JP4801133B2 - Semiconductor device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a semiconductor device in which a plurality of semiconductor elements are laminated, which is high in the degree of freedom of selection of a chip size of a semiconductor element and wiring connection and is excellent in reliability and high speed of signal transmission between semiconductor elements. <P>SOLUTION: The semiconductor device 100 includes a lower side semiconductor chip 104, an upper side semiconductor chip 106, and a silicon spacer 108 which is located between the lower side semiconductor chip 104 and upper side chip 106 and has an extensive portion extended in an outward direction more than the outer circumference of the upper side semiconductor chip 106. The silicon spacer 108 has through-electrodes 190a, 190b and rewiring 128a, 128b. <P>COPYRIGHT: (C)2009,JPO&amp;INPIT

Description

  The present invention relates to a semiconductor device.

  In recent years, for the purpose of high integration of semiconductor elements, three-dimensional mounting in which semiconductor elements such as LSIs are stacked in the vertical direction has been vigorously developed. There exists a thing of patent document 1 as this kind of technique. The semiconductor device described in this document is shown in FIGS. FIG. 10 is a cross-sectional view showing a multi-stage chip stack structure using a conventional flip chip type connection method. FIG. 11 is an enlarged view of a region surrounded by a one-dot chain in FIG.

  As shown in FIG. 10, in the semiconductor device 1001, a second semiconductor chip 1003 is mounted on a die pad 1004 formed of a metal such as nickel via an adhesive 1006 such as a resin. A semiconductor chip for control as the first semiconductor chip 1002 is further laminated on the semiconductor chip 1003. In addition, the first semiconductor chip 1002, the second semiconductor chip 1003, and the die pad 1004 are sealed by a resin package 1005 formed using a thermosetting resin such as an epoxy resin.

  As shown in FIG. 10, the first semiconductor chip 1002 has a predetermined circuit (not shown) and a wiring pattern for driving these circuits integrally formed on the main surface 1002a side. It further has a plurality of first terminal pads 1020a that are electrically connected to the circuit through this wiring pattern (not shown). The circuit and the wiring pattern are covered with an insulating film (not shown) so that the first terminal pads 1020a are exposed, and only the first terminal pads 1020a are the first. The semiconductor chip 1002 can be electrically connected to the outside.

  The second semiconductor chip 1003 has a larger planar view area than the first semiconductor chip 1002, and a predetermined circuit (similar to the first semiconductor chip 1002) is provided on the main surface 1003 a side. (Not shown) is integrated. Further, as shown in FIG. 10, a plurality of second terminal pads 1030a, 1031a, 1032a are further formed on the second semiconductor chip 1003, and these second terminal pads 1030a, 1031a, An insulating film (not shown) is formed so as to expose 1032a.

  As shown in FIG. 11, the second terminal pad 1031a is formed in the side region 1003b of the first semiconductor chip in a state where the first semiconductor chip 1002 is stacked on the second semiconductor chip 1003. It is electrically connected to the second terminal pad 1030a through the wiring portion 1033. The wiring portion 1033 is formed simultaneously with the terminal pads 1030a, 1031a, 1032a and the wiring pattern. Further, bumps 1031b are formed of gold or the like on the upper surfaces of the terminal pads 1031a, and signal terminal portions 1031 are formed by the signal terminal pads 1031a and the bumps 1031b.

  Here, the wiring portion 1033 is formed by plating with metal so as to connect the upper surfaces of the second terminal pad 1030a and the signal terminal pad 1031a, or by depositing metal. Is done. Of course, you may form by sticking the metal foil made into the predetermined shape.

  Further, the signal terminal portion 1031 and the second terminal portion 1032 of the second semiconductor chip 1003 are connected to the external connection terminal 1040 via wires W, respectively. The external connection terminal 1040 is used when the semiconductor device 1001 is mounted on a predetermined circuit board or the like. The internal lead 1041 as an internal terminal portion enclosed in the resin package 1005 and the internal lead are provided. The external lead 1042 as an external terminal portion is formed outside the resin package 1005 and is continuous with 1041.

