WO1999018030A2 - Diamond-based composites with high thermal conductivity - Google Patents

Abstract

A diamond based composite comprising diamond powder distributed in a non-metallic matrix material, the filling factor being between about 15 % and about 75 % and the conductivity of the comprising being substantially greater than the conductivity of the matrix material without diamond. The composite is dielectric and can include magnetic or fibrous materials. It can be used in biochips, electronic devices, and residential and industrial materials where thermal load management is critical.

Description

DIAMOND-BASED COMPOSITES WITH HIGH THERMAL CONDUCTIVITY

BACKGROUND OF THE INVENTION

The invention relates to high thermal conductivity diamond-based composites (DBCs) having a low electrical conductivity. The DBCs according to the invention offer better thermal and mechanical parameters than prior materials and are much more practical and cheaper to use.

Practical and theoretical obstacles have led prior researchers away from the present invention. Diamonds are known to have desirable properties of high thermal conductivity, about 22 Watts/cm/ °K. Diamond has been used in films and in sintered forms in an effort to take advantage of its thermal properties. However, such approaches are of limited applicability and are costly.

For example, U.S. Patent No. 5,270,114 is to Herb et al. is directed to high thermal conductivity diamond/ non-diamond composite materials which are deposited on a silicon wafer using a CVD technique. The thermal conductivity achieved by the composite materials taught by Herb et al. is greater than 17 Watts/cm/°K at about 20°C. According to Herb et al. , high conductivity diamond material is deposited by vapor deposition onto non-diamond particles.

A diamond composite heat sink for use with semiconductor devices is disclosed in U.S.

Patent No. 5,008,737 to Burnham et al. where the diamond particles are embedded in a metal matrix. Depending upon the purity of the diamond, the thermal conductivity ranges from about 9 to 23 W/cm/°C.

A polycrystalline diamond of improved thermal conductivity is discussed in U.S.

Patent No. 5,540,904 to Bovenkerk et al. which consists essentially of 99.5 wt-% isotopically- pure carbon- 12 or carbon-13. In many branches of industry or technology, thermal management is essential.

Thermal bottlenecks present serious limitations in many applications, including modern biomedical devices, microfluidics supports, and high speed/ very large integration electronics.

For example, the heat created inside chips cannot escape through the plastic or ceramic packaging; the clock speed and complexity of many VLSI circuits is limited by these heat removal problems. Thus, dielectric materials with better thermal conductivity are highly sought after and commercially desirable.

Modern microdevices used in medical diagnostics, especially microfluidics, involve material transfer under high voltage gradient or high pressure. The miniaturization of devices for electrophoresis, capillary electrophoresis (CE) and high pressure chromatography (HPLC) require a combination of high dielectric breakthrough, excellent thermal conductivity and mechanical properties which are simply nonexistent in all previously tested material. The ability to mold or otherwise economically manufacture with high precision is necessary. On the other hand, similar properties are advantageous in other miniaturized devices including microelectronics, electrical devices (electric motors and transformers), aerospace industry and automotive industry.

Currently, the new wave of in vitro diagnostics is most often in silico. New miniaturized analytical devices developed for biomedical applications have borrowed manufacturing techniques from the semiconductor industry. Typically they use high quality silicon wafers and steps of lithographic masking, laser machining and chemical etching to produce devices in which a few micron structures are reproducibly produced. However, silicon based diagnostic chips are quite expensive. Furthermore, in many applications, silicon is not the best material, e.g. it scatters the light when laser based read-out schemes are used. Also, techniques of derivatization of silicon are more difficult than derivatization of plastic, and there are limitations on the density of biological materials which can be attached to a silicon surface.

The majority of biomedical devices use either silicon or "unmodified" plastics. Because biological fluids contaminate silicon, i.e. are nonspecifically attached to the silicon surface, the chips cannot be reused and have to be disposable. VLSI quality silicon wafers are too expensive to be thrown out, and glass or plastic chips are preferred. However, while "unmodified" plastics, typically thermoplastics such as polyethylene, can be produced cheaply, their thermal, mechanical and electrical properties are far from optimal. That is, "unmodified" plastic has low thermal conductivity and is not efficient in heat diffusion from local heating spots. The same is true for glass. Also, glass devices tend to break under such conditions. SUMMARY OF THE INVENTION

Diamond-based composites according to the invention satisfy a long felt need for a versatile, economical, mechanically workable material having the property of high thermal conductivity but low electrical conductivity. The compositions of the invention succeed where previous efforts at providing economical, versatile thermally conductive but electrically non- conductive materials have failed. The invention solves previously unrecognized limitations in the theoretical analysis and production of such compositions.

The invention succeeds despite many prior failures in a crowded and mature art. The invention eliminates elements employed in the prior art, such as the use of pure or almost pure diamond and complex processing steps, while providing improved performance. The differences from the prior art, in materials used, processes employed, and uses for the composites, were not previously known or suggested. The compositions of the invention provide advantages that were not previously appreciated.

Conventional theories of thermal conductivity predict a low increase in conductivity for a composite with a low proportion of diamond powder with respect to its volume as a whole.

According to the invention a diamond-based composite having a thermal conductivity pathway includes diamond powder and a non-metallic matrix material where the diamond powder has a filling factor and a percolation threshold, and the filling factor is greater than the percolation threshold. The ratio of the filling factor to the percolation threshold is preferably less than about 3. The ratio can be twice as great as the percolation threshold or in the range of 1.1 to 1.5 times the threshold depending on the application of the composite.

The diamond powder of the composite according to invention may be essentially randomly distributed within the matrix material and preferably accounts for at least about 15 % of the volume of the composite, to less than about 75 % .

The diamond powder may include essentially spherical grains with a size range selected from the group consisting of a diameter smaller than 50 microns, a diameter smaller than 20 microns, and at least two fractions of grains with different size which are chosen tα increase the filling factor. Alternatively, the grains may be essentially non-spherical in shape, selected from the group consisting of grains with a large aspect ratio (e.g. ellipsoids), grains with a large aspect ratio and that are essentially two dimensional (e.g. laminas and/or flakes), grains with a large aspect ratio and that are essentially one dimensional (e.g. rods, fibers and/or needles), and grains with a highly irregular shape that maximizes the surface to volume ratio.

The composite according to the invention may include diamond grains with a large aspect ratio that are distributed along the direction of their longer axis so that a composite material with nonhomogeneous thermal properties is produced. Further, the diamond powder may include at least two fractions of grains with a substantially different shape. Moreover, the diamond powder may be produced synthetically.

The matrix of the composite according to the invention may be granular with granules smaller than that of the diamond powder. In another embodiment, the matrix may be a plastic.

According to the invention, the matrix of the composite is a dielectric and the diamond powder is distributed inside the dielectric material. In a preferred embodiment, the matrix material can be selected to satisfy one of the following constraints on the melting point: a) wherein the melting point is between 50 and 100 degree Celsius; and b) wherein the melting point is higher than 100 degree Celsius.

In other embodiments, the dielectric may be an organic material with good mechanical properties, the dielectric being one of a thermosettable plastic and a material which hardens due to chemical processes. The organic material may be a two component polymer, with at least one component being liquid at room temperature.

The dielectric may be a polymer selected from the group including an epoxy, an acrylic resin, a cyanolit based glue, Teflon or Mylar, and PVC, polystyrene or other plastic.

In yet another embodiment, the diamond powder may be distributed into a highly elastic material. Further, the dielectric may be highly viscous but not solid at temperatures between 0 and 100 degree Celsius.

A method for making a diamond based composite having a thermal conductivity pathway according to the invention includes the steps of obtaining a diamond powder and a matrix material having a lower thermal conductivity than the diamond powder, determining the percolation threshold for the diamond powder by direct measurement by either thermal difϊusivity or conductivity, and mixing the diamond powder and the matrix material to achieve a proportion of diamond powder to matrix material greater than or less than the percolation threshold thereby producing a thermally conductive diamond based composite.

In a preferred embodiment, the mixing step further includes removing all air bubbles by at least one of processing in a vacuum or under reduced pressure, repetitive mixing and stirring, use of the rolling method, and annealing at an elevated temperature.

