PORE SEALING PASTES FOR POROUS MATERIALS

Abstract

Embodiments of the present invention disclosed herein use innovative pastes to fill surface pores (cavities) and flatten (planarize) surfaces of porous materials. A method for making a heat transfer apparatus comprises making a paste comprising particles of a first heat transfer material, a vehicle, and a binder, filling cavities on an external surface of a second heat transfer material with the paste, and drying the paste filled in the cavities so that an external, surface of the dried paste in a cavity is substantially planar with the external surface of the second heat transfer material.

Description

This application claims priority to U.S. Provisional Patent Application Ser. No. 61/655,749, which is hereby incorporated by reference herein.

BACKGROUND AND SUMMARY

Advanced thermal management materials are becoming critical for today's power electronics, photonics, and photovoltaics industries. Powerful packaging requires materials capable of effectively dissipating heat while maintaining compatibility with the package and die. Most traditional low coefficient of thermal expulsion (“CTE”) materials, such as tungsten or copper, have thermal conductivities that are no better than those of aluminum alloys (approximately 200 W/m-K). On the other hand, the outstanding thermal and mechanical properties of carbon allotropes have driven considerable interest in the development of novel thermal transfer materials based on carbon. These carbon materials are expected to be superior over others with respect to CTE, thermal conductivity, and density. However, the as-fabricated graphitic blocks are often highly porous, with pores ranging in the nanometer to millimeter diameters. These pores, especially the pores for cavities) on the surfaces, strongly inhibit surface functionalization processes (e.g., plating and dielectric coatings), limiting the graphitic materials' potential applications, and also have a negative impact on the thermal and mechanical properties between the thermal interface material and heat sink attachment parts. Therefore, highly thermal conductive and with good CTE matched filling material is urgently needed to fill up the surface cavities on graphitic carbon matrix material substrates.

Embodiments of the present invention disclosed herein use innovative pastes to fill surface pores (cavities) and flatten (planarize) surfaces of porous materials.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1A shows a digital image of an untreated and porous surface.

FIG. 1B shows a digital image of the surface shown in FIG. 1A after being treated in accordance with embodiments of the present invention.

FIG. 2A illustrates a poor interface contact due to air gaps (cavities) caused by surface porosity in a material.

FIG. 2B illustrates an improved interface contact due to minimized air gaps after filling the pores (cavities) to reduce the surface porosity in the material.

FIG. 3 shows a digital image of powdered carbon dust collected from a machining process.

FIG. 4 shows surface profile graphs of an untreated and porous graphite surface (upper graph) and the graphite surface after being treated in accordance with embodiments of the present invention (lower graph).

FIG. 5 illustrates a flow diagram configured in accordance with embodiments of the present invention.

DETAILED DESCRIPTION

Thermal management deals with moving heat energy from high concentration to low concentration. Thermal management is especially important in electronic devices where small, discrete components can generate incredible amounts of heat (up to megawatt heat loads) in small areas. These devices must be cooled with an overall solution to thermal management that may include the heat source, heat spreader, heat sink, liquid cooling heat sink, and interfaces between each of these independent layers. The quality of the interface between two adjacent layers or materials is important for efficient energy transfer between the layers. Other energy transfer could be magnetic or optical.

Thermal conductivity is a physical value that determines how much heat can be dissipated based on the gradient in temperature between the heat source and the heat dissipation layer or heat sink. The thermal conductivity is derived from the total heat transferred as a function of the gradient in temperature according to Fourier's Law. Thermal conductivity may be expressed as κ=−q_x/(δT_x/δx), where κ is the thermal conductivity in W/mK, q_x is the heat flux in W/m², and (δT_x/δx ) represents the temperature gradient in the direction of the heat flux expressed in K/m.

The overall heat transfer (described above as heat flux, q) is the rate of heat transfer as a function of area. Therefore, the larger the area to transfer heat the faster heat can be moved from one material to another. The area does not just include the physical surface area of the part, but has to take into account the contact area between two parts. If heat is transferred into an atmosphere, the limit of this description equates to the total surface area. If heat is transferred between two physical parts, the contact area between the two parts is the limiting factor.

Most good electrical conductors are also good thermal conductors. Phonons (crystal lattice vibrations) are the primary mechanism for heat transfer. Tightly packed, regular lattice structures in metals provide efficient thermal phonon conduction—hence, metals typically have high thermal conductivity values. Some non-metallic materials can also have high thermal conductivity. For example, aromatic carbons, including graphite and other high-order carbon materials, can have excellent thermal phonon conduction due to the extended carbon-carbon, pi-bond network, which also makes these materials excellent electrical conductors.

