- Email: [email protected]
aluminum laminate composites with ultrahigh thermal conductivity for thermal management applications
Materials and Design 90 (2016) 508–515
Contents lists available at ScienceDirect
Materials and Design journal homepage: www.elsevier.com/locate/jmad
Graphite film/aluminum laminate composites with ultrahigh thermal conductivity for thermal management applications Yu Huang, Qiubao Ouyang ⁎, Qiang Guo, Xingwu Guo, Guoding Zhang, Di Zhang State Key Laboratory of Metal Matrix Composites, Shanghai Jiao Tong University, Shanghai 200240, PR China
a r t i c l e
i n f o
Article history: Received 12 August 2015 Received in revised form 23 October 2015 Accepted 27 October 2015 Available online 28 October 2015 Keywords: Graphite films Aluminum composites Ultrahigh thermal conductivity Thermal management
a b s t r a c t In order to meet the requirements in thermal management, a novel kind of carbon/metal composites, namely graphite film/aluminum laminate composites with excellent thermal properties, was developed. The effects of oxide films on interfacial structure and properties of the composites were identified, and the microstructures and properties of the composites were studied as a function of the volume fraction of graphite films. An ultrahigh thermal conductivity of 902 W/mK for aluminum matrix composites was obtained. The measured in-plane thermal conductivities of the composites are all over 80% of the rule of mixtures prediction. What's more, these composites are shown to have more competitive thermal conductivities as compared to other kinds of carbon/aluminum composites, which makes these composites promising candidates to be used for thermal management. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Effective thermal management is becoming increasingly important for the reliability of electronic components owing to continuous developments of the electronic industry [1–3]. Therefore, it is essential to develop thermal management materials with high thermal conductivity (TC) and low coefficient of thermal expansion (CTE) [1,4]. Carbon materials, such as graphite, diamond, carbon fibers (CFs), carbon nanotubes (CNTs) and graphene, have proved to be promising thermal management materials due to their excellent thermal properties [2,5–7]. These materials can not only be directly used in thermal management applications, but also combined with other materials, such as metals, to form thermal management composites [8]. These composites offer the possibility of tailoring the properties of a metal by adding an appropriate carbon reinforcement to meet the requirements of thermal management. However, the poor wettability and possible harmful interfacial chemical reactions between carbon materials and metals would decrease the reinforcement effect of the carbon materials [9–11]. Therefore, the fabrication process of these composites should be well-designed and well-controlled in order to obtain carbon/metal (C/metal) composites with fine microstructures and high thermal properties [9–11]. To date, diamond/metal composites, CF/metal composites and graphite flake/metal composites are the mostly studied C/metal composites that can be used in thermal management applications. Diamond/metal composites have high TC ranging from 350 W/mK to 780 W/mK, and have been commercialized [8,9,12]. Unfortunately, ⁎ Corresponding author. E-mail address: [email protected] (Q. Ouyang).
http://dx.doi.org/10.1016/j.matdes.2015.10.146 0264-1275/© 2015 Elsevier Ltd. All rights reserved.
their high-cost and poor machining properties strongly limit their application [8]. Due to the anisotropic thermal properties of CFs and graphite flakes, excellent properties are only achieved in one or two axial directions of their composites [8]. For example, if carbon fibers with high TC (also high price) were used, the TC of long carbon fiber (LCF)/metal composites can reach 500–700 W/mK along the axial direction while the TC of short carbon fiber (SCF)/metal composites are 208 W/mK in the in-plane direction [13,14]. Also, as reported by Chen et al. [15], the in-plane TC of the graphite flake/metal composites can reach 324 W/mK to 783 W/mK. However, it is difficult to control the exact direction of the graphite flakes as well as the infiltration process [8,15]. In order to promote infiltration, other reinforcements such as SiC [8,16], Si [10] and CF [8] were often added into graphite flake/ metal composites, which, unfortunately, would inevitably lead to a reduction in the TC of the resulting composites. Recently, CNT/metal composites and graphene/metal composites have drawn much attention in thermal management due to outstanding thermal properties of CNTs and graphene, while, unfortunately, it has been found that the improvement of thermal properties of these composites is hindered by many technical problems, such as the control of the distribution and orientation of the reinforcements and high interfacial thermal resistance [11]. Recently, graphite films, such as artificial graphite films fabricated from polyimide films [7,17] and graphene oxide films [2,18], have attracted much attention in thermal management due to their excellent thermal properties. Specially, artificial graphite films with high TC (1100–1600 W/mK) have become commercially available for heat removal in electronic devices containing integrated circuits, such as personal computers and mobile phones [7]. Unfortunately, so far, little attention has been paid to the potential of these graphite films being used as effective reinforcements in composites. Actually, if these
Y. Huang et al. / Materials and Design 90 (2016) 508–515
graphite films are used as reinforcements to form graphite film/metal composites and fabrication processes of the composites are welldesigned and well-controlled, high TC and low CTE can be expected. Furthermore, if we use graphite films and metal foils as raw materials to fabricate graphite film/metal laminate composites, we can easily control the distributions and orientations of the reinforcements by controlling the stacking of graphite films and metal foils, which is easier and more effective as compared to CF/metal composites, graphite flake/ metal composites, CNT/metal composites and graphene/metal composites, which may simplify the fabrication process and meanwhile obtain composites with higher thermal properties. In this study, we successfully fabricated these novel graphite film/ aluminum laminate composites with high TC and low CTE by a vacuum hot pressing process. The effects of oxide films on the surface of aluminum foils and volume fraction of graphite films on microstructures and properties of the composites were investigated and discussed. The microstructures of the composites were characterized at length scales from the macro down to the nanoscale to investigate the distribution and orientation of the graphite films, and the graphite/Al interfacial structure and the presence of harmful reaction product. Furthermore, the thermal conductivities of the as-fabricated composites were compared with the predicted values calculated by rule of mixture and results of other kinds of carbon/aluminum (C/Al) composites reported in the literatures. 2. Experimental Graphite films and pure aluminum foils (99.9% in purity) were used as raw materials. The graphite films were acquired from Dasen Electronic Material Co., Ltd. in China. The morphologies of the graphite films are shown in Fig. 1 (a). These graphite films have a rough surface with many ripples on it. The inset shows the cross-section of the graphite films. The average thickness of the graphite films was measured to be 29.5 μm. Fig. 1 (b) shows the X-ray diffraction (XRD) pattern of the graphite films, where only peaks of graphite can be detected from this pattern, and the Kα1 and Kα2 peaks of (006) peaks are separated very well, which indicates that the graphite films have high purity and crystallinity [19]. The in-plane and out-of-plane TC values of the graphite films were measured to be 1222.6 (± 30.3) W/mK and 17.5 (± 2.5) W/mK, respectively. Graphite film/aluminum laminate composites were fabricated by a vacuum hot pressing process. Four sets of composite samples with different volume fractions of graphite films were obtained by controlling the thickness of the aluminum foils. The volume fractions of graphite films of sample #1, #2, #3 and #4 are 17.4%, 36.2%, 53.2% and 69.4%, respectively. The detailed fabrication processes of these laminate composites are listed as following: 1). Graphite films and aluminum foils were
509
washed by acetone and then dried. 2). In order to eliminate the oxide films on the surface of the aluminum foils, they were first washed by 15 g/L NaOH solution, then washed by HNO3 with volume fraction of 25%, and finally washed by alcohol and dried. 3). The pre-treated graphite films and aluminum foils were carefully filled into a graphite mold layer-by-layer and then heated to 655 °C in high vacuum condition (less than 5 × 10−3 Pa) and kept for 100 min, while an uniaxial pressure of 22 MPa was applied. In order to figure out the effect of oxide films on microstructures and properties of the composites, a composite sample named #3′ was fabricated without the oxide film eliminating process (process 2)). Since the thickness of the oxide films (1–5 nm) [20,21] is much smaller than that of the aluminum foils (26 μm), the effect of them on volume fraction can be neglected. Thus, the graphite volume fraction of sample #3′ is considered to be the same with sample #3. The in-plane thermal diffusivities of the graphite films (Φ25.4 mm) and the composite samples (10 × 10 × 4 mm) at room temperature were measured by a laser flash technique using a NETZSCH LFA447 thermal analyzer. The specific heat capacities of the samples at room temperature were measured by PerkinElmer differential scanning calorimeter (DSC) 8000. A speed of 20 °C/min was used in the temperature range of 15–35 °C. The densities of the samples were measured by the Archimedes method. The in-plane TC values of the samples were calculated by the product of the density, thermal diffusivity and specific heat capacity. The in-plane CTE values of the specimens (10 × 2 × 25 mm) were measured by a dilatometer (NETZCH DIL 402C) from room temperature up to 300 °C at a speed of 5 °C/min. XRD patterns of the graphite films were obtained by a D/max-2550 instrument (Cu Kα). A scan speed of 4°/min was used in the range of 10–90°. Microstructures of the graphite films and composite samples were characterized by scanning electron microscopy (SEM) operated at 20 kV using a FEI Quanta FEG 250 electron microscope and transmission electron microscope (TEM) using a JEOL 2100F Field Emission Electron Microscope. The energy dispersive spectroscopy (EDS) element line scanning across the aluminum-graphite interface was characterized using the aforementioned SEM. 3. Results and discussion 3.1. Effect of oxide films on microstructures and properties of the composites According to the results reported in graphite fiber/aluminum composites [22] and diamond/aluminum composites [23], during the fabrication process of the composites, diffusion may happen between aluminum and carbon materials to form a strong interfacial bonding. However, in this study where aluminum foils are used for raw material, there are thin oxide films with thickness of 1–5 nm on the surface of aluminum foils due to their exposure to air [20,21], which would generally
Fig. 1. (a) Morphologies of the surface and cross-section (inset, between the two arrows) of the graphite films; (b) XRD patterns of the graphite films.
