JP2016039206A - Semiconductor device manufacturing method and semiconductor device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve radiation performance without an increase in size of a semiconductor device.SOLUTION: A semiconductor device manufacturing method comprises: a laminate formation process of forming a laminate including a first heat sink 21 and a second heat sink 52 which are arranged in parallel with each other on a first member 11 including a first insulator 11b, a third heat sink 83 arranged above the first heat sink 21 and the second heat sink 52, a first semiconductor element 31 which is physically and electrically connected between the first heat sink 21 and the third heat sink 83 and a second semiconductor element 72 physically and electrically connected between the second heat sink 52 and the third heat sink 83; a mold process of molding a mold body in which the laminate is encapsulated with a resin 100 while exposing a principal surface 11d of the first member 11; a cutting process of cutting the mold body so as to expose a principal surface 83a of the third heat sink 83 on a top face of the mold body; and a fixing process of fixing a second member 92 including an insulator to the exposed principal surface 83a of the third heat sink 83.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a method for manufacturing a double-sided heat radiation type semiconductor device and the semiconductor device.

  Conventionally, semiconductor devices including semiconductor elements such as IGBTs (insulated gate bipolar transistors) and diodes have been widely used. Furthermore, in order to efficiently cool the semiconductor elements in the semiconductor device, there is known a “double-sided heat dissipation type semiconductor device” provided with a pair of heat dissipation plates for radiating heat from both main surfaces of the semiconductor elements.

  More specifically, as shown in FIG. 9, the conventional double-sided heat dissipation type semiconductor device 900 includes a first heat dissipation plate 901, a second heat dissipation plate 902, a first semiconductor element 903, a second semiconductor element 904, and A third heat radiating plate 905 is provided. The first semiconductor element 903 is disposed on the first heat radiating plate 901, and the second semiconductor element 904 is disposed on the second heat radiating plate 902. The third heat sink 905 is disposed on the first semiconductor element 903 and the second semiconductor element 904. Further, the semiconductor device 900 includes a first insulator 907 integrated with the first fixing plate 906 and the first fixing plate 906, a second fixing plate 908 and a second fixing plate integrated with the second fixing plate. 2 insulators 909, spacers 910, and the like.

  When the semiconductor device 900 is manufactured, first, a structure in which a semiconductor element, a heat sink, an insulator, and the like are stacked is created in the above-described manner. Then, as shown in FIG. 9, the structure is housed in a “molding die 980 composed of an upper die 981 and a lower die 982” and molded with a resin 950 in the die 980. In this case, if the thicknesses of the semiconductor element, the heat sink, the insulator, and the like are different, the structure may be compressed by the molding die 980 when the mold 980 is clamped. There is a possibility that stress is generated in the constituent members and the members are damaged.

  Therefore, in one conventional manufacturing method, the first insulator 907 and the second insulator 909 are made of uncured thermosetting resin. As a result, the first insulator 907 and the second insulator 909 relieve the stress generated when the mold 980 is clamped by deformation, so that the members constituting the structure are damaged by the stress. Can be prevented (see, for example, Patent Document 1). Or the proposal which comprises the material (silicon rubber) which has the 1st insulator 907 and the 2nd insulator 909 with elasticity is also made | formed. Also by this, the same effect is anticipated (for example, refer patent document 2).

JP2013-093631A Japanese Patent No. 4479121

  However, in the conventional manufacturing method described above, in order to absorb the variation in the thickness of the structure, the thickness of the first insulator 907 and the second insulator 909 must be increased to some extent. As the thickness of the first insulator 907 and the second insulator 909 increases, the thermal resistance of these insulators increases. Therefore, according to the conventional manufacturing method, there is a problem that it is difficult to manufacture a semiconductor device having high heat dissipation performance and a reduced size.

  The present invention has been made in view of the above problems, and an object of the present invention is to provide a semiconductor device manufacturing method capable of improving heat dissipation performance without increasing its size in a double-sided heat dissipation type semiconductor device, and the semiconductor. To provide an apparatus.

  A method for manufacturing a semiconductor device according to the present invention includes: a stacked body forming process for forming a stacked body; a molding process for molding the stacked body with a resin to form a mold body; a cutting process for cutting the mold body; A fixing step of fixing a plate-like member to the mold body cut in the cutting step.

  The laminated body formed by the laminated body forming step includes a plate-like first member including a first insulator, a first heat radiating plate disposed on a first main surface of the first member, and the first A second heat sink disposed on the first main surface of one member, a third heat sink disposed on the opposite side of the first heat sink and the first member of the second heat sink, Is provided.

  Further, the laminate is disposed between the first heat radiating plate and the third heat radiating plate, and an electrode formed on a surface facing the first heat radiating plate is electrically connected to the first heat radiating plate. And an electrode formed on a surface facing the third heat radiating plate includes a first semiconductor element electrically connected to the third heat radiating plate.

  In addition, the laminate is disposed between the second heat radiating plate and the third heat radiating plate, and an electrode formed on a surface facing the second heat radiating plate is electrically connected to the second heat radiating plate. And an electrode formed on a surface facing the third heat radiating plate is provided with a second semiconductor element electrically connected to the third heat radiating plate.

  In the molding step, the laminated body is accommodated in the molding die such that the second main surface (main surface opposite to the first main surface) of the first member is in contact with the molding surface of the molding die. Supplying a resin into the mold, and molding the mold body by molding the resin so that the laminate is embedded in the resin while the second main surface of the first member is exposed. is there.

  In the cutting step, the first and second semiconductor elements of the pair of main surfaces of the third heat radiating plate are formed on a surface of the mold body opposite to the second main surface of the first member. Is a step of cutting the mold body so that the opposite main surface is exposed. The fixing step is a step of fixing a plate-like second member including a second insulator to the exposed main surface of the third heat radiating plate.

  In the present invention, in the molding step, the resin is placed in the molding die in a state where the laminate is accommodated in the molding die so that the second main surface of the first member is in contact with the molding surface of the molding die. (See FIGS. 4B and 4C). At this time, since the surface of the laminate opposite to the second main surface of the first member does not come into contact with the mold, stress due to mold clamping is not applied to the laminate. A mold body in which the laminated body is embedded in the resin is formed while the second main surface (11d) of the first member is exposed (see FIG. 4D). Thereafter, the main surface of the mold body opposite to the second main surface of the first member is opposite to the first and second semiconductor elements of the pair of main surfaces of the third heat sink. The mold body is cut so that is exposed (see FIGS. 5A and 5B).

