Pérez-Bustamante, R., Pérez-Bustamante, F., Estrada-Guel, I., Licea-Jiménez, L., Miki-Yoshida, M., & Martínez-Sánchez, R. (2013). Effect of milling time and CNT concentration on hardness of CNT/Al 2024 composites produced by mechanical alloying.Materials characterization,75, 13-19.
Acknowledgments
This research was supported by CONACYT (106658). Thanks to Redes Temáticas de Nanotecnología y Nanociencias, Reg. 0124623. Thanks to Red de Ciencia y Tecnología Espaciales Regs. 0170962 and 0170617. Thanks to E. Torres-Moyé,W. Antúnez-Flores, K. Campos Venegas, O. Solis-Canto, and C. E. Ornelas-Gutiérrez for their technical assistance.
Abstract
Carbon nanotube/2024 aluminum alloy (CNT/Al2024) composites were fabricated with a combination of mechanical alloying (MA) and powder metallurgy routes. Composites were microstructurally and mechanically evaluated at sintering condition. A homogeneous dispersion of CNTs in the Al matrix was observed by a field emission scanning electron microscopy. High-resolution transmission electron microscopy confirmed not only the presence of well dispersed CNTs but also needle-like shape aluminum carbide (Al4C3)crystalsinthe Almatrix. The formation of Al4C3 was suggested as the interaction between the outer shells of CNTs and the Al matrix during MA process in which crystallization took place after the sintering process. The mechanical behavior of composites was evaluated by Vickers micro hardness measurements indicating a significant improvement in hardness as function of the CNT content. This improvement was associated to a homogeneous dispersion of CNTs and the presence of Al4C3 in the aluminum alloy matrix.
1. Introduction
Metal matrix composites (MMC) are an important group of structural materials that can be used in order to fulfill some important requirements from the modern industry such as low density, high specific strength, high modulus, increased wear resistance, higher service temperature and in general, a better mechanical performance compared with conventional monolithic materials. Al-based composites are an important area to study due to their attractive properties for diverse industrial and potential applications in the field of MMC.
In this matter, CNTs [1] emerged as materials with outstanding properties exceeding those of any conventional material that make CNTs ideal candidates to reinforce composite materials in order to increase both stiffness and strength [2–5]. There are certain key requirements to be fulfilled as a uniform dispersion of the CNTs in the Al matrix in order to achieve an optimum mechanical performance and good interfacial bonding between the reinforcement and the Al metal matrix. Because of this, MA is a technique frequentlyusedtointegrateaveryfine distribution of hardening particles into the metal matrix by solid-state powder processing [6,7], which is difficult to be obtained in the traditional way of liquid metallurgy. Additionally, a high dislocation density and small a sub grain size in the matrix can be obtained, resulting in an improvement on the final composite properties. MA starts with a solid-state, high energy ball milling of the matrix powder with the hardening phases at room temperature producing a homogeneous composite with a fine microstructure and a good distribution of dispersed particles.
CNT dispersion for production of composites after progressive milling time has been reported, indicating, by scanning electron (SEM) micrographs, that the CNT appears with no damage after being milled after 48 h [8]. However in the case of CNT/Al composites, being produced with high energy mill, a notable damage of the CNTs has been observed after 30 h of milling with the transmission electron microscopy (TEM) [9].Inbothcases,the formation of compounds from CNTs and Al in the as-milling condition has not been observed.
Therefore, it is important to investigate the effect of milling time and CNT concentration in composites produced by MA after sintering condition by studying the effect on the microstructure and mechanical behavior of a homogeneous dispersion of CNTs into an Al2024 as well as the Al4C3 formation and their dependence of milling time for its formation.
2. Experimental
The 2024 aluminum alloy (Al2024) powder having the chemical composition of (in wt.%) 4.5 Cu, 1.5 Mg, 0.6 Mn, 0.15 Ti and 0.25 Zn; balance Al, was produced from elemental powders and CNTs were used as the starting materials to produce CNT/Al2024 composites by MA. Fig. 1 shows a secondary electron micrograph obtained by field emission scanning electron microscopy (FESEM) of a bundle dense array of CNTs used in the CNT/Al2024 composites. CNTs of about 80 nm in diameter can be observed.
