (N), d is single carbon fiber diameter (m). The average diameter of carbon fiber took 7 μm.

J. Compos. Sci. 2021, 5, 111 3 of 12

Polyacrylonitrile-based carbon fibers (T700-12000, Toray, Tokyo, Japan) were selected as original fibers for the growth of CNTs. The surface of desized carbon fiber was pretreated by electrochemical anodic oxidation (EAO). The mixed solution of cobalt nitrate hexahy- drate (Co(NO3)2·6H2O, 0.05 mol/L, alcohol solution) and thiourea (CH4N2S, 0.02 mol/L, alcohol solution) was chosen as the precursor of catalyst. A tube furnace was used as the reactor, with a 30 cm long effective heating zone, and the diameter was 60 mm. In addition, the reducing agent for the catalyst was H2, the carbon source was C2H2, and N2 was adopted as the shielding gas. In the course of the experiment, the growth temperature ◦ was 600 C. H2 and C2H2 were inserted at the same time, after the furnace temperature reaching about 600 ◦C, to make the reduction of the catalyst and the growth of CNTs be carried out synchronously, which was called the one-step method, while the N2 valve was opened from the beginning of the experiment. Carbon fibers were driven by pulleys running at a directional and uniform speed. The moving speed of carbon fibers was adjusted by tuning the running speed of pulleys to control CNTs growth time, and the surface morphology and mechanical properties of CNTs/CF were studied. In addition, a variety of CNTs/CF was prepared by changing the flow rate of H2 and C2H2 in the growth atmosphere. The flow rate of C2H2 remained 0.3 L/min, while H2 flow rate increased as manifested in Table1, influencing the volume ratio of H2 to C2H2 and gas composition in the reactor.

Table 1. Different gas compositions in the growth atmosphere (The unit of ﬂow rate is L/min).

Gas Flow Rate 1 Flow Rate 2 Flow Rate 3 Flow Rate 4 Flow Rate 5

H2 0 0.15 0.3 0.6 0.9 C2H2 0.3 0.3 0.3 0.3 0.3

2.2. Characterization The surface morphology of carbon fibers and multiscale reinforcements and the microstructure of CNTs was observed by field emission scanning electron microscope (SEM, SU-70) and high-resolution transmission electron microscope (HRTEM, JEM-2100). The degree of graphitization on the surface of multiscale reinforcements was characterized using the LabRAM-HR800 Raman spectrometer. The test condition was He-Ne light source, the wavelength was 633 nm, the resolution was 1 cm−1, and the test range was 500–2000 cm−1. The change of graphite microcrystals on the surface of carbon fiber was analyzed by X-ray diffraction (XRD, Rigaku D/max-RC). The single-filament tensile strength of carbon fiber was tested according to ASTM D3822-07 standard. Samples were made according to Figure2, and a single carbon fiber of appropriate length was fixed on the paper frame cut by a piece of mark paper with strong glue. After the glue was cured, two ends of the sample were clamped on the chuck of the XQ-1C fiber strength tester, and the paper pieces on both sides of the carbon fiber were cut with scissors to start the test. Stretching rate was 2 mm/min during the test. At least 40 effective samples were prepared for each condition, and the average value was taken. The single-filament tensile strength was calculated according to the following Formula.

4F δ = (1) πd2 In the Formula, δ is carbon fiber single-filament tensile strength (Pa), F means carbon fiber single-filament fracture load (N), d is single carbon fiber diameter (m). The average diameter of carbon fiber took 7 µm. J. Compos. Sci. Sci. 2021,, 5,, x 111 FOR PEER REVIEW 4 4of of 12

FigureFigure 2. SchematicSchematic for for the the single single fiber ﬁber tensile testing.

