- Email: [email protected]
Characterization of poly(butylene terephthalate) composites prepared by using various types of sized carbon fibers
Materials and Design 87 (2015) 318–323
Contents lists available at ScienceDirect
Materials and Design journal homepage: www.elsevier.com/locate/jmad
Characterization of poly(butylene terephthalate) composites prepared by using various types of sized carbon fibers N. Gamze Karsli a, Cem Ozkan b, Ayse Aytac a,b,⁎, Veli Deniz a,b a b
Department of Chemical Engineering, Kocaeli University, Engineering Faculty, 41380 Kocaeli, Turkey Department of Polymer Science and Technology, Kocaeli University, 41380 Kocaeli, Turkey
a r t i c l e
i n f o
Article history: Received 3 February 2015 Received in revised form 8 August 2015 Accepted 10 August 2015 Available online 12 August 2015 Keywords: Carbon fiber Poly(butylene terephthalate) Fiber/matrix adhesion Mechanical testing
a b s t r a c t This paper aims to study the effects of sizing on properties of differently sized carbon fiber (CF) reinforced poly(butylene terephthalate) (PBT) composites by comparing them to unsized CF reinforced composites. Contact angle analysis was used to evaluate the wettability of CFs and the work of adhesion between the sizing agent and PBT matrix. It was found that wettability of PU sizing material by PBT matrix was better than that of other sizing materials by PBT matrix. Tensile and dynamic mechanical analysis (DMA) tests were performed to investigate the effect of sizing agent type on mechanical and thermomechanical properties. According to tensile test results PHE and PU sized CF reinforced PBT composites gave higher tensile strength and modulus values than the others. DMA revealed that PU sizing material gave better adhesion strength than other sizing materials. It was found that electrical conductivity values of all composites are about 10−2 S/cm. SEM analysis showed that PU sized CF surface covered a layer of PBT matrix in accordance with other test results. As a conclusion of all results, it can be suggested that PU is a proper sizing material to be used for CF surface for PBT matrix. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Thermoplastic polymers exhibit good properties such as high impact strength, and fracture resistance. Because of this reason, they are widely used in polymer engineering industry [1]. Poly(butylene terephthalate) (PBT) is one of these thermoplastic polymers and it has been mostly used in automotive, electrical, and electronic industries for many years due to its exceptional electrical insulation properties [1–3]. PBT is considered as the main competitor for polyamides, but it has the advantage over polyamides due to its much lower moisture uptake and better dimensional stability properties under wet conditions [4]. Moreover, PBT has been widely used in fiber-reinforced composites because of its good mechanical properties and properties of easy molding and fast crystallization [5]. When compared to glass fiber reinforced PBT composites; carbon-fiber-reinforced PBT composites exhibit higher tensile strength and modulus due to exceptional mechanical effects of carbon fibers [5]. On the other hand, mechanical properties of fiber reinforced composites not only depend on the intrinsic characteristics of the matrix and fiber, but also on the fiber/matrix interface adhesion. Since the stress transfer phenomena from the matrix to the fiber during deformation is affecting mechanical resistance of composites, better interfacial adhesion militates in favor of better stress transfer and
⁎ Corresponding author at: Kocaeli University, Chemical Engineering Department, Umuttepe Campus, 41380 Izmit, Kocaeli, Turkey. E-mail address: [email protected] (A. Aytac).
http://dx.doi.org/10.1016/j.matdes.2015.08.047 0264-1275/© 2015 Elsevier Ltd. All rights reserved.
mechanical properties. Therefore, it is crucial to control the interfacial adhesion to get better mechanical properties [6,7]. The interfacial adhesion becomes more important when using thermoplastic matrix because of the poor chemical functionality of this matrix type [8]. Besides, carbon fibers contain very few reactive groups due to their carbonization process which occurs at elevated temperatures, since these functional groups on the fiber surface are eliminated at elevated temperatures [9]. The lack of reactive groups on the fiber surface and thermoplastic matrix leads to weak adhesion between carbon fiber and thermoplastic matrix. In the case of weak adhesion, the matrix cannot transfer the stress to fiber efficiently and fiber cannot perform its duty of carrying the applied load [8]. Two methods are generally used for obtaining better adhesion and stress transfer. These methods are matrix modification and fiber surface modification. One of the fiber surface modification methods is sizing and in this method, carbon fibers are usually coated with a sizing material which consists of a proper polymer or resin [7,10,11]. The sizing material includes functional groups and these groups change the surface free energy and wettability of carbon fiber surfaces [12]. Also they can react to or interact with the polymer matrix. This reaction or interaction enhances the fiber–matrix adhesion and correspondingly the ultimate properties of composites [13]. In the literature, a few studies about carbon fiber-reinforced PBT composites have been undertaken. One of these studies was reported by Ng et al. [14]. They added boron nitride (BN) and carbon fiber (CF) into PBT matrix and investigated the hybrid filler effect on the properties of PBT composites. They found that, while the combination of BN and CF in PBT significantly reduced electrical conductivity of the
N. Gamze Karsli et al. / Materials and Design 87 (2015) 318–323
composites, the usage of mixed fillers did not lead to an improvement in thermal conductivity with respect to PBT/BN composites. They also observed that hybrid composites exhibited better tensile properties and processability than PBT/BN composites at the same total filler content. Another study was performed by Wiedmer et al. [15]. They studied the effect of electron beam radiation on carbon fiber reinforced PBT, PPS and PA composites. They prepared composites with and without the presence of crosslinking agent to clarify whether crosslinking could occur. They examined the thermal and mechanical properties of the composites before and after exposure to irradiation. Consequently, they observed that properties of PBT, PPS and PA46 composites did not change significantly in the presence of crosslinking agent after irradiation. On the other hand, CF/PA66 composites exhibited changes in some of their properties in the presence of a crosslinking agent after irradiation. Chen et al. [16] investigated the properties of recycled carbon fiber (RCF) reinforced PBT matrix composites. They also treated RCF surfaces with the solution of diglycidyl ether of bisphenol A in order to improve the interfacial adhesion between the RCF and PBT matrix. Their results showed that, surface treated RCF significantly improved the mechanical properties, heat distortion temperature, and thermal stability of the composites. The morphology studies of fracture surfaces also indicated that the RCF homogeneously dispersed in the PBT matrix. Although there are a few studies in the literature which were focused on the effects of sizing agent type, molecular weight and concentration on the properties of some of polymeric matrix based composites [17–19], there has been no study about the effects of sizing agent type on the properties of carbon fiber reinforced PBT composites to the best of authors' knowledge. This study focuses on the effects of sizing material types on the properties of CF reinforced PBT matrix composites. For this work, CFs which were unsized and which were sized with five different types of sizing agent were used as reinforcement material. The properties of composites were analyzed by mechanical, thermomechanical, electrical, morphological tests and contact angle measurements. 2. Materials and methods Matrix material PBT (Tecodur®) was supplied by Interplast (Turkey). Unsized CF and CFs sized with polyurethane (PU), polyamide (PA), polyimide (PI), phenoxy (PHE) and epoxy/phenoxy (EPO_PHE) were supplied by Akkök Group (Turkey) and used as reinforcement. PBT was dried in a vacuum oven at 120 °C for 8 h before the compounding process. PBT granules and 30 wt.% of 6 mm chopped carbon fibers were compounded in a laboratory scale co-rotating twin-screw mini extruder at 255 °C, 100 rpm. All the compounds were subsequently injection molded using a laboratory scale injection molding machine with a barrel temperature of 255 °C, mold temperature of 130 °C and injection pressure of 10 bars. Contact angles of test liquids against the sizing materials and PBT matrix material were measured by the sessile drop method with Attention Theta Lite contact angle tensiometer. The used test liquids were ethylene glycol, diiodomethane and deionized water. Surface tension of test liquids was given in Table 1 [20]. Standard test specimen of neat PBT was produced for contact angle analysis using a twin-screw mini extruder and laboratory scale injection molding machine. In addition, thin films of each sizing material were cast onto clean microscope