  Therefore, the first member does not need elasticity to absorb the thickness variation of the laminated body. That is, the thickness of the first member does not need to take into account the variation in the thickness of the laminated body (that is, the dimensional tolerance in the laminating direction (also referred to as “lamination tolerance”), and can ensure the insulation performance. Therefore, according to the present invention, since the thickness of the first member can be made thinner than that of the conventional structure, the thickness of the laminated body can be made thinner.

  Furthermore, the first member including the first insulator does not need to be deformed in order to relieve the stress in the molding die in the molding process as in the conventional manufacturing method described above. Therefore, according to this invention, it is also possible to increase the quantity which adds the filler which has heat conductivity to a 1st member. When more filler is added to the first member, the thermal conductivity of the first member can be increased as compared with the conventional member, and as a result, the heat dissipation performance can be further improved.

  In addition, the second member including the second insulator does not need elasticity to absorb the thickness variation of the laminated body, like the first member. That is, since the thickness of the second member can be reduced as compared with the conventional structure, the thermal resistance of the second member can be reduced. Further, it is possible to increase the amount of the filler having thermal conductivity added to the second member. As described above, according to the present invention, the thickness of the first member and the second member can be reduced. Therefore, according to the manufacturing method according to the present invention, it is possible to manufacture a semiconductor device having a reduced thickness and improved heat dissipation performance.

  A semiconductor device according to the present invention is manufactured by the manufacturing method of the present invention, and includes a stacked body, a resin in which the stacked body is embedded, and a plate-like member.

  The laminated body includes a plate-shaped first member including a first insulator, a first heat radiating plate disposed on a first main surface of the first member, and the first main surface of the first member. And a third heat dissipating plate disposed on the opposite side of the first heat dissipating plate and the first member of the second heat dissipating plate.

  Further, the laminate is disposed between the first heat radiating plate and the third heat radiating plate, and an electrode formed on a surface facing the first heat radiating plate is electrically connected to the first heat radiating plate. And an electrode formed on a surface facing the third heat sink includes a first semiconductor element electrically connected to the third heat sink.

  In addition, the laminate is disposed between the second heat radiating plate and the third heat radiating plate, and an electrode formed on a surface facing the second heat radiating plate is electrically connected to the second heat radiating plate. And an electrode formed on a surface opposed to the third heat radiating plate includes a second semiconductor element electrically connected to the third heat radiating plate.

  The resin exposes the second main surface of the first member and the main surface of the pair of main surfaces of the third heat sink opposite to the first and second semiconductor elements. Then, the laminated body is embedded. The plate-like member is a plate-like second member including a second insulator, and is fixed to the exposed surface of the third heat radiating plate.

  In the semiconductor device according to the present invention configured as described above, since the thickness of the first member and the second member can be reduced as described above, the thermal resistance of the first member and the second member is reduced. As a result, the semiconductor device of the present invention can be a semiconductor device that is thinner than conventional devices and has improved heat dissipation performance.

FIG. 1 is a cross-sectional view schematically showing the configuration of the semiconductor device according to the first embodiment of the present invention. FIG. 2 is a circuit diagram of the semiconductor device according to the first and second embodiments of the present invention. FIG. 3 is a diagram showing an outline of a “stacking process” in which the constituent members of the semiconductor device are stacked to form a stacked body in the manufacturing process of the semiconductor device according to the first embodiment. FIG. 4 is a view showing an outline of a “molding process” in which a laminated body is resin-molded and a mold body is formed in the manufacturing process of the semiconductor device according to the first embodiment. FIG. 5 is a view showing an outline of a “cutting process” for cutting the mold body in the manufacturing process of the semiconductor device according to the first embodiment. FIG. 6 is a diagram showing an outline of a “fixing step” for fixing the second member to the mold body cut in the cutting step in the manufacturing process of the semiconductor device according to the first embodiment. FIG. 7 is a cross-sectional view schematically showing a modification of the semiconductor device according to the first embodiment of the present invention. FIG. 8 is a cross-sectional view schematically showing another modification of the semiconductor device according to the first embodiment of the present invention. FIG. 9 is a cross-sectional view schematically showing a configuration of a semiconductor device according to the prior art.

<First Embodiment>
A method for manufacturing a semiconductor device (hereinafter also referred to as “first device”) according to a first embodiment of the present invention and the semiconductor device will be described below with reference to the drawings.

(Device configuration)
First, the configuration of the first device will be described. A cross-sectional view schematically showing the configuration of the first device is shown in FIG. Hereinafter, for convenience of explanation, the upper side of each drawing is referred to as “upper” and the lower side of the drawing is referred to as “lower”. Accordingly, the upper surface of each member shown in each drawing is referred to as “upper surface”, and the lower surface of each member is referred to as “lower surface”.

  The semiconductor device (first device) 10 includes a first member 11, a first heat sink 21, a first semiconductor element 31, a first spacer 41, a second heat sink 52, a second spacer 62, a second semiconductor element 72, 3 heat radiation plate 83, second member 92, and resin 100.

  The first member 11 is a thin plate. The planar shape of the first member 11 is substantially rectangular. The first member 11 includes a first metal plate 11a and a first insulator 11b. The first metal plate 11a is made of metal such as copper and aluminum. The first insulator 11b is made of a material such as an epoxy resin, a silicon resin, or PET, and a thermally conductive filler is added. The lower surface of the first metal plate 11a (that is, the “second main surface 11d” of the first member 11) is exposed from the resin 100. The first insulator 11b is stacked on the first metal plate 11a. Therefore, the upper surface of the first metal plate 11a is in contact with the lower surface of the first insulator 11b. In addition, the upper surface of the first metal plate 11a and the lower surface of the first insulator 11b are in close contact with each other by “molding pressure” in a molding process described later.

  The first heat radiating plate 21 is a thin plate. The thickness of the first heat radiating plate 21 is larger than the thickness of the first member 11. The planar shape of the first heat radiating plate 21 is substantially rectangular, and is approximately half the size of the first member 11 in plan view. The first heat sink 21 is made of a metal conductor such as copper and aluminum. The first heat sink 21 is stacked on the first insulator 11 b of the first member 11. Therefore, the upper surface of the first insulator 11b (that is, the “first main surface 11c” of the first member 11) and the lower surface of the first heat radiating plate 21 are in contact with each other. The upper surface of the first insulator 11b and the lower surface of the first heat radiating plate 21 are in close contact with each other by “molding pressure”.

