- Email: [email protected]
Fabrication of 2014 aluminum matrix composites reinforced with untreated and carboxyl-functionalized carbon nanotubes
Accepted Manuscript Fabrication of 2014 aluminum matrix composites reinforced with untreated and carboxyl-functionalized carbon nanotubes Xian Zhu, Yu-Guang Zhao, Min Wu, Hui-Yuan Wang, Qi-Chuan Jiang PII:
S0925-8388(16)30593-X
DOI:
10.1016/j.jallcom.2016.03.036
Reference:
JALCOM 36924
To appear in:
Journal of Alloys and Compounds
Received Date: 21 November 2015 Revised Date:
17 February 2016
Accepted Date: 7 March 2016
Please cite this article as: X. Zhu, Y.-G. Zhao, M. Wu, H.-Y. Wang, Q.-C. Jiang, Fabrication of 2014 aluminum matrix composites reinforced with untreated and carboxyl-functionalized carbon nanotubes, Journal of Alloys and Compounds (2016), doi: 10.1016/j.jallcom.2016.03.036. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Fabrication of 2014 aluminum matrix composites reinforced with untreated and carboxyl-functionalized carbon nanotubes Xian Zhu, Yu-Guang Zhao, Min Wu, Hui-Yuan Wang∗, Qi-Chuan Jiang
RI PT
Key Laboratory of Automobile Materials of Ministry of Education & School of Materials Science and Engineering, Nanling Campus, Jilin University, No. 5988 Renmin Street, Changchun 130025, PR China Abstract
SC
2014 aluminum matrix composites reinforced with untreated (P-CNTs) and carboxyl-functionalized carbon nanotubes (C-CNTs) were successfully fabricated by
M AN U
the combination of sintering and hot extrusion. It is revealed that the C-CNTs are easier to disperse compared to the P-CNTs. Interfacial analysis indicated that a thin layer of aluminum carbide (Al4C3) exists at the interface between the matrix and the
TE D
P-CNTs. However, no obvious transition layers of Al4C3 were observed at the interface of C-CNTs and 2014Al matrix in composites. Also, no preferred interfacial crystallographic orientation relationship between the C-CNTs and aluminum matrix
EP
was observed. Furthermore, C-CNTs have a larger effective interfacial contact with
AC C
aluminum matrix compared to the untreated CNTs, due to large amount of –COOH group and defects on the surface of the C-CNTs interacting with Al by strong chemical and physical interactions. These aspects in turn affect the mechanical properties of the composites. The ultimate tensile strength of the composites were raised from 530 to 600 and 630 MPa with 0.5 wt.% of P-CNTs and C-CNTs,
∗Corresponding author. Tel/fax: +86 431 8509 4699 E–mail address: [email protected] (H.-Y. Wang)
1
ACCEPTED MANUSCRIPT respectively. It reveals that the C-CNTs are more effective in strengthening the aluminum matrix than the P-CNTs. Keywords: Metal matrix composites; Carbon nanotubes; powder metallurgy;
RI PT
Dispersion; Interface; Strength
1 Introduction
SC
Metal matrix composites (MMCs) have been receiving great attention due to their
M AN U
high tensile strength, hardness and modulus, as well as their high wear resistance compared to the matrix [1-4]. Carbon nanotubes (CNTs) are considered to be ideal reinforcements of composites due to their exceptional strength, modulus and heat conductivity [5-7]. However, because CNTs have tremendous specific surface area
TE D
and very strong van der Waal’s force attraction between them which lead to the formation of CNTs clusters. How to homogeneously disperse the CNTs in a metal matrix and achieve excellent interfacial bonding between the CNTs and metal matrix
EP
are still unsolved problems [2, 8-11]. Therefore, only limited studies have been
AC C
reported on MMCs. The metal matrix of the reported CNTs-MMCs which mainly focused on the dispersion and alignment in matrix of CNTs, were essentially limited to pure aluminum [3, 4, 9, 11-13]. However, pure aluminum matrix is usually not eligible for applications due to its low mechanical strength. Al-Cu based 2014Al alloy is heat treatable aluminum alloy, which offers high strength at low specific weight and are extensively employed as structural components, particularly in the aircraft industry [14, 15]. CNTs reinforced 2014Al 2
ACCEPTED MANUSCRIPT composites are thus be considered to have more application potential. However, homogeneously disperse the CNTs in the matrix and achieve excellent interfacial bonding between the CNTs and the matrix are still needed to be addressed.
RI PT
The functionalized treatment is a common way to obtain a homogeneous dispersion of CNTs by introducing carboxyl or hydroxyl groups on the surface of the CNTs. Calculations show that the addition of carboxyl or hydroxyl groups adds
SC
around 0.2 nm to the graphitic layer roughness [16]. Higher roughness engenders
M AN U
lower van der waal’s force by reducing the proximity of interacting surfaces and by increasing the distance between their atoms [17, 18]. In the case of interfacial bonding, the amorphous carbon generated during the chemical vapor deposition [19] could affect the interfacial bonding status between CNTs and Al matrix to some extent,
TE D
because the amorphous carbon tends to react with aluminum to form aluminum carbide [20]. Meanwhile, the carboxyl treatment could remove carbon impurities and introduce defects on the surface of the CNTs [21]. However, the influence of the
AC C
studied.
EP
surface characteristic on the strengthening of CNTs in MMCs has not been well
In view of these facts, in the present work, the 2014Al matrix composites reinforced with untreated carbon nanotubes (P-CNTs) and carboxyl-functionalized carbon nanotubes (C-CNTs) by the combination of sintering and hot extrusion were fabricated to clarify the influence of surface characteristic on the dispersion of the CNTs and the interfacial bonding status between the CNTs and the 2014Al matrix. 2 Experimental details 3
ACCEPTED MANUSCRIPT CNTs (with an average diameter of ~25 nm and length of ~1.25 µm) synthesized by a chemical vapor deposition (CVD) method were selected as the reinforcements (Fig. 1a). Carboxyl functional treated CNTs (C-CNTs) (with an average diameter and
RI PT
length of about 25 nm, 1.25 µm) were provided by Chengdu Organic Chemistry Co., Ltd., China. 2014 aluminum alloy powders with an average powder size of about 10µm were employed as the matrix (Fig. 1b).
