JP2011181651A - Heat dissipation substrate and method for manufacturing the same - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem with a conventional heat dissipation substrate, wherein an increase in the content of an inorganic filler to improve thermal conductivity easily causes coarse surface of a heat transfer layer, resulting in deterioration in bondability between the heat dissipation substrate and a lead frame embedded therein or a wiring substrate, thereby causing influence on electric insulation. <P>SOLUTION: The heat dissipation substrate 101 includes a metal plate 107, a sheet-like heat transfer layer 102 which is formed on the metal plate 107 and contains a crystalline epoxy resin, and a wiring substrate 104 and a lead frame 103 which are fixed to the heat transfer layer 102. In the substrate, the content of the inorganic filler 127 in the heat transfer layer 102 is ≥66 vol.% and ≤90 vol.% and the wiring substrate 104 and the lead frame 103 are connected to each other via a solder portion 108. Thus, the heat dissipation substrate has highly accurate positioning and superior heat dissipation performance. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a heat dissipation substrate used for various electronic components that require heat dissipation and a method for manufacturing the same.

  In order to reduce the size of electronic devices, the heat generation density of power semiconductors, high-performance semiconductors, light-emitting elements, and the like has been improved. A heat dissipation board having a heat transfer layer is used for heat dissipation of such electronic components. However, the conventional heat transfer layer produced by adding an inorganic filler to a resin has the following problems. When the content of the inorganic filler in the heat transfer layer is increased, the thermal conductivity increases. However, the surface roughness of the heat transfer layer increases and the moldability decreases. Further, the surface of the heat transfer layer and internal voids increase, the gloss of the surface decreases, and the electrical insulation properties decrease.

  Such a problem is more likely to occur as the content of the inorganic filler in the heat transfer layer is increased. This is because the resin component content in the heat transfer layer is reduced. That is, it is difficult to wet and fix the inorganic filler whose content is increased with the resin. Next, the relationship between the volume fraction of the inorganic filler in the conventional heat transfer layer and the thermal conductivity and surface roughness at that time will be described with reference to FIG.

  FIG. 21 shows the relationship between the volume fraction of inorganic filler (unit: Vol%), thermal conductivity at that time (unit: W / m · K), and surface roughness Rmax (unit: Å). is there.

  In FIG. 21, (1) is a theoretical curve (Bruggeman's equation), (2) is Rmax of a conventional product, (3) is the glossiness at that time (the measured 20 ° glossiness is zero, at most 5) Is as follows).

  It can be seen from FIG. 21 that the surface roughness increases when the volume fraction of the inorganic filler is 60 Vol% or more, and further 66 Vol% or more. Further, at 70 Vol%, the surface roughness Rmax exceeds 7500 mm (0.75 μm). As the volume fraction of the inorganic filler increases and the surface roughness Rmax increases, the glossiness decreases rapidly, and cracks, cracks, voids and the like are likely to occur on the surface and inside of the heat transfer layer. In addition, the binding force of the inorganic filler itself is reduced, or the inorganic filler exposed on the surface tends to fall off. Therefore, a problem arises in the adhesion and fixability of the heat-generating component to the heat transfer layer surface.

  This will be described in more detail with reference to FIG. FIGS. 22A and 22B are cross-sectional views schematically illustrating the problems of the conventional heat transfer layer when the content of the inorganic filler contained in the heat transfer layer is high.

  As the content of the inorganic filler increases, an infinite number of voids 5 are generated on the surface and inside of the free surface 4 of the heat transfer layer 3 composed of the inorganic filler 1 and the thermosetting resin 2 as shown in FIG. is doing. Due to the void 5, the surface roughness of the free surface 4 increases rapidly and the glossiness decreases. The free surface 4 means a natural surface that is not in contact with another individual and is not polished or cut.

  As described above, when the surface roughness of the free surface 4 increases and voids 5 are generated inside or on the surface, the thermal conductivity of the heat transfer layer 3 does not increase so much even if the content of the inorganic filler 1 is increased. In addition, the heat conduction efficiency may decrease due to a decrease in adhesion with a heating element (not shown). Furthermore, the moldability of the heat transfer layer 3 itself is reduced. Alternatively, the heat transfer layer 3 itself becomes brittle and easily chipped.

  Furthermore, as shown in FIG. 22B, when the lead frame 6 or the substrate 7 is embedded in the heat transfer layer 3, the interface between the heat transfer layer 3 and the lead frame 6, or the substrate 7 and the heat transfer. The adhesiveness at the interface with the layer 3 is low, and the gap 8 is likely to occur.

  Further, due to the gap 8 or insufficient adhesion, the thermal conductivity from the lead frame 6 to the heat transfer layer 3 or from the substrate 7 to the heat transfer layer 3 is reduced, and further, electrical insulation (for example, leakage current at high temperature). Decreases.

  Further, the positioning accuracy when the lead frame 6 and the substrate 7 are embedded in the heat transfer layer 3 is also lowered, because the fluidity of the heat transfer layer 3 is lowered.

  Conventionally, for example, it has been proposed to increase the heat dissipation effect by improving the adhesiveness of the heat transfer layer 3 by using, for example, a resin having an excellent insulating property such as a polyimide film and a high glossiness. ing. As an application example thereof, a metal core substrate has been proposed.

  Next, a conventional metal core substrate will be described with reference to FIG. FIG. 23 is a cross-sectional view of a conventional metal core substrate.

  FIG. 23 is a cross-sectional view of a conventional metal core substrate 9. In FIG. 23, an electrical insulating layer 11 is formed on the metal plate 10. A copper foil 12 is laminated on the electrical insulating layer 11. On top of this, the electronic component 14, the semiconductor 15, the terminal 16, and the like are mounted using the solder 13. As the thermal conductivity of the electrical insulating layer 11 is higher, heat from the electronic component 14 and the semiconductor 15 can be transmitted to the metal plate 10 and temperature rise can be suppressed. Although the material which added the inorganic filler to the film-like resin sheet can be used as the electrical insulating layer 11, it is difficult to raise the content rate of an inorganic filler (for example, patent document 1).

  As a method for increasing the thermal conductivity, a method of using a material having a high thermal conductivity as the filler of the electrical insulating layer 11 or increasing the filling amount of the filler is often used. It is also effective to increase the thermal conductivity of the resin. Such a configuration will be described with reference to FIGS. For example, it has been proposed to use a crystalline resin as means for increasing the thermal conductivity of the resin (for example, Patent Document 2).

  24A to 24C are explanatory diagrams of a conventional crystalline resin. The use of a crystalline resin aims at polymerizing monomers 18 having mesogenic groups 17 with each other to obtain electrical insulation and excellent thermal conductivity. In order to obtain high heat dissipation (or high thermal conductivity) using a crystalline resin, it is necessary to increase the crystallization rate of the crystalline resin. However, in a crystalline resin, the higher the crystallization rate, the harder and brittle the resulting substrate. In other words, it breaks without bending or is easily cracked or cracked. Therefore, even if a heat dissipation substrate for mounting an electronic component is manufactured using a crystalline resin, its use is greatly limited.

  For this reason, even if an epoxy resin having a mesogenic group is used, it is difficult to solve the problem described with reference to FIG. Further, as the epoxy resin is crystallized, the problem described with reference to FIG. 22 is likely to occur. Furthermore, such a problem easily occurs from a region where the content of the inorganic filler is lower than 66 Vol%.

  As a method for improving the strength, a method of blending a curing agent that easily takes a network structure has been proposed. However, in the case of a crystalline epoxy resin, by making a steric bond, the crystallinity is often inhibited and high thermal conductivity cannot be obtained in many cases.

  Furthermore, the crystalline resin has a weak adhesion to a metal plate or the like. Alternatively, stress is likely to concentrate inside the crystallized resin or at the interface between the crystallized resin and the metal plate, and the interface between the crystallized resin and the metal plate may be peeled off due to the stress. For this reason, when a heat dissipation substrate is manufactured, there may be a problem in impact resistance such as a drop test.

Japanese Patent No. 3255315 JP-A-11-313162

  In these conventional heat dissipation boards, when the content of the inorganic filler is increased to increase the thermal conductivity, the surface of the heat transfer layer is likely to be roughened, and the heat transfer layer and the lead frame or wiring board embedded in the heat transfer layer There was a possibility that the adhesion between the two would be reduced and the electrical insulation would be affected. In addition, it has been difficult to electrically connect the lead frame and the wiring board by burying the lead frame and the wiring board in the heat transfer layer with high precision in a state of having highly accurate positional information.

  The present invention has been made in consideration of the above situation. That is, according to the present invention, even when the content of the inorganic filler is increased, the surface of the heat transfer layer is prevented from being roughened, the electrical insulation is improved, and the lead frame is formed so that each of the heat transfer layers is substantially flat. Another object of the present invention is to provide a heat dissipation board in which wiring boards are embedded while being electrically connected while maintaining high positional accuracy, and a method for manufacturing the same.

  In order to solve the above-described problems, the present invention provides a heat transfer comprising a metal plate, a sheet-shaped cured thermosetting resin including a crystalline epoxy resin, and an inorganic filler. A heat dissipation board having a layer, a wiring board fixed to the heat transfer layer, and a lead frame, wherein the content of the inorganic filler in the heat transfer layer is 66 Vol% or more and 90 Vol% or less. In addition, a part or more of the side surface of the wiring board and a part or more of the side surface of the lead frame are a heat dissipation board connected via one or more of a solder part or a mechanical connection part.

  The present invention greatly improves the heat dissipation or heat transfer characteristics of the heat dissipation substrate and provides a heat dissipation substrate having excellent electrical insulation at high temperatures and a method for manufacturing the same, thereby providing a high-temperature semiconductor (eg, silicon-based). Other than SiC-based semiconductor power devices) can be operated efficiently, contributing to the environment, energy saving, power consumption, carbon dioxide gas reduction, and the like.

  Further, positioning or alignment of the lead frame and the wiring board, electrical connection thereof, or heat spread (or thermal diffusion) can be efficiently performed via the solder portion.

Sectional drawing which shows a mode that the heat sink in Example 1 of this invention was seen from diagonally (A) (B) is sectional drawing explaining the contact part provided between the wiring board and the lead frame. (A) (B) is sectional drawing explaining a mode that the side electrode provided in the side surface of the wiring board is utilized together (A) and (B) are cross-sectional views for explaining a state in which the heat dissipation substrate is viewed obliquely, and a cross-sectional view in which a part thereof is enlarged Top view of another example of heat dissipation board (A), (B) is sectional drawing explaining a mode that a heat sink is manufactured together Schematic explaining how to form an uncured composition into a sheet Diagram explaining the gloss evaluation method of heat transfer layer Schematic diagram explaining the evaluation method of surface roughness of heat transfer layer (A) is a schematic diagram for explaining the surface of the heat transfer layer in an enlarged manner, (B) is a cross-sectional view for schematically explaining the state of convection in the heat transfer layer. Schematic diagram explaining the cross section near the surface of the heat transfer layer (A) is a photomicrograph of the surface of the heat transfer layer according to this example, and (B) is a schematic diagram thereof. (A) is a perspective view showing the surface of the heat transfer layer three-dimensionally, (B) is a schematic diagram thereof. (A) is a micrograph of the surface of the heat transfer layer of the conventional example, (B) is a schematic diagram thereof. (A) is a perspective view showing the surface of the heat transfer layer three-dimensionally, (B) is a schematic diagram thereof. The graph which shows the temperature characteristic of the leakage current of the heat-transfer layer by a present Example, and the heat-transfer layer of a prior art example as an example of an electrical property Sectional drawing explaining how thermosetting resin convects on the side and bottom of a lead frame or wiring board embedded in an uncured composition The figure explaining an example of the experimental result about the relationship between the volume fraction of the inorganic filler in a heat-transfer layer, and thermal conductivity (A) and (B) are perspective views of a chip-like wiring board having wiring portions insulated from each other vertically and horizontally. (A) and (B) are perspective sectional views for explaining a state in which a chip-like wiring board is inserted between a plurality of lead frames, respectively, and a sectional view in which a part thereof is enlarged. The figure which shows the relationship between the volume fraction of an inorganic filler, the thermal conductivity at that time, and surface roughness Rmax (A) (B) is sectional drawing which illustrates typically the subject of the conventional heat-transfer layer in case the content rate of an inorganic filler is high Sectional view of a conventional metal core substrate (A)-(C) are explanatory drawing of the conventional crystalline resin

  Embodiments of the present invention will be described below with reference to the drawings. In each embodiment, components having the same configuration as that of the preceding embodiment are denoted by the same reference numerals, and detailed description may be omitted. The present invention is not limited to the following examples.