  The first semiconductor element 31 is an insulated gate bipolar transistor (hereinafter referred to as “IGBT”). Therefore, hereinafter, the first semiconductor element 31 is referred to as an IGBT 31. The IGBT 31 has a thin plate shape. The thickness of the IGBT 31 is smaller than the thickness of the first member 11. The planar shape of the IGBT 31 is substantially rectangular and is smaller than the first heat sink 21 in plan view. The IGBT 31 is made of a material mainly made of silicon. The IGBT 31 includes a collector electrode 31a, an emitter electrode 31b, and a gate electrode 31c.

  The collector electrode 31a is formed on the lower surface (one main surface) of the IGBT 31. The emitter electrode 31b and the gate electrode 31c are formed on the upper surface (the other main surface) of the IGBT 31. The IGBT 31 is stacked on the first heat radiating plate 21 via the bonding material 111. As the bonding material 111, for example, a tin-based solder paste (including an alloy made of tin, silver, copper, or the like) is used. Therefore, the collector electrode 31 a of the IGBT 31 is electrically connected to the first heat sink 21. Note that bonding materials 121, 131, 142, 152, and 162 described later are also made of the same material as the bonding material 111. The gate electrode 31c of the IGBT 31 is connected to a first control terminal (not shown) via a bonding wire. The bonding wire is made of, for example, an aluminum wire or a gold wire.

  The first spacer 41 is a thin plate. The thickness of the first spacer 41 is larger than the thickness of the first member 11 and smaller than the thickness of the first heat sink 21. The planar shape of the first spacer 41 is substantially rectangular and is smaller than the IGBT 31 in plan view. The first spacer 41 is made of a metal conductor such as copper and aluminum. The first spacer 41 is laminated on the upper surface of the IGBT 31 with a bonding material 121 interposed therebetween. Accordingly, the first spacer 41 is electrically connected to the emitter electrode 31b.

  The second heat sink 52 is the same member as the first heat sink 21. Therefore, the second heat radiating plate 52 is a thin plate. The thickness of the second heat sink 52 is substantially equal to that of the first heat sink 21. The planar shape of the second heat radiating plate 52 is substantially rectangular, is approximately half the size of the first member 11 in plan view, and is substantially equal to the first heat radiating plate 21. The second heat sink 52 is made of a metal conductor such as copper and aluminum. The second heat radiating plate 52 is stacked on the first insulator 11 b of the first member 11. Therefore, the upper surface of the first insulator 11b and the lower surface of the second heat radiating plate 52 are in contact with each other. The 1st heat sink 21 and the 2nd heat sink 52 are arrange | positioned on the 1st member 11 at predetermined distance so that the opposing edge | side may become mutually parallel.

  The second spacer 62 is the same member as the first spacer 41. Therefore, the second spacer 62 is a thin plate. The thickness of the second spacer 62 is substantially equal to that of the first spacer 41. The planar shape of the second spacer 62 is substantially rectangular, and is substantially equal to the size of the first spacer 41 in plan view. The second spacer 62 is made of a metal conductor such as copper and aluminum. The second spacer 62 is laminated on the second heat radiating plate 52 via a bonding material 142 and is electrically connected to the second heat radiating plate 52.

  The second semiconductor element 72 is the same IGBT as the first semiconductor element 31. Therefore, hereinafter, the second semiconductor element 72 is referred to as an IGBT 72. The IGBT 72 is a thin plate. The thickness of the IGBT 72 is substantially equal to the thickness of the IGBT 31. The planar shape of the IGBT 72 is substantially rectangular, is smaller than the second heat radiating plate 52 in plan view, and is substantially equal to the IGBT 31. The IGBT 72 is made of a material mainly made of silicon. The IGBT 72 includes a collector electrode 72a, an emitter electrode 72b, and a gate electrode 72c.

  The collector electrode 72 a is formed on the upper surface (one main surface) of the IGBT 72. The emitter electrode 72b and the gate electrode 72c are formed on the lower surface (the other main surface) of the IGBT 72. The IGBT 72 is stacked on the second spacer 62 via a bonding material 152 provided on the emitter electrode 72b. Therefore, the emitter electrode 72 b of the IGBT 72 is electrically connected to the second spacer 62. The gate electrode 72c of the IGBT 72 is connected to a second control terminal (not shown) via a bonding wire.

  The third heat radiating plate 83 is a thin plate. The thickness of the third heat sink 83 is larger than the thickness of the first member 11 and is substantially equal to the first heat sink 21 and the second heat sink 52. The planar shape of the third heat radiating plate 83 is substantially rectangular and is approximately the same as or slightly smaller than the first member 11 in plan view. The third heat sink 83 is made of a metal conductor such as copper and aluminum. The third heat radiating plate 83 is laminated on the first spacer 41 and the IGBT 72 via a bonding material 131 and a bonding material 162, respectively, and is electrically connected to the first spacer 41 and the collector electrode 72a of the IGBT 72.

  The second member 92 is a thin plate. The planar shape of the second member 92 is substantially the same rectangular shape as the first member 11, and is slightly larger than the third heat radiating plate 83 in plan view. The second member 92 includes a second metal plate 92a and a second insulator 92b. The second metal plate 92a is made of metal such as copper and aluminum. The second insulator 92b is made of a material such as a thermosetting epoxy resin or an ultraviolet curable epoxy resin and an acrylic resin, and is added with a heat conductive filler.

  The second insulator 92b is laminated on the third heat radiating plate 83, and is bonded to the upper surface 83a of the third heat radiating plate 83 by thermosetting the second insulator 92b. Therefore, the upper surface 83a of the third heat radiating plate 83 is in contact with the lower surface of the second insulator 92b. The second metal plate 92a is laminated on the second insulator 92b, and is bonded to the upper surface of the second insulator 92b by thermosetting the second insulator 92b. Therefore, the upper surface of the second insulator 92b is in contact with the lower surface of the second metal plate 92a.