SC
The experiment process was illustrated in Fig. 2. Firstly, the two kinds of CNTs
M AN U
(0.25 g) were dispersed in 250 ml ethanol, sonicated for 60 minutes, and then poured into 200ml ethanol solution that contained 49.75 g 2014Al alloy powders, respectively. Secondly, the composite mixtures were then vigorously stirred by magnetic stirring for 30 minutes followed by drying at 333 K for 24 h in an oven. Thirdly, ball milling
TE D
were performed in a planetary ball mill at a constant speed of 300 rpm for 6 hours under argon atmosphere with ball to power ratio of 8:1 with 0.5 g stearic acid addition into the dried mixture as process control agent. Finally, the ball-milled powders were
EP
packed and cold compacted at a pressure of 50 MPa for 2 minutes, and subsequently
AC C
the compacts were hot vacuum sintered for 50 minutes under 793 K and then pressed at a pressure of 50 MPa for 10 minutes in a self-made vessel, with furnace cooling to room temperature. Hot extrusion was conducted at 693 K with an extrusion ratio of about 15:1. The extrudates were then solution treated at 775 K in air for 2 hours and naturally aged for more than 96 hours before test. The two kinds of CNTs were characterized by X-ray diffraction (XRD) (D/Max 2500PC, Rigaku, Japan) using Cu Kα radiation in the range of 20°~80° with a 4
ACCEPTED MANUSCRIPT scanning speed of 4 deg/min and an acquisition step of 0.02° (2θ). Raman spectroscopy was obtained by REINSHAW in Via Raman spectroscope with an excitation laser wavelength of 633 nm in the range of 1000~2000 cm-1. X-ray
RI PT
photoelectron spectroscopy (XPS) was conducted on a VG ESCALAB MK Ⅱsystem. Different CNTs samples for Transmission electron microscopy (TEM) observation were prepared by dispersing the corresponding CNTs in ethanol, sonicating for
SC
several minutes, and subsequently dripping one drop of the liquid on a holey carbon
M AN U
TEM support grid. After dried overnight, the samples are observed by TEM (TEM; JEM-2100F, Japan and Tecnai G2 20 American) at an acceleration voltage of 200 KV. Samples for interface observation were prepared by ion beam thinner (Leica, EMRES101), then studied by TEM. The pre-mixed composite powders as well as
TE D
fractured surface were characterized by a Field emission scanning electron microscopy (FESEM) (JSM-6700F, Japan). Tensile specimens with a gauge length of 10 mm, a width of 4.0 mm, and a thickness of about 1 mm were wire-cut from the
EP
extruded composites parallel to the extrusion direction. Tensile tests were conducted
AC C
at a strain rate of 3×10-4s-1 at room temperature using a MTS 810 (American) system. 3 Results and discussion The surface of the pristine CNT (P-CNT) is partially covered with 1-2 nm thick amorphous carbon layer, as shown in Fig. 3a, while for the carboxyl-functionalized CNT (C-CNT), no amorphous carbon was observed on the surface. In addition, there are a lot of defects on the surface of the C-CNT, as shown in Fig. 3b. These defects increase the roughness of the surface [16], which is believed to be beneficial for stress 5
ACCEPTED MANUSCRIPT transfer from the matrix to the CNTs. This will be meticulously investigated hereinafter. The crystallinities of the P-CNTs and the C-CNTs were analyzed using XRD. Fig. 4
RI PT
presents the XRD patterns of P-CNTs and the C-CNTs. The four peaks are related to graphitic characteristic in CNTs, corresponding to the (002), (100), (004), and (110) reflection planes. The (002) and (004) peaks denote interlayer spacing between
SC
adjacent graphene layers, while the (100) and (110) peaks indicate the in-planar
M AN U
graphitic structure [21]. Compared with the XRD pattern of the P-CNTs, the (002) peak of the carboxyl treated CNTs is slightly shifted to lower angles, whereas the (004), (100) and (110) peaks remain at a position similar to that of the pristine CNTs. The nearly identical (004) peaks indicate that the C-CNTs maintain a highly ordered
TE D
tube structure and the similar (100) and (110) peaks suggest negligible damage of the in-plane graphene structure of the C-CNTs [21]. As Raman spectrum is sensitive to the disordered graphite structure, raman
EP
spectroscopy was obtained to estimate the quality of the P-CNTs and the C-CNTs.
AC C
There are two important bands of raman spectrum, ie. the D band (at about 1350 cm-1) and the G band (at about 1580 cm-1) [22]. The D band corresponds to the A1g breathing mode of sp3-hybridized carbon atoms of defects or amorphous carbon atoms in CNTs, while the G band corresponds to E2g tangential mode of sp2-hybridized carbon atoms. The intensity ratio of the D band and the G band (ID/IG) is often used to assess the quality of the internal CNTs [13]. Fig. 5 gives raman spectrum of the P-CNTs and the C-CNTs, respectively. No obvious D band or G band shift was 6
ACCEPTED MANUSCRIPT observed. Moreover, the ratio of ID/IG for P-CNTs is 1.50 and that for C-CNTs increases to 1.71. This means that although the amount of defects apparently increased in the C-CNTs after carboxyl functional treatment, there was no significant
RI PT
structural change in C-CNTs, as manifested in Fig. 3b. To get information about surface elements of the CNTs, the C-CNT and P-CNT were measured with X-ray photoelectron spectroscopy (XPS). The results were shown
SC
in Fig. 6. Appearance of strong O1s peak in the spectra of C-CNTs proves the
M AN U
attachment of oxygen groups of –COOH to the sidewalls of CNTs after carboxyl treatment [23], as shown in Fig. 6a. The comparison of O1s of the spectra of C-CNT and P-CNT is shown in Fig. 6b.
Figure 7 shows the FE-SEM images of the P-CNTs/2014Al and C-CNTs/2014Al
TE D
mixed powders after sonicated for 60 minutes and mechanically stirred for 30 minutes. The P-CNTs are difficult to disperse, and some large agglomerated P-CNTs are observed in the mixed powders, as shown in Fig. 7a and b. However, the C-CNTs are
EP
well dispersed in the 2014Al powders compared to the P-CNTs, although some
AC C
slightly agglomerated C-CNTs are observed in the mixed powders, as shown in Fig. 7c and d. This is probably because of the introducing of a lot of defects and the functional group of –COOH on the surface of the C-CNTs by carboxyl treatment. The presence of a large amount of defects on the surface of the C-CNT increases the C-CNT surface roughness compared to the relatively smooth surface of P-CNT. Higher roughness engenders lower Van der Waals' force by reducing the proximity of interacing surface and by increasing the distance between their atoms, thus facilitating 7
ACCEPTED MANUSCRIPT their ease of dispersion [18, 24]. Moreover, during wet mixing process, hetero-aggregation had occurred between the 2014Al powders and C-CNT in ethanol solution. Functional groups produce negative charges on the C-CNT surface during
RI PT
the ongoing ion exchange with the solution, resulting in the electrostatic stability required for uniform dispersion. The surface of the 2014Al powder possesses positive charges in solution. When the two solutions are mixed, an attractive electrostatic
SC
potential prevails between the oppositely charged C-CNT and 2014Al powders,
M AN U
leading to a result that the C-CNT are attached to the surface of the 2014Al powder [10, 21].
The dispersion of the CNTs in the matrix is presented in Fig. 8. Fig. 8a is for P-CNT/2014Al composites and b is for C-CNT/2014Al composites. It can be seen
TE D
that the CNTs were uniformly dispersed in the matrix except for some entangled small clusters in P-CNT reinforced composites. The results indicate that the ball milling improved the dispersion of the P-CNTs. These observations are consistent with the
EP
FESEM results of the pre-mixed composite powder shown in Figure 7.
AC C
Figure 9 presents the grain structure of the composites. Because of the same preparation process and the same CNTs loading, the composites reinforced by P-CNT (Fig. 9a) and C-CNT (Fig. 9b) show similar average grain sizes. To understand the interfacial bonding between 2014Al matrix and CNTs, TEM analysis was performed to reveal the characteristic differences of interfacial combination status between pristine CNTs-2014Al and carboxyl treated CNTs/2014Al in composites. Fig. 10a is the typical TEM image of the pristine CNTs/2014Al 8
ACCEPTED MANUSCRIPT interface. Careful inspection indicated the existence of a transition layer with a thickness of about 2-3 nm between P-CNTs and 2014Al matrix. The microstructural details of the transition layer were reconstructed using fast fourier transform (FFT) to
RI PT
enhance the detailed micropattern, as shown in the inset in Fig. 10a. The reconstructed pattern of the transition layer shows a cross lattice with lattice spacing of 0.28 nm and 0.21 nm corresponding to (012) and (0012) planes, respectively, which are consistent
SC
with those of Al4C3. In our case because of the low free energy of formation, Al4C3 is
M AN U
easily formed from the reaction of amorphous carbon and Al at the sintering conditions [20, 25]. It is known that the excessive compounds of Al4C3 would be detrimental to the interfacial bonding of the composites. In our study, however, the compounds of Al4C3, appearing in the form of thin layer at the interface between
TE D
P-CNTs and 2014Al matrix, would contribute to the interfacial bonding between CNTs and 2014Al matrix in the composites [26]. However, no obvious transition layers of Al4C3 were observed at the interface of C-CNTs and 2014Al matrix in
EP
composites (Fig. 10b). Also, no preferred interfacial crystallographic orientation
AC C
relationship between the CNTs and aluminum matrix was observed. It has been reported that the functional groups on the surface of the CNTs would produce strong chemical interactions with Al, such as covalent bond [13]. Moreover, there exists large thermal mismatch between the C-CNTs and the matrix during the processing, which may contribute to the mechanical adhesion of the C-CNTs and the aluminum matrix [3]. In addition, the defects on the surface of the C-CNTs would
9
ACCEPTED MANUSCRIPT result in generating a disordered several atom layer of Al near the interface [4], which can substantially improve the interfacial bonding between CNTs and matrix. Fig. 11 presents the representative engineering stress–strain curves of the
RI PT
samples reinforced by P-CNTs and C-CNTs, respectively. The detailed data of the ultimate tensile strength and total elongation are listed in Table 1. The ultimate tensile strength of the composites are increased from 530 to 600 and 630 MPa by addition of
SC
0.5 wt.% of untreated and carboxyl treated carbon nanotubes, respectively. However,
M AN U
the total elongation of the C-CNTs and P-CNTs reinforced composites exihibits a slight decrease compared to that of the 2014Al matrix. It reveals that the C-CNTs are more effective in strengthening the matrix than the P-CNTs. It has been reported that the strengthening mechanisms of CNT reinforced Al-Cu alloy are ascribed to load
TE D
transfer mechanism and grain refinement [27]. Therefore, the difference in mechanical properties of the composites reinforced by P-CNT and C-CNT may be attributed to the difference in the dispersion of the CNTs and the interfacial bonding
EP
status between the two kinds of CNTs and the 2014Al matrix in the composites.