3.3. Results Results and and Discussion Discussion 3.1. Effect of Moving Speeds on Surface Morphology of Multiscale Reinforcements 3.1. Effect of Moving Speeds on Surface Morphology of Multiscale Reinforcements Figure3 shows the surface morphology of CNTs/CF multiscale reinforcements de- Figure 3 shows the surface morphology of CNTs/CF multiscale reinforcements deposited at different moving speeds, respectively. It can be seen from the figure that due to posited at different moving speeds, respectively. It can be seen from the figure that due to the short deposition time, CNTs on the carbon fiber surface were sparse, short and uneven the short deposition time, CNTs on the carbon fiber surface were sparse, short and uneven when the moving speed was 10 cm/min. When it came to 6 cm/min, the distribution of when the moving speed was 10 cm/min. When it came to 6 cm/min, the distribution of CNTs on the surface of carbon fibers became uniform, the length was longer, and CNTs CNTs on the surface of carbon fibers became uniform, the length was longer, and CNTs between the two adjacent fibers were entangled with each other. At 4 cm/min or 3 cm/min, between the two adjacent fibers were entangled with each other. At 4 cm/min or 3 cm/min, the growth time of CNTs was increased even more, making carbon atoms in the catalyst the growth time of CNTs was increased even more, making carbon atoms in the catalyst particles be able to diffuse and precipitate thoroughly on the surface of carbon fibers. particles be able to diffuse and precipitate thoroughly on the surface of carbon fibers. Con- Consequently, the length of CNTs became longer and CNTs were distributed more uniform sequently, the length of CNTs became longer and CNTs were distributed more uniform and denser. As the growth time was prolonged to 15 min, a large amount of amorphous andcarbon denser. was formedAs the growth on the surfacetime was of carbonprolonge fibers,d to resulting15 min, a inlarge the amount hardening of ofamorphous the CNTs carbonlayer. Inwas the formed box indicated on the surface by the of yellow carbon arrow fibers, in resulting Figure3 e,in thethe fracturehardening surface of the ofCNTs the layer.carbon In fibers the box could indicated be meshed by the with yellow each other, arrow indicating in Figure that 3e, carbonthe fracture fibers surface were initially of the carbonconnected fibers by could the entanglement be meshed with of the each CNTs other, layer, indicating and then that the carbon brittle fracturefibers were occurred initially in connectedthe connecting by the area entanglement in the process of of the sample CNTs preparation. layer, and then Plenty the of brittle amorphous fracture carbon occurred was infilled the inconnecting the gap of area the in CNTs the process layer in of Figure samp3lef, whichpreparation. made CNTsPlentybond of amorphous with each carbon other, wasleading filled to in the the binding gap of ofthe the CNTs CNTs layer layer in andFigure the 3f, increase which inmade brittleness CNTs bond at the with adjacent each other,carbon leading fiber joints. to the binding of the CNTs layer and the increase in brittleness at the adjacent carbonFigure 4fiber shows joints. the changes of Raman curves and R values of CNTs/CF multiscale reinforcementsFigure 4 shows grown the at changes various movingof Raman speeds. curves The and two R peaksvalues of of CNTs/CF CNTs/CF exhibited multiscale in reinforcementsFigure4 represent grown the disorderedat various moving structure spee at aboutds. The 1350 two cm peaks−1 (D of peak)CNTs/CF and exhibited the ordered in Figuregraphite 4 represent structure the at about disordered 1600 cmstructure−1 (G peak), at about respectively. 1350 cm−1 (D Usually, peak) theand higher the ordered the R −1 graphite(ID/IG) value structure is, the at worse about the 1600 graphitization cm (G peak), degree respectively. of the sample Usually, is [31 ].the It canhigher be foundthe R (Ithat,D/IG) from value 10 is, cm/min the worse to 3the cm/min, graphitization with the degree decrease of the of thesample moving is [31]. speed, It can the be growth found that,time from of CNTs 10 cm/min became to longer, 3 cm/min, and with the Rthevalue decrease of samples of the moving went smaller speed, the gradually. growth time This ofmanifested CNTs became that under longer, the and repairing the R value effect of CNTs,samples the went graphite smaller microcrystalline gradually. This structure mani- festedon the that fiber under surface the was repairing perfected, effect the of surface CNTs, defects the graphite of carbon microcrystalline fibers decreased, structure bringing on theout fiber an increase surface in was the orderingperfected, of the carbon surface atoms. defects As depositionof carbon timefibers was decreased, extended bringing further, outthe an number increase of CNTsin the andordering the density of carbon of CNTsatoms. layer As deposition increased, time the defectswas extended on the further, surface theof carbon number fibers of CNTs due toand EAO the were density repaired, of CNTs which layer improved increased, the the degree defects of graphitization.on the surface ofWhen carbon the fibers moving due speed to EAO was were 4 cm/min repaired, or which 3 cm/min, improved the R thevalue degree was of similar, graphitization. and the WhenR value the decreased moving speed to the was lowest 4 cm/min level, illustratingor 3 cm/min, that the the R value degree was of similar, graphitization and the on R valuethe surface decreased of the to samplesthe lowest was level, the illustrati highest.ng The thatR thevalue degree suddenly of graphitization increased to on 0.791 the surfacein 2 cm/min of the, indicatingsamples was that the the highest. samples The had R a pronouncedvalue suddenly reduction increased of graphitization to 0.791 in 2 cm/min,degree. Thisindicating was mainly that the due samples to the had fact a pronounced that the CVD reduction process tookof graphitization too long, so degree. a large Thisnumber was ofmainly catalyst due particles to the fact lost that their the catalytic CVD process activity. took Thus, too long, active so carbon a large atoms number were of catalystexcessive particles and fallen lost ontheir the catalytic sample activity. surface. Thus, active carbon atoms were excessive and fallen on the sample surface.

J. Compos. Sci. 2021, 5, 111 5 of 12 J. J.Compos. Compos. Sci. Sci. 2021 2021, 5, ,5 x, xFOR FOR PEER PEER REVIEW REVIEW 5 of5 of12 1 2

FigureFigure 3. 3.3. SEM SEMSEM micrographs micrographs micrographs of of ofCNTs/carbon CNTs/ CNTs/carboncarbon fiber fiber ﬁber (CF) (CF) (CF)mu multiscaleltiscale multiscale reinforcements reinforcements reinforcements prepared prepared prepared at at dif- dif- at ferent moving speeds: (a) 10 cm/min; (b) 6 cm/min; (c) 4 cm/min; (d) 3 cm/min; (e) 2 cm/min and differentferent moving moving speeds speeds:: (a) ( a10) 10 cm/min cm/min;; (b) (6b )cm/min; 6 cm/min; (c) 4 ( ccm/min;) 4 cm/min; (d) 3 ( dcm/min;) 3 cm/min; (e) 2 cm/min (e) 2 cm/min and (f()f )the the expanding expanding image image of of ( e().e). The The flow flow rate rate of of H H2 and2 and C C2H2H2 was2 was 0.3 0.3 L/min. L/min. and (f) the expanding image of (e). The ﬂow rate of H2 and C2H2 was 0.3 L/min.

FigureFigure 4. 4 Raman.Raman Raman spectra spectra of of CNTs/CF CNTs/CF multiscale multiscale reinforcementsreinfo reinforcementsrcements preparedprepared prepared at at at different different different moving moving moving speeds. speeds.Thespeeds ﬂow The. The rate flow flow of H rate 2rateand of of CH H22H 2and 2andwas C C2H2 0.3H2 was2 L/min.was 0.3 0.3 L/min. L/min.