Table 1 Surface tension values of test liquids. Surface tension
Diiodomethane
Ethylene glycol
Deionized water
γTOT L γLW L γAB L γ− L γ+ L
50.8 50.8 – – –
48 29 19 47 1.92
72.8 21.8 51.0 25.5 25.5
319
slides for contact angle analysis of sizing materials. Ten measurements were conducted to obtain average contact angle values. Tensile tests were performed using Instron 4411 universal testing machine. The dimension of the test samples was 4 mm width, 2 mm thickness and 30 mm length. Average tensile strength and modulus values of composites were determined using 5 dumbbell-shaped samples for each composition at a constant crosshead speed of 5 mm/min. DMA was performed in tension mode by using Metravib DMA50 analyzer. Composites were tested under the condition of 1.5 × 10−5 m dynamic displacement and 1 Hz frequency. The scanning was carried out from 25 °C to 200 °C at 1 °C/min heating rate. Specimen dimensions were 10 mm height, 2 mm thickness and 4 mm width. Electrical resistivity values of composites were measured with 2-point-probe test by using Haoyue M890G Digital Multi Meter. For obtaining good electrical contact in this technique, copper wires were attached to both ends of test sample with silver paste. After the hardening of silver paste, resistivity measurements were performed by contacting probes with these copper wires. After that, electrical conductivity values of composites were calculated using obtained resistance values of composites as in the following formula [21,22]: Sample Thickness ðcmÞ : Electrode Area ðcm2 Þ Resistance ðΩÞ
ð1Þ
Morphologies of the tensile fractured surfaces of composites were examined using a scanning electron microscope (JEOL JSM-6335F). Before the examinations, tensile fracture surfaces of composites were sputter coated with gold and palladium. 3. Results and discussion 3.1. Surface energy analysis Sizing process which was applied to carbon fiber surface changes the surface energy and wettability of the surface and affects the fiber matrix adhesion performance. One of the methods to be used for analyzing adhesion performance of fiber surfaces is the contact angle measurement. Contact angle concept which was first developed and formulized by Young, became the main concept for the development of later approaches. One of these approaches was developed by Van Oss, Good and Chaudhury, which aimed at estimating the surface energy values of solids [23,24]. According to their approach, a solid surface consists of two terms: one of them is the Lifshitz–van der Waals interactions, γLW, which includes dispersion, dipolar and induction interactions, and the other one is the acid–base interaction term, γAB, which includes + all the electron donor (γ− S ) and electron acceptor (γS ) interactions. By combining this approach with the Young–Dupré equation, the general contact angle equation is obtained: qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi LW γþ γLW γS γþ γL ð1 þ cos θÞ ¼ 2 L : S γL þ S γL þ
ð2Þ
The value of γLW is determined from the contact angle of an apolar S liquid on the solid in which case Eq. (2) reduces to: qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi LW γLW γL ð1 þ cos θÞ ¼ 2 S γL
ð3Þ
when γLW S is known and the contact angles are obtained using different liquids on the solid, one can get two equations similar to Eq. (2), and − these equations can be solved simultaneously for γ+ S and γS . In addition to this, by writing Eq. (2) as below, work of adhesion between solid and liquid can be calculated: qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi LW γþ γLW γS γþ Wa ¼ γL ð1 þ cos θÞ ¼ 2 L : S γL þ S γL þ
ð4Þ
320
N. Gamze Karsli et al. / Materials and Design 87 (2015) 318–323
Table 2 Contact angle values. Sizing & matrix
PU EPO_PHE PI PHE PA Unsized CF PBT matrix
Table 4 Work of adhesion (Wa) values of composites. Contact angle (θ°) Diiodomethane
Ethylene glycol
Deionized water
19 34 32 31 54 40 57
51 55 23 37 46 32 43
37 56 53 52 60 55 81
In this study, in order to calculate the work of adhesion between carbon fiber and PBT matrix, surface tension values of sizing materials and matrix material were measured using three test liquids. Contact angle values and surface energies (polar and dispersive) of unsized CF, sizing materials and PBT matrix are given in Tables 2 and 3. In addition, surface tension value of unsized CF was obtained from literature [25]. In order to obtain good wetting of fiber by matrix, surface energy of matrix should be lower than that of fiber [26]. It can be seen from Table 3 that while PBT matrix exhibits the lowest surface energy value, the maximum surface energy is obtained for PU sizing material. According to these results, it can be concluded that, wetting between PU sized CF and PBT is better than that of other sizing materials and PBT matrix. While wetting of fiber by polymer is the prior condition of good fiber matrix adhesion, functional groups on fiber surface also plays an important role for good adhesion. This idea was first developed by Fowkes [27]. According to Fowkes, hydrogen bonding interactions are important for fiber–matrix adhesion and these interactions can be evaluated by investigating the polar components of surfaces. Therefore, when choosing the proper sizing material for used matrix material, the polar component amount of sizing material should also be taken into account. It can be seen from Table 3 that in our study PU sizing material exhibited the highest amount of polar component among all samples. Fiber–matrix interface strength can also be calculated as “work of adhesion” (Wa) value by using Eq. (4) [28]. The Wa value can be separated into two components. One of them is Wa12 and it represents the work of adhesion value between unsized CF and sizing material. The other one is Wa23 and it represents the work of adhesion value between sizing material and matrix material. In the case when Wa12 N Wa23, it can be said that the required energy to separate unsized CF and sizing material is higher than the required energy to separate sizing and matrix materials. In such a case, the sizing contributes to the fiber–matrix adhesion while it protects the fiber from breakage during composite preparation process. On the other hand, when Wa23 N Wa12, it can be said that the required energy for the separation of the sizing material and matrix material is higher than the required energy to separate unsized CF and sizing material. In such a case, sizing cannot protect the fiber from breakage satisfyingly. In our study, the calculated Wa12 and Wa23 values were given in Table 4.
Sizing type
Work of adhesion between unsized CF and sizing material Wa12 (mJ·m−2)
Work of adhesion between sizing material and PBT matrix Wa23 (mJ·m−2)
PU EPO_PHE PI PHE PA
108.6 95.2 92.8 104.8 80.4
102.4 90.2 87.4 88.8 77.5
As can be seen from Table 4, Wa12 values are higher than that of for all cases. It can be concluded that the adhesion strength between all sizing materials and unsized CF is higher than the adhesion strength between sizing material and PBT matrix. Hence, it can also be said that all sizing materials used in this study protect the fiber from breakage and contribute to the fiber matrix adhesion during processing. Table 4 also shows that the maximum Wa values (W a12 and Wa23) were obtained for PU sizing material. This result can be attributed to the highest surface energy of this sizing material. Thus, PU, owing to its high surface energy, exhibits better interfacial adhesion with both unsized CF and PBT matrix.
