Fabrication and thermal conductivity of short graphite fiber/Al composites by vacuum pressure infiltration

Abstract

Short graphite fiber/Al composites were fabricated by a modified two-step vacuum pressure infiltration technique. Copper-coated graphite fibers preform was infiltrated with liquid aluminum at 800℃ under infiltration pressure of 1 MPa and solidification pressure of 30 MPa for 30 min. The effects of surface modification and the processing parameters of vacuum pressure infiltration on relative density and thermal conductivity of the composites were systematically studied. The results show that short graphite fiber/Al composite with relatively high density of 99.1% and thermal conductivity of 208 W·m⁻¹·K⁻¹ was successfully fabricated. Through the application of copper coating onto the graphite fibers, the in-plane thermal conductivity of the composite was effectively enhanced from 117 W·m⁻¹·K⁻¹ to 208 W·m⁻¹·K⁻¹ as a result of improved interfacial bonding. The obtained short graphite fiber/Al composites are promising materials for electronic packing applications.

Keywords

Short graphite fibers vacuum pressure infiltration aluminum matrix composites relative density thermal conductivity

Introduction

Metal matrix composites with high thermal conductivity and tailorable coefficient of thermal expansion (CTE) has attracted considerable attention because of the increasing requirement imposed on heat dissipate materials microelectronics and semiconductors.^1,2 High-performance mesophase pitch-based graphite fiber is an outstanding reinforcement for thermal management composites because of its high thermal conductivity (1100 W·m⁻¹·K⁻¹) and low CTE (−1.5 × 10⁻⁶ K⁻¹) in the axial direction.³ At the same time, aluminum has been widely used as metal matrix due to its high specific thermal conductivity (thermal conductivity divided by density) and ease of machinability.^4–6 Therefore, graphite fiber reinforced aluminum matrix composites, if processed properly, will offer a well combination of low CTE and high thermal conductivity.

Recently, graphite fiber (carbon fiber)/Al composites have been fabricated by powder metallurgy, stir casting and liquid metal infiltration method.^7–9 Liquid metal infiltration, in the form of gas infiltration, CVD infiltration, ultrasonic infiltration and vacuum pressure infiltration, as an efficient method, has been widely used recently. Vacuum pressure infiltration is a well-known procedure for the fabrication of complex-shaped components at high production rates and low cost. This technique possesses two prominent advantages. One is that infiltration process carried out under vacuum and high pressure conditions can eliminate shrinkage defects in composites. The other is that undesirable interfacial reactions can be avoided by controlling processing parameters, such as temperature and pressure, due to its low demand in equipment and simple operation. Most researchers focused on developing composites for ameliorating mechanical properties, such as tensile strength, friction and wear properties.^10–13 However, few works deal with the detailed manufacturing process and thermal property of the composites. Metal Matrix Cast Composite Inc. produced graphite fiber reinforced Al composites with thermal conductivity of 200–230 W·m⁻¹·K⁻¹ to be used as matching stock for electronic heat sink and substrate components.^14,15 Nevertheless, the specific manufacturing process and the effects of processing parameters on thermal property of the composites were not given in detail. Therefore, acting as heat management materials, the graphite fiber/Al composites need to be further studied in order to improve the thermal conductivity and economically fabricated to meet the demand of further industrial applications.

In the present work, short graphite fiber/Al composites were fabricated by vacuum pressure infiltration technique, simultaneously a modified two-step pressure infiltration was designed to raise the relative density of composites. Additionally, copper coating was applied onto graphite fibers by electroless plating method to improve the wettability and avoid undesirable interfacial reaction between graphite fiber and liquid aluminum. Then, the effects of processing parameters of vacuum pressure infiltration and surface metal coating on relative density and thermal conductivity were systematically investigated. And the most efficient process was optimized and put forward.

Experimental

Materials

Mesophase pitch-based graphite fibers, supplied by Nippon Graphite Fiber Corp., Japan, with a length of 100–500 µm (mean length 200 µm), diameter of 10 µm, density of 2.225 g·cm⁻³ and thermal conductivity (axial) of 900 W·m⁻¹·k⁻¹, have been selected as the reinforcements (shown in Figure 1(a)). In addition, graphite fibers coated with copper prepared by electroless plating have been also used in this work. The electroless plating of copper consists of four stages: roughening, sensitization, activation and electroless plating. The optimized working condition based on lots of experiments by our group and results of other researchers^16,17 is shown in Table 1. Figure 1(b) shows the microstructures of copper-coated graphite fibers. As seen from it, the copper coating is continuous and compact, spreading on the graphite fibers uniformly. The high-purity aluminum (99.9%) has been chosen as the matrix of short graphite fiber/Al composites.