510
Y. Huang et al. / Materials and Design 90 (2016) 508–515
Fig. 2. Schematic diagram of the effect of oxide films on fabrication of the composites: (a) without the oxide film eliminating process; and (b) after eliminating the oxide films.
act as diffusion barriers [24,25] to impede the interfacial diffusion between aluminum and carbon during the hot pressing process, as shown in Fig. 2 (a). Therefore, if these oxide films are eliminated before the hot pressure process of the composites, it may be easier for the diffusion to take place to form an interface with good interfacial bonding, as illustrated in Fig. 2 (b). To evaluate the effect of the oxide films, the microstructures of the interfaces of the composite sample #3′ and #3 were examined and are shown in Fig. 3 (a) and (b), respectively. Interfacial debonding is much easier to occur at the interface of sample #3′ after mechanical polishing, which may indicate that the interface with oxide films would result in a weaker interfacial bonding strength. In contrast, if the oxide films are eliminated, fine interface with no interfacial debonding will be obtained as shown in Fig. 3 (b). In order to have a better understanding of the reason for that interfacial debonding, the elemental line distribution across typical interface of sample #3′ and #3 were measured by EDS and the results are shown in Fig. 4. As shown in Fig. 4 (c), a decaying concentration profile of carbon atoms across the C/Al interface was found, indicating that the diffusion of carbon into aluminum may take place during the processing of the composites, most likely during the high temperature (655 °C) step, which led to a strong adhesion at the interface of sample #3. More specifically, the diffusion speed k (μm/min) of carbon in aluminum can be calculated by the Arrhenius diffusion function [26]: k = A exp (− Qa/RT), where A is a temperature-independent pre-
exponential, Qa is the activation energy for diffusion, R is the gas constant and T is the absolute temperature. Referring to the parameters reported in reference [26], the diffusion speed k of carbon in aluminum at the fabrication temperature of 655 °C (928 K) can be calculated to be ~ 2.2 × 10−2 μm/min. By multiplying time that the processing step lasted (100 min), the diffusion range of carbon atoms into aluminum matrix can be estimated to be about 2.2 μm, which is consistent with the range of carbon concentration profile found experimentally (2–3 μm, Fig. 4 (d), for sample #3 without the oxide films). On the other hand, affected by the oxide films, the range of carbon distribution on the aluminum side in sample #3′ was found to be less than 1 μm (Fig. 4 (a) and (b)). By comparing the ranges that carbon atoms diffused across the C/Al interface into the aluminum side, we may come to the conclusion that the oxide films served as an effective barrier for carbon diffusing into aluminum, giving rise to a weak interfacial strength in sample #3′. The comparison of the thermal properties between sample #3′ and sample #3 is shown in Fig. 5 and the specific values of their thermophysical properties can be found in Table 1. Due to the weaker interfacial bonding and existence of interfacial debonding as illustrated in Figs. 3 and 4, the in-plane TC of sample #3′ is measured to be much lower than that of sample #3 and the in-plane CTE of sample #3′ is much higher than that of sample #3, which indicates that the composite sample fabricated with the oxide films eliminating process have much better thermal properties than sample fabricated without this process. Therefore, the oxide films on the aluminum foils may act as diffusion barriers to impede the interfacial diffusion between carbon and aluminum, which leads to formation of interface with weak interfacial bonding strength. Interfacial debonding would easily take place at this kind of interface, which results in relatively low thermal properties. By eliminating these oxide films, interfaces with strong interfacial bonding strength will be obtained and interfacial debonding will be restrained, which leads to an improvement in thermal properties of the composites. 3.2. Effect of volume fractions of graphite films on microstructures and properties of the composites Four sets of composites with different volume fractions of graphite films were studied in this study to figure out the effect of volume fractions of graphite films on microstructures and properties of the composites. Fig. 6 shows the microstructures of the composite samples and it is obvious that the volume fraction of the graphite films increases from (a) to (d). As expected, these composites have lamellar microstructures. The graphite films are highly homogeneous and the orientations of these films are well controlled. Therefore, the distribution and
Fig. 3. Microstructures of the composite samples fabricated by different fabrication process: (a) sample #3′ fabricated without the oxide film eliminating process; and (b) sample #3 fabricated with the oxide film eliminating process.
Y. Huang et al. / Materials and Design 90 (2016) 508–515
511
Fig. 4. Microstructure and elemental line distribution results across typical interface of (a–b) sample #3′ and (c–d) sample #3.
orientation control of the reinforcements in graphite film/aluminum laminate composites are much more effective as compared to other C/Al composites, such as CF/metal composites [13], graphite flake/ metal composites [15], CNT/metal composites and Graphene/metal composites [11], which may lead to high thermal properties. In order to have a better understanding of the interfacial bonding between graphite films and aluminum, TEM was used to characterize the interfacial structure of the composites. The results are shown in Fig. 7. Fig. 7 (a) shows the overall view of the aluminum matrix, graphite films and their interface while Fig. 7 (b) gives a high magnification image of the interface. The inset on the bottom left of Fig. 7 (b) shows the indexed [233] zone axis selected area diffraction (SAD) pattern of the aluminum. The inset on the top right of Fig. 7 (b) shows the SAD pattern of graphite films, where (002), (100), (004) and (006) diffraction
Fig. 5. Comparison of the CTE and TC values between composite sample #3′ and sample #3.
rings can be identified. The opening angle of the (002) arcs (marked by small circles) is very small, indicating that the graphite planes in the graphite films are well-oriented [27]. A clear row of spots (marked with discontinuous lines) can be detected from the diffraction pattern, which implies the existence of preferred orientation in the selected area of graphite films [27]. As shown in Fig. 7, the composites have a clear interface and no reaction products can be detected along the whole interface. Typical high resolution TEM (HRTEM) image of the interface of the composites is shown in Fig. 8 (a). Indexed Fast Fourier Transform (FFT) results of some selected areas of graphite films and aluminum are shown in Fig. 8 (b), (c), (d) and (e). A thin interfacial layer, with amorphous-like structure and thickness of nearly 2 nm, was observed to tightly adhere to the aluminum matrix. Its FFT pattern (Fig. 8 (b)) indicated that it is likely to have amorphous structure. Such an amorphous interfacial layer is frequently observed in carbon (in the form of graphite or diamond)/aluminum composites such as graphite flakes/aluminum composites and diamond/aluminum composites. For example, Zhou et al. [10] reported that there exists an amorphous graphite layer at the interface of graphite flake/aluminum composites. Beffort et al. [28, 29], Khalid et al. [30] and Jiang et al. [31] also found an amorphous layer at the interface of diamond/aluminum composites. Because of the low thickness of the amorphous layer, it was difficult to accurately determine its composition. Fortunately, Beffort et al. [29] found a thicker amorphous layer (about 50–200 nm) in diamond/ aluminum composites fabricated at 770–800 °C and measured its chemical composition by EDS. The interfacial layer was found to be mainly comprised of C, Al and O [29]. It was found that O may come from the oxide films of the aluminum melt during the fabrication process or from the raw carbon material, and these authors proposed that carbide-forming elements (such as aluminum) are particularly active in catalyzing the reconstruction of the diamond surface to form an amorphous layer [29]. Therefore, it may be concluded that the
512
Y. Huang et al. / Materials and Design 90 (2016) 508–515
Table 1 Thermo-physical properties of graphite film/aluminum laminate composites with different volume fractions of graphite films. Sample
Thickness design (μm)
VG (%)
ρ (g/cm3)
Cp (J/g/K)
αa (mm2/s)
TCa (W/mK)
CTEa (ppm/K)
#1 #2 #3 #3′ #4 Al Gr film
Al (140) + Gr film (29.5) Al (52) + Gr film (29.5) Al (26) + Gr film (29.5) Al (26) + Gr film (29.5) Al (13) + Gr film (29.5) – –
17.4 36.2 53.2 53.2 69.4 – –
2.54 2.39 2.25 2.23 2.10 2.7 1.96
0.836 0.818 0.798 0.798 0.746 0.88 0.709
167.445 262.133 409.594 377.487 575.701 99.7 879.8
356 512 735 672 902 237 [14] 1222.6
20.6 10.58 7.22 11.94 3.55 23.2 [14] −1.5 [10]
a
The TC and CTE of graphite film/aluminum laminate composites are all anisotropic, and here we only list the values along the in-plane directions.
