Mechanical reinforcement of PBO fibers by dicarboxylic acid functionalized carbon nanotubes through in situ copolymerization

Abstract

To make stronger poly(p-phenylene benzobisoxazole) (PBO) fibers, we designed and synthesized a series of CNTs & PBO copolymer fibers based on dicarboxylic acid functionalized CNTs (CNTs I–III). The aim of introducting three dicarboxylic acids including two flexible (I, II) and one rigid chain (III) functionalized CNTs is to overcome the limited dispersivity and reactivity of CNTs in the PBO polymerization system. The CNTs I–III as well as pure terephthalatic acid (TPA) were further covalently incorporated with the monomer 4,6-diaminoresorcinol dihydrochloride (DADHC) via in situ polymerization to obtain copolymers of CNTs I & PBO, CNTs II & PBO and CNTs III & PBO, respectively. Then continuous CNTs I–III & PBO copolymer fibers have been fabricated through a dry-jet wet spinning process. The chemical structure and morphology of CNTs I–III & PBO copolymer fibers were characterized by FTIR, WAXD, SEM and TEM. The mechanical and thermal properties of CNTs I–III & PBO copolymer fibers were characterized by TG and tensile testing, compared with PBO, and the properties have been enhanced. As a result, dicarboxylic acids on CNTs improves the dispersion of CNTs in the PBO matrix, provide high reaction activity and an efficient load transfer between CNTs reinforcement and PBO matrix, which may be the main reasons for improving the mechanical and thermal properties.

---

1. Introduction

Poly(p-phenylene benzobisoxazole) (PBO) is a representative of the strongest organic polymer with aromatic heterocyclic rings, which has superior mechanical properties, excellent thermal and thermo-oxidative stability, and good flame retardance.^1–3 PBO fibers have been used in the fields of aerospace, industrial materials, sports and protective clothing.⁴ However, even for commercial high modulus PBO fibers, there are still large gaps between the actual tensile modulus (280 GPa) with the theoretical tensile modulus (460–478 GPa),^5,6 which limits the applications of PBO fibers. To solve the problems, the main methods are modifying the molecular structures,⁷ adding the inorganic nanoparticles⁸ and improving the molecular orientation.^9,10

Carbon nanotubes (CNTs) were first reported by Iijima in 1991.¹¹ CNTs have a unique combination of mechanical, electrical and thermal properties, which make them excellent candidates as reinforcing agents in the fabrication of polymer nanocomposites.^12,13 Extensive investigations also have been carried out in the CNTs/PBO composite with count papers in the past few decades. The CNTs/PBO polymer composites were prepared via in situ polymerization with pristine CNTs,^14,15 carboxylated SWCNTs and MWCNT,^16–19 oligohydroxy-amide modified MWCNTs,²⁰ or bisAPAF terminated PHA modified MWCNTs and pyrene modified SWCNTs.²¹ The above-mentioned fabrication methods of CNT–PBO composites, most of these works focused on introducing pure CNTs or single carboxylic groups functionalized CNTs into the process of PBO polymerization. Introduction of pure CNTs into PBO results in the blending composites, which could not form good chemical interface between CNTs and PBO. The universal method is to introduce the single carboxylic groups functionalized CNTs, but the single carboxylic groups might seal ends of long PBO chains, which further partially restrict improvements in molecular weight as well as properties. That is to say, the potential of CNTs as reinforcement for PBO has not been fully explored.

Based above-mentioned analysis, we have deliberately synthesized three dicarboxylic acids functionalized CNTs including two flexible chains and one rigid chain as reinforcing agents, and PBO as the polymer matrix to prepare a serial of CNTs & PBO copolymers. Firstly, CNTs were dispersed ultrasonically in a mixture of H₂SO₄ and HNO₃ to form monocarboxylated CNTs. Then carboxylated CNTs were acyl chloride and grafted with amino dicarboxylic acid including L-aspartic acid (I), L-glutamic acid (II) and 5-amino-isophthalic acid (III) to obtain dicarboxylic functionalized CNTs, respectively. Finally, the functionalized CNTs I–III were further covalently incorporated with the monomer of PBO by in situ polycondensation in the presence of polyphosphoric acid (PPA) and P₂O₅. Then high weight molecular CNTs I–III & PBO copolymers were prepared. Continuous long copolymer fibers have been fabricated using a dry-jet wet-spinning technique. The different functionalized CNTs (I–III) reinforced PBO fibers were characterized using Fourier-transform infrared (FT-IR), wide angle X-ray diffraction (WAXD), scanning electron microscopy (SEM), thermogravimetry (TG) and tensile testing. Dispersion of CNTs in PBO fibers was characterized by transmission electron microscopy (TEM). These results show that an optimized interfacial interaction between three types of dicarboxylic acids functionalized CNTs fillers and the PBO matrix, decrease steric hindrance, provide high reaction activity and high degree of covalent modification.