AC C
As we know that the CNTs enhancement is derived from homogeneous dispersion of CNTs in 2014Al matrix. In this study, the C-CNTs shows better dispersity compared to the P-CNTs, because of the introducing of a lot of defects and the functional group of –COOH on the surface of the C-CNTs by carboxyl treatment. In the case of interfacial bonding status, the compounds of Al4C3 appearing in the form of thin layer at the interfcae between P-CNTs and 2014Al matrix would contribute to the interfacial bonding between CNTs and 2014Al matrix in composites. 10
ACCEPTED MANUSCRIPT For the interations between C-CNT and the matrix, as mentioned before, covalent bond between C-CNT and the matrix, large thermal mismatch between the C-CNTs and the matrix, and the disordered atom layers of Al near the interface caused by the
RI PT
Al interaction with the defects on the surface of the C-CNTs substantially contribute to the interfacial bonding between CNTs and matrix.
To evaluate the contribution of the interfacial bonding to mechanical properties,
SC
the characteristic of the pulled-out CNTs on the fracture surface of composites was
M AN U
considered as an indicator of the interface adhesion strength. Fig. 12 shows the tensile fractographs of the composites reinforced with untreated and carboxyl treated carbon nanotubes. Note that the fractured surfaces of both composites are typical of ductile fractures, and the pulled-out CNTs are all observed on the fracture surface. However,
TE D
the pulled-out length of the P-CNTs on the fracture surface (marked by the black arrows in Fig. 12a) of composites with P-CNTs was about 100~150 nm, and some large agglomerated P-CNTs (marked by white elliptical dotted line) were observed in
EP
Fig. 12a. While, only the tips of the C-CNTs emerge on the fracture surface of
AC C
composites with C-CNTs, as marked by black arrows in Fig. 12b. In addition, large agglomerated C-CNTs are rarely observed on the fracture surface. For the short fiber-reinforced composites, the applied force can be transferred from the matrix to the short fibers by shear stress that develops along the fiber–matrix interfaces. In this case, it generates variation in the stress along the length of the fiber, and the stress of the short fibers increases proportionally from the fiber end to reach a maximum value at the mid-length. Thus, the CNT-2014Al interfaces at the site of the 11
ACCEPTED MANUSCRIPT mid-length failed easily and then cracks developed along the interfaces. In this study, the average length of the pulled-out P-CNTs was found to be about 100–150 nm which is smaller than the average length of 1250 nm for P-CNT. This means that most
RI PT
of the P-CNTs are embeded in the 2014Al matrix, implying the interfacial bonding being strong between 2014Al matrix and P-CNT in the composite. However, the pulled-out of the P-CNT result in inadequate load transfer from the matrix to the
SC
P-CNT [11]. For the composite with carboxyl treated carbon nanotubes, the evidence
M AN U
of pulled-out of C-CNTs is hardly discovered on the fracture surface in Fig. 9b. The presence of the few tips of CNTs reveals further strong interface adhesion in the composite. This is mainly because the rough surface of C-CNTs immerses into the matrix to produce an interlocking effect [28, 29], and generates a larger effective
TE D
interfacial contact with the aluminum matrix, causing the interface at tips of C-CNTs to be damaged due to the higher stress concentration. It has been found that the CNTs
matrix [30].
AC C
4 Conclusions
EP
with damaged tip interfaces could still transfer load and further strengthen the mteal
In this study, 2014 aluminum matrix composites reinforced with untreated (P-CNTs) and carboxyl-functionalized carbon nanotubes (C-CNTs) were successfully fabricated by the combination of sintering and hot extrusion. The following conclusions can be drawn: The C-CNTs are found easier to disperse in the aluminum matrix compared to the P-CNTs. Although a transition thin layer of Al4C3 was observed at the interface 12
ACCEPTED MANUSCRIPT between the matrix and the P-CNT, the interface is not strong enough for load transfer from the matrix to the P-CNT. Compared to the P-CNT, no obvious transition layers of Al4C3 were observed at the interface of C-CNTs and 2014Al matrix in composites.
RI PT
C-CNTs have a larger effective interfacial contact with aluminum matrix due to large amount of –COOH group and defects on the surface interacting with Al by strong chemical and physical interactions, which would produce an interlocking effect and
SC
guarantee an adequate load transfer from the matrix to the C-CNT. Ultimate tensile
M AN U
strength of the composites is increased from 530 to 600 and 630 MPa with addition 0.5 wt. % of P-CNTs and C-CNTs, respectively. While, the total elongation of the composites slightly decreases compared to the matrix. Acknowledgement
TE D
Financial supports from the National Basic Research Program of China (973 Program, No. 2012CB619600), the Natural Science Foundation of China (No. 51474111) and Doctoral Fund of Ministry of Education of China (20120061110031)
EP
are greatly acknowledged. Partial financial supports come from the Fundamental
AC C
Research Funds for the Central Universities (JCKY-QKJC02) and the Foundation of Jilin University for Distinguished Young Scholars.
Reference
[1] C.D. Li, X.J. Wang, K. Wu, W.Q. Liu, S.L. Xiang, C. Ding, X.S. Hu, M.Y. Zheng, Distribution and integrity of carbon nanotubes in carbon nanotube/magnesium composites. J. Alloys Compd. 612 (2014) 330–336. 13
ACCEPTED MANUSCRIPT [2] B. Chen, S. Li, H. Imai, L. Jia, J. Umeda, M. Takahashi, K. Kondoh, Carbon nanotube induced microstructural characteristics in powder metallurgy Al matrix composites and their effects on mechanical and conductive properties. J. Alloys
RI PT
Compd. 651 (2015) 608-615. [3] R. Pérez-Bustamante, C.D. Gómez-Esparza, I. Estrada-Guel, M. Miki-Yoshida, L. Licea-Jiménez, S.A. Pérez-García, R. Martínez-Sánchez, Microstructural and
M AN U
milling. Mater. Sci. Eng. A 502 (2009) 159-163.
SC
mechanical characterization of Al–MWCNT composites produced by mechanical
[4] H. Li, J. Kang, C. He, N. Zhao, C. Liang, B. Li. Mechanical properties and interfacial analysis of aluminum matrix composites reinforced by carbon nanotubes with diverse structures. Mater. Sci. Eng. A 577 (2013) 120-124.
TE D
[5] S.R. Bakshi, D. Lahiri, A. Agarwal, Carbon nanotube reinforced metal matrix composites - a review. Int. Mater. Rev. 55 (2010) 41-64. [6] A.M.K. Esawi, M.M. Farag, Carbon nanotube reinforced composites: Potential
EP
and current challenges. Mater. Des. 28 (2007) 2394-2401.
AC C
[7] T. Laha, Y. Chen, D. Lahiri, A. Agarwal, Tensile properties of carbon nanotube reinforced aluminum nanocomposite fabricated by plasma spray forming. Compos Part A. 40 (2009) 589-594. [8] H. Li, J. Fan, X. Geng, B. Li, C. Liang, H. Wang, Alumina powder assisted carbon nanotubes reinforced Mg matrix composites. Mater. Des. 60 (2014) 637-642.
14
ACCEPTED MANUSCRIPT [9] B. Chen, S. Li, H. Imai, L. Jia, J. Umeda, M. Takahashi, An approach for homogeneous carbon nanotube dispersion in Al matrix composites. Mater. Des. 72 (2015) 1-8.