Wa23
3.2. Tensile tests The tensile strength value of neat PBT is 63.9 MPa and the addition of unsized CF increases this value to approximately 125 MPa. Also, tensile test results of sized CF reinforced PBT composites were evaluated in Fig. 1. Fig. 1 shows that the maximum tensile strength values were obtained for PHE and PU sized CF reinforced PBT composites, respectively. According to tensile strength values of composites, it can be said that while the maximum enhancement in tensile strength was observed as 111% with addition of PHE sized CF to PBT matrix, this enhancement was observed as 107% with addition of PU sized CF to the PBT matrix. As is known, when the adhesion between fiber and matrix is good, applied load can efficiently be transferred from the matrix to the fiber during tensile test, and also fiber can bear more load in such a case. Consequently, it can be concluded that adhesion between PHE and PU sized CF and PBT matrix is better than the others. It was supposed that proton donor hydroxyl groups of PHE and proton acceptor carbonyl groups of PBT may react via hydrogen bonding and this reaction could improve the adhesion between PHE and PBT [29]. There might be also a chemical interaction between isocyanate groups of PU and carboxyl groups of PBT. Thus, it is possible to obtain better adhesion between PU sized CF and PBT by means of these chemical interactions [9]. Consequently, it can be considered that improved tensile strength values arose from these chemical interactions.
Table 3 Surface energy values of samples. Sizing & matrix
Polar component γsAB (mJ·m−2)
Dispersive component γsLW (mJ·m−2)
Total surface energy γsTot (mJ·m−2)
PU EPO_PHE PI PHE PA Unsized CF⁎ PBT matrix
21.8 9.5 5.5 6.8 4.4 4.8 4.9
48 42.2 43.2 43.6 32 39.6 30
69.8 51.7 48.7 50.4 36.4 44.4 34.9
⁎ The surface tension value of unsized CF was obtained from literature [25].
Fig. 1. Tensile strength values of composites.
N. Gamze Karsli et al. / Materials and Design 87 (2015) 318–323
321
Fig. 2. Modulus values of composites.
Fig. 3. Tan delta values of composites.
Effect of sizing agent type on the modulus values of composites is given in Fig. 2. Modulus value of neat PBT is 2512 MPa. It can be seen from Fig. 2 that modulus values of all composites are higher than this value. This is an expected result and can be explained by using the rule of mixture theory in the literature [10,11,30]. Additionally, PU sized CF reinforced composites exhibit the maximum modulus values. It has been explained in our previous study that the ultimate fiber length in the composite affected the modulus value of a composite [30]. When ultimate fiber length is higher, the contribution of the fibers to the modulus of the composites is more prominent. It is known that in the case of proper sized CF usage; less fiber breakage occurs during composite preparation process. Consequently, higher modulus values can be attributed to the better adhesion between fiber and matrix.
semiconductive materials varies between 10− 8–102 S/cm. The minimum fiber loading level at which a non-conductive polymer becomes a semi conductive polymer (10−8 S/cm) is called “percolation threshold.” After this point, a continuous carbon fiber network occurs in the polymer matrix and electrons transfer easily from a fiber to another by means of this network. It can be seen from Fig. 4 that electrical conductivity values of all composites are about 10−2 S/cm. As mentioned above, conductivity value of 10−2 S/cm is considered as a very high conductivity value comparatively for commodity composites, therefore these composites are considered as semi-conductive. Moreover, it can be said that used CF loading level in this study (30 wt.%) is higher than percolation threshold [22]. The maximum electrical conductivity value was measured for PU sized CF reinforced composites and electrical conductivity value of unsized CF reinforced composites was the second. When non-conductive polymeric material coated conductive particles are used in composites, this polymeric layer acts as a non-conductive gap which hinders electron transfer between conductive particles [36]. However, our test results showed that electrical conductivity of PU sized CF reinforced composites are higher than that of unsized CF reinforced composites. In this case, dispersion of fiber in polymer matrix comes into play as an important factor. In a study, Choi et al. [22] found that, electrical conductivity values of CF reinforced composites mainly depended on the dispersion of carbon fiber in polymer matrix. According to their results, when adhesion between fiber and matrix is weak, there should be a phase separation while agglomeration occurs in the composite to affect electron transfer negatively. It can be inferred from this results that, due to the better adhesion tendency between PU sized CF and PBT matrix,
3.3. Dynamic mechanical analysis (DMA) There are lots of techniques for evaluating the interfacial adhesion in polymer matrix composites. DMA has some advantages over the others, because it provides a sensitive and non-destructive detection. DMA measures two types of responses of polymeric materials against to low strain periodic deformation. One of these responses is the elastic response and it represents the stiffness of the composite. The other one is the damping response and it represents the energy dissipation in the composite [31]. This damping factor gives clues about all kinds of molecular motions in the composite including motions which take place at the interface. It also gives clues about potential of the material in dissipating mechanical energy [32,33]. Ashida and Noguchi [34] reported that there was a relation between damping factor (tan delta, tanδ) peak magnitude and interfacial adhesion in short fiber reinforced composites. According to this theory, the lower the tan delta peak magnitude, the better the obtained interfacial adhesion. Because, when the adhesion between fiber and matrix is poor, more energy dissipation takes place due to the non-efficient stress transfer between fiber and matrix. And this results in higher peak magnitude [35]. Fig. 3 shows the effect of sizing agent type on the damping factor values of composites as a function of temperature. It can be seen from Fig. 3 that the lowest tan delta peak magnitude was obtained for PU sized CF reinforced composites. As it is shown in Fig. 3, the lowest peak magnitude obtained demonstrates the fact that PU sized CF and PBT matrix give the best adhesion strength among all composites. 3.4. Electrical conductivity analysis The effect of sizing agent type on the electrical conductivity of composites is given in Fig. 4. If a conductive material such as carbon fiber is added into a non-conductive polymer, electrical conductivity property of this polymer will change. Electrical conductivity value of
Fig. 4. Electrical conductivity values of composites.
322
N. Gamze Karsli et al. / Materials and Design 87 (2015) 318–323
Fig. 5. SEM micrographs of composites (a) unsized, (b) PI, (c) PU.
dispersion of PU sized fibers is better than that of unsized fibers in the PBT matrix and this results in better electrical conductivity values.
other test results. As a conclusion of all results, the authors propose that PU is an appropriate sizing material to be used for CF surface for PBT matrix.