  The resin 100 has a substantially rectangular parallelepiped shape. The resin 100 is made of an epoxy resin insulating material having thermosetting properties, and contains, for example, an inorganic filler such as silica and alumina. The resin 100 includes the first member 11, the first heat radiation plate 21, the IGBT 31, the first spacer 41, the second heat radiation plate 52, the second spacer 62, the IGBT 72, the third heat radiation plate 83, and the bonding materials 111, 121, 131, 142. , 152 and 162 are embedded (sealed). However, the lower surface of the first member 11 (that is, the lower surface of the first metal plate 11 a) and the upper surface (83 a) of the third heat radiating plate 83 are exposed from the resin 100. In other words, the resin 100 exposes the lower surface 11 d of the first metal plate 11 a and the upper surface 83 a of the third heat radiating plate 83 and embeds members other than the second member 92 in the first device 10.

  With the above configuration, the first device 10 has a “double-sided heat radiation type” structure. That is, part of the heat generated in the IGBT 31 is directed to the first heat radiating plate 21 via the bonding material 111, and further from the first heat radiating plate 21 through the first insulator 11b. Heat is dissipated through the path toward 11a. The rest is directed to the third heat radiating plate 83 via the bonding material 121, the first spacer 41, and the bonding material 131, and further from the third heat radiating plate 83 through the second insulator 92b to the second metal plate 92a. The heat is dissipated through the path toward On the other hand, part of the heat generated in the IGBT 72 is directed to the third heat radiating plate 83 via the bonding material 162, and further from the third heat radiating plate 83 through the second insulator 92b to the second metal plate. Heat is dissipated through the path toward 92a. The rest is directed to the second heat radiating plate 52 via the bonding material 152, the second spacer 62, and the bonding material 142, and further from the second heat radiating plate 52 through the first insulator 11b to the first metal plate 11a. The heat is dissipated through the path toward

  As described above, the first heat radiating plate 21 and the second heat radiating plate 52 are arranged at a predetermined distance, and the gap between the first heat radiating plate 21 and the second heat radiating plate 52 is filled with the resin 100. Yes. Therefore, insulation between the first heat sink 21 and the second heat sink 52 is ensured.

  Although not shown, the first heat radiating plate 21, the second heat radiating plate 52, and the third heat radiating plate 83 are exposed to the outside from the side of the resin 100 through a conductive portion (not shown). Furthermore, a part of each of the first control terminal and the second control terminal is exposed to the outside of the resin 100. As a result, the semiconductor device 10 can be electrically connected to an external circuit.

(Circuit configuration of the device)
Next, the circuit configuration of the first device 10 will be described. The 1st apparatus 10 is applied to inverters, such as a hybrid vehicle and an electric vehicle. An inverter including the first device 10 and its peripheral circuit diagram are shown in FIG. In FIG. 2, the same parts as those in FIG. 1 are denoted by the same reference numerals.

  The circuit shown in FIG. 2 includes a storage battery B1, a smoothing capacitor C1, an inverter 200, and a motor generator MG. The inverter 200 is a three-phase output inverter having a U phase, a V phase, and a W phase, and is represented as a region surrounded by a broken line in FIG. The inverter 200 includes semiconductor elements (IGBTs) T1 to T6, diodes D1 to D6, a positive terminal P1 and a negative terminal N1 connected to the positive and negative electrodes of the storage battery B1, respectively, and an output terminal connected to the motor generator MG. U1, V1 and W1. The semiconductor elements T1 to T6 and the diodes D1 to D6 are connected in antiparallel.

  The semiconductor element T1 to which the diode D1 is connected in antiparallel and the semiconductor element T2 to which the diode D2 is connected in antiparallel are connected in series to constitute a U-phase circuit of the inverter 200. This configuration is referred to as a “leg”. Similarly, the circuit composed of the semiconductor elements T3 and T4 and the diodes D3 and D4 constitutes the V-phase “leg” of the inverter 200, and the circuit composed of the semiconductor elements T5 and T6 and the diodes D5 and D6 is the W of the inverter 200. It constitutes the “leg” of the phase. One semiconductor element in the inverter 200 and one diode connected in antiparallel to the element constitute one “arm”. Each circuit including the semiconductor element T1 and the diode D1, the semiconductor element T3 and the diode D3, and the semiconductor element T5 and the diode D5 is referred to as an “upper arm”. Each circuit including the semiconductor element T2 and the diode D2, the semiconductor element T4 and the diode D4, and the semiconductor element T6 and the diode D6 is referred to as a “lower arm”.

  In FIG. 2, a region surrounded by an alternate long and short dash line is a portion of only one of the “legs” included in the inverter 200 and includes only the semiconductor elements T <b> 1 and T <b> 2, and the semiconductor device (first device) 10 illustrated in FIG. Equivalent to. That is, the semiconductor element T1 and the semiconductor element T2 correspond to the IGBT 31 and the IGBT 72 shown in FIG.

  In the semiconductor device (first device) 10, the collector electrode 31 a of the IGBT 31 is electrically connected to the “positive terminal 10 a for connecting to the positive terminal P <b> 1 of the inverter 200” via the first heat radiating plate 21. The emitter electrode 31 b of the IGBT 31 is electrically connected to the third heat radiating plate 83 of the semiconductor device (first device) 10. The gate electrode 31 c of the IGBT 31 is electrically connected to the first control terminal 10 b of the semiconductor device (first device) 10.

  The collector electrode 72 a of the IGBT 72 is electrically connected to the third heat radiating plate 83 of the semiconductor device (first device) 10. The emitter electrode 72 b of the IGBT 72 is electrically connected to the “negative electrode terminal 10 c for connecting to the negative electrode terminal N 1 of the inverter 200” via the second heat dissipation plate 52. The gate electrode 72 c of the IGBT 72 is electrically connected to the second control terminal 10 d of the semiconductor device (first device) 10. The output terminal 10 e is electrically connected to the third heat sink 83 of the semiconductor device (first device) 10.

  In the semiconductor device (first device) 10 having such a circuit configuration, different voltages are applied or generated to the first heat radiating plate 21, the second heat radiating plate 52, and the third heat radiating plate 83, respectively. Therefore, insulation is required between the first heat radiating plate 21, the second heat radiating plate 52, and the third heat radiating plate 83. In particular, as shown in FIG. 1, the first heat radiating plate 21 and the second heat radiating plate 52 are adjacent to each other and have a large potential difference, so that the first heat radiating plate 21 and the second heat radiating plate 52 are high. Dielectric strength is required.