  Here, in the technique using the anisotropic conductive resin 1007, the resin component 1070 is in a molten state, and the first terminal portion 1020 of the first semiconductor chip 1002 is replaced with the second terminal of the second semiconductor chip 1003. Lamination is performed by pressing the first semiconductor chip 1002 against the second semiconductor chip 1003 in correspondence with the portion 1030.

  At this time, since the resin component 1070 is in a molten state and the conductive balls 1071 are dispersed in the resin component 1070, the resin component 1070 between the first terminal portion 1020 and the second terminal portion 1030 is The conductive ball 1071 is interposed between the first terminal portion 1020 and the second terminal portion 1030 by being pushed away. The main surface 1002a of the first semiconductor chip and the main surface 1003a of the second semiconductor chip are mechanically connected by thermosetting the resin component 1070. In addition, since the conductive ball 1071 is interposed between the first terminal portion 1020 and the second terminal portion 1030, the first terminal portion 1020 and the second terminal portion 1030 are electrically connected. The

  According to this configuration, the first semiconductor chip 1002 and the second semiconductor chip 1003 are arranged to face each other such that the first terminal portion 1020 and the second terminal portion 1030 face each other. Then, a signal terminal portion 1031 is formed at a lateral position of the first semiconductor chip 1002 as an external terminal of the first semiconductor chip 1002. Further, it is described that the signal terminal portion 1031 can be used as a wire bonding part to be connected to the external connection terminal 1040 via the wire W.

  Moreover, there exists a thing of patent document 2 as this kind of technique. A semiconductor device described in this document is shown in FIG. FIG. 12 is a cross-sectional view showing a chip-on-chip structure using a conventional insulating film.

  This chip-on-chip structure includes a first semiconductor chip 411 and a second semiconductor chip 417. An insulating film 414 is sandwiched between the first semiconductor chip 411 and the second semiconductor chip 417. The insulating film 414 has a structure in which a wiring pattern 416 and a wiring pattern 420 are provided in the film 415.

  A connection portion 423 is provided on the lower surface of the wiring pattern 416. The connection part 423 is connected to the bump 413a of the first semiconductor chip provided on the surface 412 of the first semiconductor chip. A connection portion 424 is provided on the upper surface of the wiring pattern 416. The connection part 424 is connected to the bump 419a of the second semiconductor chip provided on the surface 418 of the second semiconductor chip.

  A connection portion 427 is provided on the lower surface of the wiring pattern 420. The connection part 427 is connected to the bump 413c of the first semiconductor chip provided on the surface 412 of the first semiconductor chip. A connection portion 426 is provided on the upper surface of the wiring pattern 420. The connection part 426 is connected to the bump 419b of the second semiconductor chip provided on the surface 418 of the second semiconductor chip.

  Patent Document 2 describes that according to this structure, it is possible to provide a chip-on-chip type semiconductor device having a high degree of freedom in design, in which semiconductor chips having different electrode arrangement pitches and arrangement positions can be overlapped and joined together. ing.

Japanese Patent Laid-Open No. 2000-22074 JP 2000-252408 A

  However, the prior art described in the above literature has room for improvement in the following points.

  First, in the semiconductor device 1001 described in Patent Document 1, the degree of freedom of the combination of chip sizes of the first semiconductor chip 1002 and the second semiconductor chip 1003 is limited. For example, when general-purpose semiconductor chips are used as the first semiconductor chip 1002 and the second semiconductor chip 1003, the chip size of the second semiconductor chip 1003 is set according to the chip size of the first semiconductor chip 1002. It is usually difficult to change freely.

  There may be a case where the chip size of the first semiconductor chip 1002 is larger than that of the second semiconductor chip 1003. In this case, it is difficult to form an electrode pad in a lateral region of the first semiconductor chip 1002, and it is difficult to externally connect the first semiconductor chip 1002.

  Further, the connection between the first semiconductor chip 1002 and the second semiconductor chip 1003 is made through the conductive ball 1071, and the resin component 1070 is mixed in the gap between the conductive balls 1071. For this reason, there is room for further improvement in terms of reliability and high speed of signal transmission between the first semiconductor chip 1002 and the second semiconductor chip 1003.

  Second, the chip-on-chip structure described in Patent Document 2 is difficult to apply to wire bonding because the insulating film 414 does not have rigidity. Further, since the insulating film 414 does not have rigidity, it has been difficult to produce the film with sufficient dimensional stability.