3.5. SEM analysis Acknowledgment Tensile fracture surface morphology analysis is widely used to characterize the adhesion between fiber and matrix in polymer composites. Fig. 5 illustrates the micrographs of tensile fracture surfaces of unsized, PI and PU sized CF reinforced PBT matrix composites respectively. It can be seen from Fig. 5(a, b) that there are some voids between the fibers and matrix while fiber surfaces are clean and smooth. This indicates that there is a weak adhesion between unsized and PI sized CFs and PBT matrix. On the other hand, Fig. 5(c) illustrates that PU sized CF surfaces are covered by a layer of PBT matrix as expected. This can be considered as another evidence for the chemical interaction and stronger interfacial adhesion between PU sized CF and PBT as mentioned above. 4. Conclusions In this study, effects of different sizing agent types on the mechanical, thermomechanical, electrical and morphological properties of CF reinforced PBT matrix composites were investigated. Contact angle analysis was also performed to evaluate the wettability performance of carbon fibers by PBT matrix. Contact angle analysis test results showed that the best wetting by PBT matrix was observed in the case of PU sizing material. Mechanical test results revealed that PU and PHE sized CF reinforced composites gave better tensile properties. In order to investigate the thermo-mechanical properties of composites, DMA was performed with the comparison of tan delta peak magnitude values of composites and evaluation of fiber matrix adhesion. It was found that PU sized CF reinforced composites exhibited the best adhesion strength among all composites. According to the results of electrical conductivity measurements, it was found that PU sized CF reinforced composites gave the highest conductivity value and it was attributed to the better adhesion between PU sized CF and PBT matrix. The results of morphological analysis of all composites studied are accordant with
This study was supported by the Ministry of Science, Industry and Technology of Turkey under project number 01020.STZ.21011-2. References [1] C.C. Hsu, J.L. Doong, C.P. Fung, Fracture toughness of injection-molded poly(butylene terephthalate) composites, Polym. Plast. Technol. 44 (2005) 335–350. [2] J.Y. Kim, The effect of carbon nanotube on the physical properties of poly(butylene terephthalate) nanocomposite by simple melt blending, J. Appl. Polym. Sci. 112 (2009) 2589–2600. [3] S.H. Jang, Y.H. Kim, S. Lim, G.D. Choi, S.H. Kim, W.N. Kim, Effects of fiber characteristics on the mechanical and rheological properties of poly(butylene terephthalate)/ glass fiber composites, J. Appl. Polym. Sci. 116 (2010) 3005–3012. [4] S. Thomas, P.M. Visakh, Handbook of Engineering and Specialty Thermoplastics, United States of America, 2010. [5] J. Scheirs, T.E. Long, Modern Polyesters: Chemistry and Technology of Polyesters and Copolyesters, Wiley, Chichester, 2003 (England). [6] H. Zhou, K. Han, M. Yu, Preparation of long glass fiber reinforced poly(butylene terephthalate) composites with chemical bonding interphase, J. Macromol. Sci. A 43 (2006) 1835–1851. [7] J. Yang, J. Xiao, J. Zeng, L. Bian, C. Peng, F. Yang, Matrix modification with silane coupling agent for carbon fiber reinforced epoxy composites, Fibers Polym. 14 (2013) 759–766. [8] L. Dányádi, J. Gulyás, B. Pukánszky, Coupling of carbon fibers to polycarbonate: surface chemistry and adhesion, Compos. Interfaces 10 (2003) 61–76. [9] X. Qian, L. Chen, J. Huang, W. Wang, J. Guan, Effect of carbon fiber surface chemistry on the interfacial properties of carbon fibers/epoxy resin composites, J. Reinf. Plast. Compos. 32 (2013) 393–402. [10] M.R. Piggott, The interface in carbon fibre composites, Carbon 27 (1989) 657–662. [11] J.O. Iroh, W. Yuan, Surface properties of carbon fibres modified by electrodeposition of polyamic acid, Polymer 37 (1996) 4197–4203. [12] Z. Dai, B. Zhang, F. Shi, M. Li, Z. Zhang, Y. Gu, Chemical interaction between carbon fibers and surface sizing, J. Appl. Polym. Sci. 124 (2012) 2127–2132. [13] W. Chen, Y. Yu, P. Li, C. Wang, T. Zhou, X. Yang, Effect of new epoxy matrix for T800 carbon fiber/epoxy filament wound composites, Compos. Sci. Technol. 67 (2007) 2261–2270. [14] H.Y. Ng, X. Lu, S.K. Lau, Thermal conductivity, electrical resistivity, mechanical, and rheological properties of thermoplastic composites filled with boron nitride and carbon fiber, Polym. Compos. 26 (2005) 66–73.
N. Gamze Karsli et al. / Materials and Design 87 (2015) 318–323 [15] S. Wiedmer, D. Gellner, K. Friedrich, Effect of electron-beam radiation on thermoplastic composites, Plast. Rubber Compos. 34 (2005) 76–84. [16] Y. Chen, X. Wang, D. Wu, Recycled carbon fiber reinforced poly(butylene terephthalate) thermoplastic composites: fabrication, crystallization behaviors and performance evaluation, Polym. Adv. Technol. 24 (2013) 364–375. [17] R.L. Zhang, Y.D. Huang, D. Su, L. Liu, Y.R. Tang, Influence of sizing molecular weight on the properties of carbon fibers and its composites, Mater. Des. 34 (2012) 649–654. [18] N.G. Karsli, C. Ozkan, A. Aytac, V. Deniz, Effects of sizing materials on the properties of carbon fiber-reinforced polyamide 6,6 composites, Polym. Compos. 34 (2013) 1583–1590. [19] R.L. Zhang, Y.D. Huang, N. Li, L. Liu, D. Su, Effect of the concentration of the sizing agent on the carbon fibers surface and interface properties of its composites, J. Appl. Polym. Sci. 125 (2012) 425–432. [20] E.C. Botelho, L. Figiel, M.C. Rezende, B. Lauke, Mechanical behavior of carbon fiber reinforced polyamide composites, Compos. Sci. Technol. 63 (2003) 1843–1855. [21] O. Koysuren, S. Yesil, G. Bayram, Effect of surface treatment on electrical conductivity of carbon black filled conductive polymer composites, J. Appl. Polym. Sci. 104 (2007) 3427–3433. [22] M.H. Choi, B.H. Jeon, I.J. Chung, The effect of coupling agent on electrical and mechanical properties of carbon fiber/phenolic resin composites, Polymer 41 (2000) 3243–3252. [23] H.Y. Erbil, Surface Chemistry of Solid and Liquid Interfaces, UK, 2006. [24] M.D. Doganci, C.E. Cansoy, I.O. Ucar, H.Y. Erbil, E. Mielczarski, J.A. Mielczarski, Combined XPS and contact angle studies of flat and rough ethylene-vinyl acetate copolymer films, J. Appl. Polym. Sci. 124 (2012) 2100–2109. [25] Y. Luo, Y. Zhao, Y. Duan, S. Du, Surface and wettability property analysis of CF300 carbon fibers with different sizing or without sizing, Mater. Des. 32 (2011) 941–946.
323
[26] E. Mäder, Study of fibre surface treatments for control of interphase properties in composites, Compos. Sci. Technol. 57 (1997) 1077–1088. [27] F.M. Fowkes, Role of acid–base interfacial bonding in adhesion, J. Adhes. Sci. Technol. 1 (1987) 7–27. [28] K. Tsutsumi, S. Ishidal, K. Shibata, Determination of the surface free energy of modified carbon fibers and its relation to the work of adhesion, Colloid Polym. Sci. 268 (1990) 31–37. [29] P. Bourgeois, T. Davidson, Plasma modifications to the interface in carbon fiber reinforced thermoplastic matrices, J. Adhes. 45 (1994) 73–88. [30] N.G. Karsli, A. Aytac, V. Deniz, Effects of initial fiber length and fiber length distribution on the properties of carbon-fiber-reinforced-polypropylene composites, J. Reinf. Plast. Compos. 31 (2012) 1053–1060. [31] S. Dong, R. Gauvin, Application of dynamic mechanical analysis for the study of the interfacial region in carbon fiber/epoxy composite materials, Polym. Compos. 14 (1993) 414–420. [32] S. Mohanty, S.K. Verma, S.K. Nayak, Dynamic mechanical and thermal properties of MAPE treated jute/HDPE composites, Compos. Sci. Technol. 66 (2006) 538–547. [33] C. Guo, L. Li, Q. Wang, Investigation on the compatibilizing effect of m-isopropenyla, a-dimethylbenzyl isocyanate grafted polypropylene on polypropylene and wood flour composites, Wood Sci. Technol. 46 (2012) 257–270. [34] M. Ashida, T. Noguchi, Effect of matrix's type on the dynamic properties for short fiber-elastomer composite, J. Appl. Polym. Sci. 30 (1985) 1011–1021. [35] M. Ashida, T. Noguchi, Dynamic moduli for short fiber-CR composites, J. Appl. Polym. Sci. 29 (1984) 661–670. [36] V.S. Mironov, M. Park, C. Choe, J. Kim, S. Lim, H. Ko, Influence of carbon fiber surface treatments on the structure and properties of conductive carbon fiber/polyethylene films, J. Appl. Polym. Sci. 84 (2002) 2040–2048.