Table 1.

The optimized parameters of copper electroless plating.

Stage	Concentration of chemical solution	Temperature/^oC	Time/min	Condition
Roughening	–	400	20	In an open oven
Sensitization	30 g/L SnCl₂	Room temperature	15	High-speed stirring
Sensitization	60 ml/L HCl	Room temperature	15	High-speed stirring
Activation	0.25 g/L PdCl₂	Room temperature	15	High-speed stirring
Activation	10 ml/L HCl	Room temperature	15	High-speed stirring
Electroless coating	15 g/L CuSO₄·5H₂O	45	10	pH 12.5
	14 g/L Sodium potassium tartrate
	14.5 g/L Na₂EDTA
	0.02 g/L Bipyridyl
	0.01 g/L Potassium ferrocyanide
	NaOH for adjusting pH

Composites fabrication

The fabrication of short graphite fiber/Al composites involves two main steps: preparation of short graphite fiber preform and pressure infiltration of molten aluminum into porous preform. In order to investigate the influential factors on the properties of graphite fiber/Al composites directly and credibly, the preforms with approximate 40 vol.% graphite fibers were selected for all samples. The infiltration was done by an apparatus designed for this study. After loading short graphite fiber preform, aluminum and piston in sequence from bottom up into graphite die, the system was highly evacuated to a level of 0.1 Pa. Then the system was heated to 700–840℃ to study the effects of infiltration temperature on the relative density and thermal conductivity of the composites. Subsequently, the piston moved downwards to make aluminum liquid infiltrate into graphite fiber preform. Two kinds of pressurization methods were used in this paper containing conventional one-step and modified two-step pressure infiltration techniques. The schematic illustration of the two-step pressure infiltration process is shown in Figure 2. This modified pressure infiltration process includes two steps: (1) a very low pressure for infiltration (i.e. infiltration pressure) is applied to ensure the steady infiltration of melt-preform system; (2) after infiltration, a very high pressure for solidification (i.e. solidification pressure) is forced to obtain the fully densification of composite materials.

Characterization

For microstructural characterization of the fabricated composites, some analysis methods were utilized. X-ray diffraction (XRD) patterns recorded by a D5000 Siemens diffractometer were used to study the phase composition of the composites. The microstructures and fracture surfaces of graphite fiber/Al composites were observed by scanning electron microscope (SEM, LEO-1450). In this study, the fracture was made by using plier clamping both ends of the sample and breaking it in two. Energy dispersive spectroscopy (EDS) element line scanning across the interface was studied by field emission SEM (FE-SEM, LEO JSM-7001F). The relative density of the composites was measured by Archimedes’ drainage method¹⁸ and compared with the theoretical density. Thermal diffusivity (α) was measured by JR-3 thermal physical testing instrument. Specific heat of composites ( $C_{p}$ ) was represented by following expression:

C_{p} = \frac{C_{f} \times V_{f} \times ρ_{f} + C_{m} \times V_{m} \times ρ_{m}}{V_{f} \times ρ_{f} + V_{m} \times ρ_{m}}

(1)

where V is the volume fraction of each component, f and m stand for fiber and matrix. Thermal conductivity was calculated by the equation:

K = ρ \times α \times C_{p}

(2)

where K and

ρ

are the thermal conductivity and density of composites, respectively.

Results and discussion

Infiltration densification of the composites

Table 2 shows the effects of processing parameters of vacuum pressure infiltration and graphite fiber surface copper coating on the relative density of graphite fiber/Al composites. The data of relative density given in the table is the average value of three experimental results. It can be seen that two-step pressure infiltration process can improve the relative density efficiently by comparing property of sample No. 1 with that of No. 2. The two-step pressure infiltration can not only avoid the reinforcement preform deform and liquid aluminum flow out of the mould under high infiltration pressure but also eliminate defects such as internal porosities and shrinkage cavities in composites during solidification. Therefore, short graphite fiber/Al composites with higher relative density can be fabricated by the use of two-step pressure infiltration technique. This result is similar to the research of Hu¹⁹ who fabricated SiCw/ZK51A magnesium matrix composite by two-step squeeze casting. They found that the two-step squeeze casting as an effective route for fabrication of magnesium matrix composites was advantageous over the conventional squeeze casting.

Table 2.

Effects of pressure infiltration process and graphite fiber surface copper coating on relative density of the composites.

