Characterization of Electrolessly Plated Graphite Foams with Particle Additions

Characterization of Electrolessly Plated Graphite Foams with Particle Additions

Michael Asaro, Jennifer Mueller, Patrick Dykema

Virginia Polytechnic Institute and State University, Department of Materials Science & Engineering 213 Holden Hall, Blacksburg, VA 24061


Abstract

With a low density and high bulk thermal conductivity, graphite foams are ideal for thermal management systems such as computer heat sinks, radiators, and heat exchangers. Previous work has shown it is possible to improve the foams performance by opening the porosity with nanoparticle additions in the foams precursor, an oil based mesophase pitch.[1] The open porosity allows more fluid, such as air or water, to pass through the foam and carry heat away. The original study, performed by Jennifer Mueller at Oak Ridge National Labs (ORNL), considered the concentrations of nanoparticle’s used. The present study looked to determine the effects of using different types of nanoparticles at a range of sizes from the nano- to micro- level. The study began by adding, respectively, silver, ceria, alumina, tungsten, and nickel to different batches of mesophase pitch at a single weight percent concentration to create graphite foams with a significant amount of continuous porosity. The pitch was foamed, carbonized, and graphitized. The final foam products were then measured in a variety of ways including thermal conductivity, permeability, and scanning electron microscope (SEM). As a side project, an electroless copper plating solution was passed through the foams to determine if a continuous and uniform copper coating could be built up. The copper coating that was eventually obtained coated the foam walls without filling the open porosity and may help to increase the foams solderability, strength, durability, and corrosion resistance.

Keywords: Carbon, Foam, Thermal, Electroless

1. Introduction

Graphite foams are an excellent material for use in thermal management systems such as computer heat sinks, radiators, and heat exchangers. The graphite ligaments themselves have a thermal conductivity of >1700 W/mK.[2] As can be seen in Figure 1, they are usually aligned parallel to the pore structure, which gives the bulk foam a thermal conductivity on par with aluminum. The foams also exhibit a very low density of approximately 0.5g/cm3, one-fifth that of aluminum.[2]

Previous work has shown that the addition of carbon nanoparticles to the foam precursor, an oil based mesophase pitch, will act as nucleation sites for pore windows during the foaming process. The open porosity will create a more permeable foam with a greater capacity to let fluid move through it; this will create a better performing system with greater overall efficiency. Whereas the previous work looked at the effects of varying the concentrations of the nanoparticles, the new work looked to determine how particles of different materials and varying sizes at the same concentration, 5 weight percent, would effect the pore structure.

2. Procedure

2.1 Foam Preparation

Ceria, nickel, tungsten, silver, and alumina particles were each added to the mesophase pitch precursor, at 5 weight % concentrations, and processed to create graphite foam. Their measured particle size can be seen in Table 1 where the particles ranged from 11um to 50nm (the analyzer used was unable to get an accurate reading on the 50nm alumina because the particles were too small). The pitch-particle mixture was heated in an oxygen-free atmosphere to approximately 50℃ above the pitch’s softening point. The pressure and temperature in the furnace were then raised after the pitch had melted. In its molten stage, the pitch formed bubbles at nucleation sites, and mesophase crystals formed. The mesophase pitch began to polymerize and the bubbles were trapped in place creating a foam structure. After the foaming process the material was carbonized by slowly ramping a furnace up to 1000℃. Finally, the carbonized foam was graphitized at 2800℃ in an inert atmosphere of nitrogen and argon.

2.2 Permeability

Tests were conducted at Oak Ridge National Laboratory on an apparatus developed to measure the face velocity of air flowing into and out of the system, thus giving a pressure drop. A (2 × 2 × 0.25 inch) block of graphite foam was placed into the system and the lid tightly sealed. The flow rate was manually adjusted to obtain face velocities varying from 0 to 300 Ipm and the corresponding pressures were re5corded.

2.3 Scanning Electron Microscopy (SEM)

Samples were viewed with a LEO 1550 Field Emission SEM with an accelerating voltage of 5 kV to view the graphitized samples. Since the carbon foam is already electrically conductive, the samples did not need to be sputtered with a metallic coating before being analyzed.

2.4 Electroless Plating

Electroless plating consists of three major steps: sensitizing, activating, and plating. Sensitizing helps to gather all the readily oxidized material on the surface. The activating step lays down a layer that provides catalytic nucleation sites. The final plating stage builds a uniform autocatalytic layer on the surface that will continue to thicken as long as new solution is passed over the surface.