In yet another modification, the filling factor of the diamond powder may be increased by at least one of pressure comptification (e.g., sintering), sedimentation of the diamond powder in the viscous matrix, and centrifugation of the diamond powder in the viscous matrix.

In yet another method according to the invention, the diamond based composite may be fabricated into a device by one of stamping, injection molding, and thermosetting.

This invention further discloses a method of transferring heat in a device using a diamond based composite according to the invention.

DBCs according to the invention may be used in recently developed miniaturized devices for "diagnostic chips" and microfluidics. DBCs can improve device performance without making them too expensive.

BRIEF DESCRIPTION OF THE DRAWINGS

Further objectives and advantages will become apparent from a consideration of the following description and drawings in which: Figure 1 is a schematic representation of a cross sectional view of a diamond based composite according to the invention; and

Figure 2 is a graph showing thermal conductivity in a composite based on the filling factor of the diamond based composite according to the invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

In describing preferred embodiments of the present invention illustrated in the drawings, specific terminology is employed for the sake of clarity. However, the invention is not intended to be limited to the specific terminology so selected, and it is to be understood that each specific element includes all technical equivalents which operate in a similar manner to accomplish a similar purpose. Each reference cited herein is hereby incorporated by reference as if it was individually incorporated by reference.

Percolation is defined as diffusion by gradual spreading or penetration. The term "percolation theory" refers to a number of general and powerful theories of conduction in inhomogeneous media which have been developed to describe the behavior of a random network of mixtures. According to this theory, if bonds are removed from a network, the conductivity of the network becomes zero when the fraction of the remaining bonds falls below a critical value which is known as the percolation threshold. The value of the percolation threshold or "percolation factor" depends upon the connectivity and dimensionality of the lattice. Here the percolation threshold strongly depends on the shape, size and orientation of the diamond powder.

Thermal diffusivity refers to the quality of heat passing through an area in a direction x over a time dt. Thermal conductivity, as discussed below, is a linear function of diffusivity dependent on the specific heat and density of a material. According to the invention, a thermal conductivity pathway is a network which permits the propagation of phonons from one surface to another, dramatically increasing thermal conductivity. The pathway is not necessarily physically connected but the spacing is close enough to allow phonon transmission. This is determined by the percolation factor. Thus, without intending to limit the scope of the invention, this pathway is analogous to "stepping stones" as the grains of the diamond powder are dispersed in the polymer matrix without necessarily touching one another.

As shown schematically in Figure 1 (without regard to scale or propoer ^steion), the invention describes a new class of materials composed of a filler of natural or synthetic diamond powder, either spherical 10 or assymetric 11 , which is embedded in a matrix 12 of an appropriate dielectric material. The proportion of diamond to matrix (filling factor) is higher than the percolation threshold but lower than previously imagined as useful. DBCs according to the invention have excellent thermal conductivity comparable to metals but are electrical insulators. The composition may further comprise components such as fibers 14 or ferromagnetic particles 16.

The existence of the effect of enhanced thermal conductivity at low filling factors is surprising. Normally one expects

^composite ^{= a} "-filler ^"■" (l~^α) ^matrix (l^a and

"^•composite'-' natnx ^— 1 +

~1) (lb)

wherein α is the filling factor in volume and K, are the thermal conductivities of the composite, filler and matrix respectively. In the following we consider the case when the matrix is epoxy and filler is diamond powder. With ~ 0.1 and

- 10, formula (lb) gives K_coπφosι K_epoxy ~ 1.9, i.e. , only a modest increase of thermal conductivity. One of ordinary skill in the art would expect no strong dependence on the shape of the filler material.

Surprisingly, experiments show ^'K_composιJK_epmy - 3-8 depending on the shape of the powder crystals. The optimal shape of diamond powder seems to be elongated ellipsoids with axial ratios of a few. The K_c0mposite/K_epo-_y vs. filling factor dependence is highly non-linear.

Experimental and theoretical studies established that:

1. Percolation is important for thermal conductivity of composites above a critical filling factor (α > 10%);

2. Considerable thermal conductivity enhancement is possible; thermal conductivity enhancements of a factor of ten are possible in epoxy loaded with diamond powder. The percolation threshold for phonon transport is controlled by the spectrum of sizes and shapes of the diamond filler and by the aspect ratio of the particles. For example, use of diamond powder with at least three size fractions improves considerably the thermal conductivity. Furthermore, at least 50% of the diamond particles should be essentially non- spherical.

A wide variety of matrix materials may be used according to the invention. These include thermosettable polymers and/or chemosettable resins; silicon grease and rubber; matrices with high magnetic permeability; and electrically resistant varnish settable by evaporation of the solvent, ( e.g. 7071 GE varnish).

The disclosed diamond based composites offer several innovative and unique opportunities for the control of thermal properties of the composite. One of the advantages of using synthetic diamond is the ability to select and control grain dimensions and shapes. This permits an increase in the filling factor and thus the enhancement of the thermal conductivity of the composite.

The repeatability of the geometrical and thermal properties of the synthetic diamond powder provides for good definition of the properties of the composite. Another possibility is offered by using isotopically pure synthetic diamond powder which conducts heat about 60% more efficiently than natural diamond powder which contains two different carbon isotopes according to their natural abundances. For example, the diamond powder may be isotopically enriched to more than 99.5% of one of the carbon isotopes.

An example according to the invention is an epoxy matrix loaded with about 20% (by volume) of diamond powder filler which shows a surprising increase in thermal conductivity. Preferably, the diamond powder particle diameter is less than 500 μm, and the diamond powder properties (spectrum of sizes, shapes and filling factor) are selected to optimize the material properties. The invention encompasses methods of fabricating composites with optimized properties, using various diamond and matrix materials. DBCs may include tertiary composites using a fiber-like matrix for high tensile strength materials. The invention also encompasses devices comprising DBCs.

According to the invention, tertiary components may be used, wherein the third component permits improvement of the mechanical and/or magnetic properties of the composite. The diamond-based composites of the invention have all of the properties necessary to fabricate high performance yet low cost miniaturized biomedical chips. For example, the thermal conductivity of a diamond loaded epoxy is about tenfold better than that of plastic or ceramic. Thus, a chip composed of a diamond loaded epoxy may be a factor of 5-10 faster than plastic or ceramic chips. As a result, the diamond loaded epoxy chip can be operated at higher voltage which leads to better performance.

More specifically, the use of DBCs in VLSI packaging permits faster operation of microprocessors, Random Access Memory (RAM) and fast Analog-to-Digital Convertors (ADCs). Another important application is in construction of the voltage convertors and power supplies, e.g. AC-DC and DC-DC convertors. Furthermore, the composites according to the invention can be used in the production of motherboards with high thermal conductivity which can act as distributed, large surface radiators. There are examples of possible applications of insulators with better heat conductivity in the design of high performance machines, e.g. in the automotive industry. It will be especially important in the case of composite materials used in aviation, wherein the majority of structural composites have highly anisotropic properties and often poor thermal conductivity. This often leads to large temperature gradients across the composites and thus induces severe nonlinear stress effects; material fatigue and microcracking may severely limit the lifetime and performance of these composites. An admixture of diamond powder leads to an increase in thermal conductivity and permits materials which have almost isotropic properties. The thermal properties of some high thermal conductivity materials are as follows:

Table 1: Thermal Conductivity of Some Materials

Materials Diffusivity (cm²/sec) Conductivity (Watts/cm/ °K)

0.07% C-13 Diamond 18.5 33.2

0.5 % C-13 Diamond 14.5 26

1.0% C-13 Diamond 12.4 22.3

Natural Diamond 12.2 21.9

BN 7.6

SiC 4.9

Cu 1.25 4.0

BeO 3.7

A1N 3.2

Si 0.86 1.6 The C-13 data were obtained by measurement. The other data are from T.D. Anthony,

"On the thermal conductivity of di.amond under changes to its isotopic composition", (1990). It can be observed that the thermal conductivity normalized to Si is 20.75, 16.25, 13.94 and

13.69 for the 0.07% C-13, 0.5% C-13, 1 % C-13 and natural diamond respectively.