FigureFigure 55 5manifestsmanifests manifests XRDXRD XRD curvescurves curves ofof of CNTs/CFCNTs/CF CNTs/CF multiscalemultiscale multiscale reinforcementsreinforcements reinforcements preparedprepared prepared atat at differentdifferent moving moving speeds. speeds. It ItIt can cancan be bebe seen seenseen from fromfrom the the figure figurefigure that that each each curve curve had had an an obvious obvious ◦ diffractiondiffraction peak peak at at 2 2θθ = =25.3°, 25.3 25.3°,, corresponding correspondingcorresponding to to to the the the (002) (002) (002) crystal crystal crystal plane plane of of carbon carbon fiber. fiber. fiber. The The FWHMFWHM (full-width (full-width at at half-maximum) half-maximum) value value of of the the (002) (002) crystal crystal plane plane diffraction diffraction peak peak was was analyzedanalyzed to toto characterize characterizecharacterize the the grain grain size size size of of of the the the samples, samples, samples, shown shown shown in in Table in Table Table 2. 2. 2According .According According to to the tothe ScherrertheScherrer Scherrer formula, formula, formula, the the higher the higher higher the the value thevalue value was, was, was, the the smaller thesmaller smaller the the grain thegrain grain size size sizewas, was, was,indicating indicating indicating that that thethatthe (002) the (002) (002) crystal crystal crystal plane plane plane was was was damaged damaged damaged more more more se severely.verely. severely. With With With the the the increase increase in inin the thethe moving moving speed,speed,speed, the the FWHM FWHM value value decreased decreased at at first first first andan andd then then raised. raised. When When the the moving moving speed speed was was 4 4cm/min, cm/min cm/min, ,the the curve curvecurve became becamebecame the thethe steepest steepeststeepest and andand the thethe FWHM FWHMFWHM value value value reached reached reached the the minimum. minimum. The The resultsresultsresults showed showed thatthat that inin in thethe the process process process of of theof the the carbon ca carbonrbon nanotubes’ nanotubes' nanotubes growth, growth,' growth, activated activated activated carbon carbon carbon atoms atat- continued to accumulate on the carbon fiber surface, and graphite microcrystal defects that

J. Compos. Sci. 2021, 5, x FOR PEER REVIEW 6 of 12

J. Compos. Sci. 2021, 5, 111 6 of 12

oms continued to accumulate on the carbon fiber surface, and graphite microcrystal defects that were oxidized and etched on the carbon fiber surface were repaired, which was were oxidized and etched on the carbon ﬁber surface were repaired, which was helpful to helpful to improve the mechanical properties of carbon fiber. improve the mechanical properties of carbon ﬁber.

FigureFigure 5. XRDXRD curves curves of of CNTs/CF CNTs/CF multiscale multiscale reinforcements reinforcements prepared prepared at at different different moving moving speeds speeds.. The flow ﬂow rate of H 2 andand C C22HH22 waswas 0.3 0.3 L/min. L/min.

Table 2. FWHM values at 22θθ = 25.3° 25.3◦..

SampleSample FWHM FWHM 2 cm/min 0.960 2 cm/min 0.960 33 cm/min 0 0.934.934 44 cm/min 0 0.928.928 66 cm/min 0 0.975.975 1010 cm/min 1.114 1.114

3.2.3.2. Effect Effect of of Moving Moving Speeds Speeds on on Tensile Tensile Strength of Multiscale Reinforcements FigureFigure 66 shows the effect effect of of moving moving speeds speeds on on the the single single-filament-filament tensile tensile strength strength of CNTs/CF.of CNTs/CF. It is Itworth is worth noting noting that thatthe mechanical the mechanical properties properties of EAO of treated EAO treated carbon carbon fibers wasfibers the was poorest the poorest compared compared to other to other samples. samples. This This was was mainly mainly because because the the surface surface of desizeddesized carbon fiber fiber was etched and damaged during the EAO treatment, contributing to thethe decrease decrease of tensile strength, lower than that of the desized carboncarbon fibers.fibers. AtAt 44 cm/min,cm/min, tensiletensile strength strength reached reached the the maximum value value of of 4.46 GPa, which was about 5.4% higher thanthan that that of of the the desized carbon fiber fiberss and 8.8% higher than that of the electrochemically treatedtreated carbon fiber fibers.s. This This was was because because,, with the the prolonged prolonged CVD process time, the defects onon the surface of carboncarbon fibersfibers werewere repaired repaired and and made made up up gradually gradually by by a greata great number number of ofnewly newly deposited deposited CNTs CNTs of the of highestthe highest quality. quality. Thus, Thus, the strength the strength of the of fibers the wasfibers enhanced, was en- presenting a more effective reinforcing effect of the CVD process, which corresponded to hanced, presenting a more effective reinforcing effect of the CVD process, which corre- the result of SEM and Raman images given above. However, when carbon fibers moved sponded to the result of SEM and Raman images given above. However, when carbon quicker than 4 cm/min, the tensile strength of the samples was decreased correspondingly fibers moved quicker than 4 cm/min, the tensile strength of the samples was decreased with the increase in moving speeds. Because of the short deposition time, CNTs/CF correspondingly with the increase in moving speeds. Because of the short deposition time, synthesized at 10 cm/min with unsatisfactory surface morphology presented very low CNTs/CF synthesized at 10 cm/min with unsatisfactory surface morphology presented mechanical properties, almost equivalent to the EAO treated fiber. As the moving speed very low mechanical properties, almost equivalent to the EAO treated fiber. As the mov- went down to 3 cm/min, the tensile strength of the samples was reduced slightly compared ing speed went down to 3 cm/min, the tensile strength of the samples was reduced slightly to the highest strength. While the continuous carbon fibers passed through the reactor at the compared to the highest strength. While the continuous carbon fibers passed through the speed of 2 cm/min, the value was about 5% lower than the biggest one. This phenomenon reactor at the speed of 2 cm/min, the value was about 5% lower than the biggest one. This was mainly related to the life of the catalyst, which led to the rising R value in Figure4 . phenomenon was mainly related to the life of the catalyst, which led to the rising R value Catalyst particles were deactivated one after another when the CVD process took a long intime, Figure and 4 the. Catalyst deactivated particles catalyst were particles deactivated would one no after longer another facilitate when the the growth CVD ofprocess CNTs. tookAs a result, a long defects time, and could the not deactivated be obviously catalyst repaired, particles so the would tensile no strength longer offacilitate multiscale the growthreinforcements of CNTs. was As no a longer result, improved.defects could If the not CVD be timeobvious continuedly repaired, to be so added the tensile at this time, a great quantity of amorphous carbon would be deposited on the surface of the

J. Compos. Sci. 2021, 5, x FOR PEER REVIEW 7 of 12

J. Compos. Sci. 2021, 5, 111 7 of 12

strength of multiscale reinforcements was no longer improved. If the CVD time continued to be added at this time, a great quantity of amorphous carbon would be deposited on the carbonsurface ﬁbers, of the which carbon not fibers, only could which not not cause only an could increase not in cause the tensile an increase strength in of the the tensile ﬁbers butstrength also adverselyof the fibers affect but thealso preparation adversely affect of samples. the preparation of samples.