2. Experimental section

2.1 Materials

One monomer of PBO, 4,6-diaminoresorcinol dihydrochloride (DADHC), was synthesized in our laboratory from 1,2,3-trichlorobenzene by a modified method.^22,23 Another monomer–pure terephthalic acid (PTA) powder (GR) was purchased from Korea SK Chemicals. Polyphosphoric acid (PPA) was purchased from Shanghai Reagents Company. The CNTs were provided by Nanjing NANO XFNANO Materials Tech Co., Ltd, the purity is over 90% and the diameter distribution is about 5–15 nm with the length of 10–30 μm. Methanesulfonic acid (MSA) was purchased from Sigma Aldrich Chemical Company.

2.2 Synthesis of CNTs I–III

(i) Oxidation of CNTs (CNTs–COOH). Firstly, the 2 g CNTs and 100 mL oleum (20% SO₃) were added into three-necked flask with magnetic stirring for 72 h in order to ensure the CNTs dispersed well in the oleum. Another 50 mL oleum (20% SO₃) and 50 mL concentrated HNO₃ were slowly added into the CNTs/oleum dispersion mixture to make fuming sulfuric acid and nitric acid reach 3 : 1 (V/V). After 30 min, the mixture was further raised to 65 °C and stirred for 2 h. The warm solution was cooled to room temperature and poured into ice water (600 mL) to dilute, centrifugal filtration and washed repeatedly with deionized water until the pH of the wash water was neutral, then freeze-dried at −60 °C for 24 h. Thereby, the CNTs–COOH was obtained. TGA weight loss: 10.2% (296.2 °C); XPS analysis: C1s 91.3%, O1s 8.7%.

(ii) Preparation of acyl chloride CNTs (CNTs–COCl). A typical procedure for preparing of CNTs–COCl is depicted as follows. 1.5 g CNTs–COOH was refluxed in the mixture solution of 50 mL SOCl₂ and 1 mL pyridine at 75 °C for 24 h. After completed, the SOCl₂ was removed by the reduced pressure distillation. The collected solid was washed repeatedly with THF to removed the residual SOCl₂ and dried under vacuum at room temperature to yield acyl chloride functionalized CNTs (CNTs–COCl). TGA weight loss: 26.3% (282.2 °C); XPS analysis: C1s 93.27%, O1s 5.3%, Cl1s 1.43%.

(iii) Synthesis of acid-grafted CNTs (CNTs I–III). A typical procedure for synthesizing of CNTs I, II, III is shown in Scheme 1. Firstly, 1.5 g CNTs–COCl and 1 g L-aspartic acid (I) (or L-glutamic acid (II), or 5-amino-isophthalic acid (III)) were added into the mixture solution of 60 mL DMAc and 1 mL pyridine at 75 °C for 24 h. Then, the mixture solution were stirred and refluxed at 60 °C for 12 h. After completed, the mixture solution was filtered and washed repeatedly with DMAc and deionized water. Finally, the obtained solid was freeze-dried at −60 °C for 24 h. CNTs I, TGA weight loss: 20.5% (265.2 °C); XPS analysis: C1s 86.62%, O1s 12.51%, N1s 0.87%. CNTs II, TGA weight loss: 24.5% (211.1 °C); XPS analysis: C1s 83.62%, O1s 14.11%, N1s 2.07%. CNTs III, TGA weight loss: 16.1% (229.3 °C); XPS analysis: C1s 88.84%, O1s 9.86%, N1s 1.30%.

Scheme 1 Synthesis routs of functionalized CNTs (CNTs I–III).