Particle Size
Mean Size (um) Median Size (um)
Alumina (0.05) Not Accurate
Alumina (0.30) 1.95 0.27
Ceria 5.39 4.99
Nickel 12.69 9.69
Silver 0.62 0.47
Tungsten 11.84 10.72
Figure 3. Particle size data

2.5 Thermal Conductivity

Thermal diffusivity was measured using a xenon flash diffusivity technique and thermal conductivities were calculated using Equation 1.[5] The technique involves a bulb flashing for less than 1 millisecond and releasing a small amount of heat on one side of the sample. An IR camera at the other side measures the change in temperature.

Where:
α is the thermal diffusivity
κ is the thermal conductivity
ρ CΡ is the heat absorbed by the material

3. Results and Discussion

3.1 Particle Additions

The foams that were processed at higher processing pressures had a higher pressure drop, meaning there was less open porosity, compared to the foams processed at lower pressures for a given material (Figure 4). Interestingly though, the ceramic particles caused more open porosity than the metal particles at the two lower processing pressures, but were about equal at the 400psi processing pressure.

At the 400 psi pressure level, very little change can be seen in the thermal conductivity of the foams. The 200 and 300 psi runs, however, show significant differences between the metal and ceramic particles. See Figure 5.

The larger of the alumina and silver particles had respective median sizes of 0.3um and 0.5um. The nickel and tungsten particles had a median size of approximately 10um. As can be seen in Figure 5 below, the trend in these numbers is not consistent with the trend in thermal conductivity. This means that cheaper particles of larger sizes can be used to increase the thermal conductivity just as effectively as the smaller particles.

3.2 Copper Plating

As can be seen in Figure 6 a copper coating was obtained. The EDS analysis in Figure 7 shows that the pores are indeed copper while the ligaments remain pure carbon. Figure 8 shows a picture of the copper plated foam taken with the SEM. The non-coated pores could have occured either because no plating solution reached the pore or the copper fell out during the cutting process.

4. Conclusions

Adding particles to the foam precursor helped to increase the amount of open porosity and thus allow more heat to be dissipated into the atmosphere more quickly. The specific properties of the particles that affect the amount of porosity have yet to be determined, but there is a very clear trend that at lower processing pressures (less than 400 psi), ceramic particles have a much greater effect than metal ones. Thermal conductivity of the foams was also aided by the addition of particles at the lower processing pressures, but the specific properties that are most important are still unknown.

EDS analysis showing the pore as 100% copper and the ligament as 99.96% carbon Figure 7. EDS analysis showing the pore as 100% copper and the ligament as 99.96% carbon
An SEM image showing numerous pores coated with copper Figure 8. An SEM image showing numerous pores coated with copper

The feasibility of coating the foam with a uniform layer of copper was shown. Other plating materials should be possible but at this point are untested.

Acknowledgements

We would like to thank graduate student, Ben Poquette for his help with numerous aspects of the project and most importantly his support in the lab during the copper plating experiments. The work was done under the supervision of Dr. Steve Kampe with the aid and ideas of Dr. Gary Pickrell and Dr. Marie Paretti. Their input was invaluable.

References

[1] Gallego, Nidia C. and Klett, James W. 2003. Carbon Foams for Thermal Management. Carbon 41: 1461-1466.

[2] Klett, James W. High Thermal Conductivity Foams One Pager. Available at: http://www.ms.ornl.gov/researchgroups/CMT/cmt.htm, Accessed on October 12, 2006.

[3] Klett, James and McMillan, April. Modeling Convective Heat Transfer with Graphite Foams. Available at: http:// www.ms.ornl.gov/researchgroups/cmt/foam/Convective%20Heat%20Transfer.pdf. Accessed on April 2, 2007.

[4] Klett, James W. High Conductivity Graphite Foams and Applications.” Available at: http://www.ms.ornl.gov/researchgroups/CIMTECH/foam/one-pagers.pdf, Accessed on April 2, 2007

[5] Klett, J. W., McMillan, A. D., Gallego, N. C., Walls, C. A. 2004. The Role of Structure on the Thermal Properties of Graphitic Foams Journal of Materials Science 39: 3659-3676.

About the Author

Michael Asaro

Michael Asaro is a senior at Virginia Polytechnic Institute and State University in Blacksburg, VA. After graduation this spring he will begin graduate work in Materials Science and Engineering at Virginia Tech as well as a career at Nanosonic, Inc., a company also located in Blacksburg.