Example 1
FIG. 1 is a cross-sectional view showing a state in which the heat dissipation board according to the first embodiment of the present invention is viewed from an oblique direction.

  In FIG. 1, 101 is a heat dissipation board, 102 is a heat transfer layer, 103 is a lead frame, 104 is a wiring board, 105 is a wiring part, 106 is an insulating part, 107 is a metal plate, 108 is a solder part, 109 is a connecting part. is there.

  The solder portion 108 is not only connected using a commercially available solder material, but also lead-free solder for mounting electronic components, alloy bonding used for bonding conductive paste, semiconductor, etc., gold-gold bonding, ultrasonic wave This includes bonding, metal bonding, and the like.

  Further, by connecting a part or more of the side surface of the lead frame 103 and a part or more of the side surface of the wiring board 104 via the solder portion 108, the surface of the heat dissipation board 101 is smoothed and the connection area is reduced. It can be expanded in the thickness direction without affecting the shape. It may be connected by mechanical or physical connection means other than the solder part 108. For example, a part or more of the side surface of the lead frame 103 and a part or more of the side surface of the wiring board 104 are fixed using mechanical means (or physical means) such as welding, rivet, screwing, adhesive, Alternatively, it may be connected. The wiring board 104 and the lead frame 103 are connected to each other by using a part 109 that is electrically connected to each other by using a part or more of the side surfaces, thereby smoothing the surface of the heat dissipation board 101 and reducing the connection area. Can be spread. Moreover, the heat spread effect can be enhanced by electrically connecting each other.

  The heat dissipation board 101 includes at least a sheet-like heat transfer layer 102, a wiring board 104 embedded and fixed in the heat transfer layer 102, and a lead frame 103.

  The wiring board 104 and the lead frame 103 are connected, positioned, or electrically or thermally connected via the solder portion 108.

  As the wiring substrate 104, a commercially available glass epoxy substrate (the glass epoxy substrate is an interlayer connection made between an insulating portion 106 in which a glass wire is impregnated with an epoxy resin and a wiring portion 105 made of copper foil through a through hole 112 or the like. It is a glass epoxy multilayer board) or hybrid IC (for example, a wiring part 105 such as silver or copper is provided on an insulating part 106 made of a ceramic plate such as alumina, and has heat resistance and heat dissipation. It is desirable to use

  For example, the lead frame 103 of the heat dissipating board 101 in FIG. 1 can cope with a large current exceeding 100 A, and further has an effect as a heat spreader, so that it can be used for mounting or dissipating power semiconductors (not shown). Is preferred.

  Since the wiring board 104 embedded in the heat transfer layer 102 of the heat dissipation board 101 in FIG. 1 can cope with fine wiring, it can be used for high-density mounting of a semiconductor for controlling a power semiconductor (not shown) and various chip parts. Is preferred. For example, by using a glass epoxy substrate for the wiring substrate 104, it is possible to cope with high-density mounting of the control circuit. Further, by using a ceramic substrate such as an alumina substrate for the wiring substrate 104, further excellent heat dissipation can be obtained.

  This will be described in more detail. The reinforced insulation can be satisfied by setting the thickness of the heat transfer layer 102 to 0.6 mm or more. Here, reinforced insulation means a single insulation that provides protection against electric shock equal to or better than that of double insulation. Double insulation is insulation formed by basic insulation and additional insulation. By satisfying the reinforced insulation in this way, the safety of the product known by the UL standard, the IEC standard, etc. can be satisfied, and a primary side electric circuit (for example, a power supply circuit) is formed on the heat dissipation substrate 101. Can do. UL means Underwriters' Laboratories Incorporated, IEC means International Electrotechnical Commission, and corresponds to Japanese JIS.

  In addition, when using the thermal radiation board | substrate 101 for a secondary side circuit, it is also possible to make the thickness of the heat-transfer layer 102 into less than 0.6 mm. The secondary circuit includes a car power source such as a DC-DC converter. The DC-DC converter is a power converter that creates a DC output from a DC input.

  The content of the inorganic filler 127 in the heat transfer layer 102 is 66 Vol% or more and 90 Vol% or less.

  In addition, although the metal plate 107 is being fixed to one surface of the sheet-like heat-transfer layer 102, it is not necessary to limit to this form. That is, it is also useful to embed the metal plate 107 in the sheet-like, rod-like, or plate-like heat transfer layer 102. For example, it is also useful to embed the metal plate 107 on one side of the heat transfer layer 102 or embed the metal plate 107 inside the heat transfer layer 102.

  The metal material used for the metal plate 107 is gold, silver, copper, nickel, aluminum, or a composite thereof (a composite member includes an alloy of these metal materials, plating, a deposit, a clad material, etc.). Useful. This is because gold, silver, copper, nickel, and aluminum can be used alone or as a composite to have excellent characteristics such as thermal conductivity useful for the metal plate 107.

  The surface of the heat transfer layer 102 preferably has a so-called piano black-like clean gloss with a 20-degree gloss of 70 or more. For example, it is desirable that the surface be a smooth surface having a jet black depth.

  The color of the heat transfer layer 102 need not be limited to black. In FIG. 4 and the like, which will be described later, the inventors added a black pigment such as carbon black to the heat transfer layer 102, so that the color is only black. For example, if pigments such as red, blue, and green are added, a smooth surface having gloss in each color tone can be obtained. By adding a white pigment such as alumina or titanium oxide, a white glossy surface or a glossy surface like a white grand piano can be obtained. By setting it to be white as described above, the heat dissipation substrate 101 of Example 1 can be used as a mounting substrate for a semiconductor light emitting element represented by a light emitting diode (LED) or the like. In this case, since the glossy surface of the white heat transfer layer 102 functions as a light reflector of the LED, it can be used as domestic LED lighting or a vehicle headlight.

  Although it is useful to make the surface part itself that is the surface of the heat transfer layer 102 transparent, when making the surface part white, a white pigment having a particle diameter smaller than that of the inorganic filler may be added.

  Depending on the application, the gloss of the surface of the heat transfer layer 102 may not be preferable. In such a case, the planarity (or smoothness) of the surface can be improved by polishing a part of the surface of the heat transfer layer 102. Furthermore, as will be described later with reference to FIG. 9 and the like, in the heat transfer layer 102 of each example, the surface layer preferably has a low inorganic filler content and is easily polished. That is, the surface of the heat transfer layer 102 is thinly polished (for example, a surface layer 128 shown in FIG. 10B described later is polished and removed), and a heat generating component is fixed thereon with an adhesive having high thermal conductivity. It is also useful.

  The solder part 108 provided between the wiring board 104 and the lead frame 103 will be described with reference to FIGS.

  2A and 2B are cross-sectional views for explaining a contact portion provided between the wiring board and the lead frame.

  In FIG. 2, 110 is a bonding wire and 112 is a through hole. As shown in FIG. 2A, wiring with the bonding wire 110 is performed between the wiring portion 105 provided on the surface of the wiring substrate 104 and the lead frame 103 using a bonding apparatus (not shown). Can do. As the bonding wire 110, an aluminum wire, a gold wire, a copper wire, or the like is used. These have high thermal conductivity and low electrical resistance. Note that it is useful to perform gold plating or the like in advance on the surfaces of the wiring portion 105 and the lead frame 103 in order to improve bonding properties.

  Note that the lead frame 103 and the wiring board 104 need not have the same thickness.

  As shown in FIG. 2 (B), it is useful to provide a through hole 112 for interlayer connection of a plurality of wiring portions 105 provided on the wiring substrate 104.

  FIGS. 3A and 3B are cross-sectional views illustrating a state in which side electrodes provided on the side surface of the wiring board are used.

  In FIG. 3, reference numeral 111 denotes a side electrode, for example, a through hole (particularly a through hole formed using the through hole) formed in the wiring substrate 104 is divided into half (sometimes referred to as a half-hole). ) Etc. As shown in FIG. 3A, the side electrode 111 is formed on the side surface of the wiring board 104, the leading end portion of the lead frame 103 is inserted (or pressed) into the half hole, and the solder portion 108 is further formed. Since the alignment function can be exhibited at the time of alignment, the alignment accuracy can be further improved. It is useful to process the shape of the tip portion of the lead frame 103 so that the lead frame 103 can be in close contact with the side electrode 111.

  4A and 4B are a cross-sectional view illustrating a state in which the heat dissipation substrate is viewed obliquely and a cross-sectional view in which a part thereof is enlarged.

  As shown in FIG. 4A, a half hole (not numbered) is formed around the wiring board 104, and the lead frame 103 is fixed by using this half hole, The connection unit 109 is configured by being electrically connected via the network 108.

  FIG. 5 is a top view of another example of the heat dissipation board. As shown in FIG. 5, by connecting the lead frame 103 and the wiring 105 provided on the surface of the wiring substrate 104 with bonding wires 110, the electrical continuity and thermal continuity (that is, the heat spread effect) are further enhanced. be able to.

(Example 2)
Next, as a second embodiment, an example of a method for manufacturing the heat dissipation substrate 101 shown in FIG. 1 and the like will be described with reference to FIGS.

  FIGS. 6A to 6B are cross-sectional views illustrating how the heat dissipation substrate 101 is manufactured. Reference numeral 114 denotes a mold, and the temperature can be adjusted using a temperature adjusting device (not shown). Reference numeral 115 denotes a protective film that prevents resin or the like from adhering to the surface of the mold 114.

  Here, the wiring substrate 104 is a resin substrate using a commercially available glass epoxy resin, a hybrid IC using a ceramic substrate, or the like. Reference numeral 116 denotes an uncured composition, and the uncured composition 116 is thermally cured to become the heat transfer layer 102. 117 is an arrow.

  When the plurality of lead frames 103 are integrated at the periphery as shown in FIG. 4 and FIG. 5 (or the shape in which the plurality of lead frames 103 protrudes inward from the frame-shaped peripheral portion). In this case, the lead frame 103 can be positioned at the periphery (or the frame portion). However, in the case of the wiring board 104, it is in a floating island state (that is, a state independent from the outside), and accurate positioning and fixing are difficult. There is a case. In such a case, positioning the wiring board 104 is facilitated by fixing the wiring board 104 and the lead frame 103 with one or more solder portions 108.

  As shown in FIG. 6A, the uncured composition 116 is disposed on the metal plate 107. The uncured composition 116 may be in the form of a sheet, a pellet, or a round bar.