Experimental results suggest a 50-60% increase in the diffusivity of isotopically enriched diamond, implying a corresponding decrease in photon-isotope scattering. Heuristic arguments predict that the thermal conductivity of isotopically pure diamond should increase by at most about 5% at room temperature. Even more impressive are the advantages of diamond at a lower temperature. The thermal conductivity of natural diamond increases to over 100 Won ¹ K^"1 at about 77°K, B.H. Armstrong, The Physics of Si0₂... (Pergamon Press, NY 1878), in sharp contrast to the thermal conductivity of Cu or other metals, which remain essentially flat at temperatures between liquid nitrogen and room temperature. Preliminary estimates suggest that isotopically pure diamond can reach a thermal conductivity of about 1000 Wcm^"1 K^"1 at a low temperature. Furthermore, while the local theory of heat conductivity seems appropriate for natural diamond down to liquid helium temperatures, non-local heat conductivity may dominate for isotopically enriched diamond at liquid nitrogen temperatures and below.

Heat Conductivity in Diamond Based Composites

The outstanding heat conductivity of diamonds, especially of isotopically enriched diamonds, is used to provide excellent properties to a class of binary and ternary composites loaded with diamond powder. Two types of theories of heat and electrical conductivity in random media have been developed, the so-called effective media theories and percolation theories. Typically, effective media theory works well for the case of thermal conductivity when the ratio of conductivity between the matrix and the filler is small, where a linear relationship between the filling factor and conductivity exists. And percolation theory works well for electrical conductivity where metals conducts a thousand times better than insulators.

The assumptions of the percolation theory suggest that it is not applicable for the case of diamond-loaded composites. Based on the prior art, it was believed that the behavior of such a composite would be described by the effective media theory and not by percolation theory. Surprisingly, the composites of the invention show clear "percolation like" characteristics. Due to the irregularity of grain shapes and imperfections in mixing, the parameters/,, (percolation threshold) and t have to be established empirically for each type of diamond based composite.

A percolation theory predicts σ = A { f ~ f_c ) ^t (2 ) where A and t are material dependent constants, and f and f, are respectively the diamond filling factor and percolation threshold or 'filling factor' , when the conductivity component first forms a continuous percolation path across an otherwise insulating medium. Strictly speaking, percolation theory applies only forf = f_c ± e and can only hold if the conductivity ratios are very large, e.g. when electric conductivity is studied in dispersions of metal powders inside insulating medium. Experimental data show that the exponent can be large, e.g. t ~ 2.5 was found in sintered nickel powder. See N. Deprez and D.S. McLachlan, Solid State Communications, vol. 66, pp. 869-872 (1988). Due to the exponential characteristic of the percolation theory, conduction increases dramatically when the filling factor exceeds the percolation threshold.

The measurement of thermal conductivity by conventional methods is subject to large errors for materials with high thermal diffusivity. In the case of measurements of diamond, the flash diffusivity method is the most reliable. In this method the thermal conductivity is calculated on the basis of the measured thermal diffusivity using the following formula:

K = Cpα (3)

Wherein C is the (separately measured) specific heat, p is the density and α the measured thermal diffusivity of the sample.

If a pulse of radiant energy Q is instantaneously and uniformly absorbed within a small depth at the front surface x=0 of a thermally insulated solid of uniform thickness L, the rear surface (x=L) temperature history can be expressed as:

_T { L, t ) = -A [ 1 +2∑ ( -l pexp ( ^~n ^ ) ] (4)

QCL L ²

The thermal diffusivity could be determined as a fitting parameter of the aforementioned equation or a simple numerical formula for the diffusivity could be used namely:

T 2 = 0 . 139 -=— (5) ^fc0 . 5 where ^ ₅ is the time required for the back surface to reach one-half the maximum temperature rise.

Fig. 2 shows a hypothetical sigmoidal curve for thermal conductivity of a DBC as a function of filling factor. Pure matrix material, filling factor = 0% , has low conductivity which does not increase much up to "a", at or near the percolation threshold. Conductivity increases rapidly in the range of "b", and levels off again at "c" well below 100% . Thus, surprisingly, filling factors in the range between "a" and "c" are adequate for composites of the invention.

EXAMPLES

The validity of the Percolation theory and characteristic parameters A, f_c and t should be established for each composite sample. The measurement step takes only a few minutes but for better reliability it has to be repeated several times; a few hours per sample is typical including computations and sample handling time. The data acquisition and analysis can be considerably improved by using a Pentium^® computer and PC-compatible plug-in oscilloscope card. A total data acquisition/analysis/reporting time may thus be diminished to about 10 minutes per sample.

The experimental setup is both simpler and more sensitive that other currently existing methods. A sample of material is placed between a radiation source (Xenon flash lamp) and an infrared detector. The heat pulse from the source is absorbed by the front face of the sample and conducted through the material to the back side. The IR detector measures the temperature of the re.ar side of the sample and stores it as a function of time on a digital scope. This digitized trace is then used by the computer to calculate the thermal diffusivity. Thus the required equipment can be conveniently divided into:

• fast trigger electronics and Xenon flash lamp • low noise preamplifiers

• fast ADC and data acquisition

Hardware requirements are that the pulse duration of the light source and the IR detector response time be as short as possible. The sampling time of the setup is about 10 μsec and is limited by the time necessary to integrate the signal from an IR detector to achieve a good signal to background ratio. Furthermore, the use of the Xenon flash light limits the amount of energy and timing.

A series of binary diamond-loaded composites was produced according to the invention, using a two component epoxy. The synthetic diamonds had sizes between 15 and

25 μm, and were observed to have a diversity of shapes, dominated by cubic and hexagonal polyander. The preliminary results are shown in Table 2. They tend to confirm the sigmoidal relationship shown in Fig. 2.

TABLE 2: Thermal Diffusivity Coefficient of Diamond Composites Filling Factor Thermal Diffusivity Coefficient (cm²/sec^

(Diamond)

0% 0.0009

10% 0.002

15% 0.0022 20% 0.0024

24% 0.0048

28% 0.0078

37.5% 0.0105

50% 0.0123 The filling factor is by weight. The calculation of the filling factor by volume requires that the densities of epoxy (ca. 0.9 g/cc) and diamond (ca. 3.51 g/cc) be taken into account. At a filling factor f ~ 25% the conductivity jumps to o>_DBC/o_Epoxy = 8 and at/ ~ 50% it reaches σ_DBC/σ_Epoxy ~ 15. These data support increasing the thermal conductivity at room temperature by a factor of at least ten. At low temperature, say liquid nitrogen, a thermal conductivity gain of fifty or more is expected. Furthermore, when isotopically enriched diamond is used a factor of one thousand improvement is possible. Finally, the data suggests that percolation onset occurs at/ ~ 25 % . However the small number of data points with/ >/ does not allow estimate of the exponent t.

According to the prior art electrical percolation theories, one would not expect to obtain this relationship where the conductivity ratio of filler /matrix material is less than 100. The prior art indicates that a percolation-type relationship can only hold if the conductivity ratios are very large.

In one embodiment according to the invention, epoxy is the matrix material. Epoxy has a thermal conductivity of about 0.5 W/cm/°K. The synthetic or natural diamonds of the diamond powder have approximately a thermal conductivity of 20 W/cm/°K. This is a conductivity ratio of 40 at the most. Nevertheless, surprisingly, the heat conductivity of DBCs according to the invention are consistent with the electrical conductively predictions of the percolation theory, as described above. However, percolation theory does not work well for the case of epoxy loaded with alumina. The ratio of thermal conductivity of aluminum to epoxy is about 5.

The thermal properties of silicon grease loaded with diamond powder were measured and the results are provided in the following Table. Table 3: Thermal properties of grease loaded with diamond powder

Filling factor Thermal diffusivity (cπ Vsec) 0.0% 0.00104

30% 0.00227

70% 0.00351

Again, in this case, there is a large increase of thermal conductivity when appropriately large filling factors of diamond powder are used. Taken together, these two series of experiments suggest that the properties of the filler ( . e. , diamond powder) are more important than the properties of the matrix material.