amorphous interfacial layer found in our study may come from a similar origin, i.e., aluminum element catalyzes the reconstruction of graphite film surface to form an amorphous layer, and the composition of this amorphous layer is likely to contain C, Al and O. Also, as shown in Fig. 8 (a), the lattice structure of the crystalline graphite adjacent to this amorphous layer is distorted to some extent, which is probably caused thermal stress between aluminum matrix and graphite film [32]. As a result, according to the lattice structure illustrated by the HRTEM, the as-fabricated composites have fine interface with good interfacial bonding and free of harmful reaction product. Table 1 reports the thermo-physical properties of the composite samples and raw materials. As the volume fraction of the graphite films increases from 17.4% to 69.4%, the in-plane TC of composites increases from 356 to 902 W/mK and the in-plane CTE of the composites decreases from 20.6 to 3.55 ppm/K. Therefore, the addition of graphite films has brought significant improvement of thermal properties to
the aluminum matrix. However, similar to graphite flake/aluminum composites [15], if the volume fraction of the graphite films is too high, some local regions in the composites would be short of aluminum during the fabrication process which would lead to formation of pores. These pores will inevitably degrade the properties of the composites. Therefore, the volume fraction of the graphite films is limited to 70% in this study in order to guarantee stable properties. In a word, graphite film/aluminum laminate composites with tailorable and stable high thermal properties were fabricated in this study and these composites have a great potential for applications in thermal management. 3.3. Comparisons with predictions and other kinds of C/Al composites For laminate composites, their theoretical in-plane TC can be simply calculated using rule of mixture as in this case no interface effects are envisaged [4,14]. The material parameters for the calculation can be
Fig. 6. Microstructures of the composite samples: (a) sample #1 with volume fraction of 17.4%; (b) sample #2 with volume fraction of 36.2%; (c) sample #3 with volume fraction of 53.2%; (d) sample #4 with volume fraction of 69.4%. Gr represent graphite.
Y. Huang et al. / Materials and Design 90 (2016) 508–515
513
Fig. 7. Typical TEM image of the interface of the composites: (a) overall view; and (b) Al/Graphite interface. The two insets are the SAD patterns of the graphite and aluminum.
found in Table 1. Fig. 9 (a) shows the experimental and predicted inplane TC of the composite samples. The four samples with different volume fractions all exhibit in-plane TC over 80% of the predictions. Moreover, samples with high volume fraction, such as 53.2% and 69.4%, exhibit high in-plane TC over 90% of the predictions. Therefore, the graphite films can act as effective reinforcements in aluminum matrix composites. The deviation between the experimental results and predictions is probably due to the existence of interfacial thermal resistance [9,14]. The out-of-plane thermal conductivity of laminate composites can be estimated by the equation given by reference [16], assuming an aluminum/graphite interfacial thermal conductance of 4.5 × 107 W/m2K [10]. Calculation revealed that as the volume fraction of the graphite films increases from 17.4% to 69.4%, the calculated out-of-plane thermal
conductivity would decrease from 73 to 23.8 W/mK, which is much less than the in-plane thermal conductivities as shown in Table 1 (from 356 to 902 W/mK). Therefore, the thermal conductivity of the as-fabricated composites is highly anisotropic. If these composites are used in heatremoving applications, heat would be much more quickly removed by being conducted along the in-plane direction rather than the out-ofplane direction, and thus the in-plane thermal conductivity dominates the heat-removing process and subsequently the overall thermal conductivity of the composite. As a result, we will only focus on the in-plane thermal conductivity in the following discussions of this paper. Fig. 9(b) shows the comparison of the TC of the graphite film/ aluminum laminate composites fabricated in this study to the TC of other kinds of C/Al composites that can be used in thermal management. In most cases, graphite film/aluminum laminate composites
Fig. 8. Typical HRTEM image of the interface of the composites: (a) HRTEM image; (b) FFT results of the selected area of amorphous layer; (c–d) FFT results of the selected area of crystalline graphite; (e) FFT results of the selected area of aluminum.
514
Y. Huang et al. / Materials and Design 90 (2016) 508–515
Fig. 9. Thermal conductivities of the graphite film/aluminum laminate composites with different volume fractions of graphite films fabricated in this work: (a) compared with the predicted values calculated by rule of mixture; (b) compared with other kinds of thermal management C/Al composites fabricated in the reference, as a function of volume fraction of carbon. Gr represent graphite. Graphite flakes and SiC (or Si) reinforced aluminum composites are treated as C/Al composites due to the fact that graphite flake is the main reinforcement and SiC (or Si) is used to produce space for infiltration [8,15,16].
fabricated in this study exhibit higher in-plane TC than that of the C/Al composites reported in these references. This is likely caused by the following facts: 1). The measured in-plane TC of the starting graphite films is very high, exceeding 1200 W/mK, which makes graphite films promising reinforcements for thermal management composites. 2). For graphite film/aluminum laminate composites, it is very easy to control the orientation and distribution of the graphite films by controlling the stacking of graphite films and metal foils. Therefore, it is easier to bring the enhancement effect of the graphite films into full play than other kind of reinforcements such as graphite flakes and CFs. 3). Graphite film/aluminum laminate composites fabricated in this study have fine microstructures and interfaces as shown in Figs. 6, 7 and 8, which makes the TC of the composites approach the theoretical values. 4. Conclusions In conclusion, a novel kind of carbon/metal composites, namely graphite film/aluminum laminate composites with high TC and low CTE, has been fabricated by a vacuum hot pressing process. The effects of oxide films on the interfacial structure and thermal properties of the composites were identified. The result shows that these oxide films are harmful to the interfacial bonding of the composites, which leads to interfacial debonding and low thermal properties. As the volume fraction of the graphite films increases from 17.4% to 69.4%, the in-plane TC of the composites increases from 356 to 902 W/mK and the in-plane CTE of the composites decreases from 20.6 to 3.55 ppm/K. The as-fabricated composites have fine microstructures with welldistributed and well-oriented graphite films and clean interface, which leads to excellent thermal properties. The in-plane TC values of the composites with different graphite volume fractions are all over 80% of the predictions calculated by the rule of mixture. What's more, these composites are shown to have more competitive TC as compared to other kinds of C/Al composites, which makes these composites promising candidates to be used for thermal management. Finally, it is worth mentioning that graphite films can be used as reinforcements in other matrixes such as Cu and polymer. Moreover, the morphologies of the graphite films added to the metal matrix are not limited to films, but may also be strips and flakes, which makes microstructures and properties of their composites more designable. In addition, the fabrication process is suitable for other carbon film/metal composites, such as graphene oxide film/metal composites and so on. Acknowledgments The authors would like to acknowledge the financial support of the National Basic Research Program of China (973 Program, No.