Despite being good thermal conductors, metals are often not the best choice for applications in electronics or high temperature devices. Metals also have high thermal coefficients of expansion (“CTE”), and this can cause significant stress between adjoining layers as the temperature increases. The high CTE is due to low interatomic (intermolecular) forces (see, “Properties of Materials: Anisotropy, Symmetry, Structure: Anisotropy . . . ,” Robert F. Newnham, Oxford University Press. Nov. 11, 2004, pp 82-84). In contrast, materials with high intermolecular forces typically have low CTE. For example, graphitic carbon materials have high intermolecular forces, and thus a low CTE. Due to the high thermal conductivity and low CTE, this makes graphitic materials an excellent candidate for thermal management interface and heat sink materials.

Graphite can have extremely high thermal conductivity in one plane in a direction parallel to the direction of the graphitic plane (in-plane thermal conductivity can be as high as 1100 W/mK). Orthogonal to the crystal plane, the thermal conductivity is poor due to lack of orbital overlap in the extended pi-bond structure, (out-of-plane thermal conductivity can be less than 10 W/mK). In order to provide a more uniform thermal conductivity throughout the multi-dimensional solid, some anisotropy must be built into the solid. In this case, a graphitic composite matrix will have uniform thermal conductivity in the X, Y, and Z planes, yet the maximum thermal conductivity will not be as high as graphite in-plane values.

Graphitic carbon matrix materials can be made with many methods. The most common method is to take graphitic particles (e.g., from natural sources or needle coke products) and mix them with other carbonaceous materials as a binder. The mixture is then subjected to intense heat (e.g., up to 3000° C.) and pressure (e.g., up to 10 MPa) such that the carbon-based binder materials (sp- and sp² Carbon) are converted to graphitic carbon (sp³) as well. The average thermal conductivity of a graphitic carbon matrix will be above 300 W/mK in all three dimensions. The local domains of the initial graphitic particles can be aligned during the pressurization process to provide preferential thermal and electrical conductivity in one axis of direction, with such preferential thermal conductivity as high as 425 W/mK. However, due to changes in uniformity, the average value is often approximately 360 to 380 W/mK.

A disadvantage with most graphitic carbon materials is the porosity in the matrix. The pores are a result of mass and volume loss conversion of carbon-based binders into graphitic-like carbon (sp³). Porosity is an open void space within the graphitic carbon matrix, such as shown in the image of FIG. 1A. The pores can cause the graphitic matrix to be gas and liquid permeable. The density of the carbon can dictate the overall thermal conductivity. Higher density carbon (e.g., density of approximately 1.6 to 2.1 g/cm³) is desired such that it has reduced void volume (i.e., reduced porosity) and a measured increase in thermal conductivity. The assignee has used high thermal conductivity carbon materials where a density of 1.75 g/cm³ has a thermal conductivity of 300 W/mK, and another carbon material that has a density of 1.80 g/cm³ and a thermal conductivity of 350 W/mK.

The pores or cavities effectively decrease the contact surface area between the carbon-based thermal conductor and the heat sources. Recall that total heat flux is area dependent. Voids within a material reduce the overall contact area, and thus reduce the total heat flux between two adjacent layers. Embodiments of the present invention provide materials and methods to reduce the pore volume and provide increased contact area between a carbon-based heat transfer material and an adjacent heat source. Referring to FIG. 2A, if two substrates A and B are to be adjoined for the purpose of transfer of energy (e.g., heat) between the substrates, but at least one of the substrates B is made from a porous material, the cavities on the surface of the substrate B will result in air gaps, reducing the overall contact area, and thus reducing the total energy (e.g., heat) flux between the two adjacent layers. Referring to FIG. 2B, if the substrate(s) that is porous has its cavities filled in accordance with embodiments of the present invention, the air gaps at the interface are reduced or minimized, increasing the overall contact area, and thus increasing the total energy (e.g., heat) flux between the two adjacent layers.

Embodiments of the pore filling material have several characteristics: low volume reduction after curing, matched CTE to the parent material, and high thermal conductivity. The volume reduction is low during the curing process such that the pores are filled so that the top surface remains substantially level to the surrounding external surface of the heat transfer material. Any shrinkage will partially fill a pore leaving no surface for contact to an adjacent interface surface (e.g., of a heat source). In contrast, pores may be overfilled and then planarized (e.g., using well-known methods) after curing such that a smooth, planar surface is achieved. The CTEs of the binder and filler should match the host material. If the CTE is too large, this will create stress that may damage the host heat transfer material or push it off the surface of the adjacent heat source interface as the temperature increases. If the CTE is too small, additional pores, cavities, or voids might be created as the temperature increases. The thermal conductivity of the material should be high. The thermal conductivity of air (from an unfilled pore) is typically 0.024 W/mK. Thermal greases can have thermal conductivity up to 1-2 W/mK. While this is a significant improvement (greater than 100×), the thermal grease will coat the interface between the adjacent materials in addition to filling the pores and removing the air gap. In embodiments of the present invention, only the pores are filled with high thermal conductivity materials, leaving the host material in contact with the thermal source interface.