RI PT
[10] L. Jiang, G. Fan, Z. Li, X. Kai, D. Zhang, Z. Chen, An approach to the uniform dispersion of a high volume fraction of carbon nanotubes in aluminum powder. Carbon. 49 (2011) 1965-1971.
SC
[11] H.J. Choi, J.H. Shin, D.H. Bae, The effect of milling conditions on
M AN U
microstructures and mechanical properties of Al/MWCNT composites. Compos Part A. 43 (2012) 1061-1072.
[12] Q. Liu, L. Ke, F. Liu, C. Huang, L. Xing, Microstructure and mechanical property of multi-walled carbon nanotubes reinforced aluminum matrix composites
TE D
fabricated by friction stir processing. Mater. Des. 45 (2013) 343-348. [13] S.K. Singhal, R. Pasricha, S. Teotia, G. Kumar, R.B. Mathur, Fabrication and characterization of Al-matrix composites reinforced with amino-functionalized carbon
EP
nanotubes. Compos. Sci. Technol. 72 (2011) 103-111.
AC C
[14] P. Venkatachalam, S. Ramesh Kumar, B. Ravisankar, V. Thomas Paul, M. Vijayalakshmi, Effect of processing routes on microstructure and mechanical properties of 2014 Al alloy processed by equal channel angular pressing. Trans. Nonferrous Metals Soc. 20 (2010) 1822-1828. [15] Y. Birol, Age hardening of EN AW 2014 alloy extruded in the semi-solid state. Mater. Chem. Phys. 131 (2012) 694-697.
15
ACCEPTED MANUSCRIPT [16] H.C. Schniepp, J.L. Li, M.J. McAllister, H. Sai, M. Herrera-Alonso, D.H. Admason, R.K. Prud’homme, R. Car, D.A. Saville, I.A. Aksay, Functionalized Single Graphene Sheets Derived from Splitting Graphite Oxide. J. Phys. Chem. B 110 (2006)
RI PT
8535-8539. [17] Y.I. Rabinovich, J.J. Adler, A. Ata, R.K. Singh, B.M. Moudgil, Adhesion between Nanoscale Rough Surfaces. J. Colloid Interface Sci. 232 (2000) 10-16.
SC
[18] F.W. Delrio, M.P. de Boer, J.A. Knapp, E. David Reedy Jr., P.J. Clews, Dunn ML.
M AN U
The role of van der Waals forces in adhesion of micromachined surfaces. Nat Mater. 4 (2005) 629-634.
[19] M. Paradise, T. Goswami, Carbon nanotubes – Production and industrial applications. Mater. Des. 28 (2007) 1477-1489.
TE D
[20] L. Ci, Z. Ryu, N.Y. Jin-Phillipp, M. Rühle, Investigation of the interfacial reaction between multi-walled carbon nanotubes and aluminum. Acta Mater. 54 (2006) 5367-5375.
layers
in
the thermal
conductivity of
a multi-walled
carbon
AC C
interfacial
EP
[21] S. Cho, K. Kikuchi, A. Kawasaki, On the role of amorphous intergranular and
nanotube–copper matrix composite. Acta Mater. 60 (2012) 726-736. [22] J.H. Lehman, M. Terrones, E. Mansfield, K.E. Hurst, V. Meunier, Evaluating the characteristics of multiwall carbon nanotubes. Carbon. 49 (2011) 2581-2602. [23] K.A. Wepasnick, B.A. Smith, K.E. Schrote, H.K. Wilson, S.R. Diegelmann, D.H. Fairbrother, Surface and structural characterization of multi-walled carbon nanotubes following different oxidative treatments. Carbon. 49 (2011) 24-36. 16
ACCEPTED MANUSCRIPT [24] A. Fallah, Y. Nakayama, Effect of gas phase oxidation on the structure and intertube adhesion force of a brush-like assembly of carbon nanotubes. Carbon. 50 (2012) 1879-1887.
RI PT
[25] H. Kwon, D.H. Park, J.F. Silvain, A. Kawasaki, Investigation of carbon nanotube reinforced aluminum matrix composite materials. Compos. Sci. Technol. 70 (2010) 546-550.
SC
[26] L. M. Tham, M. Gupta, L. Cheng, Effect of limited matrix–reinforcement
M AN U
interfacial reaction on enhancing the mechanical properties of aluminium–silicon carbide composites. Acta Mater. 49 (2001) 3243–3253.
[27] D.H. Nam, S.I. Cha, B.K. Lim, H.M. Park, D.S. Han, S.H. Hong, Synergistic strengthening by load transfer mechanism and grain refinement of CNT/Al–Cu
TE D
composites. Carbon. 50 (2012) 2417-2423.
[28] C. Cui, W. Qian, M. Zhao, F. Ding, X. Jia, F. Wei, High strength composites
102-108.
EP
using interlocking carbon nanotubes in a polyimide matrix. Carbon. 60 (2013)
AC C
[29] E.M. Byrne, M.A. McCarthy, W.A. Curtin, Pullout of rough multiwall carbon nanotubes: A parametric study. Compos Part A. 56 (2014) 93-102. [30] Z.Y. Liu, B.L. Xiao, W.G. Wang, Z.Y. Ma, Developing high-performance aluminum matrix composites with directionally aligned carbon nanotubes by combining friction stir processing and subsequent rolling. Carbon. 62 (2013) 35-42.
17
ACCEPTED MANUSCRIPT Figure captions Fig. 1 FESEM micrographs of the as-received carbon nanotubes (a) and the 2014 aluminum alloy powders (b).
RI PT
Fig. 2 Schematic illustration of the procedure to fabricate the CNTs/2014Al composites.
C-CNTs.
M AN U
Fig. 4 XRD patterns of the P-CNTs and the C-CNTs.
SC
Fig. 3 TEM micrographs of the surface morphologies of the CNTs: (a) P-CNTs and (b)
Fig. 5 Raman spectra taken from the P-CNTs and the C-CNTs. Fig. 6 XPS spectra: (a) full surveys of the P-CNT and C-CNT, (b) enlarged O1s regions of (a). Note that the O1s peaks in (b) have been normalized to the
TE D
accompanying C1s peak in each case.
Fig. 7 FE-SEM images of the composites powders after sonication and mechanical stirring: (a) and (c) show spherical morphologies of an aluminum alloy particles
EP
contain P-CNTs and C-CNTs, respectively, while (b) and (d) are the enlarged images
AC C
of the regions in (a) and (c) marked by the blank rectangle, respectively. Fig. 8 Dispersion of CNTs in the matrix: (a) P-CNT/2014Al composites and (b) C-CNT/2014Al composites. Fig. 9 Grain structures of the composites: (a) P-CNT/2014Al composites and (b) C-CNT/2014Al composites.
18
ACCEPTED MANUSCRIPT Fig. 10 TEM images of 2014Al composites reinforced by CNTs. (a) P-CNTs and (b) C-CNTs, respectively. The reconstructed Fast Fourier Transformation (FFT) insert in (a) is the Al4C3 phase.
RI PT
Fig. 11 Typical engineering stress–strain curves of the 2014 Aluminum alloy and the CNTs/2014Al composites.
Fig. 12 FESEM micrographs of the fracture morphology of the composites after
SC
tensile test: (a) P-CNTs/2014Al composites, (b) C-CNTs/2014Al composites.
M AN U
Table caption
Table 1 Ultimate tensile strength and the total elongation of the 2014Al alloy and the
AC C
EP
TE D
2014Al composites reinforced by the P-CNTs and the C-CNTs.
19
ACCEPTED MANUSCRIPT Table 1 Ultimate tensile strength and the total elongation of the 2014Al alloy and the 2014Al composites reinforced by the P-CNTs and the C-CNTs. Samples
UTS (MPa)
EL (%)
530 ± 5
13.5 ± 0.8
P-CNT/2014Al
600 ± 12
12.7 ± 0.9
C-CNT/2014Al
630 ± 2
12.0 ± 1.1
AC C
EP
TE D
M AN U
SC
RI PT
2014Al
1
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Highlights: 1. Effect of carbon nanotubes surface on the dispersion and strengthening effect was investigated.