More than 100 samples were produced and each sample was characterized under an optical microscope. About 20% of samples passed the optical quality test and their thermal properties were measured. Non-reproducibility of the thermal properties from sample to sample may result from microscopic bubbles of air. Furthermore, when diamond powder samples with a wide distribution of particle sizes are used, preferential dispersion of grains of the same size may lead to thermal conductivity inhomogeneities. Best results are obtained after selecting the grain size. Furthermore, the homogeneity of diamond powder dispersion depends on the composite production method, and a change of mixing procedure can lead to large thermal conductivity changes. Material Selection Criteria

As those skilled in the art well know, heat transfer in diamond occurs through propagation of phonons in the lattice. That is, phonons are essentially lattice vibrations which are quantitized in energy. The efficiency of heat transfer (phonon propagation) is limited by a number of independent effects. Scattering of phonons can be induced by other phonons, the walls of the crystal, grain boundaries, chemical impurities, vacancies, dislocations and isotopes. In most dielectrics and semiconductors the main source for the scattering of phonons are static defects (isotopes and impurities) which results in the reduction in heat conductivity.

In currently used composites, the heat conductivity is always by thermal phonons. In diamond- based composites, especially if isotopically enriched diamond is used or in low temperatures, the ballistic phonons carry a considerable fraction of energy. The thermal conductivity due to ballistic phonons can be extremely large in low temperatures. A technically challenging problem is how to use this property in diamond-based composites. In this case the bottleneck is the phonon mismatch on the grain surface. Thus, the use of a diamond powder with large aspect ratio increases the heat conductivity. Also, the use of powder having grain shapes with large surface to volume ratio is advantageous.

The thermal properties of diamond-based composites are influenced by: • filling factor — a range of 15-75% by volume is preferred;

• diamond powder size distribution — the use of diamond powder selected by size and powder with a broad size distribution is preferred;

• diamond particle shape ~ the use of roughly hexagonal and highly asymmetric (e.g. needle shaped) diamond powders is preferred.

As mentioned above, the temperature at which the DBC material of the invention is expected to operate is important. Low temperature properties of DBCs are of importance, as are properties at temperature much above room temperature. One example is in biomedical devices, wherein one of the steps of PCR process is at about 95O . Another example is the huge automotive products market, e.g. car radiators and brake pads.

Ternary composites are preferred, e.g. for structural applications with both mineral fibers and diamond powder used in the appropriate glue matrix. For ternary compounds the application of percolation theory is unhelpful and empirical studies may be used by a person of ordinary skill based on the principles disclosed herein. In binary composites, the thermal properties are influenced by filling factor, size and shape of the diamond powder. For example, with diamond grain diameters of 15-25 μm at filling factors between 10% and 50%, the percolation limit was reached at a filling factor of about 25% . Diamond powders with grain size of 5-15 μm, 15-25 μm, 25-35 μm and 35- 50 μm are commercially available. Furthermore, by appropriate calibrated filters and sedimentation in oil, grains with narrow size distribution can be obtained. For example, composites may have narrow distribution with grain diameters of φ = 25-28 μm, or a wide size distribution (φ = 5-50 μm).

A preferred ternary composite consists of epoxy (α = 30% or 50%), diamond (β = 30-50%) and kevlar fiber (gamma = 1-α-β).

Fabrication Methods

The process of fabricating diamond-based composites of the invention overcomes three main difficulties:

• the need to remove all air bubbles from the composite; • the need to disperse the diamond powder homogeneously inside the composite;

• the need to achieve high filling factors.

These difficulties have to be overcome simultaneously. In selecting the preferred technique, there are important quality vs. cost tradeoffs.

According to the invention, one method for making a diamond based composite (DBC) according to the invention is to obtain a diamond powder and a matrix material having a lower thermal conductivity than the diamond powder, determine the percolation threshold for the diamond powder by direct measurement using either thermal conductivity or thermal diffusivity; and mix the diamond powder and the matrix material to achieve a proportion of the diamond powder to matrix material greater than or equal to the percolation threshold thereby producing a thermally conductive diamond based composite.

The filling factor strongly depends on the shape of the diamond powder grains, as well as the size and orientation. A proportion of the diamond powder to matrix material or filling factor could be up to about three times greater than the percolation factor. A preferred proportion or filling factor would be double the percolation factor. A filling factor in the range of 1.1 to 1.5 times the percolation threshold would be even more preferred. The thermal conductivity of the diamond based composite is substantially greater than that of the matrix material, typically at least double, preferably at least about four times higher. The composite may have conductivity eight or more times higher than the matrix material. One way to remove air bubbles is to perform the composite synthesis in a vacuum.

Thus, even if voids are created in the composite, the subsequent steps of mixing and compacting under pressure removes them efficiently. In some cases, however, especially in the case of some two component polymers, the very process of polymerization leads to production of small amount of gases which tend to be trapped in the material. In this case, the processes of injection under pressure and of continuous mixing and rolling into thin films permits removal of the majority of bubbles. Air bubbles should preferably be removed when the composite is still liquid or of low viscosity. Thus, the time of polymer hardening should be made long enough to permit mechanical stirring. Preferably, the diamond powder is first mixed with one, preferably liquid monomer and only later the second monomer (hardener) added. Often, air can be removed after mixing the diamond powder with the first monomer by annealing at high temperature. However, the mixture of diamond and first monomer should then be allowed to cool before adding the hardener; in the case of many polymers the hardening time is highly temperature dependent. Typically, by slowing the process of hardening, better diamond-based composites are obtained. Another method of removing bubbles is by injecting diamond powder into the monomer(s) under pressure. Taking into account the small size of grains, many techniques used in genetically modified plants, e.g. "bacteria guns" can be used. Optionally, both plastic and grains can be injected under pressure leading to two fluid instabilities which efficiently mix the components. The rolling method is a preferred method of preparing DBCs both leads to superior homogeneity of diamond powder dispersion and is efficient in removing air bubbles, especially if performed under vacuum. In this method, the diamond powder is dispersed into pre- polymerized material with the consistency of bread dough. A slab of the composite is rolled into approximately a one millimeter thick film and folded upon itself a few times. The resulting, typically less than one centimeter thick, layer is once more rolled into a thin film. The operation is repeated five to ten times and results in highly homogeneous distribution of diamond powder. Topological arguments suggest that any initial inhomogeneity is diluted by a factor of 2^k wherein k is number of foldings. Not only the homogeneity improves but also air bubbles are efficiently removed. The rolling method can be easily automated for large scale production of high quality DBCs. An important challenge is to obtain the DBCs with very large diamond filling factor.

One of the most efficient methods is not to add a large fraction of the diamond filler in a single step but rather add a small amount of diamond filler in a number of steps. Each step of adding the filler is followed by thorough mixing. A variation of this method permits creation of voids in the mixed material, adding the diamond powder, and mixing with pressure compaction after each step.

An efficient method of obtaining high filling factors is centriftigation. The density of diamond crystals is about a factor of three higher than that of the organic or silicone based matrix. If the diamond powder is added to a not yet hardened matrix, a few minutes centriftigation will remove all diamond grains of the powder from the top of the centriftigation container and compact them at the bottom part to densities close to maximal packing for given crystal shapes. The unused matrix can be removed and the polymerization process accelerated, e.g. by heating. When Teflon centriftigation tubes are used and a polymer is a two component epoxy, the hardened sample of DBC can be easily removed from a tube. The centriftigation is also very efficient in removing air bubbles from the composite. As the time of centriftigation depends strongly on the grain radius; somewhat larger (35 to 100 microns) size selected diamond grains are preferentially used when a centriftigation step is applied. Finally, when diamond grains with large aspect ratios are used, e.g. short pieces of diamond fibers, centriftigation permits aligning them along the centriftigation axis. Thus, an interesting class of DBCs with thermal conductivity higher in one direction can be produced. Similarly, in the case of ternary composites loaded with high tensile strength fibers, centriftigation permits ordering the fiber direction and creation of materials with excellent mechanical properties in the chosen direction. The Choice of Matrix

Two new families of diamond based composites have been developed; binary (diamond + appropriate matrix) and ternary composites (diamond + high quality mineral/carbon fiber + appropriate matrix). The choice of matrix is a complicated trade-off between ease of fabrication, cost, and the properties of composite required by the application. The majority of existing high thermal conductivity composites are produced by mixing metal components with an appropriate binder. There are, however, characterized by both good thermal and electrical conductivity. The major advantage of DBCs is that they are electrical insulators, i.e. typically the matrix used in the production of DBCs should be a dielectric with good electrical properties.