Figure 6. SingleSingle-filament-filament tensile tensile strength strength test test of of CNTs/CF CNTs/CF multiscale multiscale reinforcements reinforcements prepared prepared at at different movingmoving speeds.speeds. The flowflow raterate ofof HH22 and C2H22 waswas 0.3 0.3 L/min. L/min. 3.3. Effect of Growth Atmosphere on Morphology of Multiscale Reinforcements 3.3. Effect of Growth Atmosphere on Morphology of Multiscale Reinforcements In addition, different growth atmosphere compositions also affected the morphology In addition, different growth atmosphere◦ compositions also affected the morphology of CNTs. In the tube reactor kept at 600 C, the carbon source (C2H2) was pyrolyzed at high of CNTs. In the tube reactor kept at 600 °C , the carbon source (C2H2) was pyrolyzed at temperature and under the action of the catalyst to form active carbon atoms and release hydrogen.high temperature Then the and catalyst under converted the action active of the carbon catalyst atoms to form into CNTs.active Incarbon Figure atoms7a–e, theand release hydrogen. Then the catalyst converted active carbon atoms into CNTs. In Figure flow rate of C2H2 remained the same at 0.3 L/min, and the flow rate of H2 increased from 0 7a–e, the flow rate of C2H2 remained the same at 0.3 L/min, and the flow rate of H2 in- to 0.9 L/min. It was obvious that under the condition of no H2, there were few CNTs grown oncreased the surface from 0 of to carbon 0.9 L/min. fibers. It was The precursorobvious that of catalystunder the decomposed condition of at highno H temperature2, there were tofew form CNTs metal grown oxides on the with surface low catalyticof carbon activity, fibers. T givinghe precursor rise to of insufficient catalyst decomposed reduction ofat catalyst.high temperature Moreover, to theform pyrolysis metal oxides reaction with of low the carboncatalytic source activity, was giving not restrained, rise to insuffi- so a largecient reduction number of o activef catalyst. carbon Moreover, atoms would the pyrolysis be produced reaction and of transformed the carbon source into pyrolytic was not carbonrestrained, wrapped so a large on the number surface of of active the catalyst carbon inatoms a short would time, be resulting produced in and the deactivationtransformed ofinto the pyrolytic catalyst. carbon These twowrapped reasons on affected the surface the normalof the catalyst growth in of a CNTs, short leadingtime, resulting to a small in quantitythe deactivation of CNTs of and the nonuniform catalyst. These distribution two reasons [32]. affected the normal growth of CNTs, leadingAfter to a small amountquantity ofof HCNTs2 was and injected nonuniform into the distribution growth atmosphere, [32]. the reduction of someAfter catalysts a small became amount sufficient, of H2 was and inject theed life into of thethe catalystgrowth wasatmosphere, prolonged, the which reduction was beneficialof some catalysts to the growth became of sufficient, CNTs. Therefore, and the life the of number the catalyst of CNTs was raised prolonged, significantly, which andwas thebeneficial introduction to the growth of H2 inhibited of CNTs. theTherefore, cracking the of number carbon sourceof CNTs to raise a certaind significantly, extent [33 ,and34]. Whenthe introduction the H2 flow of rate H2 increasedinhibited the to 0.3 cracking L/min, of the carbon CNTs’ source deposition to a certain began toextent be uniform. [33,34]. AndWhen the the length H2 flow and rate number increased of CNTs to 0.3 got L/min, a pronounced the CNTs increase' deposition as the began H2 flow to ratebe uniform. reached 0.9And L/min. the length Additionally, and number CNTs wereof CNTs entangled got a withpronounced each other, increase agglomerated as the orH2 connectedflow rate intoreached sheets. 0.9 L/min. Additionally, CNTs were entangled with each other, agglomerated or connectedAs shown into insheets. Figure 8, with the addition of H 2 content, the R value became smaller step by step,As demonstratingshown in Figure that 8, thewith degree the addition of graphitization of H2 content, on the the surface R value of thebecame samples smaller was promoted.step by step This, demonstrating indicated that that the the regularity degree of of graphitization carbon atoms on on the the surface CNTs got of the an increasesamples andwas thepromoted number. ofThis impurities indicated such that as the amorphous regularity carbon of carbon reduced. atoms By adjustingon the CNTs the content got an ofincrease H2 and and making the number use of theof impurities inhibitory such effect as of amorphous H2 on the cracking carbon reduce of carbond. By source, adjusting the surfacethe content morphology of H2 and andmaking graphitization use of the inhibitory degree of effect CNTs/CF of H2 multiscale on the cracking reinforcements of carbon couldsource, be the controlled surface [morphology35]. and graphitization degree of CNTs/CF multiscale reinforcements could be controlled [35].

J. J.Compos. Compos.J. Compos. Sci. Sci. Sci. 2021 20212021, 5, ,5 5x,, xFOR 111 FOR PEER PEER REVIEW REVIEW 8 of 12 8 8of of 12 12

Figure 7. SEM micrographs of CNTs/CF multiscale reinforcements prepared under different FigureFigure 7. 7.SEM SEM micrographs micrographs of CNTs/CF of CNTs/CF multiscale multiscale reinforcements reinforcements prepared prepared under different under growth different growth atmospheres, the flow rate of H2 to C2H2: (a) 0:0.3; (b) 0.15:0.3; (c) 0.3:0.3; (d)0.6: 0.3 and (e) atmospheres,growth atmospheres, the ﬂow rate the of flow H to rate C H of:( Ha2) to 0:0.3; C2H (b2): ( 0.15:0.3;a) 0:0.3; ( c()b 0.3:0.3;) 0.15:0.3; (d)0.6: (c) 0.30.3:0.3; and ((ed))0.6: 0.9:0.3. 0.3 and (e) 0.9:0.3. The moving speed of carbon2 2 fibers2 was 4 cm/min. The0.9:0.3. moving The speed moving of carbon speed ﬁbers of carbon was 4 fibers cm/min. was 4 cm/min.