RI PT
2. Carboxyl-functionalized carbon nanotubes exhibit better dispersion than the untreated one.
AC C
EP
TE D
M AN U
SC
3. Carboxyl-functionalized carbon nanotubes show more effective in load transfer.
S0925-8388(16)30593-X
DOI:
10.1016/j.jallcom.2016.03.036
Reference:
JALCOM 36924
To appear in:
Journal of Alloys and Compounds
Received Date: 21 November 2015 Revised Date:
17 February 2016
Accepted Date: 7 March 2016
Please cite this article as: X. Zhu, Y.-G. Zhao, M. Wu, H.-Y. Wang, Q.-C. Jiang, Fabrication of 2014 aluminum matrix composites reinforced with untreated and carboxyl-functionalized carbon nanotubes, Journal of Alloys and Compounds (2016), doi: 10.1016/j.jallcom.2016.03.036. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT
Fabrication of 2014 aluminum matrix composites reinforced with untreated and carboxyl-functionalized carbon nanotubes Xian Zhu, Yu-Guang Zhao, Min Wu, Hui-Yuan Wang∗, Qi-Chuan Jiang
RI PT
Key Laboratory of Automobile Materials of Ministry of Education & School of Materials Science and Engineering, Nanling Campus, Jilin University, No. 5988 Renmin Street, Changchun 130025, PR China Abstract
SC
2014 aluminum matrix composites reinforced with untreated (P-CNTs) and carboxyl-functionalized carbon nanotubes (C-CNTs) were successfully fabricated by
M AN U
the combination of sintering and hot extrusion. It is revealed that the C-CNTs are easier to disperse compared to the P-CNTs. Interfacial analysis indicated that a thin layer of aluminum carbide (Al4C3) exists at the interface between the matrix and the
TE D
P-CNTs. However, no obvious transition layers of Al4C3 were observed at the interface of C-CNTs and 2014Al matrix in composites. Also, no preferred interfacial crystallographic orientation relationship between the C-CNTs and aluminum matrix
EP
was observed. Furthermore, C-CNTs have a larger effective interfacial contact with
AC C
aluminum matrix compared to the untreated CNTs, due to large amount of –COOH group and defects on the surface of the C-CNTs interacting with Al by strong chemical and physical interactions. These aspects in turn affect the mechanical properties of the composites. The ultimate tensile strength of the composites were raised from 530 to 600 and 630 MPa with 0.5 wt.% of P-CNTs and C-CNTs,
∗Corresponding author. Tel/fax: +86 431 8509 4699 E–mail address: [email protected] (H.-Y. Wang)
1
ACCEPTED MANUSCRIPT respectively. It reveals that the C-CNTs are more effective in strengthening the aluminum matrix than the P-CNTs. Keywords: Metal matrix composites; Carbon nanotubes; powder metallurgy;
RI PT
Dispersion; Interface; Strength
1 Introduction
SC
Metal matrix composites (MMCs) have been receiving great attention due to their
M AN U
high tensile strength, hardness and modulus, as well as their high wear resistance compared to the matrix [1-4]. Carbon nanotubes (CNTs) are considered to be ideal reinforcements of composites due to their exceptional strength, modulus and heat conductivity [5-7]. However, because CNTs have tremendous specific surface area
TE D
and very strong van der Waal’s force attraction between them which lead to the formation of CNTs clusters. How to homogeneously disperse the CNTs in a metal matrix and achieve excellent interfacial bonding between the CNTs and metal matrix
EP
are still unsolved problems [2, 8-11]. Therefore, only limited studies have been
AC C
reported on MMCs. The metal matrix of the reported CNTs-MMCs which mainly focused on the dispersion and alignment in matrix of CNTs, were essentially limited to pure aluminum [3, 4, 9, 11-13]. However, pure aluminum matrix is usually not eligible for applications due to its low mechanical strength. Al-Cu based 2014Al alloy is heat treatable aluminum alloy, which offers high strength at low specific weight and are extensively employed as structural components, particularly in the aircraft industry [14, 15]. CNTs reinforced 2014Al 2
ACCEPTED MANUSCRIPT composites are thus be considered to have more application potential. However, homogeneously disperse the CNTs in the matrix and achieve excellent interfacial bonding between the CNTs and the matrix are still needed to be addressed.
RI PT
The functionalized treatment is a common way to obtain a homogeneous dispersion of CNTs by introducing carboxyl or hydroxyl groups on the surface of the CNTs. Calculations show that the addition of carboxyl or hydroxyl groups adds
SC
around 0.2 nm to the graphitic layer roughness [16]. Higher roughness engenders
M AN U
lower van der waal’s force by reducing the proximity of interacting surfaces and by increasing the distance between their atoms [17, 18]. In the case of interfacial bonding, the amorphous carbon generated during the chemical vapor deposition [19] could affect the interfacial bonding status between CNTs and Al matrix to some extent,
TE D
because the amorphous carbon tends to react with aluminum to form aluminum carbide [20]. Meanwhile, the carboxyl treatment could remove carbon impurities and introduce defects on the surface of the CNTs [21]. However, the influence of the
AC C
studied.
EP
surface characteristic on the strengthening of CNTs in MMCs has not been well
In view of these facts, in the present work, the 2014Al matrix composites reinforced with untreated carbon nanotubes (P-CNTs) and carboxyl-functionalized carbon nanotubes (C-CNTs) by the combination of sintering and hot extrusion were fabricated to clarify the influence of surface characteristic on the dispersion of the CNTs and the interfacial bonding status between the CNTs and the 2014Al matrix. 2 Experimental details 3
ACCEPTED MANUSCRIPT CNTs (with an average diameter of ~25 nm and length of ~1.25 µm) synthesized by a chemical vapor deposition (CVD) method were selected as the reinforcements (Fig. 1a). Carboxyl functional treated CNTs (C-CNTs) (with an average diameter and
RI PT
length of about 25 nm, 1.25 µm) were provided by Chengdu Organic Chemistry Co., Ltd., China. 2014 aluminum alloy powders with an average powder size of about 10µm were employed as the matrix (Fig. 1b).
SC
The experiment process was illustrated in Fig. 2. Firstly, the two kinds of CNTs
M AN U
(0.25 g) were dispersed in 250 ml ethanol, sonicated for 60 minutes, and then poured into 200ml ethanol solution that contained 49.75 g 2014Al alloy powders, respectively. Secondly, the composite mixtures were then vigorously stirred by magnetic stirring for 30 minutes followed by drying at 333 K for 24 h in an oven. Thirdly, ball milling
TE D
were performed in a planetary ball mill at a constant speed of 300 rpm for 6 hours under argon atmosphere with ball to power ratio of 8:1 with 0.5 g stearic acid addition into the dried mixture as process control agent. Finally, the ball-milled powders were
EP
packed and cold compacted at a pressure of 50 MPa for 2 minutes, and subsequently
AC C
the compacts were hot vacuum sintered for 50 minutes under 793 K and then pressed at a pressure of 50 MPa for 10 minutes in a self-made vessel, with furnace cooling to room temperature. Hot extrusion was conducted at 693 K with an extrusion ratio of about 15:1. The extrudates were then solution treated at 775 K in air for 2 hours and naturally aged for more than 96 hours before test. The two kinds of CNTs were characterized by X-ray diffraction (XRD) (D/Max 2500PC, Rigaku, Japan) using Cu Kα radiation in the range of 20°~80° with a 4
ACCEPTED MANUSCRIPT scanning speed of 4 deg/min and an acquisition step of 0.02° (2θ). Raman spectroscopy was obtained by REINSHAW in Via Raman spectroscope with an excitation laser wavelength of 633 nm in the range of 1000~2000 cm-1. X-ray
RI PT
photoelectron spectroscopy (XPS) was conducted on a VG ESCALAB MK Ⅱsystem. Different CNTs samples for Transmission electron microscopy (TEM) observation were prepared by dispersing the corresponding CNTs in ethanol, sonicating for
SC
several minutes, and subsequently dripping one drop of the liquid on a holey carbon
M AN U
TEM support grid. After dried overnight, the samples are observed by TEM (TEM; JEM-2100F, Japan and Tecnai G2 20 American) at an acceleration voltage of 200 KV. Samples for interface observation were prepared by ion beam thinner (Leica, EMRES101), then studied by TEM. The pre-mixed composite powders as well as
TE D
fractured surface were characterized by a Field emission scanning electron microscopy (FESEM) (JSM-6700F, Japan). Tensile specimens with a gauge length of 10 mm, a width of 4.0 mm, and a thickness of about 1 mm were wire-cut from the
EP
extruded composites parallel to the extrusion direction. Tensile tests were conducted
AC C
at a strain rate of 3×10-4s-1 at room temperature using a MTS 810 (American) system. 3 Results and discussion The surface of the pristine CNT (P-CNT) is partially covered with 1-2 nm thick amorphous carbon layer, as shown in Fig. 3a, while for the carboxyl-functionalized CNT (C-CNT), no amorphous carbon was observed on the surface. In addition, there are a lot of defects on the surface of the C-CNT, as shown in Fig. 3b. These defects increase the roughness of the surface [16], which is believed to be beneficial for stress 5
ACCEPTED MANUSCRIPT transfer from the matrix to the CNTs. This will be meticulously investigated hereinafter. The crystallinities of the P-CNTs and the C-CNTs were analyzed using XRD. Fig. 4
RI PT
presents the XRD patterns of P-CNTs and the C-CNTs. The four peaks are related to graphitic characteristic in CNTs, corresponding to the (002), (100), (004), and (110) reflection planes. The (002) and (004) peaks denote interlayer spacing between
SC
adjacent graphene layers, while the (100) and (110) peaks indicate the in-planar
M AN U
graphitic structure [21]. Compared with the XRD pattern of the P-CNTs, the (002) peak of the carboxyl treated CNTs is slightly shifted to lower angles, whereas the (004), (100) and (110) peaks remain at a position similar to that of the pristine CNTs. The nearly identical (004) peaks indicate that the C-CNTs maintain a highly ordered
TE D
tube structure and the similar (100) and (110) peaks suggest negligible damage of the in-plane graphene structure of the C-CNTs [21]. As Raman spectrum is sensitive to the disordered graphite structure, raman
EP
spectroscopy was obtained to estimate the quality of the P-CNTs and the C-CNTs.