  As described above, the semiconductor device (first device) 10 is sealed with the resin 100. Therefore, the resin 100 is filled between the first heat radiating plate 21 and the second heat radiating plate 52, and the initial withstand voltage is secured. However, when a voltage is applied to the positive terminal 10a and the negative terminal 10c of the semiconductor device (first device) 10, an electric field is generated between the first heat radiating plate 21 and the second heat radiating plate 52, and insulation by ion migration is performed. A decrease in breakdown voltage can occur. Ion migration is more likely to proceed as the moisture absorption amount of the electrodes (that is, the first heat radiating plate 21 and the second heat radiating plate 52) is higher.

  On the other hand, in the 1st apparatus 10, the 1st heat sink 21 and the 2nd heat sink 52 are sealed with the 1st member 11 and resin 100 which these contact | abut. As will be described later, in the manufacturing process of the first device 10, the first heat radiating plate 21 and the second heat radiating plate 52 are sealed by the first member 11 and the resin 100 in a high temperature state (in a state where moisture has evaporated). Stopped. Therefore, adhesion of moisture to the surfaces of the first heat sink 21 and the second heat sink 52 is suppressed. Furthermore, the first heat radiating plate 21 and the second heat radiating plate 52 are not exposed to the atmosphere even after their manufacture. Therefore, even when the first device 10 is in use, the adhesion of moisture to the surfaces of the first heat sink 21 and the second heat sink 52 is suppressed.

  As a result, the influence of ion migration that can occur between the first heat sink 21 and the second heat sink 52 can be suppressed. In other words, the first device 10 is less affected by ion migration when compared to a device configured such that the surfaces of the first heat sink 21 and the second heat sink 52 are exposed from the resin 100 and exposed to the atmosphere. can do.

(Production method)
Next, a method for manufacturing the first device 10 will be described with reference to FIGS. 3 to 6, the same parts as those in FIG. 1 are denoted by the same reference numerals, and the description thereof may be omitted.

The manufacturing process of the 1st apparatus 10 is divided roughly into the following four processes.
(1) “Laminate Forming Process” in which members constituting the first device 10 are laminated to form a “laminate”.
(2) “Molding process” for forming a “molded body” by molding the laminate with the resin 100
(3) “Cutting step” of cutting the “mold body” such that the main surface 83a of the third heat radiating plate 83 is exposed from the “mold body”.
(4) “Fixing process” for fixing the second member 92 to the main surface 83 a of the third heat radiating plate 83.
Hereinafter, each step will be described.

(1) Laminate Forming Step In the laminate forming step, first, as shown in FIG. 3A, the upper surface (one main surface) of the first heat radiating plate 21 and the lower surface (one main surface) of the IGBT 31. The collector electrode 31a formed on 31d is physically and electrically connected through the bonding material 111. Furthermore, the emitter electrode 31 b formed on the upper surface (the other main surface) 31 e of the IGBT 31 and the lower surface (the one main surface) of the first spacer 41 are physically and electrically connected via the bonding material 121. In addition, a bonding wire (not shown) is connected to the gate electrode 31c formed on the upper surface (the other main surface) 31e of the IGBT 31. The structure configured in this manner is referred to as “first structure S1”.

  Next, as shown in FIG. 3B, the upper surface (one main surface) of the third heat radiating plate 83 and the collector electrode 72a formed on the lower surface (one main surface) 72d of the IGBT 72 are joined together. It is physically and electrically connected through 162. Further, the emitter electrode 72b formed on the upper surface (the other main surface) 72e of the IGBT 72 and the lower surface (the one main surface) of the second spacer 62 are physically and electrically connected through the bonding material 152. In addition, a bonding wire (not shown) is connected to the gate electrode 72c formed on the upper surface (the other main surface) 72e of the IGBT 72. The structure configured in this way is referred to as “second structure S2”.

  Subsequently, as shown in FIGS. 3C and 3D, the first insulator 11b is disposed in contact with the upper surface (one main surface) of the first metal plate 11a, and the first member is arranged. 11 is formed. Furthermore, the lower surface (one main surface) of the first structure S1 and the lower surface (one main surface) of the second heat dissipating plate 52 are disposed in contact with the upper surface (one main surface) 11c of the first member 11. . In addition, after the second structure S2 is turned upside down, the second structure S2, the upper surface of the first spacer 41 of the first structure S1 (the other main surface), and the upper surface of the second heat radiation plate 52 (the other surface) Are connected physically and electrically via a bonding material 131 and a bonding material 142, respectively. In this way, a “laminate L1” as shown in FIG. 3D is formed.

(2) Molding Process In the molding process, first, as shown in FIG. 4A, the laminate L1 is accommodated in a molding die 300 composed of an upper die 310 and a lower die 320. In this case, the laminated body L1 is accommodated in the molding die 300 so that the second main surface 11d of the first member 11 is in contact with the molding surface 321 formed on the lower die 320.

  As shown in FIG. 4B, the mold body 300 is housed and the mold 300 is clamped with the surface 311 of the upper mold 310 and the surface 322 of the lower mold 320 in contact with each other. . At this time, the space surrounded by the molding die 300 is referred to as a cavity 330. A space is provided between the upper portion of the laminate L1, that is, between the main surface 83a of the third heat radiating plate 83 and the inner wall surface 312 of the upper mold.

  Next, as shown in FIG. 4 (C), the temperature of the molding die 300 is raised, and the liquid 330 at a normal temperature is injected into the cavity 330 from the injection gate 340 provided at the “upper” portion of the upper die 310. Resin 100 is injected. As the resin 100, a resin based on an epoxy-based thermosetting resin is used. Since the curing temperature of the resin 100 is about 180 ° C., the temperature of the molding die 300 is set to about 180 ° C. or higher. When the liquid resin 100 is filled into the cavity 330 from the injection gate 340, the laminate L1 is pressed against the molding surface 321 of the lower mold 320 by the molding pressure generated in the cavity 330. Therefore, the 1st insulator 11b, the 1st heat sink 21, and the 2nd heat sink 52 are fixed in the state which contact | adhered. Furthermore, since the second main surface 11d of the first member 11 and the molding surface 321 are brought into close contact with each other by the molding pressure, the second main surface 11d is not buried in the resin 100. The injection of the resin 100 is performed until the upper portion of the laminate L1, that is, the main surface 83a of the third heat radiating plate 83 is completely buried in the resin 100.