2.3 Synthesis of 4,6-diaminoresorcino functionalized CNTs I–III

In a 250 mL three necked round bottom glass flask, equipped with a mechanical stirrer and a nitrogen inlet/outlet, 21.31 g of DADHC and 0.05 g SnCl₂·2H₂O were placed into PPA solution (97.28 g, 83 wt% P₂O₅). The reaction mixture underwent dehydrochlorination under a nitrogen atmosphere at 60 °C for 2 h and subsequently at 100 °C for 10 h. After this stage, the CNTs I (or CNTs II or CNTs III) was add into the above mixture. The components of CNTs I–III were 1 wt% with respect to the polymer concentration in the polymerization. The condensation step was began to carry out at 120 °C for 2 h, 140 °C for 6 h, 160 °C for 8 h and 180 °C for 10 h. After that, the reaction resulting of CNTs I–III respectively grafted by 4,6-diaminoresorcino (DAR) (DAR–CNTs I–III) was used directly in the following copolymerization. In order to characterize DAR–CNTs I–III, a portion of reaction dope was take out and terminated by the addition of the water. DAR–CNTs I–III was precipitated by filtration. The collected solid was washed with water several times and then dried under vacuum at room temperature.

2.4 In situ copolymerization of PBO with CNTs I–III

A typical procedure for in situ copolymerization of PBO with CNTs I–III is illustrated in Scheme 2. 16.60 g PTA was added into the above mixture in 120 °C, which were maintained at 140 °C for 6 h and 160 °C for 8 h. Another P₂O₅ were added to the reaction mixture to bring the P₂O₅ concentration up to 85 wt% and resulted in a final polymer concentration of 14 wt%. The polymerizing mixture was heated to 180 °C stepwise at 5 °C h⁻¹ and kept at this temperature for 8 h with constant stirring.

Scheme 2 Synthesis routs of CNTs I–III & PBO copolymers.

2.5 Fabrication of CNTs I–III & PBO copolymer fibers

The dry-jet wet-spinning of continuous CNTs I–III & PBO copolymer fibers was carried out using a piston-driven system designed for lab-scale spinning. The prepared CNTs I–III & PBO copolymer dope was first transferred under the protection of nitrogen atmosphere to the dope tank and vacuumed to remove the bubble at 160 °C for 6 h before spinning. Then the dope tank was pressured with nitrogen to extrude the dope through with an orifice diameter of 0.4 mm. The dry-jet fibers was cooled along a 30 cm long air gap and solidified under the draw ratio 10 in a coagulated bath with phosphate acid. The fibers was washed with running water to remove the solvent PPA and wrapped on plastic spools. Subsequently, the fibers were dried under vacuum at 80 °C for 24 h.

2.6 Characterization and measurements

Fourier-transform infrared (FTIR) spectra were recorded over wavenumber range of 4000–400 cm⁻¹ for characterization of the chemical structure using a Thermo-Nicolet Nexus 670 spectrometer. High-resolution ¹³C NMR (¹³C CPMAS NMR) measurements were carried out at 9.4 T on a Varian Infinity Plus 400 spectrometer operating at a ¹³C Larmor frequency of 100.4 MHz. Transmission electron microscopy (TEM) images were observed from a JEOL JEM-2010 operating at 200 kV. The ultra-thin sections were cut with a diamond knife from the compression molded sheets of the composite and placed on a 300 mesh Cu-grid for TEM analysis. The chemical nature and elemental composition were characterized by X-ray photoelectron spectroscopy (XPS) spectra using HI 5700 ESCA System (Perkin-Elmer). WAXD patterns of PBO and CNTs I–III & PBO copolymer fibers were obtained on a multi-filament bundle by the RIGAKU D/MAX-rB rotating anode X-ray generator with Ni-filtered Cu Kα radiation operated at 100 mA and 40 kV. The surface and fractured cross-sections of PBO and the CNTs I–III & PBO copolymer fibers mounted on a copper sample board were coated with gold using an ion sputter and then observed by SEM (HITACHI S-3700). TG analyses were carried out with a thermogravimetric analyzer (NETZSCH STA-449C) from room temperature to 750 °C in air at a heating rate of 10 °C min⁻¹. Tensile testing of PBO and CNTs I–III & PBO copolymer fibers were was performed on a universal tensile tester (model WD⁻¹) using 2.00 cm gage length at a strain rate of 10 mm min⁻¹. Fifty samples were tested in each case.

3. Results and discussion

It is not easy to use pristine CNTs to improve the properties of polymer composites, it is necessary to modify CNTs.²⁰ Studies have shown that chemical functionalization of CNTs can effectively improve both the dispersion of CNTs in polymer matrix and compatibility between these two phases.²⁴ In the study, an optimized interfacial interaction between the CNTs and PBO matrix was performed via three types of dicarboxylic acids functionalized CNTs, result in well disperse, high reaction activity and an efficient load transfer.