2012CB619600), the Foundation of Shanghai Science and Technology Committee of China (No. 14DZ2261200), the Shanghai Science and Technology Committee (No. 13PJ1404000), the National Natural Science Foundation of China (No. 51471190), and the National High Technology Research and Development Program (863 Program, No. 2013AA031201). References [1] D. Chung, Materials for thermal conduction, Appl. Therm. Eng. 21 (2001) 1593–1605, http://dx.doi.org/10.1016/S1359-4311(01)00042–4. [2] G. Xin, H. Sun, T. Hu, H.R. Fard, X. Sun, N. Koratkar, et al., Large-area freestanding graphene paper for superior thermal management, Adv. Mater. 26 (2014) 4521–4526, http://dx.doi.org/10.1002/adma.201400951. [3] A. Mortensen, J. Llorca, Metal matrix composites, Annu. Rev. Mater. Res. 40 (2010) 243–270, http://dx.doi.org/10.1146/annurev-matsci-070909-104511. [4] Y.-M. Chen, J.-M. Ting, Ultra high thermal conductivity polymer composites, Carbon 40 (2002) 359–362, http://dx.doi.org/10.1016/S0008-6223(01)00112–9. [5] Z. Ma, J. Shi, Y. Song, Q. Guo, G. Zhai, L. Liu, Carbon with high thermal conductivity, prepared from ribbon-shaped mesosphase pitch-based fibers, Carbon 44 (2006) 1298–1301, http://dx.doi.org/10.1016/j.carbon.2006.01.015. [6] G. Yuan, X. Li, Z. Dong, A. Westwood, Z. Cui, Y. Cong, et al., Graphite blocks with preferred orientation and high thermal conductivity, Carbon 50 (2012) 175–182, http://dx.doi.org/10.1016/j.carbon.2011.08.017. [7] M. Inagaki, Y. Kaburagi, Y. Hishiyama, Thermal management material: graphite, Adv. Eng. Mater. 16 (2014) 494–506, http://dx.doi.org/10.1002/adem.201300418. [8] R. Prieto, J.M. Molina, J. Narciso, E. Louis, Fabrication and properties of graphite flakes/metal composites for thermal management applications, Scr. Mater. 59 (2008) 11–14, http://dx.doi.org/10.1016/j.scriptamat.2008.02.026. [9] Z. Tan, Z. Li, G. Fan, Q. Guo, X. Kai, G. Ji, et al., Enhanced thermal conductivity in diamond/aluminum composites with a tungsten interface nanolayer, Mater. Des. 47 (2013) 160–166, http://dx.doi.org/10.1016/j.matdes.2012.11.061. [10] C. Zhou, G. Ji, Z. Chen, M. Wang, A. Addad, D. Schryvers, et al., Fabrication, interface characterization and modeling of oriented graphite flakes/Si/Al composites for thermal management applications, Mater. Des. 63 (2014) 719–728, http://dx.doi.org/10. 1016/j.matdes.2014.07.009. [11] S.C. Tjong, Recent progress in the development and properties of novel metal matrix nanocomposites reinforced with carbon nanotubes and graphene nanosheets, Mater. Sci. Eng. R 74 (2013) 281–350, http://dx.doi.org/10.1016/j.mser.2013.08.001. [12] K. Mizuuchi, K. Inoue, Y. Agari, Y. Morisada, M. Sugioka, M. Tanaka, et al., Processing of diamond particle dispersed aluminum matrix composites in continuous solid– liquid co-existent state by SPS and their thermal properties, Compos. Part B 42 (2011) 825–831, http://dx.doi.org/10.1016/j.compositesb.2011.01.012. [13] T. Liu, X. He, L. Zhang, Q. Liu, X. Qu, Fabrication and thermal conductivity of short graphite fiber/Al composites by vacuum pressure infiltration, J. Compos. Mater. 48 (2013) 2207–2214, http://dx.doi.org/10.1177/0021998313495750. [14] T. Ueno, T. Yoshioka, J.-I. Ogawa, N. Ozoe, K. Sato, K. Yoshino, Highly thermal conductive metal/carbon composites by pulsed electric current sintering, Synth. Met. 159 (2009) 2170–2172, http://dx.doi.org/10.1016/j.synthmet.2009.10.006. [15] J.K. Chen, I.S. Huang, Thermal properties of aluminum–graphite composites by powder metallurgy, Compos. Part B 44 (2013) 698–703, http://dx.doi.org/10.1016/j. compositesb.2012.01.083. [16] R. Prieto, J.M. Molina, J. Narciso, E. Louis, Thermal conductivity of graphite flakes–SiC particles/metal composites, Compos. Part A 42 (2011) 1970–1977, http://dx.doi.org/ 10.1016/j.compositesa.2011.08.022. [17] M. Inagaki, N. Ohta, Y. Hishiyama, Aromatic polyimides as carbon precursors, Carbon 61 (2013) 1–21, http://dx.doi.org/10.1016/j.carbon.2013.05.035.
Y. Huang et al. / Materials and Design 90 (2016) 508–515 [18] D.A. Dikin, S. Stankovich, E.J. Zimney, R.D. Piner, G.H. Dommett, G. Evmenenko, et al., Preparation and characterization of graphene oxide paper, Nature 448 (2007) 457–460, http://dx.doi.org/10.1038/nature06016. [19] T. Takeichi, Y. Eguchi, Y. Kaburagi, Y. Hishiyama, M. Inagaki, Carbonization and graphitization of BPDA/PDA polyimide films: effect of structure of polyimide precursor, Carbon 37 (1999) 569–575, http://dx.doi.org/10.1016/S00086223(98)00223–1. [20] M. Hunter, P. Fowle, Natural and thermally formed oxide films on aluminum, J. Electrochem. Soc. 103 (1956) 482–485, http://dx.doi.org/10.1149/1.2430389. [21] L. Jiang, G. Fan, Z. Li, X. Kai, D. Zhang, Z. Chen, et al., An approach to the uniform dispersion of a high volume fraction of carbon nanotubes in aluminum powder, Carbon 49 (2011) 1965–1971, http://dx.doi.org/10.1016/j.carbon.2011.01.021. [22] M. Gu, H. Yang, W. Jiang, G. Zhang, Study on the interfacial reaction product in Gr/Al composites, Adv. Compos. Mater. 5 (1996) 119–128, http://dx.doi.org/10.1163/ 156855196X00031. [23] S. Kleiner, F. Khalid, P. Ruch, S. Meier, O. Beffort, Effect of diamond crystallographic orientation on dissolution and carbide formation in contact with liquid aluminium, Scr. Mater. 55 (2006) 291–294, http://dx.doi.org/10.1016/j.scriptamat.2006.05.003. [24] T. Rajan, R. Pillai, B. Pai, Reinforcement coatings and interfaces in aluminium metal matrix composites, J. Mater. Sci. 33 (1998) 3491–3503, http://dx.doi.org/10.1023/ A:1004674822751. [25] Y. Tang, Y. Deng, K. Zhang, L. Liu, Y. Wu, W. Hu, Improvement of interface between Al and short carbon fibers by α-Al2O3 coatings deposited by sol–gel technology, Ceram. Int. 34 (2008) 1787–1790, http://dx.doi.org/10.1016/j.ceramint.2007.05. 008.
515
[26] K. Landry, S. Kalogeropoulou, N. Eustathopoulos, Wettability of carbon by aluminum and aluminum alloys, Mater. Sci. Eng. A 254 (1998) 99–111, http://dx.doi.org/10. 1016/S0921-5093(98)00759-X. [27] C. Bourgerette, A. Oberlin, M. Inagaki, Structural and textural changes from polyimide Kapton to graphite: Part I. Optical microscopy and transmission electron microscopy, J. Mater. Res. 7 (1992) 1158–1173, http://dx.doi.org/10.1557/JMR.1992.1158. [28] O. Beffort, S. Vaucher, F. Khalid, On the thermal and chemical stability of diamond during processing of Al/diamond composites by liquid metal infiltration (squeeze casting), Diamond Relat. Mater. 13 (2004) 1834–1843, http://dx.doi.org/10.1016/j. diamond.2004.04.014. [29] O. Beffort, F. Khalid, L. Weber, P. Ruch, U. Klotz, S. Meier, et al., Interface formation in infiltrated Al (Si)/diamond composites, Diamond Relat. Mater. 15 (2006) 1250–1260, http://dx.doi.org/10.1016/j.diamond.2005.09.036. [30] F. Khalid, O. Beffort, U. Klotz, B. Keller, P. Gasser, Microstructure and interfacial characteristics of aluminium–diamond composite materials, Diamond Relat. Mater. 13 (2004) 393–400, http://dx.doi.org/10.1016/j.diamond.2003.11.095. [31] L. Jiang, P. Wang, Z. Xiu, G. Chen, X. Lin, C. Dai, et al., Interfacial characteristics of diamond/aluminum composites with high thermal conductivity fabricated by squeeze-casting method, Mater. Charact. 106 (2015) 346–351, http://dx.doi.org/ 10.1016/j.matchar.2015.06.023. [32] H. Kurita, T. Miyazaki, A. Kawasaki, Y. Lu, J. Silvain, Interfacial microstructure of graphite flake reinforced aluminum matrix composites fabricated via hot pressing, Compos. Part A 73 (2015) 125–131, http://dx.doi.org/10.1016/j.compositesa.2015. 03.013.