Contents lists available at ScienceDirect
Materials and Design journal homepage: www.elsevier.com/locate/jmad
Characterization of poly(butylene terephthalate) composites prepared by using various types of sized carbon fibers N. Gamze Karsli a, Cem Ozkan b, Ayse Aytac a,b,⁎, Veli Deniz a,b a b
Department of Chemical Engineering, Kocaeli University, Engineering Faculty, 41380 Kocaeli, Turkey Department of Polymer Science and Technology, Kocaeli University, 41380 Kocaeli, Turkey
a r t i c l e
i n f o
Article history: Received 3 February 2015 Received in revised form 8 August 2015 Accepted 10 August 2015 Available online 12 August 2015 Keywords: Carbon fiber Poly(butylene terephthalate) Fiber/matrix adhesion Mechanical testing
a b s t r a c t This paper aims to study the effects of sizing on properties of differently sized carbon fiber (CF) reinforced poly(butylene terephthalate) (PBT) composites by comparing them to unsized CF reinforced composites. Contact angle analysis was used to evaluate the wettability of CFs and the work of adhesion between the sizing agent and PBT matrix. It was found that wettability of PU sizing material by PBT matrix was better than that of other sizing materials by PBT matrix. Tensile and dynamic mechanical analysis (DMA) tests were performed to investigate the effect of sizing agent type on mechanical and thermomechanical properties. According to tensile test results PHE and PU sized CF reinforced PBT composites gave higher tensile strength and modulus values than the others. DMA revealed that PU sizing material gave better adhesion strength than other sizing materials. It was found that electrical conductivity values of all composites are about 10−2 S/cm. SEM analysis showed that PU sized CF surface covered a layer of PBT matrix in accordance with other test results. As a conclusion of all results, it can be suggested that PU is a proper sizing material to be used for CF surface for PBT matrix. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Thermoplastic polymers exhibit good properties such as high impact strength, and fracture resistance. Because of this reason, they are widely used in polymer engineering industry [1]. Poly(butylene terephthalate) (PBT) is one of these thermoplastic polymers and it has been mostly used in automotive, electrical, and electronic industries for many years due to its exceptional electrical insulation properties [1–3]. PBT is considered as the main competitor for polyamides, but it has the advantage over polyamides due to its much lower moisture uptake and better dimensional stability properties under wet conditions [4]. Moreover, PBT has been widely used in fiber-reinforced composites because of its good mechanical properties and properties of easy molding and fast crystallization [5]. When compared to glass fiber reinforced PBT composites; carbon-fiber-reinforced PBT composites exhibit higher tensile strength and modulus due to exceptional mechanical effects of carbon fibers [5]. On the other hand, mechanical properties of fiber reinforced composites not only depend on the intrinsic characteristics of the matrix and fiber, but also on the fiber/matrix interface adhesion. Since the stress transfer phenomena from the matrix to the fiber during deformation is affecting mechanical resistance of composites, better interfacial adhesion militates in favor of better stress transfer and
⁎ Corresponding author at: Kocaeli University, Chemical Engineering Department, Umuttepe Campus, 41380 Izmit, Kocaeli, Turkey. E-mail address: [email protected] (A. Aytac).
http://dx.doi.org/10.1016/j.matdes.2015.08.047 0264-1275/© 2015 Elsevier Ltd. All rights reserved.
mechanical properties. Therefore, it is crucial to control the interfacial adhesion to get better mechanical properties [6,7]. The interfacial adhesion becomes more important when using thermoplastic matrix because of the poor chemical functionality of this matrix type [8]. Besides, carbon fibers contain very few reactive groups due to their carbonization process which occurs at elevated temperatures, since these functional groups on the fiber surface are eliminated at elevated temperatures [9]. The lack of reactive groups on the fiber surface and thermoplastic matrix leads to weak adhesion between carbon fiber and thermoplastic matrix. In the case of weak adhesion, the matrix cannot transfer the stress to fiber efficiently and fiber cannot perform its duty of carrying the applied load [8]. Two methods are generally used for obtaining better adhesion and stress transfer. These methods are matrix modification and fiber surface modification. One of the fiber surface modification methods is sizing and in this method, carbon fibers are usually coated with a sizing material which consists of a proper polymer or resin [7,10,11]. The sizing material includes functional groups and these groups change the surface free energy and wettability of carbon fiber surfaces [12]. Also they can react to or interact with the polymer matrix. This reaction or interaction enhances the fiber–matrix adhesion and correspondingly the ultimate properties of composites [13]. In the literature, a few studies about carbon fiber-reinforced PBT composites have been undertaken. One of these studies was reported by Ng et al. [14]. They added boron nitride (BN) and carbon fiber (CF) into PBT matrix and investigated the hybrid filler effect on the properties of PBT composites. They found that, while the combination of BN and CF in PBT significantly reduced electrical conductivity of the
N. Gamze Karsli et al. / Materials and Design 87 (2015) 318–323
composites, the usage of mixed fillers did not lead to an improvement in thermal conductivity with respect to PBT/BN composites. They also observed that hybrid composites exhibited better tensile properties and processability than PBT/BN composites at the same total filler content. Another study was performed by Wiedmer et al. [15]. They studied the effect of electron beam radiation on carbon fiber reinforced PBT, PPS and PA composites. They prepared composites with and without the presence of crosslinking agent to clarify whether crosslinking could occur. They examined the thermal and mechanical properties of the composites before and after exposure to irradiation. Consequently, they observed that properties of PBT, PPS and PA46 composites did not change significantly in the presence of crosslinking agent after irradiation. On the other hand, CF/PA66 composites exhibited changes in some of their properties in the presence of a crosslinking agent after irradiation. Chen et al. [16] investigated the properties of recycled carbon fiber (RCF) reinforced PBT matrix composites. They also treated RCF surfaces with the solution of diglycidyl ether of bisphenol A in order to improve the interfacial adhesion between the RCF and PBT matrix. Their results showed that, surface treated RCF significantly improved the mechanical properties, heat distortion temperature, and thermal stability of the composites. The morphology studies of fracture surfaces also indicated that the RCF homogeneously dispersed in the PBT matrix. Although there are a few studies in the literature which were focused on the effects of sizing agent type, molecular weight and concentration on the properties of some of polymeric matrix based composites [17–19], there has been no study about the effects of sizing agent type on the properties of carbon fiber reinforced PBT composites to the best of authors' knowledge. This study focuses on the effects of sizing material types on the properties of CF reinforced PBT matrix composites. For this work, CFs which were unsized and which were sized with five different types of sizing agent were used as reinforcement material. The properties of composites were analyzed by mechanical, thermomechanical, electrical, morphological tests and contact angle measurements. 2. Materials and methods Matrix material PBT (Tecodur®) was supplied by Interplast (Turkey). Unsized CF and CFs sized with polyurethane (PU), polyamide (PA), polyimide (PI), phenoxy (PHE) and epoxy/phenoxy (EPO_PHE) were supplied by Akkök Group (Turkey) and used as reinforcement. PBT was dried in a vacuum oven at 120 °C for 8 h before the compounding process. PBT granules and 30 wt.% of 6 mm chopped carbon fibers were compounded in a laboratory scale co-rotating twin-screw mini extruder at 255 °C, 100 rpm. All the compounds were subsequently injection molded using a laboratory scale injection molding machine with a barrel temperature of 255 °C, mold temperature of 130 °C and injection pressure of 10 bars. Contact angles of test liquids against the sizing materials and PBT matrix material were measured by the sessile drop method with Attention Theta Lite contact angle tensiometer. The used test liquids were ethylene glycol, diiodomethane and deionized water. Surface tension of test liquids was given in Table 1 [20]. Standard test specimen of neat PBT was produced for contact angle analysis using a twin-screw mini extruder and laboratory scale injection molding machine. In addition, thin films of each sizing material were cast onto clean microscope