3.1 Characterization and properties of functionalized CNTs

The pretreated CNTs were characterized by FTIR spectra, as shown in Fig. 1. In Fig. 1b, there obvious absorption bands of carboxylic acid groups can be detected at 1639 and 3465 cm⁻¹ for the acid-treated CNT. The peak at 1639 cm⁻¹ can be attributed to C=O stretches, and the sharp peak 3465 cm⁻¹ is assigned to the O–H stretching vibration. The characteristic peaks indicate that carboxylic acid groups have been introduced to the defect sites on end and the surface of the CNTs after acid-treatment. Fig. 1c shows acyl chloride-treated CNTs. The peak at 3465 cm⁻¹ belonged to the O–H stretching vibration of carboxylic acid has almostly disappeared. The new peaks 3000–2800 cm⁻¹ can be assigned to C–H stretching mode of defect sites of CNTs. After acyl chloride-treatment, the peak assigned to the C=O stretching vibration have moved from 1639 cm⁻¹ to 1694 cm⁻¹, while the C–Cl absorption peak at 1650 cm⁻¹ is observed. All the results show that the CNTs–COOH have completely been acyl chlorided. Fig. 1d–f show CNTs functionalized with I, II and III, respectively. The C=O stretching vibration of the amide bond formed by the grafted reaction of CNT appear at 1625 cm⁻¹ (amide I band) and 1551 cm⁻¹ (amide II band). The peaks at 3427 cm⁻¹ are attributed to N–H and the O–H stretching vibrations of the attached amines and dicarboxyl of the introduction of I, II and III. All peaks show that the CNTs have been connected with I, II, III by chemical bonds.

Fig. 1 FTIR spectra of functionalized CNTs: (a) CNTs, (b) acid-treated CNTs, (c) acyl chlorided CNT, (d–f) three amines dicarboxyl acid functionalized CNTs: CNTs I, CNTs II, CNTs III.

Fig. 2 shows the TGA thermograms of CNTs I–III in air. Observed from the TGA curve, CNTs I–III show the similarly decomposes trend and have three thermal degradation steps. The first degradation step occurs in the 102 °C due to the loss of less water. The second degradation step for CNTs I–III at 233 °C can be attributed to thermal decomposition of acylamide group and I–III on the CNTs. The third degradation step at 439 °C can be attributed to the oxide degradation of graphic CNTs in air. Moreover, CNTs III has the more thermal ability than CNTs I and CNTs II due to the rigid structure.

Fig. 2 TG curves of functionalized CNTs (CNTs I–III).

Solid state NMR is a potentially powerful technique for characterizing functionalized CNTs.²⁵ To reveal if the CNTs chemically bonded with DAR of PBO monomer in this work, ¹³C NMR spectrum of DAR–CNTs III was measured and shown in Fig. 3. A characteristic ¹³C CNT resonance at 128.3 ppm and new signals at 163.8 ppm of C₇, 130.1 ppm of C₁₃ and 144.2 ppm of C₈ contributed to benzoxazole rings. The results demonstrate that DAR has reacted with the dicarboxyl groups on CNTs III. As a result, the analyses of FT-IR, TG and ¹³C NMR indicate there is an efficient covalent bonding between the CNTs and the DAR via condensation reaction during the aforementioned reaction.

Fig. 3 Solid state ¹³C NMR spectrum of DAR–CNTs III.

3.2 Structural characterizations of CNTs I–III & PBO copolymer fibers

Visual photographs of PBO and CNTs I–III & PBO copolymer fibers are exhibited in Fig. 4. Continuous long, shining golden PBO fibers and black CNTs I–III & PBO copolymer fibers could be observed. The FTIR spectra of PBO fibers, CNTs I & PBO, CNTs II & PBO and CNTs III & PBO copolymer fibers, are shown in Fig. 5. Three characteristic peaks of PBO, an aromatic C–H stretch at 2930–3067 cm⁻¹, an –C=N stretch at ∼1682 cm⁻¹ and =C–O–C stretch at ∼1060 cm⁻¹ were detected in CNTs & PBO copolymer fibers. The absorption peak at 1626 cm⁻¹ is attributed to the amide carbonyl stretch band. The peak around 1576 cm⁻¹ is associated with the N–H in plan stretch and the vibration of carbon skeleton in CNTs, and the peak around 1202 cm⁻¹ can be attributed to the stretching mode of C–N bond. The results have showed the presence of amide groups on the end and defects sites of CNTs, which link CNTs and PBO molecular chains. In addition, the peak at 1725 cm⁻¹ owned to the stretching vibration of O–C=O in carboxylic group from the excess or unreacted PTA disappears in the PBO fibers, which did not appear in CNTs I–III & PBO copolymer fibers. The result indicates dicarboxylic acids grafted on the CNTs have completed reacted with the DAR just as PTA via in situ polymerization.