  After the resin 100 is cured, the molding die 300 is cooled, and the mold body M1 is taken out from the molding die 300 as shown in FIG. In this way, the mold body M1 in which the multilayer body L1 is embedded in the resin 100 is molded while the second main surface 11d of the first member 11 is exposed.

(3) Cutting Process In the cutting process, first, as shown in FIG. 5A, the mold body M1 so that the second main surface 11d of the first member 11 of the mold body M1 contacts the fixing base 400. And hold it with a jig (not shown). Next, the cutting tool 410 is brought into contact with the main surface M1a of the mold body M1 opposite to the second main surface 11d of the first member 11. The main surface M1a is cut by bringing the cutting tool 410 into contact with the main surface M1a and moving it in the direction indicated by the arrow A1. In this way, the mold body M1 is cut until the main surface 83a of the third heat radiating plate 83 is exposed.

  After completion of the cutting, as shown in FIG. 5 (B), a molded body M1 'in which the main surface 83a of the third heat radiating plate 83 is exposed is formed. The cut main surface of the mold body M1 'after completion of cutting is referred to as M1'a. In addition, after the main surface 83a of the third heat radiating plate 83 is exposed, a predetermined amount may be further cut so that the residue of the resin 100 does not remain on the main surface 83a of the third heat radiating plate 83. Further, although not shown, the cut surface M1'a may be smoothed by etching and polishing after the cutting with the cutting tool 410.

(4) Fixing Step In the fixing step, as shown in FIG. 6A, first, the second member 92 constituted by the second insulator 92b and the second metal plate 92a is moved to the second member 92. The insulator 92b is disposed so as to contact the main surface M1′a (83a) of the mold body M1 ′. The second insulator 92b is made of an epoxy thermosetting resin to which a filler having thermal conductivity is added.

  Next, as illustrated in FIG. 6B, the mold body M <b> 1 ′ is installed so that the second main surface 11 d of the first member 11 of the mold body M <b> 1 ′ is in contact with the fixed base 500. A predetermined amount of heat is applied over a predetermined period while pressing the head 510 of the hot press machine with a predetermined pressure against the surface of the second member 92 opposite to the surface that contacts the main surface M1′a of the molded body M1 ′. Two members 92 are provided.

  As shown in FIG. 6C, the second member 92 is fixed to the mold body M1 'by curing the second insulator 92b. By performing the above steps, the semiconductor device 10 is completed.

<Overview of the first embodiment>
The manufacturing method of the semiconductor device 10 according to the present invention includes a laminated body forming step of forming the laminated body L1, a molding step of molding the laminated body L1 with the resin 100 to form the molded body M1, and cutting the molded body M1. And a fixing step of fixing a plate-like member to the mold body M1 ′ cut in the cutting step.

  The laminated body L1 formed by the laminated body forming step includes a plate-like first member 11 including the first insulator 11b and a first heat dissipation disposed on the first main surface 11c of the first member 11. The plate 21, the second heat radiating plate 52 disposed on the first main surface 11c of the first member 11, and the first heat radiating plate 21 and the second heat radiating plate 52 opposite to the first member 11 A third heat radiation plate 83 is provided.

  Furthermore, the laminated body L1 is disposed between the first heat radiating plate 21 and the third heat radiating plate 83 and is formed on a surface facing the first heat radiating plate 21 (collector electrode 31a). Is electrically connected to the first heat radiating plate 21 and an electrode (emitter electrode 31b) formed on a surface facing the third heat radiating plate 83 is electrically connected to the third heat radiating plate 83. A semiconductor element 31 is provided.

  In addition, the laminate L1 is disposed between the second heat radiating plate 52 and the third heat radiating plate 83, and is formed on an electrode (emitter electrode 72b) formed on a surface facing the second heat radiating plate 52. ) Is electrically connected to the second heat radiating plate 52 and an electrode (collector electrode 72a) formed on a surface facing the third heat radiating plate 83 is electrically connected to the third heat radiating plate 83. Two semiconductor elements 72 are provided.

In the molding step, the laminate L1 is accommodated in the molding die (cavity 330) such that the second main surface 11d of the first member 11 contacts the molding surface 321 of the molding die 300, and the resin 100 Is supplied into the mold (cavity 330).
The mold body M1 is formed by molding with the resin 100 so that the laminated body L1 is embedded in the resin 100 while the second main surface 11d of the first member 11 is exposed.

  In the cutting step, the first and second of the pair of main surfaces of the third heat radiating plate 83 are formed on the surface M1a of the mold body M1 opposite to the second main surface 11d of the first member 11. 2 The mold body M1 is cut so that the main surface 83a opposite to the semiconductor elements (31 and 72) is exposed.

  In the fixing step, the plate-like second member 92 including the second insulator 92b is fixed to the exposed main surface 83a (M1'a) of the third heat radiating plate 83.

<Effects of First Embodiment>
Thus, in the manufacturing method according to the embodiment of the present invention, the upper surface of the mold body M1 is cut after the mold body M1 is molded. Therefore, in the dimension design of the molding die 300 used in the molding process, it is not necessary to consider the thickness variation (lamination tolerance) of the laminate L1. That is, it is not necessary to make the insulator (first insulator 11b) thicker than necessary in order to absorb the stacking tolerance.

  That is, according to the present manufacturing method, the first insulator 11b only needs to have a thickness necessary and sufficient to ensure insulation, and a decrease in heat dissipation performance due to an increase in the thickness of the insulator can be suppressed. Furthermore, according to this manufacturing method, the first insulator 11b does not need to be deformed for stress relaxation in the molding process. Therefore, the thermal resistance of the first insulator 11b can be reduced by increasing the amount of filler having thermal conductivity added to the first insulator 11b.

  Therefore, according to the manufacturing method of the present invention, it is possible to manufacture a semiconductor device having a reduced thickness and improved heat dissipation performance.

<Modification of First Embodiment>
Next, a semiconductor device according to a modification of the first embodiment of the present invention will be described.