  The present invention has been made in view of the above circumstances, and an object thereof is a semiconductor device in which a plurality of semiconductor elements are stacked, and the degree of freedom of selection of the chip size of the semiconductor elements and wiring connection is improved. An object of the present invention is to provide a semiconductor device that is high in reliability and high speed in signal transmission between semiconductor elements.

  According to the present invention, the first semiconductor element, the second semiconductor element, and the first semiconductor element and the second semiconductor element are provided between the first semiconductor element, the second semiconductor element, and the outer edge of the second semiconductor element. The first semiconductor element has a first electrode pad on the surface on the plate-like body side, and the second semiconductor element is on the plate-like body side. The plate has a second electrode pad and a third electrode pad on the surface, the plate-like body connects the first electrode pad and the second electrode pad, and the second semiconductor in the overhanging portion There is provided a semiconductor device having a fourth electrode pad provided on a surface on the element side and a wiring connecting the third electrode pad and the fourth electrode pad.

  According to this configuration, since the through electrode for connecting the first electrode pad and the second electrode pad is provided on the plate-like body, the first semiconductor element and the second semiconductor element are connected to the through electrode. Therefore, the reliability and speed of signal transmission between semiconductor elements can be improved.

  Further, according to this configuration, since the wiring for connecting the third electrode pad and the fourth electrode pad is provided on the plate-like body, the chip size of the first semiconductor element and the second semiconductor element Regardless of the combination, the wiring connected to the third electrode pad on the surface of the second semiconductor element on the plate-like body side can be drawn out to the fourth electrode pad outward from the outer peripheral edge of the second semiconductor element. it can. Then, by connecting the fourth electrode pad to an arbitrary location, the third electrode pad can also be connected to an arbitrary location via the wiring and the fourth electrode pad. For this reason, the selection of the chip size of the semiconductor element and the freedom of wiring connection are improved.

  Further, in the present invention, since a plate-like body having rigidity is used, the plate-like body has rigidity when wiring is connected to the fourth electrode pad of the protruding portion of the plate-like body by a connection method such as wire bonding. Therefore, it can connect stably. For this reason, a highly flexible wiring connection can be formed stably.

  As mentioned above, although the structure of this invention was demonstrated, what combined these structures arbitrarily is effective as an aspect of this invention. Moreover, what converted the expression of this invention into the other category is also effective as an aspect of this invention.

  For example, the semiconductor device provided with the multi-stage chip stack structure provided by the present invention may be not only a two-step stack structure but also a two-step or more stack structure. That is, a three-stage laminated structure or a four-stage laminated structure may be used.

  In the present invention, the through electrode connecting the first electrode pad and the second electrode pad may be directly bonded to the first electrode pad and the second electrode pad. It is not limited to the configuration. For example, the electrodes may be electrically connected via electrode pads, bumps, wiring, other conductive members, or the like. In either case, in the present invention, the through electrode is connected to the first electrode pad and the second electrode pad.

  According to the present invention, a semiconductor device in which a plurality of semiconductor elements are stacked, the degree of freedom in selecting the chip size of the semiconductor elements and wiring connection is high, and the reliability and high speed of signal transmission between the semiconductor elements An excellent semiconductor device is provided.

  In the present invention, the through electrode may be bump-bonded to the first electrode pad and the second electrode pad.

  According to this configuration, since the distance for connecting the first semiconductor element and the second semiconductor element is shortened, the wiring resistance and the wiring capacity can be reduced. Further, the reliability of connection between the first semiconductor element and the second semiconductor element and the signal transmission speed can be improved. This enables high-speed data transfer with high reliability between semiconductor elements.

  In the present invention, the first electrode pad and the second electrode pad, and the through electrode that is bump-bonded to each other, have both end portions of the through electrode, the first electrode pad and the second electrode pad, Although a configuration may be adopted in which bonding is directly performed via conductive bumps made of a solder material or the like, the configuration is not particularly limited to this configuration. For example, another electrode pad provided at both end portions of the through electrode, the first electrode pad, and the second electrode pad may be joined via bumps. In any case, in the present invention, the through electrode is bump-connected to the first electrode pad and the second electrode pad.