CNTs and the Al2024 powder were placed in a vial allowing the dispersion of CNTs in the Al2024 matrix since the beginning of the MA process. The vial was made from hardened steel (D2) and the milling media was made from stainless steel. The fabrication of composites with MA of raw materials was performed by using a high energy mill Spex 8000. CNTs were added in several compositions (0.5, 1.0, 2.0, 3.0, 4.0, and 5.0 wt.%). The powder mass was 8.5 g and the ball-to-powder ratio was of 5:1.
All milling runs were performed with methanol as a process control agent (PCA). Argon was used as the inert milling atmosphere. Selected milling times were of 5, 10, 20 and 30 h. The identification of each composite as a function of CNT content and milling time is presented in Table 1. Milled powders were consolidated to form discs of 5 mm in diameter and 1 mm in height, under ~1500 MPa and then pressure-less sintered under Ar atmosphere during 2 h at 500 °C with a heating and cooling rate of 5 °C/min. An unreinforced Al2024 (0.0 wt.% of CNTs) for each milling time was prepared under the same MA conditions.
The crystallization behavior occurred after the sintering process was studied by means of X-ray diffraction (XRD) in a Panalytical X'Pert PRO with X'Celerator detector. The micro-structure study of the sintered samples after metallographic preparation was performed by SEM; a JEOL model JSM-7401F equipment was used. In addition, SEM analysis was made in a fracture induced in each of the specimens with the highest CNT content (5.0 wt.%), and for each of the selected milling times above displayed (5 and 30 h). Fractographies show the mechanical bond between the matrix and the individual CNT after the MA process. TEM observations were carried out in sintered products in a JEOL JEM 2200FS. Specimens for TEM were prepared focusing ion beam (FIB) in a JEOL model JEM-9320FIB. The mechanical performance of the products was evaluated with the Vickers microhardness test. The average values of at least five points of the randomly selected regions in each sample are reported.
3. Results
XRD patterns acquired from the sintered products are presented in Fig. 2. Fig. 2a shows XRD analysis reference alloys as a function of the milling time. Fig. 2b shows XRD an analysis of composites reinforced with 5.0 wt.% of CNT as a function of the milling time. The presence of Al and Al2Cu peaks can be observed for alloys and composites where composites present an additional phase identified as Al4C3.Under thesamemilling time, the reference alloys (Fig. 2a) show very sharp peaks in comparison with the composites (Fig. 2b). This increase in the broadening of the Al peaks of the composites can be attributed to the micro-strain caused in the Al lattice caused by its distortion due to the additional presence of CNTs and Al4C3.
It has been observed before that Al and Cu do not form a compound when composites are analyzed in the as-milling condition [9]. However, the formation of Al2Cu phase, which is prompt to be formed during sintering and growing during the cooling step, can be observed from the XRD analysis when the products are slowly cooled from 500 °C to room temperature at 5 °C/min. Changes in the signal intensity of this phase seem not to be affected by variations in the milling time. However, an expected decrease in their intensity by comparison with the reference alloys (Fig. 2a) is observed due to the presence of the Al4C3 phase produced from the Al and the CNTs as can be observed in Fig. 2b.
In addition, the reaction between CNTs and Al to be crystallized in the Al4C3 phase during sintering process is observed. This phase has been previously reported by otherauthors, in the synthesis of CNT/Al composites using different routes [10,11]. In this aspect, the formation of the carbide has been reported in CNT/Al composites produced by powder metallurgy routes at processing temperatures of 656 °C [11]. However, at lower temperatures of processing (500 °C) the formation of the carbide when the composite is produced by a combination of powder metallurgy routes and a low energy mill for the CNTs dispersion, has been notobservedas itwas reported by Esawi et al. [12]. They suggest that the formation of the Al4C3 compound has a strong dependence of the processing temper-ature. Nevertheless, the use of 500 °C as the sintering tempera-ture suggests that the formation of the Al4C3 phase mainly depends on the processing conditions of the fabrication of the composite due to the use of high energy mills. In this research, the increase in the milling time seems to have no considerable effect on the intensity of aluminum carbide signals even though CNTs were undergoing for 10, 20 and 30 h of milling as it is observed in Fig. 2b. A semi-quantitative analysis in composites with 5.0 wt.% of CNTs for all selected milling runs indicates the presence of about 9.0 wt.% of Al4C3 for the A50 composite and 9.2 wt.% in the D50 composite, which indicates that the amount of the aluminum carbide does not present noticeable changes due to the increase of the milling time for the fabrication of the composites. A previous investigation indicated that CNTs are being damaged as the milling time is being increased, leading as a result the formation of an amorphous carbon layer from the outer walls of the nanotubes as the milling time was increased [9]. However, the increase in the presence of amorphous carbon seems not to favor the increase in the formation of aluminum carbide phase after sintering. All these suggest that the reaction between nanotubes and the aluminum matrix during their early interactions in the mechanical milling process sets the amount of material to be crystallized in aluminum carbide during the sintering process. It can be seen that the aluminum carbide phase is stable under a fixed amount of CNTs, even though variations in the milling time are used.