  Further, the wiring board 104 and the lead frame 103 are disposed thereon, and these are pressed and embedded with the mold 114. Note that heating the mold 114 is preferable in the embedding step because the uncured composition 116 has a low viscosity.

  FIG. 6B is a cross-sectional view showing a state after the molding of the uncured composition 116 is completed. As shown by the arrow 117, the mold 114 and the protection from the uncured composition 116 (shown as the cured heat transfer layer 102 in FIG. 6B) in which the wiring substrate 104 and the lead frame 103 are embedded. Film 115 is removed. Thereafter, the uncured composition 116 is thermally cured to form the heat transfer layer 102. The thermosetting of the uncured composition 116 may be performed in the mold 114 or taken out from the mold 114.

  In this curing, it is better to peel off the protective film 115 in advance. By doing so, the surface of the heat transfer layer 102 is a free surface, and the thermosetting resin contained in the uncured composition 116 is easily convected. As a result, the 20 degree gloss of the surface portion can be effectively 70 or more, or the surface roughness Ra can be 3000 Å (0.3 μm) or less.

  As described above, the metal plate 107 includes an uncured thermosetting resin (not shown) including a crystalline epoxy resin (not shown) and an inorganic filler (not shown). An arrangement step of disposing the uncured composition 116, the lead frame 103, and the wiring substrate 104, and heating the uncured composition 116 on the metal plate 107, so that the lead frame 103 and the wiring substrate 104 are heated. And a convection step of heating at least the uncured composition 116 to a temperature equal to or higher than the crystallization temperature of the crystalline epoxy resin to form a liquid state and convection through the gaps of the inorganic filler, The content of the inorganic filler 127 is 66 Vol% by the method of manufacturing the heat dissipation substrate 101 including the thermosetting step of thermally curing the curable composition 116 to form the heat transfer layer 102. Furthermore, even less 90 vol%, can be realized radiating substrate 101 having an excellent insulating property. The lead frame 103 and the wiring board 104 are electrically connected via one or more (preferably two or more, more preferably three or more) solder portions 108 or connection portions 109 and fixed at the same time. The mounting accuracy of various electronic components on the lead frame 103 and the wiring board 104 can be improved.

  Next, preforming of the uncured composition 116 will be described with reference to FIG.

  FIG. 7 is a schematic diagram for explaining how the uncured composition 116 is formed into a sheet. Reference numeral 118 denotes a molding apparatus. For example, the uncured composition 116 can be formed from a kneading device (not shown) and simultaneously formed into a shape that can be easily formed, such as a sheet shape, a round bar shape, or a pellet shape, by using the forming device 118 or the like. it can.

  By pre-molding the uncured composition 116 into a sheet shape, a round bar shape, a pellet shape, or the like, handling becomes easy. In particular, a sheet shape is preferable. Thus, by making the uncured composition 116 into a sheet shape, handling of the uncured composition 116 itself becomes easy. For example, this is similar to a prepreg formed by impregnating glass fiber with an epoxy resin, which is an intermediate for forming a laminated substrate. Moreover, it becomes easy to affix on the surface of the metal plate 107 by making the uncured composition 116 into a sheet shape. Further, even when the metal plate 107 is bent into a curved surface or bent into an L shape, the uncured composition 116 can be easily pasted on the surfaces with a substantially constant thickness. Can do. In addition, the sheet-like uncured composition 116 attached to the surface of the metal plate 107 has a predetermined temperature (for example, 150 ° C. to 200 ° C., particularly added to the uncured composition and higher than the melting point of the crystalline epoxy resin). When heated, the thermosetting resin contained therein is softened and liquefied. And this liquefied and low-viscosity thermosetting resin convects the gaps of the inorganic filler 127, thereby improving the adhesion with the metal plate 107 as a base and suppressing the generation of voids inside and on the surface. .

  As the protective film 115, a commercially available resin film such as PET whose surface is treated with silicon, water repellent, or released can be used.

  The thickness of the uncured composition 116 that has been preformed is preferably 0.02 mm or more and 5.00 mm or less. If the thickness is less than 0.02 mm, pinholes may occur in the uncured composition 116. On the other hand, if the thickness exceeds 5.00 mm, the heat dissipation may be affected when used for the heat dissipation substrate 101. 5 is heated, the moldability of the uncured composition 116 can be improved, and the thickness variation of the finished sheet can be reduced. Then, by leaving the protective film 115 as a kind of protective sheet on the surface of the uncured composition 116 preformed into a sheet shape, it is possible to prevent dirt such as dust from adhering.

  The uncured composition 116 may be applied to the metal plate 107 using a coater, or may be molded with an extrusion molding machine.

  A solvent may be added to adjust the viscosity. You may knead | mix in the state previously melt | dissolved in the solvent. The manufacturing method is not limited to the above.

  In addition, as heat-generating components (not shown) attached to the heat dissipation substrate 101, for example, in addition to power semiconductors (power transistors, power FETs, high-power LEDs, high-power lasers), electronic components such as transformers, coils, choke coils, It may be a light emitting element such as an LED or a semiconductor laser.

  Further, these heating elements may be attached using a heat conductive adhesive or the like in order to use the lead frame 103 or the like for a heat spreader. Further, lead terminals or the like coming out of the heating element may be soldered to the lead frame 103. Note that lead-free solder, conductive paste, or the like may be used as solder (not shown).

  It is also useful to mount an electronic component (not shown) for controlling a heat generating component (not shown) on the wiring board 104.

  For the connection between the wiring substrate 104 and the lead frame 103 embedded in the heat transfer layer 102, it is useful to use a jumper wire, a 0Ω chip resistor, a wire or the like (the jumper wire etc. Not shown).

  By mounting power semiconductors, control semiconductors, etc. on the heat generation board, the power supply circuit (including the sustain circuit) of PDP televisions, which becomes a circuit where heat generation becomes a problem, DC-DC converters for passenger cars, DC for solar cells, etc. -AC converters can be manufactured. The DC-AC converter is a converter that converts direct current into alternating current, and the DC-DC converter is a converter that converts direct current into direct current.

(Example 3)
In Example 3, the surface state of the heat transfer layer 102 will be described with reference to FIGS. 8 and 9. 119 is a gloss meter, 120 is a light source, 121 is a lens, 122 is a light receiving unit, 123 is a dotted line, and 124 is a surface.

  It is useful that the surface of the heat transfer layer 102 be a glossy surface or a smooth surface. Hereinafter, the gloss of the heat transfer layer 102 will be described with reference to FIG. FIG. 8 is a diagram for explaining a method for evaluating the gloss of the heat transfer layer 102.

  A light source 120 is provided inside the gloss meter 119. The gloss meter 119 is sometimes called a gloss meter. The light emitted from the light source 120 is irradiated on the surface of the heat transfer layer 102 which is an object to be measured through the lens 121. This light is reflected by the surface of the object to be measured, and enters the light receiving unit 122 via the lens 121.

  As the glossiness, for example, the mirror glossiness of JIS K 5600-4-7 can be used. The specular gloss is the ratio of the light beam reflected from the object in the specular direction and the light beam reflected from the glass having a refractive index of 1.567 in the specular direction at the specified angle of the light source 120 and the light receiving unit 122. The angle formed between the direction perpendicular to the surface to be measured and the incident direction from the light source 120 and the angle formed between the direction perpendicular to the surface to be measured and the direction toward the light receiving unit 122 are, for example, 20 ° as shown in FIG. It can be. Hereinafter, the specular gloss of JIS K5600-4-7 measured with the angle set to 20 ° is referred to as 20 ° gloss.

  In addition, the measurement of glossiness may be easily affected by the surface condition such as the shape of the surface 124 and the presence or absence of a pattern. Therefore, the measurement point (so-called n number) is increased and the measured value at the place with the highest glossiness is adopted. This is because the place with the highest glossiness corresponds to the most unaffected surface.

  If measurement of 20 degree gloss is difficult, 60 degree gloss may be employed.

  The surface of the heat transfer layer 102 preferably has a 20 degree gloss of 70 or more. More desirably, it is 80 or more. When the gloss level is less than 70, the surface may be rough. In this case, when fixing a heat-generating component on the heat transfer layer 102, the adhesiveness with the heat-generating component to be fixed may decrease, and the heat dissipation efficiency may decrease. Also, in the case of evaluating with 60-degree gloss, a gloss of 70 or more is desirable, and more desirably 80 or more.

  Next, the surface roughness of the heat transfer layer 102 will be described with reference to FIG. FIG. 9 is a schematic diagram illustrating a method for evaluating the surface roughness of the heat transfer layer 102.

  The surface roughness is measured while scanning the surface of the heat transfer layer 102 as indicated by an arrow 117 on the surface roughness meter 125. The surface roughness meter 125 is preferably a contact type, but a non-contact type is also possible. For example, a laser type surface roughness meter 125 may be used. When such an optical surface roughness meter 125 is used, it is desirable to examine whether the obtained result is correct.

  Note that the measurement of the surface roughness may be easily influenced by the surface conditions such as the shape of the surface and the presence / absence of the pattern 126. Therefore, the number of measurement points (so-called n number) is increased, and the value of the measured value of the surface roughness at the portion with the smallest surface roughness is adopted. This is because the portion with the smallest surface roughness corresponds to the surface that is least affected by the other.

  The surface roughness Ra of the surface of the heat transfer layer 102 is preferably 3000 mm or less (0.3 μm or less). More desirably, it is 2000 mm or less. Alternatively, Rmax is preferably 15000 mm or less (1.5 μm or less). More desirably, it is 13000 mm or less (1.3 μm or less). Here, the surface roughness Ra is called arithmetic average roughness, and is defined in JIS-B-0601894 after revision. Rmax is called “maximum height” in JIS-N-0601882 before revision and corresponds to Ry (maximum height) in JIS-B-0601894 after revision.

  When the surface roughness Ra exceeds 3000 mm (greater than 0.3 μm), or when Rmax exceeds 15000 mm (greater than 1.5 μm), fixing is performed when fixing a heat-generating component on the heat transfer layer 102. Adhesiveness with the heat-generating component to be reduced may reduce heat dissipation efficiency.

  When the glossiness of the surface of the heat transfer layer 102 is set to 70 or more, or the surface roughness Ra is 3000 mm or less, or Rmaxz (or Rz) is 15000 mm or less (1.5 μm or less). It is desirable to provide a surface layer mainly composed of resin on the surface of 102. The resin constituting the surface layer is preferably composed mainly of a thermosetting resin that oozes out or convects from the gaps between the inorganic fillers contained in the heat transfer layer 102.

  Furthermore, the reason why the surface of the heat transfer layer 102 in each embodiment becomes smooth will be described with reference to FIGS. FIG. 10A is a schematic diagram illustrating the surface of the heat transfer layer 102 in an enlarged manner, and FIG. 10B is a cross-sectional view schematically illustrating the convection state of the heat transfer layer 102. 126 is a pattern, 127 is an inorganic filler, 128 is a surface layer, 129 is a main part, and 130 is a thermosetting resin.

  As described above, the surface of the heat transfer layer 102 is smooth and almost no voids remain inside. That is, the inorganic filler 127 constituting the heat transfer layer 102 is sufficiently covered with the thermosetting resin 130. As a result, in the thickness direction, the contact portion between the particles of the inorganic filler 127, or the interface between the inorganic filler 127 and the thermosetting resin 130, between the particles of the inorganic filler 127, or inside the thermosetting resin 130, Almost no bubbles remain. As a result, the leakage current is also reduced.

  As shown in FIG. 10B, a fine pattern 126 may be observed on the surface of the heat transfer layer 102. The pattern 126 is similar to, for example, a Benard cell.