  The plate-like body can be a plate-like spacer.

  According to this configuration, since the plate-like spacer having rigidity is provided, the fourth electrode pad on the protruding portion of the plate-like spacer can be stably connected to an arbitrary position by a connection method such as wire bonding. . Therefore, a semiconductor device having a high degree of freedom in wiring connection and excellent reliability can be obtained.

  The plate-like body can be a silicon spacer.

  According to this configuration, since the silicon spacer has rigidity, the second electrode pad of the protruding portion of the spacer can be connected to an arbitrary position by a connection method such as wire bonding. For this reason, a semiconductor device having a high degree of freedom in wiring connection and excellent in manufacturing stability and reliability can be obtained.

  In addition, when the plate-like body is a silicon spacer, the first semiconductor element and the second semiconductor element can be silicon semiconductor elements.

  According to this configuration, the linear expansion coefficient of the silicon spacer is approximately the same as the linear expansion coefficient of the first semiconductor element and the second semiconductor element, which are silicon-based semiconductor elements. The reliability of the semiconductor device is improved.

  The fourth electrode pad may be provided outside the outer peripheral edge of the first semiconductor element.

  According to this configuration, since the space around the fourth electrode pad is widened, the degree of freedom of connection to the fourth electrode pad is increased.

  The fourth electrode pad may be connected by wire bonding.

  According to this configuration, the fourth electrode pad can be connected to an arbitrary position by wire bonding, and the degree of freedom in selecting the connection destination of the fourth electrode pad is increased.

  The semiconductor device further includes a substrate, the first semiconductor element is provided on an upper portion of the substrate, and the second semiconductor element is provided on an upper portion of the first semiconductor element. be able to.

  According to this configuration, the fourth electrode pad provided on the surface on the second semiconductor element side in the overhanging portion of the plate-like body is provided on the upper surface of the overhanging portion of the plate-like body. become. For this reason, since the space above the fourth electrode pad is open, the degree of freedom of connection to the fourth electrode pad is further increased. For example, wire bonding can be suitably performed.

  The semiconductor device may further include a reinforcing material that supports the plate-like body at the protruding portion on the substrate.

  According to this configuration, since the strength of the overhanging portion of the plate-like body is improved, the fourth electrode pad of the overhanging portion of the plate-like body can be suitably connected to an arbitrary location by wire bonding or the like. For this reason, a semiconductor device having a high degree of freedom in wiring connection and excellent in manufacturing stability and reliability can be obtained.

  The semiconductor device may further include a reinforcing material that supports the plate-like body at the projecting portion on the first semiconductor element.

  According to this configuration, even when the protruding portion of the plate-like body is provided inside the peripheral edge portion of the first semiconductor element, it is possible to provide the reinforcing material that supports the plate-like body at the protruding portion. it can. As a result, since the strength of the protruding portion of the plate-like body is improved, the fourth electrode pad of the protruding portion of the plate-like body can be suitably connected to an arbitrary location by wire bonding or the like. For this reason, a semiconductor device having a high degree of freedom in wiring connection and excellent in manufacturing stability and reliability can be obtained.

  When the semiconductor device includes the substrate, a fifth electrode pad is provided on the upper surface of the substrate, and the fourth electrode pad is connected to the fifth electrode pad by wire bonding. It can be.

  According to this configuration, regardless of the chip size combination of the first semiconductor element and the second semiconductor element, the third electrode pad of the second semiconductor element is outside the outer peripheral edge of the second semiconductor element. It can be easily pulled out in the direction and connected to the fifth electrode pad of the substrate by wire bonding through the fourth electrode pad. Therefore, there is provided a semiconductor device having a multi-stage chip stacking structure that has a high degree of freedom in selection of a chip size and wiring connection of a semiconductor element, and is excellent in signal transmission reliability and high speed.

  The first semiconductor element may be a memory element, and the second semiconductor element may be a logic element.