Fig. 4 Photographs of (a) PBO fibers, (b) CNTs I & PBO, (c) CNTs II & PBO and (d) CNTs III & PBO copolymer fibers.

Fig. 5 FTIR spectra of fibers: (a) PBO fibers, (b) CNTs I & PBO copolymer fibers, (c) CNTs II & PBO copolymer fibers, (d) CNTs III & PBO copolymer fibers.

Preparation of high molecular weight polymer plays an important role in achieving the high performance fibers. However, it is hard to measure the molecular weight of heterocyclic polymer. It is usually to use inherent viscosity [η] to reflect the molecular weight indirectly, because the viscosity is dependent on solutions concentration and molecular weight of the dissolved polymer. The inherent viscosity was measured by Ubbelohde viscosimeter using MSA as a solvent at 30 °C in water bath. The viscosity-average molecular weights (M_v) were calculated from the five-spot method using the Mark–Houwink equation:²⁶ [η] = 2.77 × 10⁻⁷M_v^1.8. The measured intrinsic viscosities [η] and calculated molecular weights for PBO and CNTs I–III & PBO copolymers are listed in Table S1.† It can be seen that the molecular weights of CNTs I & PBO is almost equal to CNTs II & PBO. CNTs I–III & PBO copolymers are or higher than that of PBO. So, the special dicarboxyl acids functionalized CNTs do not hinder the growth of PBO molecular chains, instead it is good for high molecular weight. It may be caused by the addition and uniform dispersion of dicarboxyl acids functionalized CNTs, which improves the oxidation resistance of DAR and makes polymerization to high molecular weight possible. Therefore, the synthesis process of CNTs I–III & DAR may also provide a new method for the preparation of high molecular weight CNTs & PBO copolymers.

The outer surface and fracture surface morphologies of the fibers have been studied by SEM, as shown in Fig. 6. It can be observed that the longitudinally outer surface of CNTs I–III & PBO fibers (Fig. 6c, e and g) become rough compared with PBO fibers (Fig. 6a), which should be contributed to the introduction of modified CNTs and uniform distribution of CNTs in PBO matrix (skin layer and core layer). Fig. 6b, d, f and h show the tensile fracture surfaces of CNTs I–III & PBO fibers. Notable signs of weak microfibrillar, more compact fracture surfaces and lower breaking elongations than that of PBO fibers could be observed. The results should be related to the functionalized CNTs. The functionalized CNTs were covalently bonded with PBO to form network. In the network, CNTs also reinforced the microfibrillar interactions, decrease the microfibrillar and breaking elongations.

Fig. 6 SEM images of the longitudinal outer surfaces and fracture surface of fibers: (a and b) PBO fibers, (c and d) CNTs I & PBO, (e and f) CNTs II & PBO and (g and h) CNTs III & PBO copolymer fibers.

The CNTs dispersion in the polymer matrix was largely influenced by their dispersion in the solvent. The pretreated CNTs with dicarboxylic acids prevented the dispersed nanotubes from reaggregation. The dispersion degree of CNTs in the PBO fiber was studied by TEM observation. Fig. 7 shows TEM images of cross-sections of PBO fibers and CNTs III & PBO copolymer fibers. In Fig. 7a–c shows radial cross section of PBO fibers and CNTs III & PBO copolymer fibers. The inset images of Fig. 7a and c show electron diffraction pattern of PBO matrix and CNTs at high magnification, respectively. The obvious difference between PBO matrix and CNTs could be observed from electron diffraction. Fig. 7d shows axial cross section of CNTs III & PBO copolymer fibers. Individual CNTs in the PBO matrix is visible as darker features. From 7d, CNTs with the end diameter about 5.0 nm were observed. According to the relate reports,^27,28 we could conclude that CNTs as individual tube are distributed in the PBO matrix, which have not aggregation.