No.	Graphite fiber surface state	Pressure application pattern	Infiltration pressure/MPa	Solidification pressure/MPa	Solidification dwelling time/min	Relative density /%
No.	Graphite fiber surface state	Pressure application pattern	Infiltration pressure/MPa	Solidification pressure/MPa	Solidification dwelling time/min	Measured	Average
1	Copper-coated	Two-step	1	30	30	99.1;99.1;99.2	99.1
2	Copper-coated	One-step	30	–	–	96.0;96.2;96.3	96.2
3	Copper-coated	Two-step	1	10	30	97.1;97.3;97.5	97.3
4	Copper-coated	Two-step	1	20	30	98.4;98.5;98.5	98.5
5	Copper-coated	Two-step	1	30	10	97.7;97.8;97.9	97.8
6	Copper-coated	Two-step	1	30	20	98.7;98.7;98.8	98.7
7	Uncoated	Two-step	1	30	30	95.2;95.3;95.3	95.3

By contrast to the properties of sample No. 1, No. 3 and No. 4, it is obvious that the relative density of composites increases as the solidification pressure increases from 10 MPa to 30 MPa. For solidification pressure higher than 20 MPa, the increment becomes slow and the relative density of composite is over 99%. It also can be perceived that an increase in the solidification dwelling time results in an enhancement in the relative density of the composites by comparing the properties of sample No. 1, No. 5 and No. 6. It is thus clear that an appropriate increase in the solidification pressure and solidification dwelling time can eliminate shrinkage defects in composites more efficiently leading to amelioration of relative density of graphite fiber/Al composites. Moreover, it is worth noting that the copper coating on the surface of graphite fiber is beneficial to improve the wettability between Al matrix and graphite fiber reinforcements resulting in the enhancement of relative density from 95.3% (No. 7) to 99.1% (No. 1).

The microstructures of sample No. 1 and No. 2 are shown in Figure 3. The composite made using one-step pressure infiltration technique displays porosities and interfacial cracks are observed clearly (shown by the arrow) in Figure 3(c) and (d). In contrast, no obvious defects are seen and a uniform distribution of graphite fibers is clearly visible no matter in perpendicular to press direction (i.e. X-Y direction) or in parallel to press direction (i.e. Z direction) image (Figure 3(a) and (b)), which further verifies the conclusion that short graphite fiber/Al composites with high relative density can be fabricated by the use of modified two-step vacuum pressure infiltration technique. Besides, the graphite fiber/Al composite shows a strong anisotropy microstructure between the perpendicular and parallel directions due to the press process. It is expected that this orientation of graphite fibers leads to anisotropic properties such as thermal conductivity. Thus, much attention was focused on the thermal conductivity in X-Y direction in this paper.

Effect of infiltration temperature on properties of the composites

For all experimental runs in this part, the infiltration pressure, solidification pressure and solidification dwelling time were selected as 1 MPa, 30 MPa and 30 min, respectively. The copper-coated graphite fiber/Al composites were fabricated under the infiltration temperature within the temperature range of 700–840℃. Figure 4 illustrates the effect of infiltration temperature on relative density and thermal conductivity in X-Y direction of the composites. It can be drawn that raising infiltration temperature helps to increase the relative density and thermal conductivity. As the infiltration temperature reaches 800℃, the relative density nearly remains stable after the temperature increases, while the thermal conductivity decreases to some extent with further increment of the infiltration temperature.

Actually, Martins et al.²⁰ established an infiltration-rate parameter, $Φ$ , to assess the effects of alloy infiltrant, surface preparation modification and processing parameters such as infiltration temperature and pore size of compact can be expressed as:

Φ = \frac{r \times γ_{LV} \times \cos θ}{2 μ}

(3)

where r is the capillary radius,

γ_{LV}

is the surface tension of liquid metal,

θ

is the liquid–solid contact angle and

μ

is the viscosity of liquid metal. An increased value of

Φ

leads to a shorter time for infiltration to a given height, and thus is an indication of improved infiltration kinetics. It can be seen that a rise in the temperature reduces the viscosity of liquid aluminum. The lower viscosity which can increase the value of

Φ

facilitates the rapid compaction of the graphite fiber preforms filled with liquid aluminum. In addition, the contact angle between the liquid aluminum and the copper-coated graphite fiber depreciates with increasing temperature resulting in an increased

\cos θ

and the value of

Φ

. Meanwhile, this reduction in the contact angle improves the matrix-reinforcement interface bonding strength also leading to the enhancement in relative density of composites. Hence, the copper-coated graphite fiber/Al composites with relative density of over 99% can be obtained under a proper high-infiltration temperature.