  Even in this configuration, the third electrode pad of the second semiconductor element is connected to the second semiconductor element regardless of the combination of the memory element that has been recently reduced in size and the logic element that has been recently increased in size. It can be drawn out outward from the outer peripheral edge, and can be suitably connected to an arbitrary location by wire bonding or the like through the fourth electrode pad. Therefore, there is provided a semiconductor device having a multi-stage chip stacking structure that has a high degree of freedom in selection of a chip size and wiring connection of a semiconductor element, and is excellent in signal transmission reliability and high speed.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. In all the drawings, the same reference numerals are given to the same components, and the description will be omitted as appropriate.

<Embodiment 1>
FIG. 1 is a cross-sectional view showing the multi-stage chip stack structure of the first embodiment.

  The semiconductor device 100 includes a semiconductor chip 104 and a semiconductor chip 106 on a substrate 102. A silicon spacer 108 is sandwiched between the semiconductor chip 104 and the semiconductor chip 106.

  The semiconductor chip 104 and the semiconductor chip 106 are not particularly limited, but may be a silicon chip mainly made of the same material as the silicon spacer 108. The semiconductor chip 104 may be a memory element such as a DRAM, and the semiconductor chip 106 may be a logic element such as an ASIC.

  On the upper surface of the substrate 102, electrode pads 112a, 112b, 112c, and 112d are provided. Electrode pads 114 c and 114 d are provided on the upper surface of the semiconductor chip 104. Electrode pads 116 a, 116 b, 116 c and 116 d are provided on the lower surface of the semiconductor chip 106.

  The silicon spacer 108 is provided with through electrodes 190 a and 190 b that penetrate the silicon spacer 108. The silicon spacer 108 is provided with a rewiring 128 a and a rewiring 128 b having end portions inside and outside the outer peripheral edge of the semiconductor chip 106. On the upper surface of the silicon spacer 108, electrode pads 118a, 118b, 118c, 118d, 118e, 118f, 118g, and 118h are provided. Electrode pads 192c and 192d are provided on the lower surface of the silicon spacer.

  The electrode pad 118c is connected to the upper end portion of the through electrode 190a. The electrode pad 118d is connected to the upper end portion of the through electrode 190b. The electrode pad 192c is connected to the lower end of the through electrode 190a. The electrode pad 192d is connected to the lower end of the through electrode 190b.

  The electrode pad 118a is provided inside the outer peripheral edge of the semiconductor chip 106 and is connected to the inner end of the rewiring 128a. The electrode pad 118e is provided outside the outer peripheral edge of the semiconductor chip 106 and is connected to the outer end of the rewiring 128a. The electrode pad 118f is provided further outside than the electrode pad 118e.

  The electrode pad 118b is provided inside the outer peripheral edge of the semiconductor chip 106 and is connected to the inner end of the rewiring 128b. The electrode pad 118g is provided outside the outer peripheral edge of the semiconductor chip 106, and is connected to the outer end of the rewiring 128b. The electrode pad 118h is provided further outside than the electrode pad 118g.

  The electrode pads 114c and 114d of the lower semiconductor chip 104 are directly connected to the electrode pads 192c and 192d that are electrically connected to the through electrodes 190a and 190b of the silicon spacer 108, respectively.

  The electrode pads 116c and 116d of the upper semiconductor chip 106 are directly connected to the electrode pads 118c and 118d that are electrically connected to the through electrodes 190a and 190b of the silicon spacer 108, respectively.

  Here, the electrode pad 114c of the lower semiconductor chip 104, the electrode pad 192c of the silicon spacer 108, the through electrode 190a, the electrode pad 118c, and the electrode pad 116c of the semiconductor chip 106 are substantially collinear. Is provided.

  In the present invention, being provided on the same straight line does not mean being provided on the same straight line. That is, it means a state provided on the same straight line or in the vicinity of the same straight line and substantially on the same line. From another point of view, these members are provided at the same position or very close to each other when viewed in a plan view.

  The electrode pads 116a and 116b of the upper semiconductor chip 106 are directly connected to the electrode pads 118a and 118b that are electrically connected to the inner ends of the rewirings 128a and 128b of the silicon spacer 108, respectively.

  The electrode pads 118f and 118h of the silicon spacer 108 are connected to the electrode pads 112b and 112c provided on the upper surface of the substrate 102 by wires 120b and 120c, respectively.