Fig. 7 High resolution TEM images: (a) radial cross section of PBO fibers, (b) integral and (c) partial enlarged view of radial cross section of CNTs III & PBO copolymer fibers, (d) partial enlarged view of axial cross section of CNTs III & PBO copolymer fibers.

The WAXD patterns of PBO fibers and CNTs & PBO copolymer fibers along the equatorial directions are shown in Fig. 8. Two strong reflections located on the equator can be indexed as (200) and (010), respectively. These two reflections show a small spread along the azimuthal direction indicating that the PBO fiber has a high degree of crystal orientation along the fiber axis. Meanwhile, the spread became broader with addition of modified CNTs and the two reflections spacing of CNTs I–III & PBO copolymer fibers became narrower. It indicates that the orientation decreases. Therefore, CNTs I–III & PBO copolymer fibers with lower orientation degree show higher tensile properties than PBO fibers, we presumed the results are caused by intensified covalent interaction of PBO molecules with functionalized CNTs and well dispersion of CNTs I–III in PBO matrix. The results are similar to the related reports.¹²

Fig. 8 X-ray diffraction patterns of (a) PBO fibers, (b) CNTs I & PBO copolymer fibers, (c) CNTs II & PBO copolymer fibers, (d) CNTs III & PBO copolymer fibers and functionalized CNTs.

3.3 Mechanical properties of CNTs I–III & PBO copolymer fibers

The testing apparatus for mechanical properties is shown in Fig. 9a. Tensile modulus, the tensile strength and the tensile break elongation of the PBO and CNTs I–III & PBO copolymer fibers are listed in Table 1 and Fig. 9b. PBO fibers show an average tensile strength of 1.13 GPa and tensile modulus of 54.07 GPa. Compared with PBO fibers, tensile strengths of CNTs I–III & PBO (1%) fibers increase by 31%, 25% and 38%, respectively. The tensile moduli of CNTs I–III & PBO (1%) fibers increase by 61%, 51%, 76%, respectively. The elongation experienced a small loss as result of functionalized CNTs I–III addition. When the fiber is stretched, CNTs connect with microfibrils probably carry the most loading. From the comparative analysis of these results, increase on tensile strength and modulus of the PBO fibers by CNTs I is better than that by CNTs II. It is speculated that the flexible chains' length of dicarboxylic acids grafted CNTs affect the mechanical property. The CNTs II has longer flexible chains than CNTs I, which make the CNTs II & PBO is prone to break when loading external force. Furthermore, addition of CNTs III has the most effects on tensile strength and modulus among CNTs I–III. The results indicate that the introduction of rigid chain molecules on CNTs has the better effect on mechanical property than of flexible chain molecules, the shorter the flexible chains, the better of the PBO fibers. These data are better than the results (35%, 13%) of PBO copolymer fibers reinforced by 1% carboxylic functionalized CNTs¹⁹ and the results (23.1%, 13%) of PBO composite fibers reinforced by 5% pure SWCNTs¹⁴ in the earlier report. In the study, addition of such a low content of CNTs, the strength and modulus of PBO fibers have considerably improved. Because the special modified CNTs improve well disperse, high reaction activity and an efficient load transfer between CNTs fillers and PBO matrix. Through addition of the special modified CNTs, the PBO chains change from one to three dimensional crosslink network structure. It is a very effective method for enhancing tensile strength and modulus of PBO fiber and reducing microfibrillar.

Fig. 9 (a) Testing facility for mechanical properties, (b) tensile strength and tensile modulus of PBO and CNTs I–III & PBO fibers.

Table 1 Mechanical properties of PBO and CNTs I–III & PBO fibers

Sample fibers	Diameter (μm)	Tensile strength (GPa)	Tensile modulus (GPa)	Elongation at break (%)
PBO	39 ± 2	1.13 ± 0.05	54.07 ± 7.1	2.09 ± 0.4
CNTs I & PBO	42 ± 3	1.48 ± 0.04	87.06 ± 9.3	1.70 ± 0.3
CNTs II & PBO	42 ± 3	1.41 ± 0.04	81.50 ± 8.9	1.73 ± 0.3
CNTs III & PBO	43 ± 3	1.56 ± 0.03	95.12 ± 9.6	1.64 ± 0.3