  FIG. 10B is a cross-sectional view illustrating how the thermosetting resin 130 convects (for example, thermal convection such as Benard convection) when the uncured composition 116 is thermally cured to form the heat transfer layer 102. FIG.

  An arrow 117 in FIG. 10B indicates that the thermosetting resin 130 contained in the uncured composition 116 (corresponding to the portion illustrated as the heat transfer layer 102) has a drastic decrease in viscosity during heating, and the inorganic filler The state of convection through the gap 127 is schematically shown. Note that it is useful to use a crystalline epoxy resin as the thermosetting resin 130. The crystalline epoxy resin is easily convectioned by heat because it changes to a very low viscosity liquid when the melting point (or crystallization temperature) is exceeded. Note that FIG. 10B is a schematic diagram and is not accurate. Therefore, the pattern 126 or Benard cell may not be observed on the surface 124 of the heat transfer layer 102.

  Next, Benard convection will be described. Benard convection is a phenomenon discovered in 1900 by Henri Bernard. In this phenomenon, when the lower surface of the fluidized bed having high viscosity is heated, the heated fluid rises by buoyancy, and a cellular pattern 126 (fluid cell) is generated inside the fluid and further on the surface 124 thereof.

  Next, the surface layer 128 and the surface 124 of the heat transfer layer 102 due to Benard convection, and the mechanism in which voids are less likely to be generated inside, or the voids remaining in the surface layer and inside disappear due to Benard convection and the like will be described.

  The inventors have conducted various experiments on hundreds of different uncured compositions 116. That is, the surface 124 was observed until the uncured composition 116 was heated and thermally cured. Then, when the thermosetting resin 130 before curing contained in the uncured composition 116 forms the heat transfer layer 102 mainly composed of the inorganic filler 127 having a content rate of 66 Vol% or more, the Benard convection is actively performed. Thus, it has been found that voids in the heat transfer layer 102 are drastically reduced. Note that the fluidity of the surface layer 128 may be increased, or traces such as Benard convection may be difficult to find on the surface 124 of the heat transfer layer 102 depending on the heat curing conditions.

  Note that the thermosetting resin 130 before thermosetting contained in the uncured composition 116 includes a crystalline epoxy resin component (crystallinity before thermosetting represented by a crystalline epoxy monomer or the like). It is desirable that an epoxy resin is included.

  The uncured composition 116 before curing includes a crystalline epoxy resin in a state before thermal curing, typified by a crystalline epoxy monomer. The crystalline epoxy resin before thermosetting is in a solid state at a temperature lower than the crystallization temperature (or lower than the melting point), but becomes a low viscosity liquid state such as water at a temperature higher than the crystallization temperature (or higher than the melting point). Although the crystallized epoxy resin (or the resin member that forms a crystalline epoxy resin by thermosetting) before thermosetting exists in this state as a stable solid in the crystalline state, the crystalline state quickly melts as the melting point is reached. To a very low viscosity liquid.

  Therefore, the crystalline epoxy resin before curing contained in the uncured composition 116 exceeds the crystallization temperature in the course of heating the uncured composition 116 and becomes a low-viscosity liquid state such as water. The low-viscosity liquid crystalline epoxy resin component (uncured state) dissolves the amorphous epoxy resin component whose viscosity is not easily lowered even by heating. The thermosetting resin 130 thus reduced in viscosity (that is, the thermosetting resin 130 before being cured) increases its fluidity and positively generates Benard convection or similar convection therein. Finally, the component that becomes the crystalline epoxy resin and the amorphous epoxy resin component are thermally cured, and the convection is terminated.

  And by this convection such as Benard, the particles of the inorganic filler 127 move to a more stable position with each other, the inorganic fillers 127 closely adhere to each other, and the packing density increases. Further, the thermosetting tree 130 having a reduced viscosity penetrates into narrow gaps between the particles of the inorganic filler 127 and actively wets the surfaces of various members.

  In this way, bubbles and voids remaining on the surface and the periphery of the inorganic filler 127 can also be discharged to the outside in a short time by this Benard convection. Thus, the voids are reduced by Benard convection and the filling rate of the inorganic filler 127 is increased. Further, the surface 124 of the heat transfer layer 102 becomes smooth.

  The crystalline epoxy resin before curing exists as a stable solid at room temperature (for example, 25 ° C.). Therefore, work such as weighing is easy.

  In general, there is a correlation between the molecular weight and viscosity of the epoxy resin component before curing. Therefore, in order to lower the viscosity of the epoxy resin component, the molecular weight is generally reduced. However, when the molecular weight of the epoxy resin component is reduced, the softening point is also lowered, so that the handleability at room temperature (for example, 20 ° C. to 30 ° C.) is lowered. Moreover, Tg (glass transition temperature) of the cured epoxy resin having a reduced viscosity is lowered. Therefore, it is preferable to use a crystalline epoxy resin that is solid at room temperature even before curing.

  As described above, the heat transfer layer 102 includes at least the thermosetting resin 130 and the inorganic filler 127. And as for the content rate of the inorganic filler 127, 66 Vol% or more is desirable. In addition, a surface layer 128 mainly having a thermosetting resin 130 is provided on the surface 124 with a thickness of 20 μm or less. With this configuration, generation of voids in the heat transfer layer 102 and on the surface 124 can be suppressed. Further, the inorganic filler 127 in the heat transfer layer 102 can be highly filled, and the thermal conductivity can be increased. Furthermore, the moldability of the heat transfer layer 102 and the convenience of handling can be greatly improved.

  Although it is desirable that a Benard cell is generated as a result of generating thermal convection, it may be difficult to distinguish whether the fine convex portion of the surface 124 is a Benard cell. However, at least if it is caused by convection generated when the heat conductive material is heated, it is a kind of Benard cell.

  As the inorganic filler 127, it is desirable to use at least one selected from alumina, aluminum nitride, boron nitride, silicon carbide, silicon nitride, magnesium oxide, and zinc oxide. By using such a material having high thermal conductivity as the inorganic filler 127, the thermal conductivity of the heat transfer layer 102 can be further increased. These materials are also preferable from the viewpoint of insulation and the like. Moreover, it is also possible to combine these materials.

  The average particle size of the inorganic filler 127 is desirably in the range of 0.1 μm to 100 μm. When the average particle size is less than 0.1 μm, the specific surface area becomes large, the kneading of the uncured composition 116 becomes difficult, and the moldability of the heat transfer layer 102 may be affected. On the other hand, if the thickness exceeds 100 μm, it is difficult to make the heat transfer layer 102 thinner, which may affect the heat dissipation of the heat dissipation substrate 101 and may affect the miniaturization of the product. In order to increase the filling rate of the inorganic filler 127, a plurality of types of inorganic fillers 127 having different particle size distributions may be selected and mixed.

  FIG. 11 is a schematic diagram illustrating a cross section near the surface of the heat transfer layer 102. The heat transfer layer 102 includes a main portion 129 having a high content of inorganic filler 127 (for example, a portion formed by precipitation of the inorganic filler 127 by its own weight) and a surface layer 128 having a low content of inorganic filler 127 (for example, a supernatant portion). ). A dotted line 123 indicates a boundary portion between the main portion 129 and the surface layer 128, but a boundary or interface does not actually exist. This is because the thermosetting resin 130 constituting the main portion 129 convects and exudes from the inside as the surface layer 128. In addition, since the interface does not exist, a problem due to the presence of the interface does not occur, so that the reliability is increased.

  The surface layer 128 is a layered portion formed mainly on the thermosetting resin 130 having a thickness of 20 μm or less, desirably 10 μm or less, and more desirably 5 μm or less, formed on the surface of the heat transfer layer 102. In this embodiment, the surface layer 128 is positively provided near the surface of the heat transfer layer 102. In addition, when the thickness of the surface layer 128 exceeds 20 μm, there is a possibility that heat conduction to the heat-generating component fixed thereon is hindered.

  Further, the content of the inorganic filler 127 in the entire heat transfer layer 102 is desirably 66 Vol% or more and 90 Vol% or less. Vol% means a percentage of the volume. If it is less than 66 Vol%, the moldability and the like are excellent, but the thermal conductivity may be low.

  In addition, the thermal conductivity can be 1 W / m · K or more by covering 98% or more of the surface of the inorganic filler 127 with the thermosetting resin 130. When less than 98% of the surface of the inorganic filler 127 is covered with the thermosetting resin 130, the thermal conductivity may be lowered. Alternatively, the predetermined strength may not be obtained.

  As described above, 98% or more of the surface of the inorganic filler 127 is covered with the thermosetting resin 130, and the surface layer 128 mainly composed of the thermosetting resin 130 exists on the surface layer of the heat transfer layer 102. preferable. This prevents a current from flowing along the interface between the inorganic filler 127 and the thermosetting resin 130, and therefore the withstand voltage characteristic is unlikely to deteriorate.

  A part of the thermosetting resin 130 that covers the surface of the inorganic filler 127 is preferably provided as the surface layer 128 on the surface of the heat transfer layer 102. Thereby, the bending strength of the heat transfer layer 102 can be increased, and even if the heat transfer layer 102 is bent, cracks are hardly generated.

  In addition, after the mixture of the inorganic filler 127 and the thermosetting resin 130 is cured to form a base layer corresponding to the main portion 129, a separately prepared thermosetting resin 130 is thinly applied and cured, and the surface layer 128 is cured. A resin layer corresponding to can also be formed. However, even if a structure similar to the heat transfer layer 102 is produced in this way, an excellent heat dissipation effect is not exhibited. Further, peeling or the like may occur at the interface between the base layer and the resin layer.

  As described above, the epoxy resin (or thermosetting resin 130 or the like) included in the main portion 129 and the epoxy resin (or thermosetting resin 130 or the like) included in the surface layer 128 have substantially the same composition or substantially the same. An epoxy resin composition is desirable. By making the epoxy resin contained in the thermosetting resin 130 have substantially the same composition, it is possible to suppress the stress 133 from being concentrated on the portion shown by the dotted line 123 in FIG.

  Whether the surface layer 128 exudes from the gap between the inorganic fillers 127 at the time of thermosetting can be determined by analyzing the cross section with an analyzer such as a scanning microscope (SEM) or an FTIR using a microscope. Easy. Thus, if the thermosetting resin 130 oozes out from the gap between the inorganic fillers 127 or is convected to form the surface layer 128, the thermosetting resin 130 filling the gaps between the inorganic fillers 127 and the surface layer The thermosetting resin 130 constituting 128 is substantially the same material. Also, there are no connection surfaces or interfaces between them, or they cannot be found by analysis. A state in which the connection surface or interface cannot be detected by a normal analysis method is expressed as a oozing state.

  The content of the inorganic filler 127 in the surface layer 128 is desirably 40 Vol% or less. More preferably, it is 30 Vol% or less, More preferably, it is 20 Vol% or less. As described above, the content of the inorganic filler 127 in the surface layer 128 is reduced, and the thermosetting resin 130 is mainly used. Further, a concentration distribution gradient of the inorganic filler 127 is positively provided in the thickness direction of the heat transfer layer 102. For example, the content rate of the inorganic filler 127 is small on the surface layer side, and the content rate of the inorganic filler 127 is high on the inner layer side as 66 Vol% or more. In this way, it is possible to improve both the fixing property of the heat-generating component fixed to the surface layer 128 with a heat conductive adhesive and the heat conductivity of the heat transfer layer 102.