In the majority of applications, e.g. in miniaturized biomedical devices and electronics, the DBC should be a solid with good mechanical properties and low specific density. In some applications, the DBCs may be produced by molding and the final composite is used at relatively low temperature, say below 100°C. In this case the use of diverse waxes, e.g. paraffin wax, may be appropriate and leads to the DBCs which can be produced at low cost and recast by simple heating.

Many devices are operated in the "intermediate" range of temperatures from CPC to 100°C. More specifically, this is the case for miniaturized biomedical devices which typically use temperatures from room temperature to about 95°C (annealing temperature in PCR or DNA hybridization methods). In this case the use of thermosettable plastics is appropriate including multi-component materials which harden by chemical reactions. An excellent example of this class of matrix are organic materials which harden by polymerization. Materials in which at least one monomer is a liquid are especially important because the diamond powder can be easily mixed inside the liquid. A particular class of such multicomponent organic materials is epoxy.

In other applications, e.g. in electronics packaging, the DBC should work for long time periods in a more elevated temperature, say in 150PC. In this case, preferred matrix materials are fluorine-based polymers, e.g. Teflon, or silicone-based polymers which have decomposition temperature up to 250°C. In many applications a vacuum-tight two monomer-based plastic, e.g. epoxy with good temperature conductivity, is especially attractive. The two sides of the material may be at drastically different temperatures. The temperature gradient across the joint leads to inhomogeneous stress and often leads to cracking, and the vacuum tightness is lost. A well- known example of such situation is the breakdown of vacuum tightness of silicon seals used in the "Challenger" shuttle, which led to the explosive destruction of the shuttle. Other examples are found in many aerospace and cryogenic applications.

According to the invention, DBCs with excellent properties may be based on acrylic resins, cyanolit-based glues, Teflon, Mylar, PVC and polystyrene. The use of DBCs with acrylic resins as a matrix is especially important because it permits transparent enclosures with good thermal properties. A good example is their use as a heat removing element in electrophoretic apparatuses used by biologists for studies of structure of DNA and proteins. The use of cyanolit-based glue leads to the possibility of producing very thin, thermally- conducting bonds with excellent mechanical properties. For example, transparent DBCs may be used to join together pieces of quartz and thus considerably diminish the cost of manufacturing quartz objects with complicated shapes. Another important application is in repair of glassware and porcelain ware.

Many chemical synthesis processes which lead to heat production should be performed at a constant temperature. Thus the reagents are mixed within an appropriate container which is cooled from the outside by a forced flow of water or other coolant. In some cases, a gradient of voltage is used, e.g. in electrodissociation of materials. More specifically, the electrolysis of many salts and metalloorganic compounds is performed at elevated temperature which should be regulated with high precision. The quartzware currently used is both expensive and easy to break. The use of Teflon and Mylar matrix is especially appropriate in the production of containers for chemically corrosive materials, but pure Teflon and Mylar are very bad heat conductors. This limitation can be overcome by loading with diamond powder.

Another application of DBCs is the use of diamond powder loaded Teflon to coat domestic utensils. Currently-used pure Teflon coated frying pans and other utensils are easily scratched. A thicker layer of Teflon would help but can not be used because Teflon is an excellent heat insulator. The use of diamond powder admixed into Teflon would not only permit a thicker layer of the Teflon coating but also will make it more scratch resistant. Another use of DBCs with a Teflon matrix is in the fabrication of containers and plates for food processing which can be used in ovens and microwave heaters. This important application is made possible by the fact that such DBCs are transparent to microwaves, chemically inert and good heat conductors.

PVC and polystyrene-based DBCs are probably the lowest cost materials and are preferred in many applications. These include biomedical devices, chemical containers, containers for microwave heaters, domestic electronics and electrical appliances as well as in the automotive industry, i.e. in the applications which are highly cost-sensitive.

In many applications there is a need for shape deformable elements with high temperature conductivity. For example, this is the case in the automotive industry, wherein many hot, metallic elements are to be shielded from outside, potentially corrosive medium by rubber penumbra. Contact with these elements or just radiative heat transfer leads to a considerable temperature increase in the internal side of the rubber. The low temperature conductivity of rubber means that it cannot be efficiently cooled. Another example includes shape deformable joints with high temperature gradients between the two elements to be bound. In such and many other applications, use of rubber loaded with diamond powder is disclosed; this includes the use of silicone-based rubber.

In many applications, even deformable joints are not practical and highly viscous grease is used either as a corrosion resistant layer over metallic parts or to provide thermal contact between two objects. The grease was used to help the heat removal from spark plugs in cars and other engines. Grease loaded with diamond is a good replacement to currently used materials. In one study it increased the average lifetime of spark plugs by a factor of three. Thus use of DBCs with a matrix which is highly viscous but not solid at room temperature is one preferred embodiment.

In many electromechanical applications, e.g. in motors, the wires are insulated by an appropriate varnish. However, this means that any heat flow is restricted to one direction, i.e. along the electric conductor. The maximal current across the winding is limited by the existence of weak spots, e.g. places when wires have been bonded or are metallurgically defect. This leads to local thermal instabilities which often propagates and increase the probability of burning the winding. The use of varnish loaded with diamond powder to electrically insulate the windings permits the construction of electrical motors in which the heat flow is essentially three-dimensional and thus will permit production of more compact and lighter devices.

The majority of the above mentioned organic materials decompose at high temperature.

In applications wherein DBCs work at temperatures substantially higher than 25CPC, the use of a special, high temperature resistant matrix is necessary. The chemistry of silicon being very similar to carbon, some silicon polymers are available. As described above, the use of multi-component materials with at least one silicone based monomer is especially appropriate.

A diamond placed in anaerobic conditions survives heating up to 800PC. At this temperature even silicon-based components disintegrate or melt. DBCs according to the invention may be prepared in which the matrix is a ceramic material, e.g. porcelain or porcelite. Optionally, diamond can be loaded inside of glass or other similar amorphous solid leading to improved thermal conductivity. One application will be in the production of spark plugs for the automotive industry.

The diamond powder may be distributed inside refractory materials, e.g. boron nitride or other oxides, nitrides and carbides with very high melting point. Many of these materials, e.g. W₂C and other carbides, have very good thermal conductivity, but are conducting electrically. An admixture of diamond powder may further increase their thermal conductivity while at the same time diminishing their electric conductivity.

Ternary DBCs may have as the third component a material with very high magnetic permeability, e.g. powders of iron group metals or their alloys. Such ternary DBCs can be used in the production of transformer cores, ferrite heads and light weight magnetic field shields. In such industrial applications the absorption of variable electromagnetic field leads to material heating. With modern, high permissivity magnetic materials, e.g. powders of rare earth metals, the volume and weight of transformers are limited by the thermal bottleneck. Thus the use of specially formulated ternary DBCs enables the next generation of higher performance transformers.

Other ternary DBCs are composed of at least three components, including but not limited to :

• Diamond powder to increase thermal conductivity; • Fibers and/or laminas of material with high tensile strength used to improve the mechanical properties of the composite; • Appropriate binder material, e.g. epoxy.

Such ternary DBCs are useful applications in the production of motherboards for the electronics industry and in aerospace applications.

Examples of Applications of DBCs Due to excellent thermal properties, which can potentially be coupled with excellent mechanical properties, corrosion resistance and electrical properties, DBCs can find major applications in:

• biomedical applications;

• heat management in electronics; • structural applications wherein local temperature gradients lead to nonlinear stress and long-term degradation of material properties;

• containers for chemical and domestic applications.