Figure 8. Raman spectra of CNTs/CF multiscale reinforcements prepared under different growth FigureFigure 8 .8 Raman. Raman spectra spectra of of CNTs/CF CNTs/CF multiscale multiscale reinforcements reinforcements prepared prepared under under different different growth growth atmospheres. The moving speed of carbon ﬁbers was 4 cm/min. atmospheresatmospheres. The. The moving moving speed speed of of carbon carbon fibers fibers was was 4 4cm/min. cm/min. 3.4. CNTs Microstructure Analysis 3.4. CNTs Microstructure Analysis 3.4. FigureCNTs 9Microstructure shows the detailed Analysis microstructure of CNTs on the surface of CNTs/CF preparedFigFigureure when 9 9show show thes moving sthe the detailed detailed speed microstructure was microstructure 4 cm/min andof of CNTs CNTs H2:C 2onH on2 the =the 0.3:0.3. surface surface From of of CNTs/CF CNTs/CF SEM pre- pre- paredimages when and Figure the moving9, we can speed obtain was that 4 thecm/min diameter and of H CNTs2:C2H with2 = 0.3: excellent0.3. From character SEM wasimages and pared when the moving speed was 4 cm/min and H2:C2H2 = 0.3:0.3. From SEM images and concentrated around 9–12 nm, while the inner diameter was around 3.5–4.5 nm, and the FigureFigure 9, 9, we we can can obtain obtain that that the the diameter diameter of of CNTs CNTs with with excellent excellent character character was was concen- concen- average length of CNTs was about 40 nm. As seen in Figure9, the CNTs grown under tratedtrated around around 9 –912–12 nm nm, while, while the the inner inner diameter diameter was was around around 3.5 3.5–4.5–4.5 nm, nm, and and the the average average lengththese two of CNTs growth was conditions about 40 had nm. a largeAs seen aspect in ratio,Figure an 9, obvious the CNTs hollow grown structure under and these two tubularlength morphology,of CNTs was and about the distribution40 nm. As seen was uniformin Figure and 9, dense.the CNTs From gro Figurewn under9b,d, it these two growth conditions had a large aspect ratio, an obvious hollow structure and tubular mor- couldgrowth be clearlyconditions observed had a that large the aspect thickness ratio, of CNTs an obvious tubes was hollow even, structure the inner and diameter tubular mor- pholophology,gy, and and the the distribution distribution was was uniform uniform and and dense. dense. From From Figure Figure 9 b9 band and 9 d,9d, it it could could be be clearlyclearly observed observed that that the the thickness thickness of of CNTs CNTs tubes tubes was was even even, the, the inner inner diameter diameter and and the the outerouter diameter diameter were were coaxial coaxial. Additionally,. Additionally, t hethe graphite graphite sheet sheet on on the the tube tube wall wall, or, or t hthe elat- lat- ticetice fringes fringes of of the the CNT CNT, ,basically basically aligned aligned parallel parallel to to the the CNTs CNTs' axis' axis, ,indicating indicating the the high high crystallinity.crystallinity. This This indicated indicated that that the the CNTs CNTs grafted grafted under under these these two two conditions conditions had had fewer fewer

J. Compos. Sci. 2021, 5, 111 9 of 12

J. Compos. Sci. 2021, 5, x FOR PEER REVIEW 9 of 12 and the outer diameter were coaxial. Additionally, the graphite sheet on the tube wall, or the lattice fringes of the CNT, basically aligned parallel to the CNTs’ axis, indicating the high crystallinity. This indicated that the CNTs grafted under these two conditions had fewerwall defects, wall defects, and andthe thedegree degree of ofgraphitization graphitization of of CNTs CNTs was thethe largest largest [36 [3].6 What]. What is is more, more,the catalyst the catalyst particles particles were were located located at at the the endend ofof the the CNTs CNTs far far away away from from the surface the surface of ofcarbon carbon fiber ﬁberss and and would would fallfall offoff fromfrom the the top top of of CNTs CNTs during during the the sample sample preparation preparation pro- process,cess, resulting resulting in in an openopenstate state at at the the top top of theof CNTs.the CN ThisTs. typeThis of type structure of structure could be could be foundfound easily easily in in Figure Figure9. 9.

FigureFigure 9. 9HRTEM. HRTEM images images of CNTs of CNTs grafted grafted on the on samples: the samples 4 cm/min: 4 cm/min and H2 :Cand2H 2H=2:C 0.3:0.3.2H2 = 0.3:0.3.