  In addition, the heat transfer layer 102 and the heat dissipation substrate 101 manufactured using the heat transfer layer 102 are both highly reliable and can support a pressure cooker test. Further, even when the influence of the pressure cooker occurs on the metal plate 107 side forming the heat dissipation substrate 101, the influence of the pressure cooker hardly occurs on the heat transfer layer 102 side. This is because the surface of the heat transfer layer 102 is protected by the surface layer 128, and the interface between the main part 129 and the metal plate 107 is well wetted by the thermosetting resin 130. The pressure cooker test is one of methods for testing moisture resistance of electronic parts sealed with resin, and is standardized by IEC 68-2-66. In addition, IEC means the International Electrotechnical Commission (International Electrotechnical Commission).

Example 4
Example 4 describes the uncured composition 116.

  Next, the monomer and the curing agent of the crystalline epoxy resin, which are members constituting the uncured composition 116, will be individually described using (General Formula 1) to (Formula 8).

(General Formula 1) and (Formula 3) to (Formula 8) show a structural formula showing an example of a crystalline epoxy monomer. In (General Formula 1), X is S (sulfur) or O (oxygen), CH ₂ (methylene group), or none (single bond or direct bond). R1, R2, R3, and R4 are CH ₃ , H, t-Bu (tertiary butyl group), and the like. R1 to R4 may be the same. That is, the compounds of (Formula 4) to (Formula 8) are specific examples of (General Formula 1). This monomer having an epoxy group is also referred to as a main agent. Note that there are epoxy monomers that have the structure of (General Formula 1) but are not crystalline due to steric hindrance due to the size of R1, R2, R3, and R4. For example, when X is a single bond and R1, R2, R3, and R4 are all CH ₃ , it becomes amorphous. Such an epoxy monomer does not correspond to the crystalline epoxy resin in the present invention.

  Crystalline epoxy monomers self-align with each other to develop crystallinity. For this self-arrangement, it is useful to utilize a mesogenic structure comprising a phenyl group-phenyl group as shown in (General Formula 1) and (Formula 4).

  In order to examine the crystallized state of the crystalline epoxy resin in the thermosetting resin 130 after curing, it is useful to cut out very thinly without destroying only the resin portion on the sample surface and evaluate it with a polarizing microscope or the like. Differential scanning calorimetry (DSC) is also useful. When DSC is used, the amount of the crystalline epoxy resin contained in the thermosetting resin 130 is measured by weight% (or wt%). However, it is useful to convert this data to volume% (or Vol%). This is because the molecular weight of the crystalline epoxy monomer used varies depending on the type. In order to extract a crystalline epoxy resin (crystalline epoxy monomer) in an uncured state, a separation method such as centrifugation is useful.

  (Formula 2) is a structural formula of a curing agent used for curing the crystalline epoxy resin. X is S (sulfur), O (oxygen) or a single bond. A polymer obtained by mixing the monomers of (General Formula 1) or (Formula 3) and the curing agent of (Formula 2) and polymerizing them may be called a crystalline epoxy resin.

  The ratio between the main agent and the curing agent is calculated from the epoxy equivalent. “Epoxy equivalent” means the weight (g number) of “epoxy resin” per 1 g equivalent of epoxy group contained in the epoxy resin. For example, in the case of an epoxy resin with a known molecular structure, it can be calculated by dividing the molecular weight of the epoxy resin by the number of epoxy groups contained in one molecule of the epoxy resin. If necessary, the “epoxy equivalent” can be determined by measurement using a hydrochloric acid-dioxane method or the like. It is useful to refer to JIS K7236 “Testing method for epoxy equivalent of epoxy resin”.

  A curing agent other than (Formula 2) may be used as the curing agent. In addition, as a specific crystalline epoxy monomer, what was shown, for example in (Formula 3)-(Formula 8) can be used. The melting points of the monomers shown in (Formula 3) to (Formula 8) are about 50 to 120 ° C., and the dissolution viscosity is also low. For example, the viscosity at 150 ° C. is 6 to 20 mPa · s. Therefore, it is easy to mix and disperse with the inorganic filler 127.

  Next, the preferable ratio of the crystalline epoxy monomer in the main agent will be described. The volume ratio of the crystalline epoxy monomer in the main agent is desirably in the range of 5 Vol% or more and 100 Vol% or less. More preferably, it is the range of 10 Vol% or more and 93 Vol% or less. By setting the ratio of the crystalline epoxy monomer to 5 Vol% or more, the surface 124 of the heat transfer layer 102 can be made smooth. On the other hand, if it is less than 5 Vol%, the effect of reducing the viscosity at the time of thermosetting may not be obtained. Further, when the ratio of the crystalline epoxy monomer exceeds 93 Vol%, the viscosity may be excessively lowered at the time of thermal curing, and there is a possibility that it may ooze or affect the shape retention as the heat transfer layer 102. Further, it is more desirable that the crystalline epoxy monomer is contained in an amount of 40 Vol% or more with respect to the whole main component. If it is this range, crystallinity will be easy to express.

  In addition, as above-mentioned, another additive can be added as needed. For example, release agents such as plasticizers, acid amides, esters, paraffins, stress relieving agents such as nitrile rubber and butadiene rubber, organic flame retardants such as phosphate esters and melamine, antimony pentoxide, molybdenum oxide, boric acid Inorganic flame retardants such as zinc, tin oxide, barium metaborate, aluminum hydroxide, magnesium hydroxide, calcium aluminate, silane coupling agents, titanate coupling agents, coupling agents such as aluminum coupling agents, Colorants such as dyes and pigments, inorganic fibers such as glass fibers, boron fibers, silicon carbide fibers, alumina fibers and silica alumina fibers, organic fibers such as aramid fibers, polyester fibers and cellulose fibers, oxidation stabilizers, light Stabilizer, moisture resistance improver, thixotropy imparting agent, diluent, antifoaming agent, various other It fat, tackifiers, antistatic agents, lubricants, also be incorporated an ultraviolet absorber and the like.

  In addition, the impact resistance of a crystalline epoxy resin can be improved by adding a thermoplastic resin. The addition of the thermoplastic resin is limited due to the structure of the crystalline epoxy resin. In order to realize both thermal conductivity and impact resistance, an amount of thermoplastic resin of 0.3 volume part or more and 5.0 volume part or less is desirable with respect to 100 volume parts of the total of the main agent and the curing agent. . An acrylic resin or the like can be used as the thermoplastic resin, and impact resistance is improved particularly when a core-shell type acrylic resin is blended. Here, the core shell is pearl-like fine particles in which a core layer made of a thermoplastic resin such as an acrylic resin is covered with a shell layer made of another resin (for example, a glassy polymer or an epoxy resin). A multi-layer structure may be used depending on the application. Dispersibility can be enhanced by using particles having a primary particle size in the range of 0.05 to 1.00 μm (preferably 0.1 to 0.5 μm). In addition, even if it comprises the secondary aggregate, it can disperse | distribute easily by using a biaxial kneader etc. This is due to the core-shell structure.

  The core is made of a thermoplastic resin having a low Tg (desirably 50 ° C. or lower, more preferably 0 ° C. or lower) made of, for example, a homopolymer or copolymer of an acrylic monomer. This is because it acts as a stress concentration point in the uncured composition 116 or in the heat transfer layer 102 to improve impact resistance and relieve stress.

  When a core-shell type polymer is used, a rubber-like resin may be used as a kind of thermoplastic resin for the core portion. In the fine particle state with a primary particle size of 1.00 μm or less, this works substantially in the same way regardless of whether it is a crosslinked rubber-like resin or a non-crosslinked thermoplastic resin. This is to improve or relieve stress. For this reason, the thermoplastic resin may be a rubber-like resin (crosslinked resin).

  Note that the Tg of the shell is set high. Desirably, it is 100 ° C. or higher, more preferably 150 ° C. or higher. This is for the purpose of increasing the compatibility between the particles, the compatibility with the epoxy resin (including the crystalline epoxy resin), the curing agent, and the dispersibility.

  The thermoplastic resin such as an acrylic resin added to the uncured composition 116 is desirably added in a state where it can be extracted with acetone or the like. This is because, in the uncured composition 116, such a resin that improves impact resistance is dispersed (or scattered or dispersed) in, for example, a sea-island structure, thereby improving impact resistance. And by setting it as such a sea island structure, it can extract with acetone etc., and it is confirmed that the thermoplastic resin is disperse | distributed by the sea island structure. In the extraction, the detection accuracy of the thermoplastic resin or the like can be increased by crushing the sample.

  When a thermoplastic resin such as an acrylic resin is present in the uncured composition 116 in a state where it cannot be extracted, it is difficult to extract it using acetone or the like. For example, it is a state in which it has reacted with an epoxy resin, dissolved in an epoxy resin, or dispersed in a molecular state in an epoxy resin. When the thermoplastic resin is dissolved in the uncured composition 116 in such a state, the effect of improving the impact resistance cannot be obtained. This is because an interface (or stress concentration) does not occur between the epoxy resin and the thermoplastic resin.

  Since the amount of extraction depends on the state of the sample for extraction (such as the pulverized state), the amount of thermoplastic resin in the total resin in the sample is, for example, 0.1 Vol% or more, preferably 1 Vol% or more. . This is because it may be difficult to determine the absolute value of the thermoplastic resin by extraction.

  Next, a method for preparing the uncured composition 116 will be described. First, a crystalline epoxy monomer, an amorphous epoxy monomer, a curing agent, a thermoplastic resin, and an inorganic filler 127 are weighed and mixed according to the blending ratio.

  As the kneading apparatus, a commercially available heating kneading apparatus or a biaxial kneader is used. For example, a planetary mixer or kneader or a laboratory plast mill manufactured by Toyo Seiki Seisakusho Co., Ltd. is used. In the kneading apparatus, a stirring blade such as a Σ type or a Z type may be attached as a stirring blade. Moreover, shapes other than a blade | wing can also be used. Further, it is desirable to use a kneading apparatus that can be heated to a predetermined temperature with a heater or the like. By heating the kneading apparatus, the material that is in a solid state at room temperature can be dissolved (or liquefied), so that the affinity with other members can be increased.

  By setting the temperature of the monomer within the range of the melting temperature or higher (for example, 50 to 200 ° C.), the melt viscosity of the curable resin containing the thermoplastic resin is kept low, and the inorganic filler 127 is uniformly dispersed in the resin. Can be possible.

  It is desirable that these members have a minimum temperature that can maintain a liquefied state, for example, 100 ° C. or higher. And it is easy to make it uniform by kneading above this minimum liquefaction temperature.

  In addition, when adding a hardening | curing agent, the thermosetting reaction after addition or a time-dependent change can also be suppressed by lowering | hanging the temperature of a kneading apparatus or the material set in it. In this way, an uncured composition 116 is produced.

  In addition, in the ratio in the inorganic filler 127 and curable resin total, the inorganic filler 127 has the desirable range of 66 Vol% or more and 90 Vol% or less. That is, the curable resin is desirably in the range of less than 5 Vol% and less than 34 Vol%. Furthermore, the inorganic filler 127 is desirably 70 vol% or more and 90 vol% or less, and the resin component is desirably in the range of 12 vol% or more and 30 vol% or less. Here, the curable resin means the total of the main agent containing the crystalline epoxy monomer and the curing agent.

  When the ratio of the inorganic filler 127 is less than 66 Vol%, the thermal conductivity of the heat transfer layer 102 formed by curing the uncured composition 116 may be reduced. Moreover, when the ratio of the inorganic filler 127 becomes larger than 96 Vol%, the moldability of the thermosetting resin 130 may be affected.