In the following, some possible applications are disclosed. The use of DBCs will lead to major improvements in performance and the durability of some products. On the other hand, the use of DBCs will be marginally more expensive than currently used composites. The following examples do not attempt to illuminate the trade-offs between performance and cost. More specifically, recent progress in the production of diamond powders may decrease the prize of DBCs considerably.

Biochemical and chemical processing Many chemical and biological processes require temperature stabilized conditions. The appropriate containers should be chemically inert and mechanically sturdy. In some cases, e.g. in electrochemistry, the use of electrically insulating containers and/or membranes is necessary. Containers, vials, tubes, microtiter and other multiwell plates, capillaries and holders produced from DBCs can replace more expensive and easily breakable quartzware. In some cases, it may also replace glass-ware and current generation of plasticware. A good example is the use of DBCs in the fabrication of centriftigation containers wherein good temperature control and excellent mechanical properties are essential. Yet another application is in the use of polymerase chain reaction (PCR) wherein biological specimens placed in an appropriate container( e.g. a microtiter plate) are thermally cycled to about 95"C up fifty times.

The DBCs will be very useful in many medical and health care devices, such as:

• in surgical cast material to produce a lightweight, heat conducting and respiration removal permitting cast;

• in electrophoresis devices - DBCs are used in heat removal plates; and

• the construction of cryotips used in cryosurgery.

In all of these applications, the chemically inert diamond has considerable advantages over metal-based composites. They are also cheaper, easier to produce and have better mechanical properties than ceramics.

Microfluidics and "lab-on-chip" devices

Both miniaturization and automation are required to achieve the performance and cost required for next generation searches for drugs, when a few millions of target may be tested in a search for hits, i.e. biomolecules which strongly interact with known targets. A new generation of miniaturized biomedical devices, often referred to as "lab-on-a- chip" technology is enabled by microfluidics, i.e. the art of performing basic laboratory procedures (pour, move, mix, wash, incubate, heat/cool and separate by electrophoresis or chromatography) in tiny channels etched onto a chip. Actually, this happens in hundreds or thousands of channels concurrently. In the majority of implementations, a sophisticated network of electrical circuitry is deposited on at least one side of the biochip. Thus, the biochip itself is typically produced of a dielectric such as silicon, glass or plastic. Advantages result from producing the biochips or a part of biochip out of DBC.

The biochip has typically a plurality of functions to be performed in parallel in thousands of channels. In the following we analyze the case of a "DNA analysis" chip but a similar situation exists in the use of biochips optimized for protein analysis or bioassay tasks. A known amount of analyte is induced into the "biochip" and washed, bioreagents introduced, the sample purified, biochemically processed (e.g. DNA should be extracted), DNA amplified (e.g. by PCR), products separated by size (e.g. by capillary electrophoresis), label conjugated and read-out. On the other hand, many biochip operation require well defined cycling of temperature. The most important example is PCR, wherein about 30-40 cycles of temperature changes are performed, typically with three separate operational temperatures of about 30PC, about 55°C and over 90°C. Typically thermal processing is performed in highly stable but large temperature cyclers. Obviously, miniaturized biochips easily fit in such cyclers. However, then the whole chip will go through the temperature cycling characteristic for PCR. This will be destructive to a plurality of other processes involved in the "lab-on-the chip" concept. Thus, it seems that instead of a single biochip one need to use a set of biochips, each operated under different thermal conditions. The minimal set should consist of three chips: chip 1 = an array of input + washing + extraction elements; chip 2 = an array of PCR microchambers, and chip 3 = an array of electrophoresis + read-out elements. Each chip should have its own thermal regulation system. As can be understood, the advantage of miniaturization is quickly lost due to limitations imposed by the need for thermal insulation and temperature conditioning of specific "elementary" processes.

To solve this problem one can use a "thermal motherboard" to which the individual chips are thermally anchored at well defined temperature zones. For convenience we will define four temperature zones: low temperature (< 10°C); room temperature; temperate temperature (37°C); and variable (PCR cycle of operations). Typically, the thermal motherboard should include at least three thermal buses separated by a material with good mechanical properties but very low thermal conductivity, e.g. made from plastic foam. Each thermal bus line should include a dielectric with very high thermal conductivity and excellent mechanical properties. Each bus line has an independent miniaturized heater/cooler element and at least one temperature sensor. The system is remotely controlled by a computer. In a preferred implementation of such a "thermal motherboard, " DBC or a thin diamond film- coated plastic is used.

Another innovative implementation uses a single biochip in which a series of localized ohmic heaters and Peltier element coolers is used to locally generate the temperature profile required by miniaturized PCR chambers. To obtain the best temperature regulation and homogeneity, DBC is used as the material from which the array of miniaturized PCR chambers is produced. Preferably, due to the excellent heat conductivity of DBC, the array of active coolers can be eliminated; heat dissipation to the whole biochip via DBC is used. The "localized heater/cooler" architecture is somewhat complicated. As an alternative, in some biological applications it can be replaced by configuring a biochip with an array of heating pads and placing the whole chip in a rapidly changing magnetic field. This preferably employs a ternary DBC loaded with small, preferably sub-micron ferrite or ferryte grains. In such a ferromagnetic medium, a rapid change of magnetic field leads to "eddy currents" which rapidly decay deposit a considerable amount of heat. Each pad is thermally anchored to a PCR micro-chamber produced from DBC. Note the similarity to microwave heating. However, microwave heating leads to considerable thermal loads in any plastic and biological samples, especially in ionic fluids such as blood or urine. On the other hand, biomaterials and plastics are nonmagnetic and no harmful heating occurs.

In yet another implementation, the local source of heat can be light absorbing pads, e.g. produced by incorporating small grains of graphite in DBC. A laser optically addressing system, or an array of laser diodes under computer control can be used as a light source.

There are thermal limitations on the capillary electrophoresis part of the "lab-on-the chip" . The use of cross-linked polymers is preferred in capillary electrophoresis. However, the cross-linking increases with temperature which, leads to electrophoresis artifacts. Unfortunately, the process of electrophoresis itself generates a considerable amount of Joule heating. This heating depends strongly on the amount and the sizes of nucleic acids. Thus, a variable and highly sample dependent heating occurs. Conventional systems use a single or a small number of macroscopic capillaries and this thermal run-away process can be controlled. However, in biochips thousands of capillaries are implemented in a small volume leading to thermal run-away. For diminishing electrophoretic artifacts, the use of DBC in the construction of biochips, especially in the part of the "real estate" of the chips which are dedicated to electrophoresis, is preferred.

Electronics

Miniaturization is a trend in modern electronics, e.g. portable or laptop PCs. Thus, the amount of heat per unit volume (or unit of device surface) increases much faster than the total energy requirements. Currently, many electronic devices are limited by the need to remove heat.

These exponential increases in the heat per chip surface have led to many evolutionary improvements as use of ceramic packaging, forced air cooling or, more recently the use of liquid coolant or electrocooling. The bottlenecks appear on four levels:

• removal of heat from the VLSI itself, e.g. Si on sapphire circuits are being used in critical applications;

• removal of heat from the VLSI package, e.g. by using specially formulated ceramics and appropriate metal radiators; • removal of heat from the close neighborhood of the critical VLSI chip, e.g. by use of high conductivity chip carriers and multilayer printed boards;

• removal of heat from the device itself, e.g. by forced air or liquid circulation.

The use of diamond substrates, especially isotopically enriched diamonds, can address the first limitation. The major advantages of DBCs are on the second and third levels. Three properties of DBCs make their use very promising:

• excellent heat conductivity;

• perfect electrical insulator properties, including very low dielectric losses, e.g. when diamond loaded Teflon is used;

• simplicity of production including the possibility of injection molding techniques. The last property is very important because it may lead to both price and performance advantages over ceramics. Four illustrations with commercial potential follow. a. Chip carriers and printed circuit boards

There is a large dynamic range between heat depositions in diverse electronic chips. Typically, the power supplies, voltage regulators and highest complexity devices, e.g. microprocessors, deposit orders of magnitude more heat than other composites, e.g. TTLs, DACs etc. On the local level, chip carriers built with DBCs may replace the radiators by virtue of acting as an additional heat sink, e.g. doubling or tripling the heat capacity by efficiently removing the heat into the printed circuit board.