However, excessively high infiltration temperature will adversely affect thermal conductivity of the composites. During infiltration, Al melt contacts with copper-coated graphite fiber, copper atoms can dissolve into Al matrix and form Al-Cu supersaturated solid solution.¹⁷ For the infiltration temperature higher than 800℃, the undesirable high temperature promotes the diffusion of copper atoms, resulting in the precipitation of Al-Cu secondary phases at room temperature. This secondary phase will be further discussed in the next part. Since the thermal conductivity of Al-Cu intermetallic compound is much lower than that of graphite fiber and aluminum matrix, the solid solution has a negative impact on the thermal conductivity of the composites. Thus, there exists an optimal infiltration temperature of 800℃ for the fabrication of copper-coated graphite fiber/Al composites.

Effect of copper coating on thermal conductivity of the composites

Properties of the composites with uncoated and copper-coated graphite fibers (sample No. 7 and No. 1 in Table 1) are shown in Table 3. It can be clearly seen that the thermal conductivities of the uncoated and copper-coated composites in X-Y direction are both higher than those in Z direction due to the planar orientation of the fibers as mentioned before. It is also worth noting that the thermal conductivity of copper-coated graphite fiber/Al composite is evidently superior to that of uncoated graphite fiber/Al composite. The copper layer can also raise the relative density from 95.3% to 99.1%. The thermal conductivities of composites mainly depend on the thermal conductivities of both matrix and reinforcement, as well as the microstructures of the composites, especially the interface structure.²¹ For uncoated graphite fiber/Al composite, the interfacial structures show weak interfacial bonding. The poor wettability and undesirable interfacial reactions between uncoated graphite fiber and Al matrix have detrimental impacts on the interfacial bonding.¹² As a result, its interface thermal resistance is so high that the composite shows low thermal conductivity. When the uncoated fibers are coated with copper, the wetting between Al matrix and graphite fiber can be transformed into that between Al matrix and copper coating. This transformation can improve the interfacial bonding of the composite and reduce interface thermal resistance between graphite fiber and Al matrix, resulting in higher thermal conductivity of graphite fiber/Al composite.

Table 3.

Properties of the composites with uncoated and copper-coated graphite fiber.

Composites sample No.	Graphite fiber surface state	Relative density/%	Thermal conductivity/W·m⁻¹·K⁻¹
Composites sample No.	Graphite fiber surface state	Relative density/%	X-Y	Z
7	Uncoated	95.3	117	89
1	Copper-coated	99.1	208	146

The typical interfacial structure of the uncoated graphite fiber/Al composite is demonstrated in Figure 5. It can be seen that pores and cracks appear at the interface of uncoated graphite fiber/Al composite. Meanwhile, the presence of fiber-like phase grows from the interface into the aluminum matrix (marked by the arrow). This type of interface structure indicates that the interfacial bonding is very weak and the undesirable interface reaction has been formed between uncoated graphite fiber and aluminum matrix.²² It can be concluded from the XRD result of the uncoated graphite fiber/Al composite shown in Figure 6 that this observed needle or fiber-like phase is Al₄C₃. The presence of brittle Al₄C₃ leads to an unstable interface by destroying graphite fibers and drastically reduces thermal conductivity of the composites.

The distribution of chemical elements at the interface area of copper-coated graphite fiber/Al composite has been analyzed by EDS line scans and the results are presented in Figure 7. As seen from it, graphite fiber is well bonded to the Al matrix by the copper plating layer with the thickness of about 0.5µm, meanwhile, the copper interlayer diffuses into the matrix and forms Al-Cu intermetallic compound. The XRD pattern (Figure 8) indicates this phase is Al₂Cu. It is also noted that no Al₄C₃ phase is observed in the composite (Figure 7) and the absence of Al₄C₃ phase is also confirmed by XRD analysis (Figure 8). Therefore, copper coating can act as middle layer, strengthening the interfacial bonding which prevents graphite fibers from the formation of Al₄C₃.

The fractural surfaces of the uncoated and copper-coated graphite fiber/Al composites (sample No. 7 and No. 1 in Table 1) are shown in Figure 9. As presented in Figure 9(a), both ductile rupture of the matrix and interfacial debonding are observed, the area proportion of interfacial debonding is obviously more than that of ductile rupture of matrix. This is different from the copper-coated graphite fiber/Al composite (Figure 9(b)), the area proportion of interfacial debonding is much less than that of ductile rupture of matrix, which can be explained by the influence of the copper coating which enhances the adhesion of Al matrix to the graphite fiber. Hence, all results above show a good potential for copper coating to be used as surface metallization layer to fabricate graphite fiber/Al composite.