3.5.3.5. Effect Effect of of Growth Growth Atmosphere Atmosphere on Tensile on Tensile Strength Strength of Multiscale of Multiscale Reinforcements Reinforcements Figure 10 reflects the effect of growth atmosphere on the mechanical properties of Figure 10 reflects the effect of growth atmosphere on the mechanical properties of CNTs/CF. In the absence of H2, the single-filament tensile strength of the sample was 4.33CNTs/CF. GPa, which In the was absence about 2.8%of H higher2, the single than that-filament of the tensile desized strength carbon fibers. of the After sample the was 4.33 injectionGPa, which of H 2was, the about tensile 2.8% strength higher was than increased that withof the the desized addition carbon of H2 flow fibers rate.. After As it the injec- wenttion upof H to2 0.3, the L/min, tensile the strength tensile strengthwas increased was 4.41 with GPa, the reaching addition the maximumof H2 flow value rate. ofAs it went theupsystem, to 0.3 L/min, which was the about tensile 4.3% strength higher thanwas that4.41 of GPa, the desized reaching carbon the fibers.maximum When value the of the ratesystem, was added which to was 0.6 L/min,about 4.3% the tensile higher strength than that did notof the make desized a change carbo obviously.n fibers But. When the the rate tensilewas added strength to was 0.6 reducedL/min, tothe 4.38 tensile GPa whenstrength it came did to not 0.9 L/min.make a H change2 could restrainobviously. the But the cracking of the carbon source and reduce the production of amorphous carbon, so injecting tensile strength was reduced to 4.38 GPa when it came to 0.9 L/min. H2 could restrain the an appropriate amount of H in the growth atmosphere could improve the efficiency of the cracking of the carbon source2 and reduce the production of amorphous carbon, so inject- catalyst, which was beneficial to the reinforcement of the CVD process. However, when ing an appropriate amount of H2 in the growth atmosphere could improve the efficiency the flow rate of H2 was too large, the inhibition on the cracking of carbon source was too strong,of the andcatalyst, the number which of was active beneficial carbon atoms to the decreased reinforcement in the early of stagethe CVD of CNTs process. growth, However, whichwhen would the flow make rate the of surface H2 was of carbon too large, fiber morethe inhibition seriously etched, on the and cracking finally ledof carbon to the source decreasewas too of strong, single-filament and the tensile number strength of active of samples. carbon Additionally, atoms decreased the high in crystallinity the early stage of ofCNTs CNTs growth, shown in which TEM imageswould hadmake a positive the surface effect of on carbon improving fiber CNTs/CF more seriously mechanical etched, and properties.finally led That to the was de whycrease the of tensile single strength-filament of carbon tensile fibers strength increased of samples. significantly Additionally, after the thehigh deposition crystallinity of CNTs. of CNTs shown in TEM images had a positive effect on improving CNTs/CF mechanical properties. That was why the tensile strength of carbon fibers increased significantly after the deposition of CNTs.

J. Compos. Sci. 2021, 5, x 111 FOR PEER REVIEW 10 of 12

Figure 10. 10. SingleSingle-ﬁlament-filament tensile tensile strength strength test test of ofCNTs/CF CNTs/CF multiscale multiscale reinforcements reinforcements prepared prepared under different growth atmospheres.atmospheres. The moving speed of carbon ﬁbersfibers was 4 cm/min.

4. Conclusions Conclusions

InIn conclusion, conclusion, by by tuning tuning the the moving moving speeds speeds of of carbon carbon fibers fibers and and the the ratio ratio of ofC2 CH22H to2 Hto2 Hin2 thein thegrowth growth atmosphere, atmosphere, effects effects of deposition of deposition time time and andgrowth growth atmosphere atmosphere on the on surfacethe surface morphology morphology and andtensile tensile strength strength of CNTs/CF of CNTs/CF multiscale multiscale reinforcements reinforcements were were sys- tematicallysystematically investigated. investigated. With With the the decrease decrease of of moving moving speeds speeds,, the the thickness thickness and and density density of the the CNTs CNTs layer layer gradually gradually increased, increased, and and the the degree degree of graphitization of graphitization of CNTs/CF of CNTs/CF multiscalemultiscale reinforcements reinforcements was improved was improved at first at and first then and had then a reduction had a reduction.. When the When moving the speedmoving was speed 3 or was4 cm/m 3 orin, 4 cm/min,the distribution the distribution of CNTs was of CNTs ideal was, and ideal, the degree and the of degreegraphiti- of graphitization on the surface of the sample was the highest. In addition to reducing catalyst, zation on the surface of the sample was the highest. In addition to reducing catalyst, H2 H had a certain inhibitory effect on carbon source pyrolysis. Adding the proportion of H had2 a certain inhibitory effect on carbon source pyrolysis. Adding the proportion of H2 in2 thein the growt growthh atmosphere atmosphere could could reduce reduce the the production production of of amorphous amorphous carbon carbon and and prevent thethe deactivation ofof thethe catalyst,catalyst, which which was was beneficial beneficial to to the the growth growth of CNTs,of CNTs improving, improving the thesurface surface graphitization graphitization and and crystal crystal structure structure of CNTs/CF of CNTs/CF multiscale multiscale reinforcements. reinforcements.

Author Contributions: Conceptualization, C C.W..W. and and Y Y.W.;.W.; Data curation, C C.W.;.W.; Investigation, C C.W..W. and Y.W.; Y.W.; Methodology, Methodology, C.W.; C.W Project.; Project administration, administration, Y.W.; Y Software,.W.; Software, S.S.; Supervision, S.S.; Supervision, Y.W.; Valida- Y.W.; Validation,tion, C.W.; Visualization,C.W.; Visualization, C.W.; Writing—original C.W.; Writing—original draft, C.W. draft, and C.W S.S.;. and Writing—review S.S.; Writing— andreview editing, and editing,C.W., Y.W. C.W and., Y S.S..W. Alland authors S.S. have read and agreed to the published version of the manuscript. Funding: ThisThis research research was was funded funded by Natural by Natural Science Foundation Science Foundation in Shandong in Province Shandong (ZR2020ME039 Province , (ZR2020ME039,ZR2020ME134) andZR2020ME134) National Natural and National Science FoundationNatural Science of China Foundation (51773110). of China (51773110). Institutional Review Board Statem Statement:ent: NoNott applicable applicable.. Informed Consent Statemen Statement:t: NNotot applicable. Acknowledgments: The authors thank the editor and the anonymous reviewers for their valuable comments on on this this manuscript. manuscript. The The authors authors also also acknowledge acknowledge the the support support of technical of technical staff staff for foras- sistingassisting in inpreparing preparing samples samples and and analyzing analyzing them. them. ConflictsConﬂicts of Interest: The authors declare no conﬂictconflict of interest.interest.