---

In general, the uncontrolled CNTs may be distributed randomly within the PBO matrix, but it is also possible to arrange for them to be aligned preferentially in the axial direction of the PBO fibers when the PBO macromolecular oriented. Thus, the Halpin–Tsai formula were used to calculate the effect of the volume fraction of the CNTs filler phase,²⁹ and predict tensile modulus of composites in above two ways. The tensile modulus data are compared with those predicted according to the Halpin–Tsai equation, which is shown in Fig. S1.† It is shown that the measured values are higher than the predicted ones. A similar phenomenon has been reported and explained in the literature.^19,20 This finding is explained by the presence of highly aligned CNTs in PBO matrix and strong covalent bonding at the CNTs–PBO interface. The CNTs were modified by dicarboxylic groups and further covalently bonding with PBO molecular chains. The special modified CNTs improve the compatibility of CNTs and PBO, transfer loads from PBO to CNTs, and also protect the PBO fibers from environmental attack. So, the control of the CNTs filler by chemical modification is an important means of optimizing compatibility of CNTs fillers in PBO matrix and strengthening the composites properties.

The thermal stabilities of all PBO and CNTs I–III & PBO fibers have also been evaluated using TG thermograms under air at a heating rate of 10 °C min⁻¹, as shown in Fig. 10. The TG thermograms indicated that all fibers have excellent thermal stabilities before the temperature of 550 °C unless little decomposition. From TGA thermograms, it is observed that maximal decomposition of PBO fibers started at a temperature of 550 °C. The maximal decomposition of CNTs I–III & PBO fibers starts around 597, 609, 609 °C, respectively. The thermal stability of CNTs I & PBO, CNTs II & PBO and CNTs III & PBO fibers increase by 47, 59 and 59 °C compared with the PBO fibers. The degradation tendency of CNTs I & PBO and CNTs II & PBO showed the similar trend and a little decomposes about 300 °C, which could be caused by grafted flexible chain molecules of CNTs, resulting in a relatively weak connectivity between CNTs and PBO. The thermal decomposition trend of CNTs II & PBO is more obvious than CNTs I & PBO. It shows that addition of CNTs with more length and flexible chains will decrease the thermal stability of PBO fibers. It also have higher thermal stability when addition of CNTs III with rigid chains than with flexible chains. The results confirm there are efficient interfacial adhesion and the effect of three-dimensional framework between CNTs fillers and PBO matrix. It also provides a reference for the future research work on grafting carbon nanotube by dicarboxyl acid.

Fig. 10 TG curves of PBO and CNTs I–III & PBO copolymer fibers in air.

The mechanism of CNTs I–III reinforced PBO fibers is proposed as Fig. 11. In this research, the obtained CNTs I–III & PBO copolymer fibers have more high mechanical and thermal properties compared with PBO fibers, because: (i) the CNTs is pretreated by dicarboxyl acid, which prevents CNTs from reaggregating and obtained good dispersion in PPA solution, improved the compatibility of the CNTs and PBO. (ii) The dicarboxyl groups on the CNTs just as PTA provide active groups to react with DADHC of PBO monomer, and then further graft with PBO macromolecular via in situ polymerization. Compared with one monocarboxyl CNTs, the special dicarboxyl functionalized CNTs are unable to seal end and prevent the growth of PBO chains, result in the high molecular weight of CNTs I–III & PBO. (iii) The covalent bonding bridge the CNTs and PBO matrix, form the framework in CNTs I–III & PBO copolymers, enhance the interface interaction and restrict the slip of molecular chains. Meanwhile, framework makes CNTs reinforced microfibrils, when outside force act on the fibers, the strong interface interaction will transfer mechanical load between CNTs fillers and PBO matrix. As a result, the good dispersion of CNTs in PBO matrix, the interface interaction from the covalent bond and network in copolymer fibers are expected to make copolymer fibers have the higher mechanical and thermal properties.

Fig. 11 Proposed mechanism of CNTs I–III reinforced PBO fibers.

4. Conclusions

In this work, PBO fibers reinforced with functionalized CNTs were prepared via surface grafting of three different dicarboxyl acid onto CNTs. Three dicarboxyl acids have been used as a coupling agent to improve the dispersivity, reactivity and interfacial adhesion of CNTs in PBO matrix. The obtained CNTs I–III & PBO copolymer fibers exhibit a dramatic increase in the mechanical properties and thermal stability compared with the PBO fibers. Moreover, the effect of rigid chain dicarboxyl acid modified CNTs on mechanical properties and thermal stability of copolymer fibers are better than the other two. When the CNTs I–III & PBO fibers is subjected to tensile loading, the load could be transferred from PBO to CNTs because of perfect bonding and a strong interface between the CNTs and PBO matrix.