  The electrode pads 118e and 118g of the silicon spacer 108 are connected to electrode pads 112a and 112d provided on the upper surface of the substrate 102 by wires 120a and 120d, respectively.

  Hereinafter, the function and effect of the configuration of the semiconductor device 100 of the present embodiment will be described.

  In the semiconductor device 100, a silicon spacer 108 having a size larger than that of the upper semiconductor chip 104 is sandwiched between the semiconductor chips 104 and 106 to be stacked.

  From another point of view, the semiconductor device 100 includes a silicon spacer 108 having a protruding portion that protrudes outward from the outer peripheral edge of the semiconductor chip 104 between the semiconductor chips 104 and 106 to be stacked.

  Therefore, on both side surfaces of the semiconductor device 100, the electrode pads 116a and 116b provided on the lower surface of the upper semiconductor chip 106 are pulled out to the electrode pads 118e and 118g provided on the protruding portions of the silicon spacer 108, respectively. be able to.

  Therefore, according to the structure of the semiconductor device 100, the upper semiconductor chip 106 and the substrate 102 are bonded by wire bonding the electrode pads 118e and 118g provided on the silicon spacer 108 and the electrode pads 112a and 112b of the substrate 102, respectively. Can be connected.

  In the semiconductor device 100, a silicon spacer 108 having a size smaller than that of the semiconductor chip 106 is used. For this reason, there is an advantage that the structure of the semiconductor device 100 is stabilized structurally. As will be described later, since the space above the peripheral portion of the semiconductor chip 106 is widened, it is possible to provide a wire bonding connection by providing a separate electrode pad on the peripheral portion of the semiconductor chip 106.

  However, a silicon spacer 108 having a size larger than that of the semiconductor chip 106 may be used. Also in this case, the semiconductor chip 104 and the semiconductor chip 106 can be connected through the through electrodes 190a and 190b provided in the silicon spacer 108. In addition, if a support member for supporting the silicon spacer 108 is provided as will be described later, the semiconductor device 100 is also stable in terms of structural mechanics.

  On the other hand, in the case of the conventional multi-stage chip stack structure shown in FIGS. 10 and 11 that does not use a spacer, it is necessary to draw out wiring by providing rewiring on the upper surface of the lower semiconductor chip. . At this time, in order to provide the rewiring on the upper surface of the lower semiconductor chip, for example, the upper surface of the lower semiconductor chip is plated with metal, or the metal is deposited or formed into a predetermined shape. In some cases, a rewiring layer is formed by attaching a foil.

  In the conventional structure, a general-purpose memory element or the like cannot be used as the lower semiconductor chip in this way, and it is necessary to use a customized memory element or the like subjected to special processing. In some cases, the design period becomes longer and the manufacturing cost also increases.

  On the other hand, in the semiconductor device 100 shown in FIG. 1, since the wiring can be drawn by the rewiring 128 provided in the silicon spacer 108, like the conventional multi-stage chip stacked structure shown in FIG. 10 or FIG. In order to enable wire bonding, it is not necessary to process the upper surface of the lower semiconductor chip and provide rewiring.

  Therefore, in the semiconductor device 100 shown in FIG. 1, it is not necessary to use a customized memory element or the like as the lower semiconductor chip 104, and a general-purpose memory element or the like can be used as it is. Therefore, the degree of freedom in designing the semiconductor device 1001 can be increased and the manufacturing cost can be reduced. In addition, since the general-purpose element can be used as it is, the development period of the semiconductor device 1001 can be shortened.

  On the other hand, in the case of the conventional multi-stage chip stack structure shown in FIG. 10 and FIG. 11 where the lower semiconductor chip is smaller than the upper semiconductor chip, for example, In some cases, it is difficult to provide an electrode pad on the side, and it is difficult to connect the electrode pad of the upper semiconductor chip to the outside.

  In contrast, in the semiconductor device 100 shown in FIG. 1, the upper semiconductor chip 106 may be larger or smaller than the lower semiconductor chip 104. Therefore, in the semiconductor device 100, the semiconductor chip 106 and the substrate 102 can be easily connected without depending on the combination of the upper and lower semiconductor chip sizes.