Contents lists available at ScienceDirect
Materials and Design journal homepage: www.elsevier.com/locate/jmad
Graphite film/aluminum laminate composites with ultrahigh thermal conductivity for thermal management applications Yu Huang, Qiubao Ouyang ⁎, Qiang Guo, Xingwu Guo, Guoding Zhang, Di Zhang State Key Laboratory of Metal Matrix Composites, Shanghai Jiao Tong University, Shanghai 200240, PR China
a r t i c l e
i n f o
Article history: Received 12 August 2015 Received in revised form 23 October 2015 Accepted 27 October 2015 Available online 28 October 2015 Keywords: Graphite films Aluminum composites Ultrahigh thermal conductivity Thermal management
a b s t r a c t In order to meet the requirements in thermal management, a novel kind of carbon/metal composites, namely graphite film/aluminum laminate composites with excellent thermal properties, was developed. The effects of oxide films on interfacial structure and properties of the composites were identified, and the microstructures and properties of the composites were studied as a function of the volume fraction of graphite films. An ultrahigh thermal conductivity of 902 W/mK for aluminum matrix composites was obtained. The measured in-plane thermal conductivities of the composites are all over 80% of the rule of mixtures prediction. What's more, these composites are shown to have more competitive thermal conductivities as compared to other kinds of carbon/aluminum composites, which makes these composites promising candidates to be used for thermal management. © 2015 Elsevier Ltd. All rights reserved.
1. Introduction Effective thermal management is becoming increasingly important for the reliability of electronic components owing to continuous developments of the electronic industry [1–3]. Therefore, it is essential to develop thermal management materials with high thermal conductivity (TC) and low coefficient of thermal expansion (CTE) [1,4]. Carbon materials, such as graphite, diamond, carbon fibers (CFs), carbon nanotubes (CNTs) and graphene, have proved to be promising thermal management materials due to their excellent thermal properties [2,5–7]. These materials can not only be directly used in thermal management applications, but also combined with other materials, such as metals, to form thermal management composites [8]. These composites offer the possibility of tailoring the properties of a metal by adding an appropriate carbon reinforcement to meet the requirements of thermal management. However, the poor wettability and possible harmful interfacial chemical reactions between carbon materials and metals would decrease the reinforcement effect of the carbon materials [9–11]. Therefore, the fabrication process of these composites should be well-designed and well-controlled in order to obtain carbon/metal (C/metal) composites with fine microstructures and high thermal properties [9–11]. To date, diamond/metal composites, CF/metal composites and graphite flake/metal composites are the mostly studied C/metal composites that can be used in thermal management applications. Diamond/metal composites have high TC ranging from 350 W/mK to 780 W/mK, and have been commercialized [8,9,12]. Unfortunately, ⁎ Corresponding author. E-mail address: [email protected] (Q. Ouyang).
http://dx.doi.org/10.1016/j.matdes.2015.10.146 0264-1275/© 2015 Elsevier Ltd. All rights reserved.
their high-cost and poor machining properties strongly limit their application [8]. Due to the anisotropic thermal properties of CFs and graphite flakes, excellent properties are only achieved in one or two axial directions of their composites [8]. For example, if carbon fibers with high TC (also high price) were used, the TC of long carbon fiber (LCF)/metal composites can reach 500–700 W/mK along the axial direction while the TC of short carbon fiber (SCF)/metal composites are 208 W/mK in the in-plane direction [13,14]. Also, as reported by Chen et al. [15], the in-plane TC of the graphite flake/metal composites can reach 324 W/mK to 783 W/mK. However, it is difficult to control the exact direction of the graphite flakes as well as the infiltration process [8,15]. In order to promote infiltration, other reinforcements such as SiC [8,16], Si [10] and CF [8] were often added into graphite flake/ metal composites, which, unfortunately, would inevitably lead to a reduction in the TC of the resulting composites. Recently, CNT/metal composites and graphene/metal composites have drawn much attention in thermal management due to outstanding thermal properties of CNTs and graphene, while, unfortunately, it has been found that the improvement of thermal properties of these composites is hindered by many technical problems, such as the control of the distribution and orientation of the reinforcements and high interfacial thermal resistance [11]. Recently, graphite films, such as artificial graphite films fabricated from polyimide films [7,17] and graphene oxide films [2,18], have attracted much attention in thermal management due to their excellent thermal properties. Specially, artificial graphite films with high TC (1100–1600 W/mK) have become commercially available for heat removal in electronic devices containing integrated circuits, such as personal computers and mobile phones [7]. Unfortunately, so far, little attention has been paid to the potential of these graphite films being used as effective reinforcements in composites. Actually, if these
Y. Huang et al. / Materials and Design 90 (2016) 508–515
graphite films are used as reinforcements to form graphite film/metal composites and fabrication processes of the composites are welldesigned and well-controlled, high TC and low CTE can be expected. Furthermore, if we use graphite films and metal foils as raw materials to fabricate graphite film/metal laminate composites, we can easily control the distributions and orientations of the reinforcements by controlling the stacking of graphite films and metal foils, which is easier and more effective as compared to CF/metal composites, graphite flake/ metal composites, CNT/metal composites and graphene/metal composites, which may simplify the fabrication process and meanwhile obtain composites with higher thermal properties. In this study, we successfully fabricated these novel graphite film/ aluminum laminate composites with high TC and low CTE by a vacuum hot pressing process. The effects of oxide films on the surface of aluminum foils and volume fraction of graphite films on microstructures and properties of the composites were investigated and discussed. The microstructures of the composites were characterized at length scales from the macro down to the nanoscale to investigate the distribution and orientation of the graphite films, and the graphite/Al interfacial structure and the presence of harmful reaction product. Furthermore, the thermal conductivities of the as-fabricated composites were compared with the predicted values calculated by rule of mixture and results of other kinds of carbon/aluminum (C/Al) composites reported in the literatures. 2. Experimental Graphite films and pure aluminum foils (99.9% in purity) were used as raw materials. The graphite films were acquired from Dasen Electronic Material Co., Ltd. in China. The morphologies of the graphite films are shown in Fig. 1 (a). These graphite films have a rough surface with many ripples on it. The inset shows the cross-section of the graphite films. The average thickness of the graphite films was measured to be 29.5 μm. Fig. 1 (b) shows the X-ray diffraction (XRD) pattern of the graphite films, where only peaks of graphite can be detected from this pattern, and the Kα1 and Kα2 peaks of (006) peaks are separated very well, which indicates that the graphite films have high purity and crystallinity [19]. The in-plane and out-of-plane TC values of the graphite films were measured to be 1222.6 (± 30.3) W/mK and 17.5 (± 2.5) W/mK, respectively. Graphite film/aluminum laminate composites were fabricated by a vacuum hot pressing process. Four sets of composite samples with different volume fractions of graphite films were obtained by controlling the thickness of the aluminum foils. The volume fractions of graphite films of sample #1, #2, #3 and #4 are 17.4%, 36.2%, 53.2% and 69.4%, respectively. The detailed fabrication processes of these laminate composites are listed as following: 1). Graphite films and aluminum foils were
509
washed by acetone and then dried. 2). In order to eliminate the oxide films on the surface of the aluminum foils, they were first washed by 15 g/L NaOH solution, then washed by HNO3 with volume fraction of 25%, and finally washed by alcohol and dried. 3). The pre-treated graphite films and aluminum foils were carefully filled into a graphite mold layer-by-layer and then heated to 655 °C in high vacuum condition (less than 5 × 10−3 Pa) and kept for 100 min, while an uniaxial pressure of 22 MPa was applied. In order to figure out the effect of oxide films on microstructures and properties of the composites, a composite sample named #3′ was fabricated without the oxide film eliminating process (process 2)). Since the thickness of the oxide films (1–5 nm) [20,21] is much smaller than that of the aluminum foils (26 μm), the effect of them on volume fraction can be neglected. Thus, the graphite volume fraction of sample #3′ is considered to be the same with sample #3. The in-plane thermal diffusivities of the graphite films (Φ25.4 mm) and the composite samples (10 × 10 × 4 mm) at room temperature were measured by a laser flash technique using a NETZSCH LFA447 thermal analyzer. The specific heat capacities of the samples at room temperature were measured by PerkinElmer differential scanning calorimeter (DSC) 8000. A speed of 20 °C/min was used in the temperature range of 15–35 °C. The densities of the samples were measured by the Archimedes method. The in-plane TC values of the samples were calculated by the product of the density, thermal diffusivity and specific heat capacity. The in-plane CTE values of the specimens (10 × 2 × 25 mm) were measured by a dilatometer (NETZCH DIL 402C) from room temperature up to 300 °C at a speed of 5 °C/min. XRD patterns of the graphite films were obtained by a D/max-2550 instrument (Cu Kα). A scan speed of 4°/min was used in the range of 10–90°. Microstructures of the graphite films and composite samples were characterized by scanning electron microscopy (SEM) operated at 20 kV using a FEI Quanta FEG 250 electron microscope and transmission electron microscope (TEM) using a JEOL 2100F Field Emission Electron Microscope. The energy dispersive spectroscopy (EDS) element line scanning across the aluminum-graphite interface was characterized using the aforementioned SEM. 3. Results and discussion 3.1. Effect of oxide films on microstructures and properties of the composites According to the results reported in graphite fiber/aluminum composites [22] and diamond/aluminum composites [23], during the fabrication process of the composites, diffusion may happen between aluminum and carbon materials to form a strong interfacial bonding. However, in this study where aluminum foils are used for raw material, there are thin oxide films with thickness of 1–5 nm on the surface of aluminum foils due to their exposure to air [20,21], which would generally
Fig. 1. (a) Morphologies of the surface and cross-section (inset, between the two arrows) of the graphite films; (b) XRD patterns of the graphite films.