AC C
There are two important bands of raman spectrum, ie. the D band (at about 1350 cm-1) and the G band (at about 1580 cm-1) [22]. The D band corresponds to the A1g breathing mode of sp3-hybridized carbon atoms of defects or amorphous carbon atoms in CNTs, while the G band corresponds to E2g tangential mode of sp2-hybridized carbon atoms. The intensity ratio of the D band and the G band (ID/IG) is often used to assess the quality of the internal CNTs [13]. Fig. 5 gives raman spectrum of the P-CNTs and the C-CNTs, respectively. No obvious D band or G band shift was 6
ACCEPTED MANUSCRIPT observed. Moreover, the ratio of ID/IG for P-CNTs is 1.50 and that for C-CNTs increases to 1.71. This means that although the amount of defects apparently increased in the C-CNTs after carboxyl functional treatment, there was no significant
RI PT
structural change in C-CNTs, as manifested in Fig. 3b. To get information about surface elements of the CNTs, the C-CNT and P-CNT were measured with X-ray photoelectron spectroscopy (XPS). The results were shown
SC
in Fig. 6. Appearance of strong O1s peak in the spectra of C-CNTs proves the
M AN U
attachment of oxygen groups of –COOH to the sidewalls of CNTs after carboxyl treatment [23], as shown in Fig. 6a. The comparison of O1s of the spectra of C-CNT and P-CNT is shown in Fig. 6b.
Figure 7 shows the FE-SEM images of the P-CNTs/2014Al and C-CNTs/2014Al
TE D
mixed powders after sonicated for 60 minutes and mechanically stirred for 30 minutes. The P-CNTs are difficult to disperse, and some large agglomerated P-CNTs are observed in the mixed powders, as shown in Fig. 7a and b. However, the C-CNTs are
EP
well dispersed in the 2014Al powders compared to the P-CNTs, although some
AC C
slightly agglomerated C-CNTs are observed in the mixed powders, as shown in Fig. 7c and d. This is probably because of the introducing of a lot of defects and the functional group of –COOH on the surface of the C-CNTs by carboxyl treatment. The presence of a large amount of defects on the surface of the C-CNT increases the C-CNT surface roughness compared to the relatively smooth surface of P-CNT. Higher roughness engenders lower Van der Waals' force by reducing the proximity of interacing surface and by increasing the distance between their atoms, thus facilitating 7
ACCEPTED MANUSCRIPT their ease of dispersion [18, 24]. Moreover, during wet mixing process, hetero-aggregation had occurred between the 2014Al powders and C-CNT in ethanol solution. Functional groups produce negative charges on the C-CNT surface during
RI PT
the ongoing ion exchange with the solution, resulting in the electrostatic stability required for uniform dispersion. The surface of the 2014Al powder possesses positive charges in solution. When the two solutions are mixed, an attractive electrostatic
SC
potential prevails between the oppositely charged C-CNT and 2014Al powders,
M AN U
leading to a result that the C-CNT are attached to the surface of the 2014Al powder [10, 21].
The dispersion of the CNTs in the matrix is presented in Fig. 8. Fig. 8a is for P-CNT/2014Al composites and b is for C-CNT/2014Al composites. It can be seen
TE D
that the CNTs were uniformly dispersed in the matrix except for some entangled small clusters in P-CNT reinforced composites. The results indicate that the ball milling improved the dispersion of the P-CNTs. These observations are consistent with the
EP
FESEM results of the pre-mixed composite powder shown in Figure 7.
AC C
Figure 9 presents the grain structure of the composites. Because of the same preparation process and the same CNTs loading, the composites reinforced by P-CNT (Fig. 9a) and C-CNT (Fig. 9b) show similar average grain sizes. To understand the interfacial bonding between 2014Al matrix and CNTs, TEM analysis was performed to reveal the characteristic differences of interfacial combination status between pristine CNTs-2014Al and carboxyl treated CNTs/2014Al in composites. Fig. 10a is the typical TEM image of the pristine CNTs/2014Al 8
ACCEPTED MANUSCRIPT interface. Careful inspection indicated the existence of a transition layer with a thickness of about 2-3 nm between P-CNTs and 2014Al matrix. The microstructural details of the transition layer were reconstructed using fast fourier transform (FFT) to
RI PT
enhance the detailed micropattern, as shown in the inset in Fig. 10a. The reconstructed pattern of the transition layer shows a cross lattice with lattice spacing of 0.28 nm and 0.21 nm corresponding to (012) and (0012) planes, respectively, which are consistent
SC
with those of Al4C3. In our case because of the low free energy of formation, Al4C3 is
M AN U
easily formed from the reaction of amorphous carbon and Al at the sintering conditions [20, 25]. It is known that the excessive compounds of Al4C3 would be detrimental to the interfacial bonding of the composites. In our study, however, the compounds of Al4C3, appearing in the form of thin layer at the interface between
TE D
P-CNTs and 2014Al matrix, would contribute to the interfacial bonding between CNTs and 2014Al matrix in the composites [26]. However, no obvious transition layers of Al4C3 were observed at the interface of C-CNTs and 2014Al matrix in
EP
composites (Fig. 10b). Also, no preferred interfacial crystallographic orientation
AC C
relationship between the CNTs and aluminum matrix was observed. It has been reported that the functional groups on the surface of the CNTs would produce strong chemical interactions with Al, such as covalent bond [13]. Moreover, there exists large thermal mismatch between the C-CNTs and the matrix during the processing, which may contribute to the mechanical adhesion of the C-CNTs and the aluminum matrix [3]. In addition, the defects on the surface of the C-CNTs would
9
ACCEPTED MANUSCRIPT result in generating a disordered several atom layer of Al near the interface [4], which can substantially improve the interfacial bonding between CNTs and matrix. Fig. 11 presents the representative engineering stress–strain curves of the
RI PT
samples reinforced by P-CNTs and C-CNTs, respectively. The detailed data of the ultimate tensile strength and total elongation are listed in Table 1. The ultimate tensile strength of the composites are increased from 530 to 600 and 630 MPa by addition of
SC
0.5 wt.% of untreated and carboxyl treated carbon nanotubes, respectively. However,
M AN U
the total elongation of the C-CNTs and P-CNTs reinforced composites exihibits a slight decrease compared to that of the 2014Al matrix. It reveals that the C-CNTs are more effective in strengthening the matrix than the P-CNTs. It has been reported that the strengthening mechanisms of CNT reinforced Al-Cu alloy are ascribed to load
TE D
transfer mechanism and grain refinement [27]. Therefore, the difference in mechanical properties of the composites reinforced by P-CNT and C-CNT may be attributed to the difference in the dispersion of the CNTs and the interfacial bonding
EP
status between the two kinds of CNTs and the 2014Al matrix in the composites.