The pore filling material may be comprised of a combination of vehicle, binder, and/or filler. The vehicle material may be a transfer medium that has matched surface energy such that the pore filling material will completely fill and wet the sidewalls of the pores. The vehicle may have matched surface energy to the surface of the graphitic matrix. The matched surface energy provides low surface tension and enhances pore filling efficiency through a wicking and/or capillary action. The vehicle may be a well-known solvent. The binder may dissolve in the vehicle. The vehicle may have a viscosity modifier. The binder material should provide excellent adhesion between the sidewalls of the pores in the host material and the filler material. The binder materials should have high temperature resistance with a maximum temperature exceeding 260° C. before decomposition of phase change. This temperature range may be selected as a function of the melting temperature of many solders used in typical electronic devices. In some cases, the binder temperature should be stable up to approximately 550° C. such that high temperature operation is possible. For example, SiC based transistors can operate at several hundred degrees and up to 500° C. (e.g., a NASA 6H—SiC transistor).

The filler material should have high mass loading of typically greater than 60%, optionally about 80%. The filler material may be suspended or homogenously distributed in the vehicle or binder. After curing, the filler loading may increase as a result of removal of the vehicle through evaporation or reduction of volume of the binder due to polymerization. The filler material may conduct through physical contact between adjacent particles and physical contact between the filler and the material in the host substrate matrix. The filler material may substantially be made from grinding the parent substrate material to which it will be applied. The filler material may have a high thermal conductivity (e.g., >200 W/mK). The filler material may be made from high thermal conductivity ceramic materials (e.g., aluminum nitride, boron nitride, AlSiC, silicon carbide, or other carbides and/or nitrides), metallic fillers (e.g., aluminum, copper, copper-tungsten, copper-molybdenum, silver and/or graphitic particles (e.g., fullerenes, carbon nanotubes, graphite). The graphitic material may also be made from dust (e.g., from the sanding and machining process on a graphitic carbon matrix material) generated from the parent substrate material.

In some examples, functions of the vehicle, binder, and/or filler may be combined. For example, the binder may be a liquid prior to curing such that it undergoes a liquid to solid transition upon a curing process. The curing process may be thermal, chemical, and/or photonic. The binder and filler may be combined. For example, an electroplated metal with high adhesion may be both a binder and a filler. A molten metal may also be used as a filler. Metallic particles may be used as a filler without a binder. When the particles are melted together, they firm a continuous network with low porosity filling the surface pores of the host matrix.

The pore filling material may be applied to the host carbon matrix thermal material conductor using a variety of techniques. The pore filling material is applied to the substrate of the host material, some pressure may be used to promote complete pore filling, and the material is cured to form a solid material form. In each example, the excess material may be removed, resulting in a planar surface with high contact area and low void area. Specific embodiments are described below.

Disclosed herein are embodiments of using a graphitic particle-based pore sealing paste to fill the graphitic substrate surface pores. Referring to FIG. 5, in a first step a porous substrate or material 501 is to have at least one of its surfaces smoothed or planarized in step 502. Step 502 may be performed by collecting particles (e.g., dust or powders) from the host porous substrate 501 in a step 502A. Then, a paste is made with the particles in step 502B in accordance with the various embodiments described herein. In step 502C, the paste is applied to a surface of the host porous substrate. Subsequent steps as described herein may also be performed to complete the sealing of the surface in step 503. Optionally, any excess paste material may be removed in step 504. The paste now filled into the cavities of the substrate surface is dried in step 505. The substrate may then be adjoined so that the sealed surface is in contact with a secondary substrate in step 506. The following examples implement some or all of the foregoing steps.

The host substrate material may be a porous graphite material that has a high thermal performance, such as high thermal conductivity (e.g., approximately 300-350 W/mK). The graphitic particles may be obtained by surface grinding the host substrate material, and used to formulate the pore filling (sealing) paste due to its higher coefficient thermal expansion property and better bending strength compared to regular carbon graphite powders. The superb thermal properties and mechanical strength of ground graphitic carbon powders provide better interface interaction to the host graphitic substrate.