Conclusions

Short graphite fiber/Al composites with high relative density can be fabricated by the use of a modified two-step vacuum pressure infiltration technique. The enhanced solidification pressure and solidification dwelling time are conductive to the amelioration of relative density of short graphite fiber/Al composites. The optimum solidification pressure range is 20–30 MPa and the appropriate solidification dwelling time range is 20–30 min. The application of copper coating on graphite fibers effectively improves the interfacial bonding between graphite fibers and Al matrix by the diffusion of copper atoms and formation of Al₂Cu supersaturated solid solution in the matrix. While, higher infiltration temperature induces the formation of excess Al₂Cu phase, which is detrimental to the thermal conductivity of the composites. Thus, there exists an optimal infiltration temperature of 800℃ for the fabrication of copper-coated graphite fiber/Al composites. Copper-coated graphite fiber/Al composites with relatively high density of 99.1% and thermal conductivity of 208 W·m⁻¹·K⁻¹ are potential candidates for thermal management applications.

Footnotes

Conflict of interest

None declared.

Funding

This project is financially supported by the National Nature Science Foundation of China (Grant No. 51274040) and the Fundamental Research Funds of the Central Universities (FRF-TP-10-003B).

References

Mallik

Ekere

Best

. Investigation of thermal management materials for automotive electronic control units. Appl Therm Eng 2011; 31: 355–362.

Prieto

Molina

Narciso

. Fabrication and properties of graphite flakes/metal composites for thermal management applications. Scripta Mater 2008; 59: 11–14.

Veillère

Heintz

Chandra

. Influence of the interface structure on the thermo-mechanical properties of Cu-X (X = Cr or B)/carbon fiber composites. Mater Res Bull 2012; 47: 375–380.

Zhang

. Review of metal matrix composites with high thermal conductivity for thermal management applications. Prog Nat Sci 2011; 21: 189–197.

Ren

Guo

. Net-shape forming and properties of high volume fraction SiCp/Al composites. J Alloys Compd 2009; 484: 256–262.

Liang

Jia

Chu

. Thermal conductivity and microstructure of Al/diamond composites with Ti-coated diamond particles consolidated by spark plasma sintering. J Compos Mater 2012; 46: 1127–1136.

Surappa

. Aluminium matrix composites: challenges and opportunities. Sadhana 2003; 28: 319–334.

Daoud

. Microstructure and tensile properties of 2014 Al alloy reinforced with continuous carbon fibers manufactured by gas pressure infiltration. Mater Sci Eng A 2005; 391: 114–120.

Shalu

Abhilash

Joseph

. Development and characterization of liquid carbon fiber reinforced aluminum matrix composite. J Mater Process Technol 2009; 209: 4809–4813.

10.

Liu

Tang

. Friction and wear properties of short carbon fiber reinforced aluminum matrix composites. Wear 2009; 266: 733–738.

11.

Pelleg

Ashkenazi

Ganor

. The influence of a third element on the interface reactions in metal-matrix composites (MMC): Al-graphite system. Mater Sci Eng A 2000; 281: 239–247.

12.

Wang

Chen

. Microstructure and mechanical properties of graphite ﬁber-reinforced high-purity aluminum matrix composite. J Mater Sci 2009; 44: 4303–4307.

13.

Rams

Ureña

Escalera

. Electroless nickel-coated short carbon fibres in aluminium matrix composites. Compos Part A 2007; 38: 566–575.

14.

Zweben

. Advances in high-performance thermal management materials: a review. J Adv Mater 2007; 39: 3–10–3–10.

15.

Rawal

. Metal-matrix composites for space applications. JOM 2001; 53: 14–17.

16.

Abraham

Pai

Satyanarayana

. Copper coating on carbon fibers and their composites with aluminum matrix. J Mater Sci 1992; 27: 3479–3486.

17.

Ureña

Rams

Escalera

. Effect of copper electroless coatings on the interaction between a molten Al-Si-Mg alloy and coated short carbon fibers. Compos Part A 2007; 38: 1947–1956.

18.

Dou

Jiang

. Damping properties of aluminum matrix-fly ash composites. Mater Lett 2006; 60: 2945–2948.

19.

Wang

. Fabrication and mechanical properties of SiCw/ZK51A magnesium matrix composite by two-step squeeze casting. Mater Sci Eng A 2000; 278: 267–271.

20.

Martins

Olson

Edwards

. Modeling of infiltration kinetics for liquid metal processing of composites. Metall Trans B 1988; 19: 95–101.

21.

Ren

Shen

Guo

. Effect of coating on the microstructure and thermal conductivities of diamond-Cu composites prepared by powder metallurgy. Compos Sci Technol 2011; 71: 1550–1555.

22.

Etter

Schulz

Weber

. Aluminium carbide formation in interpenetrating graphite/aluminium composites. Mater Sci Eng A 2007; 448: 1–6.