Although, the maximum amount of 5.0 wt.% of CNTs was used in the synthesis of CNT/Al2024 composites, there are no carbon reflections presented from X-ray spectra, independently from the milling time. This is associated with a homogeneous dispersion of the CNTs into the Al2024 matrix during the milling process causing a difficulty to detect the CNT signal by XRD when dealing with structures [7].
Fig. 3 displays backscattered SEM micrographs of the cross-section of sintered products. Fig. 3(a,b) corresponds to the unreinforced A00 and D00 alloys respectively. Fig. 3(c,d) shows the A50 and D50 composites respectively. A noticeable increase in the porosity (shown in black areas) can be observed in the D50 alloy in comparison with the A50 alloy. The porosity in composites does not display notable variations by comparing them with their respective references. The increase in the porosity can be explained due to the natural hardening effect caused by the deformation and the presence of harder particles as the milling time is being increased, which leads to a decrease in the product compressibility during the consolidation process. These observations indicate that the densification of the composites is not limited when the CNT concentration in the composites is up to 5.0 wt.%, and it is mainly due to the milling time employed in the synthesis of the composite [4]. In addition to this, bright areas of two different morphologies can be observed in alloys and composites. Fig. 3(e–h) shows elemental mapping of the A50 composite; SEM analysis on the microstruc-ture reveals dark areas consisting in almost pure Al whereas bright areas contain Cu-rich phases. Al–Cu phase is observed in conglomerates of small particles of irregular shape unlike the round shape of Al–Mn–Cu phase whose behavior during sintering stage resulted in isolated and well located crystals. These phases appear homogeneously distributed for both alloys and compos-ites and they are indicated in all micrographs by a square. The formation and distribution of these phases seem to be indepen-dent from the milling time and CNT content. No clusters of CNTs were observed in the microstructure of the composites which indicates their well dispersion in the matrix. The formation of Al4C3 was not observed as part of the microstructure in composites under these conditions of analysis.
Induced fractures in composites are shown in Fig. 4 by secondary electron SEM micrographs. In the present cases, surfaces of 5.0 wt.% CNT reinforced composites milled for 5 (A50)and30(D50) after sintering are presented. It is interesting to note the ductile fracture in nature shown at microscopic scale examination by the composites under both milling conditions. The presence of a ductile behavior indicates a strong bond between the matrix particles. CNTs embedded in the matrix for A50 and D50 composites are observed (marked with white arrows). It is important to note that a cluster of CNTs was not observed in the fracture surface of the composites which corroborates a well dispersion resulted from MA process. Under this condition of examination CNTs appear with no visible damage caused by the MA process as it was reported by Esawi and Morsi [8] in their study of CNT/Al composites in the as-milled condition. There-fore, it can be observed how the subsequent sintering process does not produce a visible damage in the morphology of the CNTs by comparing them with the as-received CNTs shown in Fig. 1 according to SEM observations. The increase in the damage oftheCNTsafter the progressive millingtimecouldlead tothe formation of a strong CNT–Al bond due to the intercalation of the outer walls of the CNTs with the lattice of the Al during the continuous cycle of fracture-welding of particles in MA process. Micrographs show embedded ends of CNTs, which randomly appear oriented. No voids around CNTs are observed, which is something that suggests a good mechanical bond between the Al matrix and the nanotubes and the fracture of the nanotubes under extreme tension produced by the matrix particles to be detached from other particles.