References 1. Wazir, A.H.; A.H.; Kakakhel, Kakakhel, L. Preparation Preparation and characterization characterization of pitch-based pitch-based carbon carbon fibers. fibers. NN.. Carbon Mater. 20092009,, 24,, 83 83–88.–88, doi:[CrossRef10.1016/s1872] -5805(08)60039-6. 2. Ma, L.C.; Zhu, Y.Y.;Y.Y.; Wang, Wang. M.Z.; M.Z.; Yang, Yang, X.B.; X.B.; Song, Song, G.J.; G.J.; Huang, Huang, Y.D. Y.D. Enhancing Enhancing interfacial interfacial strength strength of epoxyof epoxy resin resin composites composites via viaevolving evolving hyperbranched hyperbranched amino-terminated amino-terminated POSS on POSS carbon on fiber carbon surface. fiber Compos. surface. Sci.Compos. Technol. Sci.2019 Technol., 170, 148–156. 2019, 170 [CrossRef, 148–156] , 3. doi:Das,10.1016/j. T.K.; Ghosh,compscitech.2018.12.001. P.; Das, N.C. Preparation, development, outcomes, and application versatility of carbon fiber-based polymer 3. Das,composites: T.K.; Ghosh, A review. P.; Das,Adv. N.C. Compos. Preparation, Mater. 2019 development,, 2, 214–233. outcomes, [CrossRef and] application versatility of carbon fiber-based polymer 4. composites:Sun, K.; Wang, A review. L.Y.; Wang, Adv. Z.X.;Compos. Wu, Mater. X.F.; Fan,2019 G.H.;, 2, 214 Wang,–233. Z.Y.; Cheng, C.B.; Fan, R.H.; Dong, M.Y.; Guo, Z.H. Flexible silver nanowire/carbon fiber felt metacomposites with weakly negative permittivity behavior. Phys. Chem. Chem. Phys. 2020, 22, 5114–5122. [CrossRef][PubMed]

J. Compos. Sci. 2021, 5, 111 11 of 12

5. Zeng, J.J.; Gao, W.Y.; Duan, Z.J.; Bai, Y.L.; Guo, Y.C.; Ouyang, L.J. Axial compressive behavior of polyethylene tereph- thalate/carbon FRP-confined seawater sea-sand concrete in circular columns. Constr. Build. Mater. 2020, 234, 117383–117399. [CrossRef] 6. Raphael, N.; Namratha, K.; Chandrashekar, B.N.; Sadasivuni, K.K.; Ponnamma, D.; Smitha, A.S.; Krishnaveni, S.; Cheng, C.; Byrappa, K. Surface modification and grafting of carbon fibers: A route to better interface. Prog. Cryst. Growth Charact. Mater. 2018, 64, 75–101. [CrossRef] 7. Van de Werken, N.; Tekinalp, H.; Khanbolouki, P.; Ozcan, S.; Williams, A.; Tehrani, M. Additively manufactured carbon fiber-reinforced composites: State of the art and perspective. Addit. Manuf. 2020, 31, 100962. [CrossRef] 8. Yuan, X.M.; Zhu, B.; Cai, X.; Qiao, K.; Zhao, S.Y.; Zhang, M. Nanoscale toughening of carbon fiber-reinforced epoxy composites through different surface treatments. Polym. Eng. Sci. 2019, 59, 625–632. [CrossRef] 9. Abbas, I.; Wang, Y.X.; Elahi, H.; Siddiqui, M.A.; Ullah, M.; Qayyum, F. Effect of MoSi2-Si3N4/SiC multi-layer coating on the oxidation resistance of carbon/carbon composites above 1770 K. J. Compos. Sci. 2020, 4, 86. [CrossRef] 10. De Greef, N.; Zhang, L.M.; Magrez, A.; Forró, L.; Locquet, J.-P.; Verpoest, I.; Seo, J.W. Direct growth of carbon nanotubes on carbon fibers: Effect of the CVD parameters on the degradation of mechanical properties of carbon fibers. Diam. Relat. Mater. 2015, 51, 39–48. [CrossRef] 11. Qian, H.; Bismarck, A.; Greenhalgh, E.S.; Kalinka, G.; Shaffer, M.S.P. Hierarchical Composites Reinforced with Carbon Nanotube Grafted Fibers: The Potential Assessed at the Single Fiber Level. Chem. Mater. 2008, 20, 1862–1869. [CrossRef] 12. Wang, Y.X.; Su, S.S.; Fan, W.X.; Qin, J.J.; Qu, C.; Wang, C.G. Metal catalyst for preparation of multiscale reinforcement of carbon nanotubes on carbon fibers. Chin. J. Mater. Res. 2018, 32, 495–501. 13. Naito, K. Tensile Properties and Fracture Behavior of Different Carbon Nanotube-Grafted Polyacrylonitrile-Based Carbon Fibers. J. Mater. Eng. Perform. 2014, 23, 3916–3925. [CrossRef] 14. Manawi, Y.M.; Ihsanullah; Samara, A.; Al-Ansari, T.; Atieh, M.A. A Review of Carbon Nanomaterials’ Synthesis via the Chemical Vapor Deposition (CVD) Method. Materials 2018, 11, 822. [CrossRef][PubMed] 15. Li, Y.-L.; Kinloch, I.A.; Windle, A.H. Direct Spinning of Carbon Nanotube Fibers from Chemical Vapor Deposition Synthesis. Science 2004, 304, 276–278. [CrossRef] 16. Anthony, D.B.; Sui, X.M.; Kellersztein, I.; de Luca, H.G.; White, E.R.; Wagner, H.D.; Greenhalgh, E.S.; Bismarck, A.; Shaffer, M.S. Continuous carbon nanotube synthesis on charged carbon fibers. Compos. Part. A Appl. Sci. Manuf. 2018, 112, 525–538. [CrossRef] 17. Tanaka, K.; Okuda, S.; Hinoue, Y.; Katayama, T. Effects of Water Absorption on the Fiber-Matrix Interfacial Shear Strength of Carbon Nanotube-Grafted Carbon Fiber Reinforced Polyamide Resin. J. Compos. Sci. 2019, 3, 4. [CrossRef] 18. Rong, H.P.; Dahmen, K.-H.; Garmestani, H.; Yu, M.H.; Jacob, K.I. Comparison of chemical vapor deposition and chemical grafting for improving the mechanical properties of carbon fiber/epoxy composites with multi-wall carbon nanotubes. J. Mater. Sci. 2013, 48, 4834–4842. [CrossRef] 19. Lee, G.; Sung, M.; Youk, J.H.; Lee, J.; Yu, W.-R. Improved tensile strength of carbon nanotube-grafted carbon fiber reinforced composites. Compos. Struct. 2019, 220, 580–591. [CrossRef] 20. Thostenson, E.T.; Li, W.Z.; Wang, D.Z.; Ren, Z.F.; Chou, T.W. Carbon nanotube/carbon fiber hybrid multiscale composites. J. Appl. Phys. 2002, 91, 6034–6037. [CrossRef] 21. An, F.; Lu, C.X.; Guo, J.H.; He, S.Q.; Lu, H.B.; Yang, Y. Preparation of vertically aligned carbon nanotube arrays grown onto carbon fiber fabric and evaluating its wettability on effect of composite. Appl. Surf. Sci. 2011, 258, 1069–1076. [CrossRef] 22. An, F.; Lu, C.X.; Li, Y.H.; Guo, J.H.; Lu, X.X.; Lu, H.B.; He, S.Q.; Yang, Y. Preparation and characterization of carbon nano-tube- hybridized carbon fiber to reinforce epoxy composite. Mater. Des. 2012, 33, 197–202. [CrossRef] 23. Zhao, N.; Ma, Z.K.; Song, H.H.; Xie, Y.G.; Zhang, M. Enhancement of bioelectricity generation by synergistic modification of vertical carbon nanotubes/polypyrrole for the carbon fibers anode in microbial fuel cell. Electrochim. Acta 2019, 296, 69–74. [CrossRef] 24. Luo, X.G.; Weng, W.; Liang, Y.X.; Hu, Z.X.; Zhang, Y.; Yang, J.J.; Yang, L.J.; Yang, S.Y.; Zhu, M.F.; Cheng, H.M. Multifunctional fabrics of carbon nanotube fibers. J. Mater. Chem. A 2019, 7, 8790–8797. [CrossRef] 25. Su, S.S.; Wang, Y.X.; Qin, J.J.; Wang, C.G.; Yao, Z.Q.; Lu, R.J.; Wang, Q.F. Continuous method for grafting CNTs on the surface of carbon fibers based on cobalt catalyst assisted by thiourea. J. Mater. Sci. 2019, 54, 12498–12508. [CrossRef] 26. Yao, Z.Q.; Wang, C.G.; Qin, J.J.; Wang, Y.X.; Wang, Q.F.; Wei, H.Z. Fex-Co1−x bimetallic catalysts for highly efficient growth of carbon nanotubes on carbon fibers. Ceram. Int. 2020, 46, 27158–27162. [CrossRef] 27. Qin, J.J.; Wang, C.G.; Wang, Y.X.; Su, S.S.; Lu, R.J.; Yao, Z.Q.; Wang, Q.F.; Gao, Q. Communication—A Technique for Online Continuous Manufacture of Carbon Nanotubes-Grown Carbon Fibers. ECS J. Solid State Sci. Technol. 2019, 8, M23–M25. [CrossRef] 28. Yao, Z.Q.; Wang, C.G.; Qin, J.J.; Su, S.S.; Wang, Y.X.; Wang, Q.F.; Yu, M.J.; Wei, H.Z. Interfacial improvement of carbon fi-ber/epoxy composites using one-step method for grafting carbon nanotubes on the fibers at ultra-low temperatures. Carbon 2020, 164, 133–142. [CrossRef] 29. Meysami, S.S.; Koos, A.A.; Dillon, F.; Dutta, M.; Grobert, N. Aerosol-assisted chemical vapour deposition synthesis of multi-wall carbon nanotubes: III. Towards upscaling. Carbon 2015, 88, 148–156. [CrossRef] 30. Jourdain, V.; Bichara, C. Current understanding of the growth of carbon nanotubes in catalytic chemical vapour deposition. Carbon 2013, 58, 2–39. [CrossRef] J. Compos. Sci. 2021, 5, 111 12 of 12