  As described with reference to FIG. 2, the semiconductor elements T <b> 1, T <b> 3, and T <b> 5 each constitute a part of the “upper arm” of the inverter 200. The semiconductor elements T1, T3, and T5 have their collectors connected in parallel and their emitters connected to the outputs U1, V1, and W1 of the respective phases of the inverter 200. On the other hand, the semiconductor elements T2, T4, and T6 each constitute a part of the “lower arm” of the inverter 200. The semiconductor elements T2, T4, and T6 have their collectors connected to the outputs U1, V1, and W1 of the respective phases of the inverter 200 and their emitters connected in parallel.

  In FIG. 2, the portion surrounded by the two-dot chain line, that is, the portion constituted by the semiconductor elements T3 and T5 is “upper” of two phases (V phase and W phase) of the three-phase “upper arms”. This is a semiconductor device 101 that constitutes a part of “arm” (a part excluding a diode). A portion surrounded by a three-dot chain line, that is, a portion constituted by the semiconductor elements T1, T3, and T5 is a semiconductor device 102 constituting a part of the three-phase “upper arm” (portion excluding the diode). is there. A portion surrounded by one-dot two-short chain line, that is, a portion constituted by semiconductor elements T2, T4, and T6 constitutes a part of a three-phase “lower arm” (a portion excluding a diode). 103.

  Hereinafter, cross-sectional structures of the semiconductor devices 101, 102, and 103 will be illustrated with reference to FIG.

(1) Two-Phase Upper Arm A semiconductor device 101 shown in FIG. 7A is a semiconductor device related to a two-phase upper arm among the three-phase upper arms of the inverter 200. The semiconductor device 101 includes the first member 11, the first heat sink 21, the IGBT 31, the first spacer 41, the second heat sink 52, the second spacer 62, the IGBT 72, the third heat sink 83, the second member 92, and the resin 100. I have.

  The semiconductor device 101 differs from the first device 10 only in that the stacking order of the IGBT 31 and the first spacer 41 is upside down. Hereinafter, the difference will be described. In the semiconductor device 101, the IGBT 31 is disposed with the collector electrode 31a on the top surface and the emitter electrode 31b and the gate electrode 31c on the bottom surface. The collector electrode 31 a is physically and electrically connected to the third heat radiating plate 83 through the bonding material 111. The emitter electrode 31 b is physically and electrically connected to the upper surface of the first spacer 41 through the bonding material 121. The lower surface of the first spacer 41 is physically and electrically connected to the upper surface of the first heat radiating plate 21 through the bonding material 131.

  That is, in the semiconductor device 101, the structure including the first heat dissipation plate 21, the first spacer 41, and the IGBT 31 and the structure including the second heat dissipation plate 52, the second spacer 62, and the IGBT 72 have the same stacked structure. In other words, in the semiconductor device 101, two parallel structures (a structure made up of a heat sink, a spacer, and an IGBT) stacked between the first member 11 and the third heat sink 83 are the same. .

  Thus, the semiconductor device 101 constitutes a circuit in which the collector electrode 31a of the IGBT 31 and the collector electrode 72a of the IGBT 72 are electrically connected.

(2) Three-Phase Upper Arm The semiconductor device 102 shown in FIG. 7B is a semiconductor device related to the three-phase upper arm of the inverter 200. The semiconductor device 102 includes a third member 11 ′, a first heat sink 21, an IGBT 31, a first spacer 41, a second heat sink 52, a fourth heat sink 53, a second spacer 62, a third spacer 63, IGBT 72, IGBT 73, 5 heat sink 83 ', 4th member 92', and resin 100 are provided.

  The semiconductor device 102 has a configuration in which two structures in parallel in the semiconductor device 101 are further added in parallel.

  The third member 11 ′ is the same as the first member 11 except that the long side is approximately 1.5 times longer than the first member 11 in plan view. The fourth heat sink 53 is the same as the first heat sink 21 and the second heat sink 52. The third spacer 63 is the same as the first spacer 41 and the second spacer 62. The IGBT 73 is the same as the IGBT 31 and the IGBT 72. The fifth heat radiating plate 83 ′ is the same as the third heat radiating plate 83 except that its long side is approximately 1.5 times longer than the third heat radiating plate 83 in plan view. The fourth member 92 ′ is the same as the second member 92 except that the long member is approximately 1.5 times longer than the second member 92 in plan view. The resin 100 ′ is made of an epoxy resin material having the same thermosetting property as the resin 100.

  Thus, the semiconductor device 102 constitutes a circuit in which the collector electrode 31a of the IGBT 31, the collector electrode 72a of the IGBT 72, and the collector electrode 73a of the IGBT 73 are electrically connected.

(3) Three-Phase Lower Arm The semiconductor device 103 illustrated in FIG. 7C is a semiconductor device according to the three-phase lower arm of the inverter 200. The three first structures S1 are arranged in parallel on the upper surface (on one main surface) of the third member 11 ′ at a predetermined interval, and are the third members 11 ′ of the three first structures S1. The fifth heat radiating plate 83 ′ is disposed on the opposite side. The upper surfaces of the three first structures S1 are both physically and electrically connected to the fifth heat radiating plate 83 ′ via conductive bonding materials.

  The semiconductor devices 101 to 103 are manufactured by the same method as the first device 10. That is, taking the semiconductor device 101 as an example, the first member 11, the first heat radiation plate 21, the IGBT 31, the first spacer 41, the second heat radiation plate 52, the second spacer 62, the IGBT 72, the third heat radiation plate 83, the bonding material. 111, 121, 131, 142, 152, and 162 are embedded in the resin 100 by a resin mold. Thereafter, the upper surface of the molded body is cut to expose the upper surface of the third heat radiating plate 83, and the second member 92 is bonded and fixed on the surface to form the semiconductor device 101.

<Other variations>
The semiconductor device 104 illustrated in FIG. 7D is a semiconductor device that is different from the first device 10 only in that it does not include the first metal plate 11a and the second metal plate 92a. With this configuration, a user can select a heat radiating body such as a metal plate, a heat sink, and a heat radiating fin in accordance with the use of the semiconductor device. Such a configuration is applicable not only to the first device 10 but also to the semiconductor devices 101 to 103.

  Next, a semiconductor device according to still another modification of the first embodiment of the present invention will be described with reference to FIG.

  A semiconductor device 1001 shown in FIG. 8A is a semiconductor device in which the first metal plate 11a in the first device 10 is changed to a first metal plate 11af with a radiation fin.