  As described above, the uncured composition 116 can be composed of the curable resin, the thermoplastic resin added to reduce the fragility, and the inorganic filler 127. The curable resin includes a main agent and a curing agent. The main agent contains 5 vol% or more and 100 vol% or less of a crystalline epoxy monomer. The thermoplastic resin is added in a range of 0.3 to 5.0 parts by volume with respect to the curable resin. The inorganic filler 127 is contained in the whole uncured composition 116 in the range of 66 Vol% or more and 90 Vol% or less. By using the uncured composition 116 having such a composition, it is possible to improve the heat dissipation and impact resistance of the heat transfer layer 102.

  Further, the thermosetting resin 130 cured by reacting the main agent and the curing agent has crystallinity, whereby the thermal conductivity can be increased.

Moreover, as shown in (Formula 2), it has reactivity with the epoxy monomer contained in the main agent because the curing agent has two OH groups or NH ₂ groups. In addition, you may have two OH groups and _two NH2 groups per molecule. Furthermore, crystallization can be promoted. Therefore, the thermal conductivity of the heat transfer layer 102 can be increased, and the heat dissipation property of the heat dissipation substrate 101 can be increased.

(Example 5)
Further, a method of adding an inorganic flame retardant as required is known. However, when an organic flame retardant is blended with structurally different resins, the amount of addition is limited because crystallization is inhibited. Therefore, flame retardancy cannot be achieved while maintaining crystallinity. In addition, in the case of an inorganic flame retardant, the thermal conductivity of the flame retardant is low. Therefore, if the amount is increased, the thermal conductivity decreases.

  Here, the problem that the crystalline resin is hard and brittle and easily burns is solved, the heat dissipation substrate 101 using the crystalline resin is compatible with high thermal conductivity and durability, and the uncured composition 116 is flame retardant. A method of providing the will be described.

  Specifically, a flame retardant epoxy resin is added to the resin component, and a flame retardant auxiliary filler is added to the inorganic filler 127. More specifically, a flame retardant of 3 to 15 parts by volume in 100 parts by volume of a curable resin composed of a main component containing 5 to 100% by volume of a crystalline epoxy monomer and a curing agent. Add a functional epoxy monomer. And the inorganic filler 127 and the flame-retardant adjuvant filler are mix | blended with the mixture of curable resin and a flame-retardant epoxy monomer, and the uncured composition 116 is prepared. The volume ratio of the inorganic filler 127 in the uncured composition 116 is 66 Vol% or more and 90 Vol% or less. The flame retardant auxiliary filler is added in an amount of 0.6 to 3.5 parts by volume with respect to 100 parts by volume of the curable resin.

  In this way, the flame retardant epoxy and the flame retardant auxiliary filler are added at a ratio that does not cause crystallization to be imparted, and flame retardancy is imparted to the heat transfer layer 102 while maintaining high thermal conductivity. Can be improved.

  Further, the addition ratio of the flame retardant auxiliary filler to the curable resin is preferably 0.6 parts by volume or more and 3.5 parts by volume or less. Since the thermal conductivity of the flame retardant aid filler is low, when it is more than 3.5 parts by volume, the thermal conductivity is drastically reduced. Moreover, if it is less than 0.6 volume part, a flame retardance cannot be provided. As the flame retardant auxiliary filler, a hydroxyl compound, a boron compound, an antimony compound, a molybdenum compound, a metal oxide salt, or the like can be used. In particular, antimony trioxide (Antimony Trioxide) is desirable because of its high flame retardancy effect. The particle size of antimony trioxide is preferably small, and is preferably 0.1 to 20 μm. More preferably, it is 0.3-10.0 micrometers. If it is smaller than 0.1 μm, it is expensive and may be difficult to disperse in the uncured composition 116. If the particle size exceeds 20 μm, the surface 124 of the heat transfer layer 102 may become rough, which may affect the adhesion with the lead frame 103.

  The particle size of antimony trioxide may affect the hygroscopicity. Therefore, when hygroscopicity becomes a problem, the smaller the particle size of antimony trioxide, the easier it is to absorb moisture. Therefore, when hygroscopicity is a problem, the particle size of antimony trioxide is desirably 0.1 μm or more. This is because fine antimony trioxide of less than 0.1 μm is highly hygroscopic.

  The particle size of antimony trioxide is matched with the particle size of the inorganic filler 127, and the particle size close to the particle size of the inorganic filler 127 can be easily dispersed in the epoxy resin or the like. In this case, when a plurality of inorganic fillers 127 having different particle diameters are used in combination, it is desirable to align the particle diameter with either a small powder or a large powder. By carrying out like this, the dispersibility of these powders can be improved. The large particle size is, for example, 1.0 to 10.0 μm, and the small particle size is, for example, 0.2 to 0.6 μm.

  In the uncured composition 116, the content of the inorganic filler 127 is preferably in the range of 66 Vol% to 90 Vol%. That is, the content of the thermosetting resin 130 made of a curable resin, a flame retardant epoxy monomer, a dispersant, or the like is in a range of more than 10 Vol% and less than 34 Vol%. When the ratio of the inorganic filler 127 is less than 66 Vol%, the thermal conductivity of the heat transfer layer 102 formed by curing the uncured composition 116 may decrease. It also affects flame retardancy. Moreover, when the ratio of the inorganic filler 127 becomes larger than 90 Vol% (and further exceeds 96 Vol%), the moldability and surface property of the heat transfer layer 102 may be lowered, and the withstand voltage characteristic may be lowered.

  By preparing the uncured composition 116 with the above composition, high thermal conductivity and flame retardancy can be realized. By using brominated epoxy as the flame retardant epoxy monomer, it is possible to achieve high thermal conductivity and flame resistance in the uncured composition 116, the heat transfer layer 102, the heat dissipation substrate 101 using these members, or the like. . The brominated epoxy resin is excellent in compatibility with other resins by selecting the molecular weight and can reduce bleed out. Bleed-out is a phenomenon in which various additives in the uncured composition 116 are aggregated due to change over time and pulverized on the surface 124, or a phenomenon such as “bleed out” of a low molecular resin or solvent generated around the cured product. It is. In particular, by reducing bleed-out, heat dissipation and flame retardancy can be improved.

In addition, since the curing agent has two OH groups and / or NH ₂ groups, the reactivity and curability of the uncured composition 116 or matching with a flame retardant is improved, and high thermal conductivity is achieved. It is possible to achieve both heat resistance and flame retardancy.

(Example 6)
Next, the coexistence of particularly high thermal conductivity, high flame retardancy, and impact resistance will be described.

  Conventional crystalline resins are limited in use for the heat dissipation substrate 101 because of their unique characteristics (hard and brittle, easily chipped, easily broken, easily burned, etc.). As a method of improving the strength, there is a method of blending a curing agent that easily takes a network structure, etc., but in the case of a crystalline epoxy resin, by creating a three-dimensional bond, the crystallinity is inhibited and high thermal conductivity is obtained. In many cases, it cannot be obtained.

  Further, the crystalline resin has a weak adhesion to the metal plate 107. Alternatively, stress is likely to be concentrated inside the crystallized resin or at the interface between the crystallized resin and the metal plate 107, and the interface between the crystallized resin and the metal plate 107 may be peeled off due to this stress. Therefore, when the heat dissipation substrate 101 is manufactured, there may be a problem in impact resistance such as a drop test.

  The addition of flame retardancy is as described above. Therefore, in order to solve the problem that the crystalline resin is hard and brittle and easily burnable, to achieve both high thermal conductivity and durability, and to impart flame retardancy, the above-described thermoplastic resin and flame retardant epoxy resin are used. And flame retardant aid filler.

  Specifically, it is difficult for 3 to 12 parts by volume with respect to 100 parts by volume of the curable resin, which is the total of the main agent containing 5 to 100% by volume of the crystalline epoxy monomer and the curing agent. Add flammable epoxy monomer. And 0.3 volume part or more and 2.5 volume part or less thermoplastic resin is added. Furthermore, the inorganic filler 127 and the flame retardant auxiliary filler are blended with these mixtures to prepare an uncured composition 116. The volume ratio of the inorganic filler 127 in the uncured composition 116 is 66 Vol% or more and 90 Vol% or less. The flame retardant aid filler is added in an amount of 0.6 to 2.5 parts by volume with respect to 100 parts by volume of the curable resin.

  In this way, flame retardancy is imparted by adding the flame retardant epoxy monomer and the flame retardant auxiliary filler at a rate that does not cause crystallization to break. And a thermoplastic resin is mix | blended in the ratio which crystallization does not collapse. As a result, it is possible to provide the required strength (hardness to crack), high thermal conductivity, and flame resistance required for a component mounting board. As a result, it is possible to achieve both high heat dissipation, robustness and flame retardancy.

  In addition, since the stress is less likely to concentrate on the uncured composition 116 (or the heat transfer layer 102 formed by curing the uncured composition 116), the adhesion with the metal plate 107 is also improved.

(Example 7)
FIG. 12A is a photomicrograph of the surface 124 of the heat transfer layer 102 according to this example, and FIG. 12B is a schematic diagram thereof. FIG. 13A is a perspective view showing the surface 124 of the heat transfer layer 102 three-dimensionally, and FIG. 13B is a schematic view thereof.

  In the heat transfer layer 102 in FIGS. 12 to 13, the content of the inorganic filler 127 was 70 Vol%. The remaining 30% by volume was mainly composed of the thermosetting resin 130, and a small amount of a dispersant, a flame retardant, a colorant and the like were added thereto as necessary.

  As shown in FIGS. 12A and 12B, when observed at a magnification of 150 times, the surface 124 of the heat transfer layer 102 (that is, the surface 124 of the surface layer 128) is smooth and is used for focusing the microscope. The attached dirt 131 remains on the surface 124 in a slightly black spot shape.

  FIG. 12A shows evaluation results of about 100 μm × about 75 μm in the planar direction and 1.2 μm in the thickness direction using a commercially available laser microscope (VK-9510 manufactured by Keyence). It can be seen that the highest thickness of the surface 124 of the sample is about 0.8 μm. Thus, the surface of this sample is extremely smooth. Further, the arithmetic average roughness Ra of the surface of this sample was 0.5 μm, the maximum height Ry was 0.90 μm, and Rz was 0.53 μm. The 20 degree gloss was 95 to 105 (n = 5), and the 60 degree gloss was 100 (n = 5).

  Next, the mechanism by which the free surface of the sample shown in FIGS. 12 and 13 becomes smooth will be described.

  The crystalline epoxy resin added to the uncured composition 116 is solid at a temperature lower than the crystallization temperature and becomes liquid at a temperature higher than the crystallization temperature. That is, the crystallized epoxy resin exists as a stable solid in the crystalline state, but as the melting point is reached, the crystalline state dissolves quickly and changes to a very low viscosity liquid. Therefore, in the case of the invention product, the thermosetting resin 130 itself changes to a very low viscosity liquid by exceeding the crystallization temperature during the heating for thermosetting. As a result, the thermosetting resin 130 can sufficiently wet the surface of the filler. Further, the thermosetting resin 130 having a reduced viscosity penetrates into a narrow gap between the fillers. As a result, in the invention product, bubbles (or voids) at the interface between the filler and the thermosetting resin 130, between the filler and the filler, or inside the thermosetting resin 130 can be eliminated.

  Next, as a conventional example, the content of the inorganic filler 127 was set to 70 Vol% as in the samples shown in FIGS. 12 (A) to 13 (B). The remainder was mainly composed of a curable resin, and a dispersant, a flame retardant, a colorant and the like were added as necessary.