510
Y. Huang et al. / Materials and Design 90 (2016) 508–515
Fig. 2. Schematic diagram of the effect of oxide films on fabrication of the composites: (a) without the oxide film eliminating process; and (b) after eliminating the oxide films.
act as diffusion barriers [24,25] to impede the interfacial diffusion between aluminum and carbon during the hot pressing process, as shown in Fig. 2 (a). Therefore, if these oxide films are eliminated before the hot pressure process of the composites, it may be easier for the diffusion to take place to form an interface with good interfacial bonding, as illustrated in Fig. 2 (b). To evaluate the effect of the oxide films, the microstructures of the interfaces of the composite sample #3′ and #3 were examined and are shown in Fig. 3 (a) and (b), respectively. Interfacial debonding is much easier to occur at the interface of sample #3′ after mechanical polishing, which may indicate that the interface with oxide films would result in a weaker interfacial bonding strength. In contrast, if the oxide films are eliminated, fine interface with no interfacial debonding will be obtained as shown in Fig. 3 (b). In order to have a better understanding of the reason for that interfacial debonding, the elemental line distribution across typical interface of sample #3′ and #3 were measured by EDS and the results are shown in Fig. 4. As shown in Fig. 4 (c), a decaying concentration profile of carbon atoms across the C/Al interface was found, indicating that the diffusion of carbon into aluminum may take place during the processing of the composites, most likely during the high temperature (655 °C) step, which led to a strong adhesion at the interface of sample #3. More specifically, the diffusion speed k (μm/min) of carbon in aluminum can be calculated by the Arrhenius diffusion function [26]: k = A exp (− Qa/RT), where A is a temperature-independent pre-
exponential, Qa is the activation energy for diffusion, R is the gas constant and T is the absolute temperature. Referring to the parameters reported in reference [26], the diffusion speed k of carbon in aluminum at the fabrication temperature of 655 °C (928 K) can be calculated to be ~ 2.2 × 10−2 μm/min. By multiplying time that the processing step lasted (100 min), the diffusion range of carbon atoms into aluminum matrix can be estimated to be about 2.2 μm, which is consistent with the range of carbon concentration profile found experimentally (2–3 μm, Fig. 4 (d), for sample #3 without the oxide films). On the other hand, affected by the oxide films, the range of carbon distribution on the aluminum side in sample #3′ was found to be less than 1 μm (Fig. 4 (a) and (b)). By comparing the ranges that carbon atoms diffused across the C/Al interface into the aluminum side, we may come to the conclusion that the oxide films served as an effective barrier for carbon diffusing into aluminum, giving rise to a weak interfacial strength in sample #3′. The comparison of the thermal properties between sample #3′ and sample #3 is shown in Fig. 5 and the specific values of their thermophysical properties can be found in Table 1. Due to the weaker interfacial bonding and existence of interfacial debonding as illustrated in Figs. 3 and 4, the in-plane TC of sample #3′ is measured to be much lower than that of sample #3 and the in-plane CTE of sample #3′ is much higher than that of sample #3, which indicates that the composite sample fabricated with the oxide films eliminating process have much better thermal properties than sample fabricated without this process. Therefore, the oxide films on the aluminum foils may act as diffusion barriers to impede the interfacial diffusion between carbon and aluminum, which leads to formation of interface with weak interfacial bonding strength. Interfacial debonding would easily take place at this kind of interface, which results in relatively low thermal properties. By eliminating these oxide films, interfaces with strong interfacial bonding strength will be obtained and interfacial debonding will be restrained, which leads to an improvement in thermal properties of the composites. 3.2. Effect of volume fractions of graphite films on microstructures and properties of the composites Four sets of composites with different volume fractions of graphite films were studied in this study to figure out the effect of volume fractions of graphite films on microstructures and properties of the composites. Fig. 6 shows the microstructures of the composite samples and it is obvious that the volume fraction of the graphite films increases from (a) to (d). As expected, these composites have lamellar microstructures. The graphite films are highly homogeneous and the orientations of these films are well controlled. Therefore, the distribution and
Fig. 3. Microstructures of the composite samples fabricated by different fabrication process: (a) sample #3′ fabricated without the oxide film eliminating process; and (b) sample #3 fabricated with the oxide film eliminating process.
Y. Huang et al. / Materials and Design 90 (2016) 508–515
511
Fig. 4. Microstructure and elemental line distribution results across typical interface of (a–b) sample #3′ and (c–d) sample #3.
orientation control of the reinforcements in graphite film/aluminum laminate composites are much more effective as compared to other C/Al composites, such as CF/metal composites [13], graphite flake/ metal composites [15], CNT/metal composites and Graphene/metal composites [11], which may lead to high thermal properties. In order to have a better understanding of the interfacial bonding between graphite films and aluminum, TEM was used to characterize the interfacial structure of the composites. The results are shown in Fig. 7. Fig. 7 (a) shows the overall view of the aluminum matrix, graphite films and their interface while Fig. 7 (b) gives a high magnification image of the interface. The inset on the bottom left of Fig. 7 (b) shows the indexed [233] zone axis selected area diffraction (SAD) pattern of the aluminum. The inset on the top right of Fig. 7 (b) shows the SAD pattern of graphite films, where (002), (100), (004) and (006) diffraction
Fig. 5. Comparison of the CTE and TC values between composite sample #3′ and sample #3.
rings can be identified. The opening angle of the (002) arcs (marked by small circles) is very small, indicating that the graphite planes in the graphite films are well-oriented [27]. A clear row of spots (marked with discontinuous lines) can be detected from the diffraction pattern, which implies the existence of preferred orientation in the selected area of graphite films [27]. As shown in Fig. 7, the composites have a clear interface and no reaction products can be detected along the whole interface. Typical high resolution TEM (HRTEM) image of the interface of the composites is shown in Fig. 8 (a). Indexed Fast Fourier Transform (FFT) results of some selected areas of graphite films and aluminum are shown in Fig. 8 (b), (c), (d) and (e). A thin interfacial layer, with amorphous-like structure and thickness of nearly 2 nm, was observed to tightly adhere to the aluminum matrix. Its FFT pattern (Fig. 8 (b)) indicated that it is likely to have amorphous structure. Such an amorphous interfacial layer is frequently observed in carbon (in the form of graphite or diamond)/aluminum composites such as graphite flakes/aluminum composites and diamond/aluminum composites. For example, Zhou et al. [10] reported that there exists an amorphous graphite layer at the interface of graphite flake/aluminum composites. Beffort et al. [28, 29], Khalid et al. [30] and Jiang et al. [31] also found an amorphous layer at the interface of diamond/aluminum composites. Because of the low thickness of the amorphous layer, it was difficult to accurately determine its composition. Fortunately, Beffort et al. [29] found a thicker amorphous layer (about 50–200 nm) in diamond/ aluminum composites fabricated at 770–800 °C and measured its chemical composition by EDS. The interfacial layer was found to be mainly comprised of C, Al and O [29]. It was found that O may come from the oxide films of the aluminum melt during the fabrication process or from the raw carbon material, and these authors proposed that carbide-forming elements (such as aluminum) are particularly active in catalyzing the reconstruction of the diamond surface to form an amorphous layer [29]. Therefore, it may be concluded that the
512
Y. Huang et al. / Materials and Design 90 (2016) 508–515
Table 1 Thermo-physical properties of graphite film/aluminum laminate composites with different volume fractions of graphite films. Sample
Thickness design (μm)
VG (%)
ρ (g/cm3)
Cp (J/g/K)
αa (mm2/s)
TCa (W/mK)
CTEa (ppm/K)
#1 #2 #3 #3′ #4 Al Gr film
Al (140) + Gr film (29.5) Al (52) + Gr film (29.5) Al (26) + Gr film (29.5) Al (26) + Gr film (29.5) Al (13) + Gr film (29.5) – –
17.4 36.2 53.2 53.2 69.4 – –
2.54 2.39 2.25 2.23 2.10 2.7 1.96
0.836 0.818 0.798 0.798 0.746 0.88 0.709
167.445 262.133 409.594 377.487 575.701 99.7 879.8
356 512 735 672 902 237 [14] 1222.6
20.6 10.58 7.22 11.94 3.55 23.2 [14] −1.5 [10]
a
The TC and CTE of graphite film/aluminum laminate composites are all anisotropic, and here we only list the values along the in-plane directions.