31. Kim, K.J.; Yu, W.-R.; Youk, J.H.; Lee, J. Degradation and Healing Mechanisms of Carbon Fibers during the Catalytic Growth of Carbon Nanotubes on Their Surfaces. ACS Appl. Mater. Interfaces 2012, 4, 2250–2258. [CrossRef][PubMed] 32. Valles, C.; Perez-Mendoza, M.; Maser, W.K.; Martínez, M.T.; Álvarez, L.; Sauvajol, J.L.; Benito, A.M. Effects of partial and total methane flows on the yield and structural characteristics of MWCNTs produced by CVD. Carbon 2009, 47, 998–1004. [CrossRef] 33. Behr, M.J.; Gaulding, E.A.; Mkhoyan, K.A.; Aydil, E.S. Effect of hydrogen on catalyst nanoparticles in carbon nanotube growth. J. Appl. Phys. 2010, 108, 053303. [CrossRef] 34. Wang, X.H.; Wang, C.G.; Wang, Z.Y.; Wang, Y.X.; Lu, R.J.; Qin, J.J. Colossal permittivity of carbon nanotubes grafted carbon fiber-reinforced epoxy composites. Mater. Lett. 2018, 211, 273–276. [CrossRef] 35. Terrado, E.; Tacchini, I.; Benito, A.M.; Maser, W.K.; Martinez, M.T. Optimizing catalyst nanoparticle distribution to produce densely-packed carbon nanotube growth. Carbon 2009, 47, 1989–2001. [CrossRef] 36. Chen, H.; Yang, J.; Shuai, Q.; Li, J.; Ouyang, Q.; Zhang, S. In-situ doping B4C nanoparticles in PAN precursors for preparing high modulus PAN-based carbon fibers with boron catalytic graphitization. Compos. Sci. Technol. 2020, 200, 108455. [CrossRef]