  The curable resin used for the conventional product was formed using an amorphous epoxy monomer as a main agent and a curing agent.

  The results evaluated in the same manner as in FIGS. 12 (A) to 13 (B) are shown in FIGS. 14 (A) to 15 (B). 14A is a photomicrograph of the surface 124 of the conventional heat transfer layer 102, and FIG. 14B is a schematic diagram thereof. FIG. 15A is a perspective view showing the surface 124 of the heat transfer layer 102 three-dimensionally, and FIG. 15B is a schematic view thereof. Reference numeral 132 denotes a protrusion, 133 denotes a crack, and the crack 133 includes a void and the like.

  The surface 124 of the conventional heat transfer layer 102 has numerous protrusions 132 and voids. 14A to 15B show that the protrusion 132 has a height of about 20 μm and the crack 133 has a depth of about 10 μm. The reason why the protrusion 132 and the crack 133 are generated in this way is that the convection described with reference to FIGS. 10A and 10B hardly occurs in the conventional product. The arithmetic average roughness Ra of the surface 124 of this sample was 0.85 μm (n = 5), the maximum height Ry was 21.50 μm (n = 5), and Rz was 21.04 μm (n = 5). The 20-degree gloss was 0 (n = 5), and the 60-degree gloss was 0 (n = 5).

  Next, the mechanism in which the free surface of the conventional product has irregularities as shown in FIG. In the re-prototype conventional product, an amorphous epoxy resin component is used as the thermosetting resin, and no crystalline epoxy resin is added.

  In general, an amorphous epoxy resin is obtained as an amorphous solid having a molecular weight distribution. The resinous material obtained as an amorphous solid state is not clear at the boundary between the liquid and the solid, and gradually softens as the temperature rises, and becomes a high-viscosity liquid. It changes to a liquid with viscosity.

  From the above, in the case of conventional products, during the heating for thermosetting, the viscosity of the thermosetting resin itself is difficult to decrease, the convection phenomenon of the thermosetting resin hardly occurs, and voids are formed inside the surface. Is likely to generate protrusions and recesses.

  Next, the results of comparison of the electrical characteristics of the heat transfer layer 102 according to this example and the heat transfer layer 102 of the conventional example will be described with reference to FIG.

  FIG. 16 is a graph showing the temperature characteristics of the leakage current of the heat transfer layer 102 according to the present embodiment and the heat transfer layer 102 of the conventional example as an example of electric characteristics. That is, in this evaluation, the heat resistance of the heat dissipation substrate 101 is evaluated from the viewpoint of leakage current.

  Here, the measurement method will be described. First, a commercially available hot plate for heating is prepared. Then, a heat-resistant ceramic substrate for ensuring insulation is placed on the hot plate, and a metal plate 107 serving as an electrode is placed on the ceramic substrate. For example, an alumina substrate having a thickness of 1 to 2 mm is used as the ceramic substrate. On this metal plate 107, a heat dissipation substrate 101 including a heat transfer layer 102 having a thickness of 0.4 mm and a square of about 5 cm is placed, and a measurement terminal extending from the leakage current measuring device is mounted thereon. Then, a voltage of 4.3 kV is applied between the metal plate and the measurement terminal, and the leakage current at that time is measured. The sample temperature is controlled by a hot plate. The surface temperature is measured by attaching a thermocouple to the surface of the sample.

  In FIG. 16, the “•” plot indicates the measurement result of the conventional product, and the “♦” plot indicates the measurement result of the sample according to this example. As apparent from FIG. 16, the leakage current of the conventional product is small at 60 ° C. or less, but when it exceeds 60 ° C., the leakage current increases as the temperature rises. And if it exceeds 110 degreeC, a leakage current will become still larger. When the temperature exceeds 160 ° C., the leakage current suddenly increases, and the insulation is broken at 190 ° C. Thus, in the conventional product, the leakage current increases as the temperature rises.

  The conventional product has innumerable irregularities on the surface of the heat transfer layer 3 as shown in FIG. Innumerable gaps 8 and voids 5 remain at the interface between the heat transfer layer 3 and the lead frame 6 or the substrate 7. Further, the inorganic filler 1 constituting the heat transfer layer 3 is not sufficiently covered with the thermosetting resin 2. As a result, in the conventional example, the inorganic fillers 1 are merely in physical contact (or point contact) in the thickness direction. As a result, it is considered that the leakage current suddenly increased with the temperature rise, and finally dielectric breakdown occurred.

  On the other hand, the leakage current of the sample according to this example is extremely small up to around 190 ° C. The leakage current gradually increases with increasing temperature from around 200 ° C., and the increase in leakage current increases from around 220 ° C. Nevertheless, dielectric breakdown does not occur even at 240 ° C. Thus, in the sample according to this example, the leakage current hardly increases even when the temperature rises. That is, the sample according to this example has excellent electrical characteristics, which is due to the reason described in FIG.

  Next, a more detailed description will be given with reference to FIG. FIG. 17 is a cross-sectional view for explaining how the thermosetting resin 130 convects on the side and bottom surfaces of the lead frame 103 and the wiring board 104 embedded in the uncured composition 116. In FIG. 17, an arrow 117 indicates a state in which the thermosetting resin 130 that has been heated to reduce the viscosity convects the gaps of the inorganic filler 127.

  As shown in FIG. 17, even if the side surface in which the wiring substrate 104 is embedded is a rough surface having irregularities, the thermosetting resin 130 with reduced viscosity wets the fine irregularities on the side surface, so that it remains on the irregularities. The air can be removed, the adhesion between the wiring substrate 104 and the uncured composition 116 is increased, and the electrical insulation and adhesion are enhanced. By roughening the side surface (so-called break surface or cut surface) of the wiring substrate 104, the adhesion area with the uncured composition 116 can be increased. As a result, the adhesion strength between the heat transfer layer 102 and the wiring substrate 104 increases, and the leakage current also decreases.

  As described above, the lead frame 103 and the wiring substrate 104 are both embedded in the heat transfer layer 102 in a state where parts of the lead frame 103 and the wiring board 104 are in contact with each other via the solder portion 108, thereby forming the lead frame 103 and the heat transfer layer. The contact area of 102 can be increased, and heat dissipation can be improved. Also, by embedding them, the thickness of the lead frame 103 protruding from the surface 124 of the heat transfer layer 102 can be reduced. Thus, even when a lead frame 103 having a thickness of 0.3 mm, for example, is used as the lead frame 103, the protrusion amount (or level difference) of the lead frame 103 from the heat transfer layer 102 is 50 μm or less by embedding in the heat transfer layer 102. Keep it down. Desirably, it is suppressed to 20 μm or less, and further to 10 μm or less. By suppressing the protruding amount in this way, it becomes easy to form a solder resist (not shown) formed on the lead frame 103, the thickness of the solder resist can be made thin and uniform, and the heat dissipation performance of the heat dissipation substrate 101 is improved. Rise.

  By cooling the heat generating component in this way, the efficiency of the heat generating component can be improved, and the life and reliability can be increased. For example, in the case of an LED, the amount of light increases. Alternatively, the efficiency of the power supply circuit is improved. Also, miniaturization is possible.

  Further, by embedding a part or more of the lead frame 103 in the heat transfer layer 102, the connection strength between the lead frame 103 and the heat transfer layer 102 is increased. In addition, the tensile strength of the lead frame 103 can be increased. Regarding the method for measuring the tensile strength, a method for evaluating the tensile strength of the copper foil in the printed wiring board 104 may be referred to.

  A part of the lead frame 103 constituting the heat dissipation substrate 101 can be bent as shown in FIG. 1 and connected to a wiring substrate 104 (not shown) prepared as required. The lead wire can be wound and soldered, or a crimp terminal can be used.

  In addition, the heat dissipation board 101 created in the example hardly generates warpage or waviness due to the thermosetting of the uncured composition 116. Warpage and undulation are caused by, for example, a difference in thermal expansion coefficient between the metal plate 107 and the uncured composition 116. However, for example, according to experiments by the inventors, it was 50 μm or less at 100 mm square. This is considered to be due to the reduction of internal stress and high density filling of the inorganic filler 127 by the void disappearance mechanism described above.

  Note that a surface layer 128 made of a thermosetting resin 130 that has oozed out from the gap between the inorganic fillers 127 is formed on the surface 124 of the cured heat transfer layer 102 provided on the heat dissipation substrate 101. In addition to the surface 124, the surface layer 128 is useful to be present at the interface between the heat transfer layer 102 and the lead frame 103, which is the wiring fixed thereon, and the heat transfer layer 102. Further, when the surface-roughened copper foil is used as the lead frame 103, the surface layer 128 is also present at the interface between the roughened surface 124 and the heat transfer layer 102.

  Also, when a roughened copper foil or the like is fixed as a wiring on the uncured composition 116 and the uncured composition 116 is cured and integrated with the copper foil, the copper foil and the heat transfer layer after curing It is desirable that a surface layer 128 is formed at the interface with 102. When a part of the copper foil is removed by etching or the like, the surface layer 128 that has fixed the copper foil remains on the surface 124 of the heat transfer layer 102. On the surface 124 of the surface layer 128 exposed on the surface 124 in this way, the shape of the surface 124 of the copper foil or the roughened surface provided on the copper foil surface 124 is transferred as it is as a replica. Even in this case, the surface layer 128 increases the reliability of the heat transfer layer 102.

(Example 8)
Next, various members used for the heat dissipation substrate 101 will be described using Example 8 of the present invention.

  The wiring made of the lead frame 103 may be made of copper foil. Depending on the application of the heat dissipation substrate 101, the copper foil and the lead frame 103 can be used properly. When the thickness of the wiring (for example, the lead frame 103) embedded in the heat transfer layer 102 requires a range of 0.002 to 0.10 mm, a copper foil is used, and a range of 0.10 to 1.00 mm is necessary The lead frame 103 may be used. The lead frame 103 is preferably made mainly of copper. For example, what is called tough pitch copper or oxygen-free copper can be used. By mainly using copper, both high heat dissipation and low resistance can be achieved. Further, by burying a part or more of the lead frame 103 in the heat transfer layer 102, a step (thickness step) due to the lead frame 103 in the heat dissipation substrate 101 can be reduced.

Example 9
Hereinafter, as Example 9, an uncured composition 116 used for manufacturing the heat transfer layer 102 will be described.

  As Example 9, a flame retardant epoxy resin of 3 Vol% or more and 12 Vol% or less with respect to the total of an epoxy resin containing 40 vol% or more of a crystalline epoxy resin, a curing agent, and the epoxy resin and the curing agent; 0.3 vol% or more and 2.5 vol% or less thermoplastic resin with respect to the epoxy resin and the curing agent, and 70 vol% or more and 88 vol% or less inorganic with respect to the epoxy resin, the curing agent and the flame retardant epoxy. The result of having experimented about the uncured composition 116 which consists of a filler, the said epoxy resin, and the flame retardant auxiliary | assistant filler of 0.6 Vol% or more and 2.5 Vol% or less with respect to the said hardening | curing agent is shown. .

  The uncured composition 116 or the heat transfer layer 102 includes at least a thermosetting resin containing a crystalline epoxy resin and an inorganic filler, and the content of the inorganic filler 127 is the uncured composition 116 or the heat transfer. It goes without saying that 66 Vol% or more and 90 Vol% or less of the layer 102 is desirable.

  First, the results of evaluating the impact resistance will be described using [Table 1].