amorphous interfacial layer found in our study may come from a similar origin, i.e., aluminum element catalyzes the reconstruction of graphite film surface to form an amorphous layer, and the composition of this amorphous layer is likely to contain C, Al and O. Also, as shown in Fig. 8 (a), the lattice structure of the crystalline graphite adjacent to this amorphous layer is distorted to some extent, which is probably caused thermal stress between aluminum matrix and graphite film [32]. As a result, according to the lattice structure illustrated by the HRTEM, the as-fabricated composites have fine interface with good interfacial bonding and free of harmful reaction product. Table 1 reports the thermo-physical properties of the composite samples and raw materials. As the volume fraction of the graphite films increases from 17.4% to 69.4%, the in-plane TC of composites increases from 356 to 902 W/mK and the in-plane CTE of the composites decreases from 20.6 to 3.55 ppm/K. Therefore, the addition of graphite films has brought significant improvement of thermal properties to
the aluminum matrix. However, similar to graphite flake/aluminum composites [15], if the volume fraction of the graphite films is too high, some local regions in the composites would be short of aluminum during the fabrication process which would lead to formation of pores. These pores will inevitably degrade the properties of the composites. Therefore, the volume fraction of the graphite films is limited to 70% in this study in order to guarantee stable properties. In a word, graphite film/aluminum laminate composites with tailorable and stable high thermal properties were fabricated in this study and these composites have a great potential for applications in thermal management. 3.3. Comparisons with predictions and other kinds of C/Al composites For laminate composites, their theoretical in-plane TC can be simply calculated using rule of mixture as in this case no interface effects are envisaged [4,14]. The material parameters for the calculation can be
Fig. 6. Microstructures of the composite samples: (a) sample #1 with volume fraction of 17.4%; (b) sample #2 with volume fraction of 36.2%; (c) sample #3 with volume fraction of 53.2%; (d) sample #4 with volume fraction of 69.4%. Gr represent graphite.
Y. Huang et al. / Materials and Design 90 (2016) 508–515
513
Fig. 7. Typical TEM image of the interface of the composites: (a) overall view; and (b) Al/Graphite interface. The two insets are the SAD patterns of the graphite and aluminum.
found in Table 1. Fig. 9 (a) shows the experimental and predicted inplane TC of the composite samples. The four samples with different volume fractions all exhibit in-plane TC over 80% of the predictions. Moreover, samples with high volume fraction, such as 53.2% and 69.4%, exhibit high in-plane TC over 90% of the predictions. Therefore, the graphite films can act as effective reinforcements in aluminum matrix composites. The deviation between the experimental results and predictions is probably due to the existence of interfacial thermal resistance [9,14]. The out-of-plane thermal conductivity of laminate composites can be estimated by the equation given by reference [16], assuming an aluminum/graphite interfacial thermal conductance of 4.5 × 107 W/m2K [10]. Calculation revealed that as the volume fraction of the graphite films increases from 17.4% to 69.4%, the calculated out-of-plane thermal
conductivity would decrease from 73 to 23.8 W/mK, which is much less than the in-plane thermal conductivities as shown in Table 1 (from 356 to 902 W/mK). Therefore, the thermal conductivity of the as-fabricated composites is highly anisotropic. If these composites are used in heatremoving applications, heat would be much more quickly removed by being conducted along the in-plane direction rather than the out-ofplane direction, and thus the in-plane thermal conductivity dominates the heat-removing process and subsequently the overall thermal conductivity of the composite. As a result, we will only focus on the in-plane thermal conductivity in the following discussions of this paper. Fig. 9(b) shows the comparison of the TC of the graphite film/ aluminum laminate composites fabricated in this study to the TC of other kinds of C/Al composites that can be used in thermal management. In most cases, graphite film/aluminum laminate composites
Fig. 8. Typical HRTEM image of the interface of the composites: (a) HRTEM image; (b) FFT results of the selected area of amorphous layer; (c–d) FFT results of the selected area of crystalline graphite; (e) FFT results of the selected area of aluminum.
514
Y. Huang et al. / Materials and Design 90 (2016) 508–515
Fig. 9. Thermal conductivities of the graphite film/aluminum laminate composites with different volume fractions of graphite films fabricated in this work: (a) compared with the predicted values calculated by rule of mixture; (b) compared with other kinds of thermal management C/Al composites fabricated in the reference, as a function of volume fraction of carbon. Gr represent graphite. Graphite flakes and SiC (or Si) reinforced aluminum composites are treated as C/Al composites due to the fact that graphite flake is the main reinforcement and SiC (or Si) is used to produce space for infiltration [8,15,16].
fabricated in this study exhibit higher in-plane TC than that of the C/Al composites reported in these references. This is likely caused by the following facts: 1). The measured in-plane TC of the starting graphite films is very high, exceeding 1200 W/mK, which makes graphite films promising reinforcements for thermal management composites. 2). For graphite film/aluminum laminate composites, it is very easy to control the orientation and distribution of the graphite films by controlling the stacking of graphite films and metal foils. Therefore, it is easier to bring the enhancement effect of the graphite films into full play than other kind of reinforcements such as graphite flakes and CFs. 3). Graphite film/aluminum laminate composites fabricated in this study have fine microstructures and interfaces as shown in Figs. 6, 7 and 8, which makes the TC of the composites approach the theoretical values. 4. Conclusions In conclusion, a novel kind of carbon/metal composites, namely graphite film/aluminum laminate composites with high TC and low CTE, has been fabricated by a vacuum hot pressing process. The effects of oxide films on the interfacial structure and thermal properties of the composites were identified. The result shows that these oxide films are harmful to the interfacial bonding of the composites, which leads to interfacial debonding and low thermal properties. As the volume fraction of the graphite films increases from 17.4% to 69.4%, the in-plane TC of the composites increases from 356 to 902 W/mK and the in-plane CTE of the composites decreases from 20.6 to 3.55 ppm/K. The as-fabricated composites have fine microstructures with welldistributed and well-oriented graphite films and clean interface, which leads to excellent thermal properties. The in-plane TC values of the composites with different graphite volume fractions are all over 80% of the predictions calculated by the rule of mixture. What's more, these composites are shown to have more competitive TC as compared to other kinds of C/Al composites, which makes these composites promising candidates to be used for thermal management. Finally, it is worth mentioning that graphite films can be used as reinforcements in other matrixes such as Cu and polymer. Moreover, the morphologies of the graphite films added to the metal matrix are not limited to films, but may also be strips and flakes, which makes microstructures and properties of their composites more designable. In addition, the fabrication process is suitable for other carbon film/metal composites, such as graphene oxide film/metal composites and so on. Acknowledgments The authors would like to acknowledge the financial support of the National Basic Research Program of China (973 Program, No.