Table 1 Surface tension values of test liquids. Surface tension
Diiodomethane
Ethylene glycol
Deionized water
γTOT L γLW L γAB L γ− L γ+ L
50.8 50.8 – – –
48 29 19 47 1.92
72.8 21.8 51.0 25.5 25.5
319
slides for contact angle analysis of sizing materials. Ten measurements were conducted to obtain average contact angle values. Tensile tests were performed using Instron 4411 universal testing machine. The dimension of the test samples was 4 mm width, 2 mm thickness and 30 mm length. Average tensile strength and modulus values of composites were determined using 5 dumbbell-shaped samples for each composition at a constant crosshead speed of 5 mm/min. DMA was performed in tension mode by using Metravib DMA50 analyzer. Composites were tested under the condition of 1.5 × 10−5 m dynamic displacement and 1 Hz frequency. The scanning was carried out from 25 °C to 200 °C at 1 °C/min heating rate. Specimen dimensions were 10 mm height, 2 mm thickness and 4 mm width. Electrical resistivity values of composites were measured with 2-point-probe test by using Haoyue M890G Digital Multi Meter. For obtaining good electrical contact in this technique, copper wires were attached to both ends of test sample with silver paste. After the hardening of silver paste, resistivity measurements were performed by contacting probes with these copper wires. After that, electrical conductivity values of composites were calculated using obtained resistance values of composites as in the following formula [21,22]: Sample Thickness ðcmÞ : Electrode Area ðcm2 Þ Resistance ðΩÞ
ð1Þ
Morphologies of the tensile fractured surfaces of composites were examined using a scanning electron microscope (JEOL JSM-6335F). Before the examinations, tensile fracture surfaces of composites were sputter coated with gold and palladium. 3. Results and discussion 3.1. Surface energy analysis Sizing process which was applied to carbon fiber surface changes the surface energy and wettability of the surface and affects the fiber matrix adhesion performance. One of the methods to be used for analyzing adhesion performance of fiber surfaces is the contact angle measurement. Contact angle concept which was first developed and formulized by Young, became the main concept for the development of later approaches. One of these approaches was developed by Van Oss, Good and Chaudhury, which aimed at estimating the surface energy values of solids [23,24]. According to their approach, a solid surface consists of two terms: one of them is the Lifshitz–van der Waals interactions, γLW, which includes dispersion, dipolar and induction interactions, and the other one is the acid–base interaction term, γAB, which includes + all the electron donor (γ− S ) and electron acceptor (γS ) interactions. By combining this approach with the Young–Dupré equation, the general contact angle equation is obtained: qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi LW γþ γLW γS γþ γL ð1 þ cos θÞ ¼ 2 L : S γL þ S γL þ
ð2Þ
The value of γLW is determined from the contact angle of an apolar S liquid on the solid in which case Eq. (2) reduces to: qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi LW γLW γL ð1 þ cos θÞ ¼ 2 S γL
ð3Þ
when γLW S is known and the contact angles are obtained using different liquids on the solid, one can get two equations similar to Eq. (2), and − these equations can be solved simultaneously for γ+ S and γS . In addition to this, by writing Eq. (2) as below, work of adhesion between solid and liquid can be calculated: qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi qffiffiffiffiffiffiffiffiffiffiffiffi LW γþ γLW γS γþ Wa ¼ γL ð1 þ cos θÞ ¼ 2 L : S γL þ S γL þ
ð4Þ
320
N. Gamze Karsli et al. / Materials and Design 87 (2015) 318–323
Table 2 Contact angle values. Sizing & matrix
PU EPO_PHE PI PHE PA Unsized CF PBT matrix
Table 4 Work of adhesion (Wa) values of composites. Contact angle (θ°) Diiodomethane
Ethylene glycol
Deionized water
19 34 32 31 54 40 57
51 55 23 37 46 32 43
37 56 53 52 60 55 81
In this study, in order to calculate the work of adhesion between carbon fiber and PBT matrix, surface tension values of sizing materials and matrix material were measured using three test liquids. Contact angle values and surface energies (polar and dispersive) of unsized CF, sizing materials and PBT matrix are given in Tables 2 and 3. In addition, surface tension value of unsized CF was obtained from literature [25]. In order to obtain good wetting of fiber by matrix, surface energy of matrix should be lower than that of fiber [26]. It can be seen from Table 3 that while PBT matrix exhibits the lowest surface energy value, the maximum surface energy is obtained for PU sizing material. According to these results, it can be concluded that, wetting between PU sized CF and PBT is better than that of other sizing materials and PBT matrix. While wetting of fiber by polymer is the prior condition of good fiber matrix adhesion, functional groups on fiber surface also plays an important role for good adhesion. This idea was first developed by Fowkes [27]. According to Fowkes, hydrogen bonding interactions are important for fiber–matrix adhesion and these interactions can be evaluated by investigating the polar components of surfaces. Therefore, when choosing the proper sizing material for used matrix material, the polar component amount of sizing material should also be taken into account. It can be seen from Table 3 that in our study PU sizing material exhibited the highest amount of polar component among all samples. Fiber–matrix interface strength can also be calculated as “work of adhesion” (Wa) value by using Eq. (4) [28]. The Wa value can be separated into two components. One of them is Wa12 and it represents the work of adhesion value between unsized CF and sizing material. The other one is Wa23 and it represents the work of adhesion value between sizing material and matrix material. In the case when Wa12 N Wa23, it can be said that the required energy to separate unsized CF and sizing material is higher than the required energy to separate sizing and matrix materials. In such a case, the sizing contributes to the fiber–matrix adhesion while it protects the fiber from breakage during composite preparation process. On the other hand, when Wa23 N Wa12, it can be said that the required energy for the separation of the sizing material and matrix material is higher than the required energy to separate unsized CF and sizing material. In such a case, sizing cannot protect the fiber from breakage satisfyingly. In our study, the calculated Wa12 and Wa23 values were given in Table 4.
Sizing type
Work of adhesion between unsized CF and sizing material Wa12 (mJ·m−2)
Work of adhesion between sizing material and PBT matrix Wa23 (mJ·m−2)
PU EPO_PHE PI PHE PA
108.6 95.2 92.8 104.8 80.4
102.4 90.2 87.4 88.8 77.5
As can be seen from Table 4, Wa12 values are higher than that of for all cases. It can be concluded that the adhesion strength between all sizing materials and unsized CF is higher than the adhesion strength between sizing material and PBT matrix. Hence, it can also be said that all sizing materials used in this study protect the fiber from breakage and contribute to the fiber matrix adhesion during processing. Table 4 also shows that the maximum Wa values (W a12 and Wa23) were obtained for PU sizing material. This result can be attributed to the highest surface energy of this sizing material. Thus, PU, owing to its high surface energy, exhibits better interfacial adhesion with both unsized CF and PBT matrix.