  “YL6120H” made by Japan Epoxy Resin as crystalline epoxy-containing resin, “YX4000H” as epoxy resin, 4,4′-diaminobiphenyl, 4,4′-dihydroxybiphenyl ether as curing agent, Japan as flame retardant epoxy resin Brominated epoxy “BREN-S” manufactured by Kayaku Co., Ltd., ZEON Kasei core-shell acrylate copolymer “F351” as a thermoplastic resin, alumina as an inorganic filler, and antimony trioxide as a flame retardant auxiliary filler were prepared.

  “P-200” imidazole (amine adduct) made by Japan Epoxy Resin as a curing accelerator is mixed and stirred, heated to a temperature at which the epoxy resin and curing agent melt, and mixed using a biaxial kneader. did. After mixing, it was molded into a shape according to various tests, and this was used as the uncured composition 116 in FIG.

  6A, an uncured composition 116, a commercially available multilayer glass epoxy resin substrate (thickness 0.6 mm), and a lead frame (thickness 0.5 mm) are set on the metal plate 107. Then, as shown in FIG. 6B, heat curing was performed under the conditions of 1170 ° C. × 5 Hour.

  A sample for a drop test was produced in a shape of 100 mm × 100 mm × 1.6 mm, and freely dropped onto the concrete from a height of 1.5 m. The result was judged by whether the sample was cracked or cracked. The blending conditions of the sample were changed by changing the ratio of the crystalline epoxy resin in the epoxy resin, the filler amount, and the ratio of the thermoplastic resin.

  As the proportions of the crystalline epoxy resin and filler increase, they tend to break, and the impact resistance is improved by adding a thermoplastic resin (decreasing in thermal conductivity). The results are shown in [Table 1].

  Impact resistance can be improved by adding 0.3% or more of thermoplastic resin.

  The thermoplastic resin such as an acrylic resin added to the uncured composition 116 is desirably added in a state where it can be extracted with acetone or the like. This is to improve the impact resistance by dispersing (or interspersing or dispersing) such a resin that improves the impact resistance in the heat conductive resin, for example, in a sea-island structure. And by setting it as such a sea island structure, it can extract with acetone etc., and it is confirmed that the thermoplastic resin is disperse | distributed by the sea island structure. In the extraction, the detection accuracy of the thermoplastic resin or the like can be increased by crushing the sample.

  In the case where a thermoplastic resin such as an acrylic resin is present in the uncured composition 116 in an unextractable state (for example, reacted with the epoxy resin, dissolved in the epoxy resin, or in the epoxy resin). In the molecular state) is difficult to extract using acetone or the like. And when a thermoplastic resin melt | dissolves in the uncured composition 116 in such a state, the improvement effect of impact resistance is not acquired. This is because an interface (or stress concentration) does not occur between the epoxy resin and the thermoplastic resin.

  Since the amount of extraction depends on the state of the sample for extraction (such as the pulverized state), the trace of the ratio of the thermoplastic resin to the total resin in the sample (for example, 0.1% or more, preferably 1% or more) To do. This is because it may be difficult to determine the absolute value of the thermoplastic resin by extraction.

  Next, the results of evaluating the flame retardancy will be described using [Table 2].

  For flame retardancy test, sample 1: 124 ± 5mm × 13 ± 0.5mm × 1.6mm, sample 2: 124 ± 5mm × 13 ± 0.5mm × 0.4mm, manufactured according to UL94 standard Based on the flame retardancy test.

  After 10 seconds of flame contact, the flame was extinguished and the combustion time (t1) until extinction was measured. After flame extinguishing, after 10 sec. Flame contact, the flame was extinguished and the combustion time (t2) until extinction was measured.

  The sample was mixed under the condition that there were many flammable crystalline epoxy resins and few fillers, and the ratio of the flame retardant epoxy and the flame retardant auxiliary filler was changed.

  Addition of flame retardant epoxy and flame retardant aid filler improves flame retardancy (decreases thermal conductivity). The results are shown in [Table 2].

  Flame retardancy can be added when 3 vol% or more flame retardant epoxy and 0.6 vol% or more flame retardant auxiliary filler are mixed.

  Next, an example of the result of evaluating the flame retardancy in a state where impact resistance is improved by adding a thermoplastic resin will be described using [Table 3].

  A sample for measuring thermal conductivity was prepared in a disk shape of φ1 / 2 inch and t1 mm, and measurement was performed using a xenon laser flash manufactured by Bruker AXS.

  The sample was mixed under the condition that there were many crystalline epoxy resins having a large thermal conductivity and a large amount of filler, and the ratio of the thermoplastic resin, the flame retardant epoxy, and the flame retardant auxiliary filler was changed. Table 3 shows the results of the thermoplastic resin.

  The thermal conductivity is reduced by adding a thermoplastic resin, and when 3% or more is added, the epoxy resin cannot take crystallinity, and the thermal conductivity is extremely lowered.

  Next, the effect of the flame retardant epoxy resin will be described using [Table 4].

  [Table 4] shows the results of the flame-retardant epoxy resin.

  Although 3% or more flame retardant epoxy resin is necessary to add flame retardancy, the addition of flame retardant epoxy resin decreases the thermal conductivity, and when 15% or more is added, the epoxy resin is crystalline. Can not be taken, and the thermal conductivity is extremely reduced.

  Next, the effect of the flame retardant aid filler will be described using [Table 5].

  [Table 5] shows the results of the flame retardant aid filler.

  A flame retardant aid filler of 0.6% or more is necessary to add flame retardancy, but the addition of the flame retardant aid filler reduces the thermal conductivity (the amount of filler does not increase drastically). When 3% or more is added, the epoxy resin cannot take crystallinity, and the thermal conductivity is extremely lowered.

(Example 10)
The results of experiments on the optimization of the content of the inorganic filler contained in the heat transfer layer 102 described in Example 1 will be described using Example 10.

  FIG. 18 is a diagram for explaining an example of an experimental result regarding the relationship between the volume fraction of the inorganic filler in the heat transfer layer and the thermal conductivity. In FIG. 18, the horizontal axis represents the volume fraction of inorganic filler contained in the heat transfer layer 102 (unit: Vol%), and the vertical axis represents the thermal conductivity of the heat transfer layer (unit: W / m · K). As shown in FIG. 18, the thermal conductivity increases as the volume fraction of the inorganic filler contained in the heat transfer layer 102 increases.

  In [Table 6], the volume content (Vol%) of the inorganic filler contained in the heat transfer layer 102 described in FIG. 18, Example 1 or the like is changed, and the thermal conductivity obtained at that time (unit is W / m · K) and withstand voltage characteristics (4.2 kV applied) are shown as an example of evaluation results.

  As shown in [Table 6], when the content of the inorganic filler 127 contained in the heat transfer layer 102 is 90 Vol% or less, a predetermined withstand voltage characteristic is obtained, but when the content is 96 Vol% or more, it is necessary. Withstand voltage characteristics cannot be obtained. Moreover, if the content rate of the inorganic filler 127 is less than 66 Vol%, the heat conductivity requested | required as a heat sink will not be obtained.

  From the above, as shown in the comprehensive evaluation results in [Table 6], it is found that the content of the inorganic filler 127 included in the heat transfer layer 102 is preferably 66 Vol% or more and 90 Vol% or less as the heat dissipation substrate 101.

  In addition, it is desirable to evaluate the content rate of the inorganic filler 127 by Vol%. This is to suppress the influence of the specific gravity of the inorganic filler 127. In order to increase the content of the inorganic filler 127, it is possible to mix inorganic fillers having different average particle diameters (for example, an average particle diameter of 10 μm and an average particle diameter of 1 μm), or to devise the shape of the inorganic filler. It is useful to use an optimization method.

(Example 11)
In Example 11, an example of connection between the lead frame 103 and the wiring board 104 via the solder portion 108 is used with reference to FIGS.

  FIGS. 19A and 19B are perspective views of a chip-like wiring board having wiring portions that are insulated from each other vertically and horizontally. By using the chip-like wiring board 104 as shown in FIG. 19 as a kind of jumper chip, a plurality of wirings can be crossed as indicated by an arrow 117.

  20A and 20B are a perspective sectional view and a partially enlarged sectional view for explaining a state in which a chip-like wiring board is inserted between a plurality of lead frames, respectively.

  As shown in FIG. 20A, for example, an alumina substrate provided with side electrodes 111 (side electrodes are not shown) on both side surfaces is sandwiched between a plurality of adjacent lead frames 103.

  As shown in FIG. 20B, by inserting a wiring substrate 104 such as an alumina substrate between adjacent lead frames 103a, 103b, 103c, the lead frames 103a, 103b, 103c are insulated from each other. , Heat spread (thermal diffusion).

  As described above, the heat transfer layer and the manufacturing method thereof according to the present invention can reduce the size, cost, and reliability of equipment that requires heat dissipation. Therefore, it can contribute to miniaturization and high efficiency of various devices such as a DC-DC converter for vehicles and a light emitting diode for illumination.

DESCRIPTION OF SYMBOLS 101 Heat dissipation board 102 Heat transfer layer 103 Lead frame 104 Wiring board 105 Wiring part 106 Insulation part 107 Metal plate 108 Solder part 109 Connection part 110 Bonding wire 111 Side electrode 112 Through hole 114 Mold 115 Protective film 116 Uncured composition 117 Arrow DESCRIPTION OF SYMBOLS 118 Molding device 119 Gloss meter 120 Light source 121 Lens 122 Light receiving part 123 Dotted line 124 Surface 125 Surface roughness meter 126 Pattern 127 Inorganic filler 128 Surface layer 129 Main part 130 Thermosetting resin 131 Dirt 132 Protrusion part 133 Crack

Claims (6)

A metal plate,
A sheet-like heat-cured resin containing a crystalline epoxy resin provided on the metal plate, and a heat transfer layer made of an inorganic filler,
A wiring board, a lead frame, fixed to the heat transfer layer,
A heat dissipation board having
The content of the inorganic filler in the heat transfer layer is 66 Vol% or more and 90 Vol% or less,
A part of the side surface of the wiring board and a part of the side surface of the lead frame are connected to each other through one or more of a solder part or a mechanical connection part. 2. The heat dissipation board according to claim 1, wherein an arithmetic average roughness Ra of the surface of the heat transfer layer is any one or more of 3000 mm or less or a maximum height Ry of 15000 mm or less. The heat dissipation substrate according to claim 1, wherein the surface of the heat transfer layer has a 20 ° gloss of 70 or more. 2. The heat dissipation substrate according to claim 1, wherein the inorganic filler includes at least one selected from alumina, aluminum nitride, boron nitride, silicon carbide, silicon nitride, magnesium oxide, and zinc oxide. 2. The heat dissipation substrate according to claim 1, wherein the inorganic filler is made of at least one selected from alumina, aluminum nitride, boron nitride, silicon carbide, silicon nitride, magnesium oxide, and zinc oxide. An arrangement step of disposing an uncured thermosetting resin containing a crystalline epoxy resin, an uncured composition containing an inorganic filler, a lead frame, and a wiring board on a metal plate,
An embedding step of heating the uncured composition on the metal plate and embedding the lead frame and the wiring board in the uncured composition;
A convection step of heating at least the uncured composition to any one or more of a crystallization temperature or a melting point of the crystalline epoxy resin to form a liquid state and convection between the inorganic fillers;
A thermosetting step of thermosetting the uncured composition to form a heat transfer layer;
A method of manufacturing a heat dissipation board comprising:
The content of the inorganic filler is 66 Vol% or more and 90 Vol% or less,
A part of the side surface of the wiring board and a part of the side surface of the lead frame are connected via one or more of a solder part or a mechanical connection part. Method.

Images