AC C
As we know that the CNTs enhancement is derived from homogeneous dispersion of CNTs in 2014Al matrix. In this study, the C-CNTs shows better dispersity compared to the P-CNTs, because of the introducing of a lot of defects and the functional group of –COOH on the surface of the C-CNTs by carboxyl treatment. In the case of interfacial bonding status, the compounds of Al4C3 appearing in the form of thin layer at the interfcae between P-CNTs and 2014Al matrix would contribute to the interfacial bonding between CNTs and 2014Al matrix in composites. 10
ACCEPTED MANUSCRIPT For the interations between C-CNT and the matrix, as mentioned before, covalent bond between C-CNT and the matrix, large thermal mismatch between the C-CNTs and the matrix, and the disordered atom layers of Al near the interface caused by the
RI PT
Al interaction with the defects on the surface of the C-CNTs substantially contribute to the interfacial bonding between CNTs and matrix.
To evaluate the contribution of the interfacial bonding to mechanical properties,
SC
the characteristic of the pulled-out CNTs on the fracture surface of composites was
M AN U
considered as an indicator of the interface adhesion strength. Fig. 12 shows the tensile fractographs of the composites reinforced with untreated and carboxyl treated carbon nanotubes. Note that the fractured surfaces of both composites are typical of ductile fractures, and the pulled-out CNTs are all observed on the fracture surface. However,
TE D
the pulled-out length of the P-CNTs on the fracture surface (marked by the black arrows in Fig. 12a) of composites with P-CNTs was about 100~150 nm, and some large agglomerated P-CNTs (marked by white elliptical dotted line) were observed in
EP
Fig. 12a. While, only the tips of the C-CNTs emerge on the fracture surface of
AC C
composites with C-CNTs, as marked by black arrows in Fig. 12b. In addition, large agglomerated C-CNTs are rarely observed on the fracture surface. For the short fiber-reinforced composites, the applied force can be transferred from the matrix to the short fibers by shear stress that develops along the fiber–matrix interfaces. In this case, it generates variation in the stress along the length of the fiber, and the stress of the short fibers increases proportionally from the fiber end to reach a maximum value at the mid-length. Thus, the CNT-2014Al interfaces at the site of the 11
ACCEPTED MANUSCRIPT mid-length failed easily and then cracks developed along the interfaces. In this study, the average length of the pulled-out P-CNTs was found to be about 100–150 nm which is smaller than the average length of 1250 nm for P-CNT. This means that most
RI PT
of the P-CNTs are embeded in the 2014Al matrix, implying the interfacial bonding being strong between 2014Al matrix and P-CNT in the composite. However, the pulled-out of the P-CNT result in inadequate load transfer from the matrix to the
SC
P-CNT [11]. For the composite with carboxyl treated carbon nanotubes, the evidence
M AN U
of pulled-out of C-CNTs is hardly discovered on the fracture surface in Fig. 9b. The presence of the few tips of CNTs reveals further strong interface adhesion in the composite. This is mainly because the rough surface of C-CNTs immerses into the matrix to produce an interlocking effect [28, 29], and generates a larger effective
TE D
interfacial contact with the aluminum matrix, causing the interface at tips of C-CNTs to be damaged due to the higher stress concentration. It has been found that the CNTs
matrix [30].
AC C
4 Conclusions
EP
with damaged tip interfaces could still transfer load and further strengthen the mteal
In this study, 2014 aluminum matrix composites reinforced with untreated (P-CNTs) and carboxyl-functionalized carbon nanotubes (C-CNTs) were successfully fabricated by the combination of sintering and hot extrusion. The following conclusions can be drawn: The C-CNTs are found easier to disperse in the aluminum matrix compared to the P-CNTs. Although a transition thin layer of Al4C3 was observed at the interface 12
ACCEPTED MANUSCRIPT between the matrix and the P-CNT, the interface is not strong enough for load transfer from the matrix to the P-CNT. Compared to the P-CNT, no obvious transition layers of Al4C3 were observed at the interface of C-CNTs and 2014Al matrix in composites.
RI PT
C-CNTs have a larger effective interfacial contact with aluminum matrix due to large amount of –COOH group and defects on the surface interacting with Al by strong chemical and physical interactions, which would produce an interlocking effect and
SC
guarantee an adequate load transfer from the matrix to the C-CNT. Ultimate tensile
M AN U
strength of the composites is increased from 530 to 600 and 630 MPa with addition 0.5 wt. % of P-CNTs and C-CNTs, respectively. While, the total elongation of the composites slightly decreases compared to the matrix. Acknowledgement
TE D
Financial supports from the National Basic Research Program of China (973 Program, No. 2012CB619600), the Natural Science Foundation of China (No. 51474111) and Doctoral Fund of Ministry of Education of China (20120061110031)
EP
are greatly acknowledged. Partial financial supports come from the Fundamental
AC C
Research Funds for the Central Universities (JCKY-QKJC02) and the Foundation of Jilin University for Distinguished Young Scholars.
Reference
[1] C.D. Li, X.J. Wang, K. Wu, W.Q. Liu, S.L. Xiang, C. Ding, X.S. Hu, M.Y. Zheng, Distribution and integrity of carbon nanotubes in carbon nanotube/magnesium composites. J. Alloys Compd. 612 (2014) 330–336. 13
ACCEPTED MANUSCRIPT [2] B. Chen, S. Li, H. Imai, L. Jia, J. Umeda, M. Takahashi, K. Kondoh, Carbon nanotube induced microstructural characteristics in powder metallurgy Al matrix composites and their effects on mechanical and conductive properties. J. Alloys
RI PT
Compd. 651 (2015) 608-615. [3] R. Pérez-Bustamante, C.D. Gómez-Esparza, I. Estrada-Guel, M. Miki-Yoshida, L. Licea-Jiménez, S.A. Pérez-García, R. Martínez-Sánchez, Microstructural and
M AN U
milling. Mater. Sci. Eng. A 502 (2009) 159-163.
SC
mechanical characterization of Al–MWCNT composites produced by mechanical
[4] H. Li, J. Kang, C. He, N. Zhao, C. Liang, B. Li. Mechanical properties and interfacial analysis of aluminum matrix composites reinforced by carbon nanotubes with diverse structures. Mater. Sci. Eng. A 577 (2013) 120-124.
TE D
[5] S.R. Bakshi, D. Lahiri, A. Agarwal, Carbon nanotube reinforced metal matrix composites - a review. Int. Mater. Rev. 55 (2010) 41-64. [6] A.M.K. Esawi, M.M. Farag, Carbon nanotube reinforced composites: Potential
EP
and current challenges. Mater. Des. 28 (2007) 2394-2401.
AC C
[7] T. Laha, Y. Chen, D. Lahiri, A. Agarwal, Tensile properties of carbon nanotube reinforced aluminum nanocomposite fabricated by plasma spray forming. Compos Part A. 40 (2009) 589-594. [8] H. Li, J. Fan, X. Geng, B. Li, C. Liang, H. Wang, Alumina powder assisted carbon nanotubes reinforced Mg matrix composites. Mater. Des. 60 (2014) 637-642.
14
ACCEPTED MANUSCRIPT [9] B. Chen, S. Li, H. Imai, L. Jia, J. Umeda, M. Takahashi, An approach for homogeneous carbon nanotube dispersion in Al matrix composites. Mater. Des. 72 (2015) 1-8.
RI PT
[10] L. Jiang, G. Fan, Z. Li, X. Kai, D. Zhang, Z. Chen, An approach to the uniform dispersion of a high volume fraction of carbon nanotubes in aluminum powder. Carbon. 49 (2011) 1965-1971.