Fig. 5 shows the results of the microstructural examination carried out by TEM in the A50 composite. Bright field TEM micrographs show the presence of isolated CNTs in three different zones as it can be observed in Fig. 5a–c. A nanobeam electron diffraction (NBD) patterns (selected area indicated by circles) taken in two of the three zones of the composite (Fig. 5a and b), show the corresponding spots to the (0 0 2) plane of the CNTs. The inner hollow space characteristic of the nanotube structure can be observed in Fig. 5b. The fringes in the inner space indicate well defined cylindrical graphitic layers with crystalline structure. Measurements in the interplanar distance of the CNT walls (inset of Fig. 5c) indicate a value of 0.379 nm according to the(00 2)planeofthe CNTs. However, close‐up to the outer walls of a CNT into the matrix is observed in the inset of Fig. 5c that indicates a slight damage in the outer walls of the CNTs that can be appreciated by a white arrow. This effect in the outer walls of the nanotube could have been caused during their synthesis by CVD method. However, it is possible that this damage could have been caused during the integration of CNTs into the aluminum matrix due to the collision between particles occurred during MA process. It can be noticed how somepartsof the outer walls of the nanotube have almost been completely detached. The discontinuity in the lattice structure of the outer walls of the CNTs suggests a strong interaction with the Al matrix.
These observations indicate that the mechanical CNT–Al bonds are strengthened as the milling time increases due to a steadily increased intercalation between the Al lattice fringes and the outer walls of the CNTs. However, the correct selection in the milling time must be taken into account in order to prevent the total destruction of the CNTs due to the accumulated damage during the MA process.
The presence of Al4C3 detected by XRD (Fig. 2b) is corroborated by TEM analysis by means of STEM micrographs in Z-contrast mode. The inset of Fig. 5d shows an ion-induced secondary electron (ISE) image of the cross-section of an indentation mark left during the hardness evaluation of the composite. Fig. 5d shows the deformed zone of analysis beneath the indentation mark. A needle-like shape aluminum carbide crystal into the aluminum matrix embedded into the aluminum matrix is observed in Fig. 5e in interaction with dislocations. The presence of CNTs and Al4C3 impeding the movements of dislocations (Orowan's mechanism) is considered as one of the main strengthening mechanisms in the improved mechanical behav-ior of CNT/Al composites produced by MA. A more detailed view of the aluminum carbide phase is displayed in Fig. 5f. Aluminum carbide lattice fringes measured with an interval of 0.83 nm are shown. The formation of this aluminum carbide is not clear at this moment and its nature is directly related with the production route of CNT/Al composites as it has been described by other authors [11,13]. Nevertheless, the closest theory in CNT/Al composites synthesized by MA is based on the loss of the periodicity of outer shells of CNTs during the initial milling stages, which have a reaction with Al matrix. However, there is no evidence of the possible co-existence of CNT–Al4C3.Shorter milling times and shorter sintering times could be useful to obtain evidence of this theory.
The influence of CNT concentration and milling time on the microhardness (Vickers scale) behavior in sintered products is given in Fig. 6. In order to compare hardness behavior, themicrohardness values of the commercial Al2024 Al alloy under different heat treatment conditions are included [14].The change in hardness is given as a function of CNT concentration. The curves indicate that the hardness in the composites reinforced with CNTs presents a rapid increase with the fraction of reinforcement up to 5.0 wt.%, which represents the maximal CNT concentration used in the composites. However, a longer milling time seems to have no considerable effect on the hardness behavior of the products. For unreinforced alloys, higher value is reached for A00 with 85.9 hardness units over B00 with 67.07 hardness units. The maximum hardness is achievedwith the B50 composite with 20 h of milling and 290.9 hardness units. The strongest composite (B50) represents an increase of ~285% over the unreinforced alloy milled with the same time (B00). As a comparison, hardness values for pure aluminum and the Al2024 commercial alloy with different heat treatments are displayed in Fig. 6. It is important to note that the hardness values showed by the commercial alloy after the heat treatment, are almost equal to the composites with 2.0 and 3.0 wt.% of CNTs. Furthermore, the addition of 4.0 and 5.0 wt.% of CNTs shows an important increase compared with the commercial alloy. These results open the possibility to obtain higher hardness values in composites after T6 temper (solution and artificial aging).