Wa23
3.2. Tensile tests The tensile strength value of neat PBT is 63.9 MPa and the addition of unsized CF increases this value to approximately 125 MPa. Also, tensile test results of sized CF reinforced PBT composites were evaluated in Fig. 1. Fig. 1 shows that the maximum tensile strength values were obtained for PHE and PU sized CF reinforced PBT composites, respectively. According to tensile strength values of composites, it can be said that while the maximum enhancement in tensile strength was observed as 111% with addition of PHE sized CF to PBT matrix, this enhancement was observed as 107% with addition of PU sized CF to the PBT matrix. As is known, when the adhesion between fiber and matrix is good, applied load can efficiently be transferred from the matrix to the fiber during tensile test, and also fiber can bear more load in such a case. Consequently, it can be concluded that adhesion between PHE and PU sized CF and PBT matrix is better than the others. It was supposed that proton donor hydroxyl groups of PHE and proton acceptor carbonyl groups of PBT may react via hydrogen bonding and this reaction could improve the adhesion between PHE and PBT [29]. There might be also a chemical interaction between isocyanate groups of PU and carboxyl groups of PBT. Thus, it is possible to obtain better adhesion between PU sized CF and PBT by means of these chemical interactions [9]. Consequently, it can be considered that improved tensile strength values arose from these chemical interactions.
Table 3 Surface energy values of samples. Sizing & matrix
Polar component γsAB (mJ·m−2)
Dispersive component γsLW (mJ·m−2)
Total surface energy γsTot (mJ·m−2)
PU EPO_PHE PI PHE PA Unsized CF⁎ PBT matrix
21.8 9.5 5.5 6.8 4.4 4.8 4.9
48 42.2 43.2 43.6 32 39.6 30
69.8 51.7 48.7 50.4 36.4 44.4 34.9
⁎ The surface tension value of unsized CF was obtained from literature [25].
Fig. 1. Tensile strength values of composites.
N. Gamze Karsli et al. / Materials and Design 87 (2015) 318–323
321
Fig. 2. Modulus values of composites.
Fig. 3. Tan delta values of composites.
Effect of sizing agent type on the modulus values of composites is given in Fig. 2. Modulus value of neat PBT is 2512 MPa. It can be seen from Fig. 2 that modulus values of all composites are higher than this value. This is an expected result and can be explained by using the rule of mixture theory in the literature [10,11,30]. Additionally, PU sized CF reinforced composites exhibit the maximum modulus values. It has been explained in our previous study that the ultimate fiber length in the composite affected the modulus value of a composite [30]. When ultimate fiber length is higher, the contribution of the fibers to the modulus of the composites is more prominent. It is known that in the case of proper sized CF usage; less fiber breakage occurs during composite preparation process. Consequently, higher modulus values can be attributed to the better adhesion between fiber and matrix.
semiconductive materials varies between 10− 8–102 S/cm. The minimum fiber loading level at which a non-conductive polymer becomes a semi conductive polymer (10−8 S/cm) is called “percolation threshold.” After this point, a continuous carbon fiber network occurs in the polymer matrix and electrons transfer easily from a fiber to another by means of this network. It can be seen from Fig. 4 that electrical conductivity values of all composites are about 10−2 S/cm. As mentioned above, conductivity value of 10−2 S/cm is considered as a very high conductivity value comparatively for commodity composites, therefore these composites are considered as semi-conductive. Moreover, it can be said that used CF loading level in this study (30 wt.%) is higher than percolation threshold [22]. The maximum electrical conductivity value was measured for PU sized CF reinforced composites and electrical conductivity value of unsized CF reinforced composites was the second. When non-conductive polymeric material coated conductive particles are used in composites, this polymeric layer acts as a non-conductive gap which hinders electron transfer between conductive particles [36]. However, our test results showed that electrical conductivity of PU sized CF reinforced composites are higher than that of unsized CF reinforced composites. In this case, dispersion of fiber in polymer matrix comes into play as an important factor. In a study, Choi et al. [22] found that, electrical conductivity values of CF reinforced composites mainly depended on the dispersion of carbon fiber in polymer matrix. According to their results, when adhesion between fiber and matrix is weak, there should be a phase separation while agglomeration occurs in the composite to affect electron transfer negatively. It can be inferred from this results that, due to the better adhesion tendency between PU sized CF and PBT matrix,
3.3. Dynamic mechanical analysis (DMA) There are lots of techniques for evaluating the interfacial adhesion in polymer matrix composites. DMA has some advantages over the others, because it provides a sensitive and non-destructive detection. DMA measures two types of responses of polymeric materials against to low strain periodic deformation. One of these responses is the elastic response and it represents the stiffness of the composite. The other one is the damping response and it represents the energy dissipation in the composite [31]. This damping factor gives clues about all kinds of molecular motions in the composite including motions which take place at the interface. It also gives clues about potential of the material in dissipating mechanical energy [32,33]. Ashida and Noguchi [34] reported that there was a relation between damping factor (tan delta, tanδ) peak magnitude and interfacial adhesion in short fiber reinforced composites. According to this theory, the lower the tan delta peak magnitude, the better the obtained interfacial adhesion. Because, when the adhesion between fiber and matrix is poor, more energy dissipation takes place due to the non-efficient stress transfer between fiber and matrix. And this results in higher peak magnitude [35]. Fig. 3 shows the effect of sizing agent type on the damping factor values of composites as a function of temperature. It can be seen from Fig. 3 that the lowest tan delta peak magnitude was obtained for PU sized CF reinforced composites. As it is shown in Fig. 3, the lowest peak magnitude obtained demonstrates the fact that PU sized CF and PBT matrix give the best adhesion strength among all composites. 3.4. Electrical conductivity analysis The effect of sizing agent type on the electrical conductivity of composites is given in Fig. 4. If a conductive material such as carbon fiber is added into a non-conductive polymer, electrical conductivity property of this polymer will change. Electrical conductivity value of
Fig. 4. Electrical conductivity values of composites.
322
N. Gamze Karsli et al. / Materials and Design 87 (2015) 318–323
Fig. 5. SEM micrographs of composites (a) unsized, (b) PI, (c) PU.
dispersion of PU sized fibers is better than that of unsized fibers in the PBT matrix and this results in better electrical conductivity values.
other test results. As a conclusion of all results, the authors propose that PU is an appropriate sizing material to be used for CF surface for PBT matrix.