Acknowledgements

The authors gratefully acknowledge the Chang Jiang Scholars Program and the Fund of the Science and Technology Department of Jilin Province (No. 20140204017GX, 20150520025JH) for financial supports.

Notes and references

1. E. W. Choe and S. N. Kim, Macromolecules, 1981, 14, 920–924.
2. J. F. Wolfe and F. Arnold, Macromolecules, 1981, 14, 909–915.
3. J. F. Wolfe, Encyclopedia of Polymer Science and Engineering, Wiley-Interscience, 2nd edn, 1988, vol. 11, pp. 601–635.
4. H. G. Chae and S. Kumar, J. Appl. Polym. Sci., 2006, 100, 791–802.
5. K. Tashiro and M. Kobayashi, Macromolecules, 1991, 24, 3706–3708.
6. S. G. Wierschke, J. R. Shoemaker, P. D. Haaland, R. Pachter and W. W. Adams, Polymer, 1992, 33, 3357–3368.
7. M. Northolt, D. Sikkema, H. Zegers and E. Klop, Fire Mater., 2002, 26, 169–172.
8. Y. Li, J. Li, Y. Song, Z. Hu, F. Zhao and Y. Huang, J. Polym. Sci., Part A: Polym. Chem., 2013, 51, 1831–1842.
9. T. Kitagawa, K. Yabuki and R. Young, Polymer, 2001, 42, 2101–2112.
10. T. Kitagawa, K. Yabuki, A. C. Wright and R. J. Young, J. Mater. Sci., 2014, 49, 6467–6474.
11. S. Iijima, Nature, 1991, 354, 56–58.
12. M. Moniruzzaman and K. I. Winey, Macromolecules, 2006, 39, 5194–5205.
13. Z. Spitalsky, D. Tasis, K. Papagelis and C. Galiotis, Prog. Polym. Sci., 2010, 35, 357–401.
14. S. Kumar, T. D. Dang, F. E. Arnold, A. R. Bhattacharyya, B. G. Min, X. Zhang, R. A. Vaia, C. Park, W. W. Adams and R. H. Hauge, Macromolecules, 2002, 35, 9039–9043.
15. T. Fukumaru, T. Fujigaya and N. Nakashima, Macromolecules, 2013, 46, 4034–4040.
16. J. Li, X. Chen, X. Li, H. Cao, H. Yu and Y. Huang, Polym. Int., 2006, 55, 456–465.
17. X. Li, Y. Huang, L. Liu and H. Cao, J. Appl. Polym. Sci., 2006, 102, 2500–2508.
18. K. Kobashi, Z. Chen, J. Lomeda, U. Rauwald, W.-F. Hwang and J. M. Tour, Chem. Mater., 2007, 19, 291–300.
19. Z. Hu, J. Li, P. Tang, D. Li, Y. Song, Y. Li, L. Zhao, C. Li and Y. Huang, J. Mater. Chem., 2012, 22, 19863–19871.
20. C. Zhou, S. Wang, Y. Zhang, Q. Zhuang and Z. Han, Polymer, 2008, 49, 2520–2530.
21. C.-W. Lin, S. L.-C. Hsu and A. C.-M. Yang, Polymer, 2012, 53, 1951–1959.
22. Z. Lysenko, US Pat., 4, 766, 244, 1988-8-23.
23. R. G. Pews, Z. Lysenko and P. C. Vosejpka, J. Org. Chem., 1997, 62, 8255–8256.
24. P.-C. Ma, N. A. Siddiqui, G. Marom and J.-K. Kim, Composites, Part A, 2010, 41, 1345–1367.
25. B. K. Gorityala, J. Ma, X. Wang, P. Chen and X.-W. Liu, Chem. Soc. Rev., 2010, 39, 2925–2934.
26. X. D. Hu, S. E. Jenkins, B. G. Min, M. B. Polk and S. Kumar, Macromol. Mater. Eng., 2003, 288, 823–843.
27. T. Uchida and S. Kumar, J. Appl. Polym. Sci., 2005, 98, 985–989.
28. C. Park, Z. Ounaies, K. A. Watson, R. E. Crooks, J. Smith, S. E. Lowther, J. W. Connell, E. J. Siochi, J. S. Harrison and T. L. St Clair, Chem. Phys. Lett., 2002, 364, 303–308.
29. J. Affdl and J. Kardos, Polym. Eng. Sci., 1976, 16, 344–352.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra16541a

---