2012CB619600), the Foundation of Shanghai Science and Technology Committee of China (No. 14DZ2261200), the Shanghai Science and Technology Committee (No. 13PJ1404000), the National Natural Science Foundation of China (No. 51471190), and the National High Technology Research and Development Program (863 Program, No. 2013AA031201). References [1] D. Chung, Materials for thermal conduction, Appl. Therm. Eng. 21 (2001) 1593–1605, http://dx.doi.org/10.1016/S1359-4311(01)00042–4. [2] G. Xin, H. Sun, T. Hu, H.R. Fard, X. Sun, N. Koratkar, et al., Large-area freestanding graphene paper for superior thermal management, Adv. Mater. 26 (2014) 4521–4526, http://dx.doi.org/10.1002/adma.201400951. [3] A. Mortensen, J. Llorca, Metal matrix composites, Annu. Rev. Mater. Res. 40 (2010) 243–270, http://dx.doi.org/10.1146/annurev-matsci-070909-104511. [4] Y.-M. Chen, J.-M. Ting, Ultra high thermal conductivity polymer composites, Carbon 40 (2002) 359–362, http://dx.doi.org/10.1016/S0008-6223(01)00112–9. [5] Z. Ma, J. Shi, Y. Song, Q. Guo, G. Zhai, L. Liu, Carbon with high thermal conductivity, prepared from ribbon-shaped mesosphase pitch-based fibers, Carbon 44 (2006) 1298–1301, http://dx.doi.org/10.1016/j.carbon.2006.01.015. [6] G. Yuan, X. Li, Z. Dong, A. Westwood, Z. Cui, Y. Cong, et al., Graphite blocks with preferred orientation and high thermal conductivity, Carbon 50 (2012) 175–182, http://dx.doi.org/10.1016/j.carbon.2011.08.017. [7] M. Inagaki, Y. Kaburagi, Y. Hishiyama, Thermal management material: graphite, Adv. Eng. Mater. 16 (2014) 494–506, http://dx.doi.org/10.1002/adem.201300418. [8] R. Prieto, J.M. Molina, J. Narciso, E. Louis, Fabrication and properties of graphite flakes/metal composites for thermal management applications, Scr. Mater. 59 (2008) 11–14, http://dx.doi.org/10.1016/j.scriptamat.2008.02.026. [9] Z. Tan, Z. Li, G. Fan, Q. Guo, X. Kai, G. Ji, et al., Enhanced thermal conductivity in diamond/aluminum composites with a tungsten interface nanolayer, Mater. Des. 47 (2013) 160–166, http://dx.doi.org/10.1016/j.matdes.2012.11.061. [10] C. Zhou, G. Ji, Z. Chen, M. Wang, A. Addad, D. Schryvers, et al., Fabrication, interface characterization and modeling of oriented graphite flakes/Si/Al composites for thermal management applications, Mater. Des. 63 (2014) 719–728, http://dx.doi.org/10. 1016/j.matdes.2014.07.009. [11] S.C. Tjong, Recent progress in the development and properties of novel metal matrix nanocomposites reinforced with carbon nanotubes and graphene nanosheets, Mater. Sci. Eng. R 74 (2013) 281–350, http://dx.doi.org/10.1016/j.mser.2013.08.001. [12] K. Mizuuchi, K. Inoue, Y. Agari, Y. Morisada, M. Sugioka, M. Tanaka, et al., Processing of diamond particle dispersed aluminum matrix composites in continuous solid– liquid co-existent state by SPS and their thermal properties, Compos. Part B 42 (2011) 825–831, http://dx.doi.org/10.1016/j.compositesb.2011.01.012. [13] T. Liu, X. He, L. Zhang, Q. Liu, X. Qu, Fabrication and thermal conductivity of short graphite fiber/Al composites by vacuum pressure infiltration, J. Compos. Mater. 48 (2013) 2207–2214, http://dx.doi.org/10.1177/0021998313495750. [14] T. Ueno, T. Yoshioka, J.-I. Ogawa, N. Ozoe, K. Sato, K. Yoshino, Highly thermal conductive metal/carbon composites by pulsed electric current sintering, Synth. Met. 159 (2009) 2170–2172, http://dx.doi.org/10.1016/j.synthmet.2009.10.006. [15] J.K. Chen, I.S. Huang, Thermal properties of aluminum–graphite composites by powder metallurgy, Compos. Part B 44 (2013) 698–703, http://dx.doi.org/10.1016/j. compositesb.2012.01.083. [16] R. Prieto, J.M. Molina, J. Narciso, E. Louis, Thermal conductivity of graphite flakes–SiC particles/metal composites, Compos. Part A 42 (2011) 1970–1977, http://dx.doi.org/ 10.1016/j.compositesa.2011.08.022. [17] M. Inagaki, N. Ohta, Y. Hishiyama, Aromatic polyimides as carbon precursors, Carbon 61 (2013) 1–21, http://dx.doi.org/10.1016/j.carbon.2013.05.035.
Y. Huang et al. / Materials and Design 90 (2016) 508–515 [18] D.A. Dikin, S. Stankovich, E.J. Zimney, R.D. Piner, G.H. Dommett, G. Evmenenko, et al., Preparation and characterization of graphene oxide paper, Nature 448 (2007) 457–460, http://dx.doi.org/10.1038/nature06016. [19] T. Takeichi, Y. Eguchi, Y. Kaburagi, Y. Hishiyama, M. Inagaki, Carbonization and graphitization of BPDA/PDA polyimide films: effect of structure of polyimide precursor, Carbon 37 (1999) 569–575, http://dx.doi.org/10.1016/S00086223(98)00223–1. [20] M. Hunter, P. Fowle, Natural and thermally formed oxide films on aluminum, J. Electrochem. Soc. 103 (1956) 482–485, http://dx.doi.org/10.1149/1.2430389. [21] L. Jiang, G. Fan, Z. Li, X. Kai, D. Zhang, Z. Chen, et al., An approach to the uniform dispersion of a high volume fraction of carbon nanotubes in aluminum powder, Carbon 49 (2011) 1965–1971, http://dx.doi.org/10.1016/j.carbon.2011.01.021. [22] M. Gu, H. Yang, W. Jiang, G. Zhang, Study on the interfacial reaction product in Gr/Al composites, Adv. Compos. Mater. 5 (1996) 119–128, http://dx.doi.org/10.1163/ 156855196X00031. [23] S. Kleiner, F. Khalid, P. Ruch, S. Meier, O. Beffort, Effect of diamond crystallographic orientation on dissolution and carbide formation in contact with liquid aluminium, Scr. Mater. 55 (2006) 291–294, http://dx.doi.org/10.1016/j.scriptamat.2006.05.003. [24] T. Rajan, R. Pillai, B. Pai, Reinforcement coatings and interfaces in aluminium metal matrix composites, J. Mater. Sci. 33 (1998) 3491–3503, http://dx.doi.org/10.1023/ A:1004674822751. [25] Y. Tang, Y. Deng, K. Zhang, L. Liu, Y. Wu, W. Hu, Improvement of interface between Al and short carbon fibers by α-Al2O3 coatings deposited by sol–gel technology, Ceram. Int. 34 (2008) 1787–1790, http://dx.doi.org/10.1016/j.ceramint.2007.05. 008.
515
[26] K. Landry, S. Kalogeropoulou, N. Eustathopoulos, Wettability of carbon by aluminum and aluminum alloys, Mater. Sci. Eng. A 254 (1998) 99–111, http://dx.doi.org/10. 1016/S0921-5093(98)00759-X. [27] C. Bourgerette, A. Oberlin, M. Inagaki, Structural and textural changes from polyimide Kapton to graphite: Part I. Optical microscopy and transmission electron microscopy, J. Mater. Res. 7 (1992) 1158–1173, http://dx.doi.org/10.1557/JMR.1992.1158. [28] O. Beffort, S. Vaucher, F. Khalid, On the thermal and chemical stability of diamond during processing of Al/diamond composites by liquid metal infiltration (squeeze casting), Diamond Relat. Mater. 13 (2004) 1834–1843, http://dx.doi.org/10.1016/j. diamond.2004.04.014. [29] O. Beffort, F. Khalid, L. Weber, P. Ruch, U. Klotz, S. Meier, et al., Interface formation in infiltrated Al (Si)/diamond composites, Diamond Relat. Mater. 15 (2006) 1250–1260, http://dx.doi.org/10.1016/j.diamond.2005.09.036. [30] F. Khalid, O. Beffort, U. Klotz, B. Keller, P. Gasser, Microstructure and interfacial characteristics of aluminium–diamond composite materials, Diamond Relat. Mater. 13 (2004) 393–400, http://dx.doi.org/10.1016/j.diamond.2003.11.095. [31] L. Jiang, P. Wang, Z. Xiu, G. Chen, X. Lin, C. Dai, et al., Interfacial characteristics of diamond/aluminum composites with high thermal conductivity fabricated by squeeze-casting method, Mater. Charact. 106 (2015) 346–351, http://dx.doi.org/ 10.1016/j.matchar.2015.06.023. [32] H. Kurita, T. Miyazaki, A. Kawasaki, Y. Lu, J. Silvain, Interfacial microstructure of graphite flake reinforced aluminum matrix composites fabricated via hot pressing, Compos. Part A 73 (2015) 125–131, http://dx.doi.org/10.1016/j.compositesa.2015. 03.013.