SC
[11] H.J. Choi, J.H. Shin, D.H. Bae, The effect of milling conditions on
M AN U
microstructures and mechanical properties of Al/MWCNT composites. Compos Part A. 43 (2012) 1061-1072.
[12] Q. Liu, L. Ke, F. Liu, C. Huang, L. Xing, Microstructure and mechanical property of multi-walled carbon nanotubes reinforced aluminum matrix composites
TE D
fabricated by friction stir processing. Mater. Des. 45 (2013) 343-348. [13] S.K. Singhal, R. Pasricha, S. Teotia, G. Kumar, R.B. Mathur, Fabrication and characterization of Al-matrix composites reinforced with amino-functionalized carbon
EP
nanotubes. Compos. Sci. Technol. 72 (2011) 103-111.
AC C
[14] P. Venkatachalam, S. Ramesh Kumar, B. Ravisankar, V. Thomas Paul, M. Vijayalakshmi, Effect of processing routes on microstructure and mechanical properties of 2014 Al alloy processed by equal channel angular pressing. Trans. Nonferrous Metals Soc. 20 (2010) 1822-1828. [15] Y. Birol, Age hardening of EN AW 2014 alloy extruded in the semi-solid state. Mater. Chem. Phys. 131 (2012) 694-697.
15
ACCEPTED MANUSCRIPT [16] H.C. Schniepp, J.L. Li, M.J. McAllister, H. Sai, M. Herrera-Alonso, D.H. Admason, R.K. Prud’homme, R. Car, D.A. Saville, I.A. Aksay, Functionalized Single Graphene Sheets Derived from Splitting Graphite Oxide. J. Phys. Chem. B 110 (2006)
RI PT
8535-8539. [17] Y.I. Rabinovich, J.J. Adler, A. Ata, R.K. Singh, B.M. Moudgil, Adhesion between Nanoscale Rough Surfaces. J. Colloid Interface Sci. 232 (2000) 10-16.
SC
[18] F.W. Delrio, M.P. de Boer, J.A. Knapp, E. David Reedy Jr., P.J. Clews, Dunn ML.
M AN U
The role of van der Waals forces in adhesion of micromachined surfaces. Nat Mater. 4 (2005) 629-634.
[19] M. Paradise, T. Goswami, Carbon nanotubes – Production and industrial applications. Mater. Des. 28 (2007) 1477-1489.
TE D
[20] L. Ci, Z. Ryu, N.Y. Jin-Phillipp, M. Rühle, Investigation of the interfacial reaction between multi-walled carbon nanotubes and aluminum. Acta Mater. 54 (2006) 5367-5375.
layers
in
the thermal
conductivity of
a multi-walled
carbon
AC C
interfacial
EP
[21] S. Cho, K. Kikuchi, A. Kawasaki, On the role of amorphous intergranular and
nanotube–copper matrix composite. Acta Mater. 60 (2012) 726-736. [22] J.H. Lehman, M. Terrones, E. Mansfield, K.E. Hurst, V. Meunier, Evaluating the characteristics of multiwall carbon nanotubes. Carbon. 49 (2011) 2581-2602. [23] K.A. Wepasnick, B.A. Smith, K.E. Schrote, H.K. Wilson, S.R. Diegelmann, D.H. Fairbrother, Surface and structural characterization of multi-walled carbon nanotubes following different oxidative treatments. Carbon. 49 (2011) 24-36. 16
ACCEPTED MANUSCRIPT [24] A. Fallah, Y. Nakayama, Effect of gas phase oxidation on the structure and intertube adhesion force of a brush-like assembly of carbon nanotubes. Carbon. 50 (2012) 1879-1887.
RI PT
[25] H. Kwon, D.H. Park, J.F. Silvain, A. Kawasaki, Investigation of carbon nanotube reinforced aluminum matrix composite materials. Compos. Sci. Technol. 70 (2010) 546-550.
SC
[26] L. M. Tham, M. Gupta, L. Cheng, Effect of limited matrix–reinforcement
M AN U
interfacial reaction on enhancing the mechanical properties of aluminium–silicon carbide composites. Acta Mater. 49 (2001) 3243–3253.
[27] D.H. Nam, S.I. Cha, B.K. Lim, H.M. Park, D.S. Han, S.H. Hong, Synergistic strengthening by load transfer mechanism and grain refinement of CNT/Al–Cu
TE D
composites. Carbon. 50 (2012) 2417-2423.
[28] C. Cui, W. Qian, M. Zhao, F. Ding, X. Jia, F. Wei, High strength composites
102-108.
EP
using interlocking carbon nanotubes in a polyimide matrix. Carbon. 60 (2013)
AC C
[29] E.M. Byrne, M.A. McCarthy, W.A. Curtin, Pullout of rough multiwall carbon nanotubes: A parametric study. Compos Part A. 56 (2014) 93-102. [30] Z.Y. Liu, B.L. Xiao, W.G. Wang, Z.Y. Ma, Developing high-performance aluminum matrix composites with directionally aligned carbon nanotubes by combining friction stir processing and subsequent rolling. Carbon. 62 (2013) 35-42.
17
ACCEPTED MANUSCRIPT Figure captions Fig. 1 FESEM micrographs of the as-received carbon nanotubes (a) and the 2014 aluminum alloy powders (b).
RI PT
Fig. 2 Schematic illustration of the procedure to fabricate the CNTs/2014Al composites.
C-CNTs.
M AN U
Fig. 4 XRD patterns of the P-CNTs and the C-CNTs.
SC
Fig. 3 TEM micrographs of the surface morphologies of the CNTs: (a) P-CNTs and (b)
Fig. 5 Raman spectra taken from the P-CNTs and the C-CNTs. Fig. 6 XPS spectra: (a) full surveys of the P-CNT and C-CNT, (b) enlarged O1s regions of (a). Note that the O1s peaks in (b) have been normalized to the
TE D
accompanying C1s peak in each case.
Fig. 7 FE-SEM images of the composites powders after sonication and mechanical stirring: (a) and (c) show spherical morphologies of an aluminum alloy particles
EP
contain P-CNTs and C-CNTs, respectively, while (b) and (d) are the enlarged images
AC C
of the regions in (a) and (c) marked by the blank rectangle, respectively. Fig. 8 Dispersion of CNTs in the matrix: (a) P-CNT/2014Al composites and (b) C-CNT/2014Al composites. Fig. 9 Grain structures of the composites: (a) P-CNT/2014Al composites and (b) C-CNT/2014Al composites.
18
ACCEPTED MANUSCRIPT Fig. 10 TEM images of 2014Al composites reinforced by CNTs. (a) P-CNTs and (b) C-CNTs, respectively. The reconstructed Fast Fourier Transformation (FFT) insert in (a) is the Al4C3 phase.
RI PT
Fig. 11 Typical engineering stress–strain curves of the 2014 Aluminum alloy and the CNTs/2014Al composites.
Fig. 12 FESEM micrographs of the fracture morphology of the composites after
SC
tensile test: (a) P-CNTs/2014Al composites, (b) C-CNTs/2014Al composites.
M AN U
Table caption
Table 1 Ultimate tensile strength and the total elongation of the 2014Al alloy and the
AC C
EP
TE D
2014Al composites reinforced by the P-CNTs and the C-CNTs.
19
ACCEPTED MANUSCRIPT Table 1 Ultimate tensile strength and the total elongation of the 2014Al alloy and the 2014Al composites reinforced by the P-CNTs and the C-CNTs. Samples
UTS (MPa)
EL (%)
530 ± 5
13.5 ± 0.8
P-CNT/2014Al
600 ± 12
12.7 ± 0.9
C-CNT/2014Al
630 ± 2
12.0 ± 1.1
AC C
EP
TE D
M AN U
SC
RI PT
2014Al
1
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
AC C
EP
TE D
M AN U
SC
RI PT
ACCEPTED MANUSCRIPT
ACCEPTED MANUSCRIPT Highlights: 1. Effect of carbon nanotubes surface on the dispersion and strengthening effect was investigated.
RI PT
2. Carboxyl-functionalized carbon nanotubes exhibit better dispersion than the untreated one.
AC C
EP
TE D
M AN U
SC
3. Carboxyl-functionalized carbon nanotubes show more effective in load transfer.