  In the semiconductor device 1002 shown in FIG. 8B, in the first device 10, the first metal plate 11a is changed to the first metal plate 11af with the radiation fins, and the second metal plate 92a has the radiation fins. The semiconductor device is changed to the second metal plate 92af.

  In the semiconductor device 1031 shown in FIG. 8C, in the semiconductor device 103, the third metal plate 11′a having a larger area than the first metal plate 11a is replaced by the third metal plate 11′af with the radiation fins. This is a changed semiconductor device.

  In the semiconductor device 1032 shown in FIG. 8D, in the semiconductor device 103, the third metal plate 11′a is changed to the third metal plate 11′af with the radiation fins and the second metal plate 92a. Further, the fourth metal plate 92′a having a large area is a semiconductor device in which the fourth metal plate 92′af with the radiation fins is changed.

  For example, in the semiconductor device 1002, part of the heat generated in the IGBT 31 and the IGBT 72 passes through the first heat radiation plate 21 and the second heat radiation plate 52, respectively, and further through a heat radiation path toward the first member 11. It reaches the first metal plate 11af with the radiation fins. Then, the heat is released from the metal plate 11af to the space. The remainder of the heat generated in the IGBT 31 and the IGBT 72 passes through the third heat radiating plate 83, passes through the heat radiating path toward the second member 92, and reaches the second metal plate 92 af with radiating fins. The metal plate 92af is discharged into the space. Therefore, according to the present embodiment, it is possible to realize a semiconductor device capable of efficiently radiating heat from both sides by the radiating fins.

  For example, the semiconductor device 1002 is different from the first embodiment by changing the molding surface 321 of the lower mold 320 so that the metal member 11af with the first heat radiating fin fits in the “molding process” described above. Similarly, the semiconductor device can be molded. Furthermore, in the “fixing step” described above, the metal member and the insulator can be fixed in the same manner as in the first embodiment by pressing the head 510 of the heat press machine against the metal member 92af with the second heat radiation fin. it can. Therefore, according to this embodiment, a double-sided heat radiation type semiconductor device with heat radiation fins can be easily manufactured.

  The present invention is not limited to the above embodiment, and various modifications can be employed within the scope of the present invention. For example, in the semiconductor device 10, the semiconductor element 31 and the semiconductor element 72 are not limited to IGBTs, and may be power MOSFETs, MOSFETs, bipolar transistors, diodes, or the like, or a combination of semiconductor elements and diodes. Good.

  Various circuit configurations other than the circuit configuration in the description of the second embodiment are conceivable. For example, the semiconductor device may have a configuration corresponding to two lower arms.

  Further, in the case of a structure that employs a diode as a semiconductor element, it is not necessary to bond a wire to the diode. Therefore, in the case of a diode, the step of turning the structure upside down after bonding the wire to the structure can be omitted. Therefore, in the manufacturing method of the present invention, a method of stacking the members upward from the first member 11 may be used without the step of turning the structure upside down.

  As described above, the bonding materials 111, 121, 131, 142, 152, and 162 are tin-based solder pastes, but may be tin-lead solder pastes containing bismuth, indium, and the like. It may be a “low melting point solder”. Moreover, not only a solder paste but a solder foil and a conductive adhesive may be used.

DESCRIPTION OF SYMBOLS 10 ... Semiconductor device, 11 ... 1st member, 11b ... 1st insulator, 21 ... 1st heat sink, 31 ... 1st semiconductor element, 41 ... 1st spacer, 52 ... 2nd heat sink, 62 ... 2nd Spacer, 72, second semiconductor element, 83, third heat sink, 92, second member, 92b, second insulator, 100, resin, 200, inverter, 300, molding die.

Claims (2)

A plate-shaped first member including a first insulator, a first heat radiating plate disposed on the first main surface of the first member, and disposed on the first main surface of the first member. A second heat radiating plate, a third heat radiating plate disposed on the opposite side of the first heat radiating plate and the first heat radiating plate from the first member, and the first heat radiating plate and the third heat radiating plate. An electrode disposed on the surface facing the first heat sink is electrically connected to the first heat sink and formed on a surface facing the third heat sink. A first semiconductor element electrically connected to the third heat sink; and a first semiconductor element disposed between the second heat sink and the third heat sink and formed on a surface facing the second heat sink. An electrode electrically connected to the second heat radiating plate and an electrode formed on a surface facing the third heat radiating plate is electrically connected to the third heat radiating plate. A stack forming step of forming a laminated body having a semiconductor element, a,
The laminated body is accommodated in the molding die so that the second main surface of the first member is in contact with the molding surface of the molding die, resin is supplied into the molding die, and the first member A molding step of molding the molded body by molding with a resin so that the laminated body is embedded in the resin while the second main surface is exposed;
A main surface of the mold body opposite to the second main surface of the first member is opposite to the first and second semiconductor elements of the pair of main surfaces of the third heat radiating plate. A cutting step of cutting the mold body such that the surface is exposed;
A fixing step of fixing a plate-like second member including a second insulator to the exposed main surface of the third heat sink;
A method for manufacturing a semiconductor device comprising: A plate-like first member including a first insulator;
A first heat radiating plate disposed on a first main surface of the first member;
A second heat radiating plate disposed on the first main surface of the first member;
A third heat sink disposed on the opposite side of the first heat sink and the first member of the second heat sink;
An electrode disposed between the first heat radiating plate and the third heat radiating plate and formed on a surface facing the first heat radiating plate is electrically connected to the first heat radiating plate and the third heat radiating plate. A first semiconductor element in which an electrode formed on a surface facing the heat sink is electrically connected to the third heat sink;
An electrode disposed between the second heat radiating plate and the third heat radiating plate and formed on a surface facing the second heat radiating plate is electrically connected to the second heat radiating plate and the third heat radiating plate. A second semiconductor element in which an electrode formed on a surface facing the heat sink is electrically connected to the third heat sink;
A semiconductor device comprising a laminate including:
The stacked layer is formed so that the second main surface of the first member is exposed and the main surface of the pair of main surfaces of the third heat radiating plate opposite to the first and second semiconductor elements is exposed. A resin that embeds the body;
A plate-like second member including a second insulator fixed to the exposed surface of the third heat sink;
A semiconductor device comprising:

Images