Typically, only about 50-80% of the surface of a printed circuit board is covered by chips. Furthermore, the majority of the surface is covered by low heat deposition elements.

Thus, the average heat per unit of surface area may be 3-5 times lower than in the "hot spots" around microprocessors, voltage regulators or power transistors. The heat conductivity of "next generation" printed boards may be up to a factor ten better than in currently used boards, i.e. forced air flow may be used to remove the heat from all the surfaces. Optionally, metal radiators or forced liquid cooling can be used to remove heat via edges of DBC printed boards.

In the use of DBCs for motherboard production, not only are the electrical and thermal properties important, but also the mechanical properties. In this case tertiary composites are advantageous, wherein the third component is an appropriate high tensile strength fiber. Alternatives include the use of DBCs sandwiched between two high mechanical strength polymer foils, and a high strength polymer-based honeycomb structure filled with DBCs. b. Power Supplies /Convenors

Typically, lower power supplies/converters are encased into a plastic with external metal insulation to provide EMI shielding. In some cases, additional radiators are provided which unfortunately require much more space, i.e. the power/ surface ratio is improved but the power /volume ratio is considerably degraded. The use of DBCs permits the production of power supplies with a factor of a few better power /volume ratio. Furthermore, as the weight of DBCs is lower than of metal radiators, a factor of ten improvement in power/ weight may be possible. The use of DBCs would increase the cost of production of a typical 40W DC/DC converter by only a few dollars, but permits development of products with a sale price much higher than currently available products. Furthermore, the small 24 pin DIL size may be used to replace larger DC/DC converters.

Typically, AC-DC and DC-DC convertors run at elevated temperature, say T> 100 °C. Thus the use of DBCs with melting points of a few hundred degrees is preferred. The maximum operational temperature of DBCs is limited by the fact that the diamond inside of the appropriate matrix transforms into graphite only at 800-1000 degrees Celsius. Thus, for all electronics and almost all other applications, the maximum operational temperature of DBCs is limited by the properties of the matrix material. For the majority of electronic applications this can be a high melting point organic material, e.g. Teflon or silicon based compounds. c. VLSI Chips

The performance of some VLSI chips, e.g. microprocessors, ADCs, DASPs, are seriously limited by heat removal. For example, special electrocoolers called "ICE CAP" were commercially introduced to permit Intel microprocessors to run T < 0° Celsius. For example, for an old 486DX25 cooling permits speeding up the chip clock from 25 to 40 MHZ. Similarly, CMOS chips may be run at very low temperature. The very efficient heat removal permits the use of voltages of up to 50 Volts, i.e. about a factor of 5 higher than the maximum tolerated by the same chips at room temperature. Thus, when using DBCs coupled to an external heat sink, e.g. a DBC-based motherboard, one may use a family of chips and/or circuits operating up to five times faster than the same chip in commercially available plastic/ceramic packaging. d. Transformers/Electrical Motors

To diminish energy losses, higher power transformers use high magnetic permeability cores. In the majority of designs these cores are produced by lamination of thin Ni/Co/Fe sheets with low dielectric loss plastics, e.g. Mylar/Teflon/PVC. However, the use of these low heat conductivity materials leads to serious thermal management problems. DBCs may permit operation at much lower temperatures and will considerably increase element life-times, and/or decrease their weight. DBCs can be used to laminate/insulate the high magnetic permeability sheets. A practical fabrication technique is to spray paint the ferromagnetic elements with a diamond based paint or varnish. When the solvent evaporates, a diamond- based composite is formed in situ. Alternatively, tertiary DBC may be used, including a dielectric matrix and filler consisting of powders of diamond and material with very high magnetic permeability.

Similarly, in multilayer transformer windings and electrical AC motors, the skin effect losses lead to considerable heating of winding; thermal tolerances are often so tight that any excursion from stationary heat flow leads to a "burned winding". In multilayer coils, the heat transfer is essentially along the wires which are a good heat conductor. However, the heat transfer across the wiring is very poor. The use of DBCs as the electric insulator on the wire may considerably improve the thermal stability of transformers and electric motors. To permit the smooth surface necessary for winding coils, an appropriate DBC may be made consisting of a dilution of essentially spherical diamond grains ( R < 10 microns) in GE varnish (e.g. General Electric 7071 varnish).

A specific case is high T_c superconducting magnets. In this case, the thermal conductivity of the superconductor, which is ceramic, is very poor. The need for external copper or silicon cladding leads to a large increase in fabrication costs as well as higher mass magnets. The use of thin film high T_c superconductors deposited on elastic films of DBCs should provide high thermal conductivity . At liquid nitrogen temperatures the diamond heat conductivity is at a maximum — about a factor of 10 higher than the heat conductivity of copper.

Household devices

DBCs have excellent heat conducting properties and at the same time are chemically and biologically inert. They have excellent mechamcal properties and can be mass-fabricated into complicated shapes, e.g. into containers. Thus, they could be used for domestic food processing replacing the current generation plastic and aluminum based objects. For example, Teflon based DBCs can be used to replace pure Teflon as a coating for cookware. Teflon and PVC based DBCs could be used to produce containers for use in microwave heaters and ovens. Finally, plastic based DBC foils can replace aluminum foils in the applications when thermally conducting wrap is used.

A very promising application of DBCs is in the production of electrical sockets for devices operating at elevated temperature. For example, light bulbs use large amounts of power, of which only a few percent is in the form of light with the rest dissipated as heat. Thus sockets for bulbs are operated quasi-permanently at elevated temperature. Traditionally, these sockets were produced from porcelain but currently to diminish weight and cost they are mass-produced from Bakelite and other plastics. Long term heating leads to a change of mechamcal properties and with time the socket becomes very brittle and often cracks when the bulb is changed. This is especially a problem when high power halogen lamps are used, both for household, industrial and automotive applications. The use of DBCs which are excellent heat conductors permits efficient heat removal and longer lifetime of electrical sockets. Other Applications

The excellent heat conductivity and excellent mechanical properties of ternary DBCs (diamond/fibers/matrix) provide advantages over many ceramics:

• the possibility of low cost fabrication (stamping, injection molding, thermosetting) • good thermal conductivity when compared with fiber/matrix composites, which usually are very anisotropic

• good mechanical properties when compared with ceramics which are brittle and stiff. Three examples from the field of automotive industry applications are:

• car radiators; • brake pads;

• spark plug cores.

These application require both operation at very high temperature and excellent mechanical properties. Organic materials cannot easily be used as a matrix. Instead the matrix is preferably selected from such materials as porcelain, glass and refractory materials. The embodiments illustrated and discussed in this specification are intended only to teach those skilled in the art the best way known to the inventors to make and use the invention. Nothing in this specification should be considered as limiting the scope of the present invention. Modifications and variations of the above-described embodiments of the invention are possible without departing from the invention, as appreciated by those skilled in the art in light of the above teachings. It is therefore to be understood that, within the scope of the claims and their equivalents, the invention may be practiced otherwise than as specifically described.

Claims

What is claimed is:

1. A diamond-based composite having a thermal conductivity pathway comprising diamond powder distributed in a non-metallic matrix material, the diamond powder having a filling factor and a percolation threshold, the filling factor being greater than the percolation threshold and the ratio of the filling factor to the percolation threshold being less than about 3.

2. The composite according to claim 1, wherein the ratio of the filling factor to the percolation threshold is less than about 2.

3. The composite according to claim 1, wherein the ratio of the filling factor to the percolation threshold ranges from about 1.1 to about 1.5.

4. The composite according to claim 1, wherein the diamond powder is isotopically enriched to more than 99.5% of one of the carbon isotopes.

5. The composite according to claim 1 , wherein the diamond powder is essentially randomly distributed within the matrix material and accounts for at least about 15% of the volume of the composite.