3.5. SEM analysis Acknowledgment Tensile fracture surface morphology analysis is widely used to characterize the adhesion between fiber and matrix in polymer composites. Fig. 5 illustrates the micrographs of tensile fracture surfaces of unsized, PI and PU sized CF reinforced PBT matrix composites respectively. It can be seen from Fig. 5(a, b) that there are some voids between the fibers and matrix while fiber surfaces are clean and smooth. This indicates that there is a weak adhesion between unsized and PI sized CFs and PBT matrix. On the other hand, Fig. 5(c) illustrates that PU sized CF surfaces are covered by a layer of PBT matrix as expected. This can be considered as another evidence for the chemical interaction and stronger interfacial adhesion between PU sized CF and PBT as mentioned above. 4. Conclusions In this study, effects of different sizing agent types on the mechanical, thermomechanical, electrical and morphological properties of CF reinforced PBT matrix composites were investigated. Contact angle analysis was also performed to evaluate the wettability performance of carbon fibers by PBT matrix. Contact angle analysis test results showed that the best wetting by PBT matrix was observed in the case of PU sizing material. Mechanical test results revealed that PU and PHE sized CF reinforced composites gave better tensile properties. In order to investigate the thermo-mechanical properties of composites, DMA was performed with the comparison of tan delta peak magnitude values of composites and evaluation of fiber matrix adhesion. It was found that PU sized CF reinforced composites exhibited the best adhesion strength among all composites. According to the results of electrical conductivity measurements, it was found that PU sized CF reinforced composites gave the highest conductivity value and it was attributed to the better adhesion between PU sized CF and PBT matrix. The results of morphological analysis of all composites studied are accordant with
This study was supported by the Ministry of Science, Industry and Technology of Turkey under project number 01020.STZ.21011-2. References [1] C.C. Hsu, J.L. Doong, C.P. Fung, Fracture toughness of injection-molded poly(butylene terephthalate) composites, Polym. Plast. Technol. 44 (2005) 335–350. [2] J.Y. Kim, The effect of carbon nanotube on the physical properties of poly(butylene terephthalate) nanocomposite by simple melt blending, J. Appl. Polym. Sci. 112 (2009) 2589–2600. [3] S.H. Jang, Y.H. Kim, S. Lim, G.D. Choi, S.H. Kim, W.N. Kim, Effects of fiber characteristics on the mechanical and rheological properties of poly(butylene terephthalate)/ glass fiber composites, J. Appl. Polym. Sci. 116 (2010) 3005–3012. [4] S. Thomas, P.M. Visakh, Handbook of Engineering and Specialty Thermoplastics, United States of America, 2010. [5] J. Scheirs, T.E. Long, Modern Polyesters: Chemistry and Technology of Polyesters and Copolyesters, Wiley, Chichester, 2003 (England). [6] H. Zhou, K. Han, M. Yu, Preparation of long glass fiber reinforced poly(butylene terephthalate) composites with chemical bonding interphase, J. Macromol. Sci. A 43 (2006) 1835–1851. [7] J. Yang, J. Xiao, J. Zeng, L. Bian, C. Peng, F. Yang, Matrix modification with silane coupling agent for carbon fiber reinforced epoxy composites, Fibers Polym. 14 (2013) 759–766. [8] L. Dányádi, J. Gulyás, B. Pukánszky, Coupling of carbon fibers to polycarbonate: surface chemistry and adhesion, Compos. Interfaces 10 (2003) 61–76. [9] X. Qian, L. Chen, J. Huang, W. Wang, J. Guan, Effect of carbon fiber surface chemistry on the interfacial properties of carbon fibers/epoxy resin composites, J. Reinf. Plast. Compos. 32 (2013) 393–402. [10] M.R. Piggott, The interface in carbon fibre composites, Carbon 27 (1989) 657–662. [11] J.O. Iroh, W. Yuan, Surface properties of carbon fibres modified by electrodeposition of polyamic acid, Polymer 37 (1996) 4197–4203. [12] Z. Dai, B. Zhang, F. Shi, M. Li, Z. Zhang, Y. Gu, Chemical interaction between carbon fibers and surface sizing, J. Appl. Polym. Sci. 124 (2012) 2127–2132. [13] W. Chen, Y. Yu, P. Li, C. Wang, T. Zhou, X. Yang, Effect of new epoxy matrix for T800 carbon fiber/epoxy filament wound composites, Compos. Sci. Technol. 67 (2007) 2261–2270. [14] H.Y. Ng, X. Lu, S.K. Lau, Thermal conductivity, electrical resistivity, mechanical, and rheological properties of thermoplastic composites filled with boron nitride and carbon fiber, Polym. Compos. 26 (2005) 66–73.
N. Gamze Karsli et al. / Materials and Design 87 (2015) 318–323 [15] S. Wiedmer, D. Gellner, K. Friedrich, Effect of electron-beam radiation on thermoplastic composites, Plast. Rubber Compos. 34 (2005) 76–84. [16] Y. Chen, X. Wang, D. Wu, Recycled carbon fiber reinforced poly(butylene terephthalate) thermoplastic composites: fabrication, crystallization behaviors and performance evaluation, Polym. Adv. Technol. 24 (2013) 364–375. [17] R.L. Zhang, Y.D. Huang, D. Su, L. Liu, Y.R. Tang, Influence of sizing molecular weight on the properties of carbon fibers and its composites, Mater. Des. 34 (2012) 649–654. [18] N.G. Karsli, C. Ozkan, A. Aytac, V. Deniz, Effects of sizing materials on the properties of carbon fiber-reinforced polyamide 6,6 composites, Polym. Compos. 34 (2013) 1583–1590. [19] R.L. Zhang, Y.D. Huang, N. Li, L. Liu, D. Su, Effect of the concentration of the sizing agent on the carbon fibers surface and interface properties of its composites, J. Appl. Polym. Sci. 125 (2012) 425–432. [20] E.C. Botelho, L. Figiel, M.C. Rezende, B. Lauke, Mechanical behavior of carbon fiber reinforced polyamide composites, Compos. Sci. Technol. 63 (2003) 1843–1855. [21] O. Koysuren, S. Yesil, G. Bayram, Effect of surface treatment on electrical conductivity of carbon black filled conductive polymer composites, J. Appl. Polym. Sci. 104 (2007) 3427–3433. [22] M.H. Choi, B.H. Jeon, I.J. Chung, The effect of coupling agent on electrical and mechanical properties of carbon fiber/phenolic resin composites, Polymer 41 (2000) 3243–3252. [23] H.Y. Erbil, Surface Chemistry of Solid and Liquid Interfaces, UK, 2006. [24] M.D. Doganci, C.E. Cansoy, I.O. Ucar, H.Y. Erbil, E. Mielczarski, J.A. Mielczarski, Combined XPS and contact angle studies of flat and rough ethylene-vinyl acetate copolymer films, J. Appl. Polym. Sci. 124 (2012) 2100–2109. [25] Y. Luo, Y. Zhao, Y. Duan, S. Du, Surface and wettability property analysis of CF300 carbon fibers with different sizing or without sizing, Mater. Des. 32 (2011) 941–946.
323
[26] E. Mäder, Study of fibre surface treatments for control of interphase properties in composites, Compos. Sci. Technol. 57 (1997) 1077–1088. [27] F.M. Fowkes, Role of acid–base interfacial bonding in adhesion, J. Adhes. Sci. Technol. 1 (1987) 7–27. [28] K. Tsutsumi, S. Ishidal, K. Shibata, Determination of the surface free energy of modified carbon fibers and its relation to the work of adhesion, Colloid Polym. Sci. 268 (1990) 31–37. [29] P. Bourgeois, T. Davidson, Plasma modifications to the interface in carbon fiber reinforced thermoplastic matrices, J. Adhes. 45 (1994) 73–88. [30] N.G. Karsli, A. Aytac, V. Deniz, Effects of initial fiber length and fiber length distribution on the properties of carbon-fiber-reinforced-polypropylene composites, J. Reinf. Plast. Compos. 31 (2012) 1053–1060. [31] S. Dong, R. Gauvin, Application of dynamic mechanical analysis for the study of the interfacial region in carbon fiber/epoxy composite materials, Polym. Compos. 14 (1993) 414–420. [32] S. Mohanty, S.K. Verma, S.K. Nayak, Dynamic mechanical and thermal properties of MAPE treated jute/HDPE composites, Compos. Sci. Technol. 66 (2006) 538–547. [33] C. Guo, L. Li, Q. Wang, Investigation on the compatibilizing effect of m-isopropenyla, a-dimethylbenzyl isocyanate grafted polypropylene on polypropylene and wood flour composites, Wood Sci. Technol. 46 (2012) 257–270. [34] M. Ashida, T. Noguchi, Effect of matrix's type on the dynamic properties for short fiber-elastomer composite, J. Appl. Polym. Sci. 30 (1985) 1011–1021. [35] M. Ashida, T. Noguchi, Dynamic moduli for short fiber-CR composites, J. Appl. Polym. Sci. 29 (1984) 661–670. [36] V.S. Mironov, M. Park, C. Choe, J. Kim, S. Lim, H. Ko, Influence of carbon fiber surface treatments on the structure and properties of conductive carbon fiber/polyethylene films, J. Appl. Polym. Sci. 84 (2002) 2040–2048.