In an embodiment, a pore filler was synthesized by grinding a graphitic carbon matrix to a powder (e.g., with resulting particle size ranging from 0.1-100 micrometers). An example of these powdered materials is shown in FIG. 3. The graphitic powders are combined with a lithium polysilicate binder material and mixed to form a paste. The binder weight percent may vary from approximately 20-40%. The powder weight percent may vary from approximately 60-80%. The ratio of powder to binder may be approximately 1:2 (wt %). The paste may be mixed using a counter rotating high-shear mixing system well known in the art.

The paste is applied to the substrate (e.g., using a bristle brush or roller). The substrate may be a graphitic carbon substrate or a porous ceramic substrate. The mechanical force of the bristles or force of the roller will help completely fill the pores of the substrate. Any excess of the paste material may be removed (e.g., with a wiping action using a dry cloth). Further surface smoothing may be accomplished (e.g., using a wet towel for water soluble binder materials). The pore filling paste is then dried (e.g., in ambient air for 30 minutes). Additional applications of the pore filling paste may be applied to ensure complete pore filling coverage. The sample may be additionally baked (e.g., in ambient air at 200° C. for up to 60 minutes) to set the binder layer.

After application of the pore filling paste material, the surface is very smooth, with no visual surface pores (cavities) (e.g., see FIG. 1B). Before sealing, the graphitic surface had a roughness of −40-+10 microns as shown by the surface profile graph (upper graph) in FIG. 4 (which may be produced using a contact profilometer trace), and after sealing, the surface had a roughness of −4-+1 microns, as shown in the surface profile graph (lower graph) of FIG. 4. This represents an order of magnitude reduction in face roughness due to porosity in the graphitic matrix. In addition to the improvement in surface roughness, the sample was no longer gas permeable nor liquid permeable.

In another embodiment, a pore filler was synthesized by grinding a graphitic carbon matrix to a powder (e.g., with resulting particle size ranging from 0.1-100 micrometers). These powdered materials are shown in FIG. 2. The graphitic powders are combined with a polyphenyl silsesquioxane (“PPSQ”) binder material and mixed to form a paste. The binder weight percent may vary from approximately 20-40%. The powder weight percent may vary from approximately 60-80%. The ratio of powder to binder may be approximately 1:2 (wt %). In some examples, a solvent such as terpineol or toluene, may be added to ensure solubility of the PPSQ binder materials. The paste may be mixed using a counter rotating high-shear mixing system.

The paste is applied to the substrate (e.g., using a bristle brush or roller). The substrate may be a graphitic carbon substrate or a porous ceramic substrate. The mechanical force of the bristles or three of the roller will help completely fill the pores of the substrate. Any excess of the paste material may be removed (e.g., with a wiping action using a dry cloth). Further surface smoothing may be accomplished (e.g., using a wet towel for water soluble binder materials. The pore filling paste is then dried at 200° C. in ambient air for 30 minutes. Due to slight binder shrinkage, Additional applications of the pore filling paste can be applied to ensure complete pore filling coverage.

In another embodiment, a pore filler was synthesized by dispersing metal particles into a vehicle made from turpineol and ethylcellulose. The metallic powders can range in size from less than 10 nm to 1-2 micrometers. Smaller particles are required to fill smaller pores. A mixture of nanoparticles and microparticles can be used to facilitate increased packing density. The vehicle weight percent can vary from 5-40%. The powder weight percent can vary from 60-95%. The preferred ratio of powder to vehicle is 4:1 (wt %). The paste is thoroughly mixed using a counter rotating high-shear mixing system.

The paste may be applied to the substrate using a bristle brush or roller. The substrate can be a graphitic carbon substrate or a porous ceramic substrate. The mechanical force of the bristles or force of the roller will help completely fill the pores of the substrate. Any excess of the paste material may be removed (e.g., with a wiping action using a dry cloth). Further surface smoothing may be accomplished (e.g., using a solvent wet towel and a wiping motion). The pore filling paste may be dried (e.g., in ambient air at 100° C. for 30 minutes). Additional applications of the pore filling paste may be applied to ensure complete pore filling coverage. The paste may be cured (e.g., by heating the metal particle filler composite to 450° C. in a reducing atmosphere containing 4% hydrogen in a balance of nitrogen).

After curing, the metal particles may be sintered creating a dense metallic network that is electrically and thermally conducting. The metals may be copper and/or silver. Aluminum particles may also be used. Copper makes a quality interface with carbon matrices.