6. The composite according to claim 1, wherein the diamond powder comprises essentially spherical grains with a size range selected from the group consisting of: a diameter smaller than 50 microns, a diameter smaller than 20 microns, and at least two fractions of grains with different size chosen to increase the filling factor.

7. The composite according to the claim 1, wherein the diamond powder comprises grains with essentially non-spherical shape, selected from the group consisting of: grains with a large aspect ratio, grains with a large aspect ratio that are essentially two dimensional, grains with a large aspect ratio that are essentially one dimensional, and grains with a highly irregular shape that maximizes the surface to volume ratio.

8. The composite according to the claim 7, wherein diamond grains with a large aspect ratio are distributed along the direction of their longer axis so that a composite material with nonhomogeneous thermal properties is produced.

9. The composite of claim 7, wherein the diamond powder consists of at least two fractions of grains with a substantially different shape.

10. The composite according to claim 1, wherein the diamond powder is produced synthetically.

11. The composite according to the claim 1 wherein the matrix is granular having granules smaller than that of the diamond powder.

12. The composite according to the claim 1 wherein the matrix is a plastic.

13. The composite according to the claim 1 wherein the matrix is a dielectric and the diamond powder is distributed inside the dielectric material.

14. The composite according to the claim 1 wherein the matrix material is selected to satisfy one of the following constraints on the melting point: a) wherein the melting point is between 50 and 100 degree Celsius; and b) wherein the melting point is higher than 100 degree Celsius.

15. The composite according to claim 13 wherein the dielectric is an organic material with good mechanical properties, being one of a thermosettable plastic and a material which hardens due to chemical processes.

16. The composite according to the claim 15 wherein the organic material is a two component polymer, with at least one component liquid at room temperature.

17. The composite according to the claim 15 wherein the dielectric is a polymer selected from the group consisting of: an epoxy; an acrylic resin; a cyanolit based glue; Teflon or Mylar; and PVC, polystyrene or other plastic.

18. The composite according to the claim 1 wherein the diamond powder is distributed into a highly elastic material.

19. The composite according to the claim 13 wherein the dielectric is highly viscous but not solid at temperatures between 0 and 100 degree Celsius.

20. The composite according to the claim 13 wherein the diamond powder is distributed into a varnish or paint, including one of GE varnish and other varnish with high dielectrical strength used to insulate electrical conductors.

21. The composite according to the claim 13 wherein the dielectric is selected from the group consisting of a silicone based compound, porcelain, porcelite or other similar ceramic material, glass or other similar amorphous solid, and a refractory material selected from the group consisting of boron nitride or other oxides, nitrides or carbides with a very high melting point.

22. The composite according to the claim 1 further comprising one of powders of iron group metals or their alloys; powders of rare earth metals; powders of ferrites; and foils of the above said powders, to provide composite with a high magnetic permeability.

23. The composite according to claim 1 wherein the matrix material comprises an appropriate binder matrix, and fibers and/or laminas of material with high tensile strength used to improve the mechamcal properties of the composite.

24. The composite according to claim 23 wherein the material having high tensile strength is one of organic fibers, graphite or boron fibers, and glass or ceramic fibers.

25. The composite according to claim 23, wherein the fibers are oriented at diverse parts of the composite, leading to a substantially multilayer structure.

26. A material comprising the diamond based composite according to claim 1 sandwiched between foils of high tensile strength material.

27. A biomedical device in which thermal load management is critical, comprising the diamond based composite of claim 1.

28. A device according to claim 27, said device being selected from the group consisting of vials, centrifugal tubes, microtiter and other multiwell plates, capillaries, heat removal plates in electrophoresis devices, and cryotips used in cryosurgery and holders.

29. A device according to claim 28, said device being a miniaturized biomedical device and selected from the group consisting of: part of a biochip setup; thermal motherboard to which individual elements of a biochip setup are thermally anchored at well defined temperature zones; and thermal bus-lines to which the individual elements of a biochip setup are thermally anchored.

30. A device having a biochip with a localized heater/cooler architecture, which uses a series of localized Joule heaters and Peltier element coolers to locally generate the temperature profile required by the biochip, said biochip being composed of a diamond-based composite according to claim 1.

31. The device according to claim 27 , comprising miniaturized PCR chambers, wherein diamond based composites are used as a material from which is produced an array of the miniaturized PCR chambers and other thermal management critical elements.

32. The device according to claim 31, wherein due to excellent heat conductivity of the diamond based composite, no active cooler is required and the heat is simply dissipated via the diamond based composite.

33. The device according to claim 27, wherein the device is a biochip with an array of heating pads heated by placing the biochip in a fast changing magnetic field.

34. A device according to claim 33, wherein a ternary diamond based composite loaded with small, sub-micron ferrite or ferryte grains is used, and wherein the heating is locally accomplished because in the ferromagnetic medium fast change of external magnetic field leads to "eddy currents" which rapidly decay while depositing a sufficient amount of heat.

35. A device according to claim 34, wherein each of the ferromagnetic diamond based composite miniaturized heating pads is thermally anchored to a PCR micro-chamber or a microcapillary produced from a diamond based composite.

36 A surgical cast material composed of the diamond-based composite according to claim 1, said surgical cast material being lightweight, heat conducting and permitting respiration removal.

37. A miniaturized electronics device comprising a diamond based composite according to claim 1, selected from the group consisting of VLSI devices; microprocessors; miniaturized AC/DC and DC/DC converters; motors; and power supplies and transformers.

38. The device of claim 37, wherein efficient heat removal from the close neighborhood of a critical VLSI chip is achieved by use of high conductivity chip carriers and multilayer printed boards produced from the diamond based composite.

39. The device of claim 37 having a structure selected from the group consisting of a ternary diamond based composite with the third component being high tensile strength fibers, a layer of diamond based composite sandwiched between two high mechamcal strength polymer foils, and a high strength polymer based honeycomb structure filled with diamond based composites.

40. The device according to claim 37, wherein the diamond based composite is ferromagnetic, and used to encase the power supplies or to produce ferromagnetic cores of the electrical transformers.

41. A power supply, transformer, or motor having windings according to claim 37, wherein a varnish loaded with diamond powder is used to electrically insulate the windings.

42. A household or industrial device comprising a diamond based composite according to claim 1, being an electric insulator, the device selected from the group consisting of containers and plates usable in ovens and microwave heaters, plastic-based diamond based foils, cookware, electrical sockets for devices operating at elevated temperature, and light bulb sockets.

43. Cookware according to claim 42, wherein Teflon is the matrix material of the diamond based composite.

44. An automotive device comprising a diamond based composite according to claim 1, said device selected from the group consisting of car radiators, brake pads and spark plug cores.

45. A method for making a diamond based composite having a thermal conductivity pathway comprising the steps of: obtaining a diamond powder and a matrix material having a lower thermal conductivity than the diamond powder: determining the percolation threshold for the diamond powder by direct measurement of either thermal diffusivity, percolation or conductivity, calculation, or use of a look-up table; and mixing the diamond powder and the matrix material to achieve a proportion of diamond powder to matrix material greater than or less than the percolation threshold thereby producing a thermally conductive diamond based composite.

46. A method of making a diamond based composite according to claim 45 , wherein the mixing step further comprises removing all air bubbles by at least one of processing in a vacuum or under reduced pressure, repetitive mixing and stirring, use of the rolling method, and annealing at an elevated temperature.

47. The method of claim 45, further comprising increasing the filling factor of the diamond powder by at least one of pressure comptification, sedimentation of the diamond powder in a viscous matrix, and centriftigation of the diamond powder in a viscous matrix.

48. The method of claim 45, further comprising fabricating the diamond based composite into a device by one of stamping, injection molding, and thermosetting.

49. A product formed by the method according to claim 45.

50. The method of transferring heat in a device, comprising using a diamond based composite according to claim 1.

51. A diamond based composite comprising diamond powder distributed in a non- metallic matrix material, the filling factor being between about 15% and about 75% and the conductivity of the comprising being substantially greater than the conductivity of the matrix material without diamond.