In this embodiment, the pores of the graphite substrate matrix are filled with metal using an electroplating process. The graphite matrix is electrically conductive and conducive to plating. The sample may be immersed in a commercial copper plating solution. The samples may be electroplated by grounding the graphite substrate and applying (+) voltage to a copper counter electrode. The samples may be plated at a constant current ranging from 0.5-1.0 mA/mm² of substrate surface area. The copper electroplates preferentially in the pores due to the increased electric field generated by the non-uniform surface structure. The external surface of the graphite matrix may also plate with copper. After plating the sample may be rinsed (e.g., with deionized water). The surface may be further sanded or polished to a smooth finish leaving a high-quality surface with low pore density.

Pore sealing pastes are not limited to treating porous material surfaces; they may also be pressed into the pores in the bulk to fill them to improve the mechanical and thermal performances of the original porous materials. In another embodiment, a pore filler was synthesized by mixing ceramic powder with resulting particle size ranging from 10 nm-10 micrometers. These ceramic powders may include boron nitride, aluminum nitride, and/or aluminum silicon carbide. The powders are combined with a lithium polysilicate binder material and mixed to form a paste. The binder weight percent may vary from approximately 20-40%. The powder weight percent may vary from approximately 60-80%. The ratio of binder to polymer may be greater than 1:2 (wt %). The paste may be mixed using a counter rotating high-shear mixing system.

The paste is applied to the substrate (e.g., using a bristle brush or roller). The mechanical force of the bristles or force of the roller will help completely fill the pores of the substrate. Any excess of the paste material may be removed (e.g., with a wiping action using a dry cloth). Further surface smoothing may be accomplished (e.g., using a wet towel for water soluble binder materials). The pore filling paste may be then dried (e.g., in ambient air for 30 minutes). Additional applications of the pore filling paste may be applied to ensure complete pore filling coverage. The sample may be additionally baked (e.g., in ambient are at 200° C. for up to 60 minutes) to set the binder layer.

Claims

1. (canceled)

2. (canceled)

3. (canceled)

4. (canceled)

5. (canceled)

6. (canceled)

7. (canceled)

8. (canceled)

9. An energy transfer apparatus comprising a first energy transfer material, wherein cavities on an external surface of the first energy transfer material are filled with a dried paste comprising a second energy transfer material and a binder suitable for making the paste.

10. The apparatus as recited in claim 9, wherein the first energy transfer material is a graphitic carbon matrix.

11. The apparatus as recited in claim 9, wherein the first energy transfer material is a porous ceramic material.

12. file apparatus as recited in claim 9, wherein the second energy transfer material is the same as the first energy transfer material.

13. The apparatus as recited in claim 9, wherein the second energy transfer material is graphite powders.

14. The apparatus as recited in claim 9, wherein the second energy transfer material is metal particles.

15. The apparatus as recited in claim 9, wherein the second energy transfer material is ceramic powders.

16. (canceled)

17. (canceled)

18. (canceled)

19. (canceled)

20. (canceled)

21. (canceled)

22. (canceled)

23. (canceled)

24. (canceled)

25. The apparatus as recited in claim 9, wherein the energy is thermal energy.

26. The apparatus as recited in claim 9, wherein the dried paste comprises particles of a same composition as the first energy transfer material, and a binder suitable for creating the paste.

27. The apparatus as recited in claim 26, wherein the first transfer material is a porous graphite, and the particles are graphite powders.

28. The apparatus as recited in claim 27, wherein the binder comprises lithium polysilicate.

29. The apparatus as recited in claim 27, wherein the binder comprises polyphenyl silsesquioxane.

30. The apparatus as recited in claim 27, wherein a ratio of the particles to binder is about 1:2 wt %.

31. The apparatus as recited in claim 9, further comprising a second energy transfer material having a planar surface in physical contact with the external surface of the first energy transfer material.

32. The apparatus as recited in claim 12, further comprising a third energy transfer material having a planar surface in physical contact with the external surface of the first energy transfer material.

33. The apparatus as recited in claim 26, wherein the first energy transfer material has a density of about 1.75 g/cm3 and a thermal conductivity of about 300 W/mK.

34. The apparatus as recited in claim 26, wherein the first energy transfer material has a density of about 1.80 g/cm3 and a thermal conductivity of about 350 W/mK.

35. A heat transfer apparatus comprising a first heat transfer material, wherein pores in the first heat transfer material are filled with a dried paste obtained from a composition comprising particles of the first heat transfer material, a vehicle, and a binder.

36. The heat transfer apparatus as recited in claim 35, wherein a source of the particles of the first heat transfer material is the first heat transfer material.

37. The heat transfer apparatus as recited in claim 36, wherein the first heat transfer material is a graphitic carbon.