  Further, in the semiconductor device 100 shown in FIG. 1, the through electrodes 190a and 190b of the silicon spacer 108 are bumped via the electrode pads 116c and 116d of the upper semiconductor chip 106 and the electrode pads 118c and 118d of the silicon spacer 108, respectively. Connected. In FIG. 1, solder bumps between the electrode pads are not particularly illustrated. The same applies to the following drawings.

  The through electrodes 190a and 190b of the silicon spacer 108 are bump-connected to the electrode pads 114c and 114d of the lower semiconductor chip 104 via the electrode pads 192c and 192d of the silicon spacer 108, respectively.

  For this reason, the upper semiconductor chip 106 and the lower semiconductor chip 104 are connected with high conductivity through a substantially shortest path. As a result, the reliability and transmission speed of signal transmission between the upper semiconductor chip 106 and the lower semiconductor chip 104 are remarkably improved as compared with the case of wire bonding connection or the like.

  Here, the thickness of the silicon spacer 108 can be, for example, not less than 50 micrometers and not more than 100 micrometers. If the thickness is within this range, the silicon spacer 108 can be made thin while having sufficient rigidity, so that a thin multistage chip stack structure can be obtained.

  In particular, when a silicon spacer 108 having a thickness within this range is used, in a semiconductor device having a structure such as COC, MCP, and three-dimensional SiP, wire bonding connection to the outside and the semiconductor chip are independent of the combination of semiconductor chip sizes. A thin package in which semiconductor elements are highly integrated can be realized while high-speed signal transmission is possible.

  Furthermore, since the semiconductor device 100 has the lower semiconductor chip 104 and the upper semiconductor chip 106 connected in a face-down manner, the upper semiconductor chip 106 does not require wire bonding. For this reason, the height of the wire is reduced, the total thickness of the multi-stage chip stacked structure is further reduced, and a smaller package can be obtained. Further, the signal path length between the two chips is shortened, the signal transmission speed is improved, and the characteristics of the upper and lower semiconductor chips can be extracted more efficiently.

  On the other hand, in the conventional chip-on-chip structure shown in FIG. 12, since a flexible film insulator is used, even if an electrode pad is provided on the film insulator, it is difficult to connect by wire bonding.

  On the other hand, since the silicon spacer 108 is a plate-like body and has rigidity, it can be suitably wire-bonded to the electrode pads 118e, 118f, 118g, and 118h.

  Further, as in the semiconductor device 100 shown in FIG. 1, the silicon spacer 108 has a linear expansion coefficient of a lower semiconductor chip when the lower semiconductor chip 104 and the upper semiconductor chip 106 are both silicon chips. 104 and the upper semiconductor chip 106, the difference in linear expansion coefficient is small.

  For this reason, even if the temperature changes, the lower semiconductor chip 104, the silicon spacer 108, and the upper semiconductor chip 106 are compared with the chip-on-chip structure using the conventional film insulator shown in FIG. It is possible to suppress a decrease in contact property or peeling between electrode pads provided on the substrate.

  In the semiconductor device 100 shown in FIG. 1, the upper semiconductor chip 106 can be a circuit element such as an ASIC, and the lower semiconductor chip 104 can be a memory element such as a DRAM or an SRAM.

  Here, the size of a circuit element such as an ASIC is often defined by the width and pitch of each wiring integrally formed therein. For this reason, it can be said that the size of circuit elements such as ASICs tends to be reduced as the wiring pattern becomes finer. On the other hand, it can be said that memory elements such as DRAMs and SRAMs have a tendency to increase in size in the future because a large capacity is desired.

  When a semiconductor device is configured by stacking such a small circuit element and a large memory element, the small circuit element is arranged on the large memory element, and the terminal portions thereof are arranged opposite to each other so as to be flip-chip type. If the semiconductor device is configured as described above, the structural mechanical stability is improved.

  However, at this time, when the plate-shaped body having the above-described configuration is not used, the main surface of the upper circuit element (formation surface of the terminal portion) faces downward, and the main memory element of the lower memory element The main surface of the circuit element is hidden by the surface (formation surface of the terminal portion). For this reason, it becomes difficult to form a terminal portion as a wire bonding portion on the main surface of the upper circuit element.

  As long as it is a terminal portion for exchanging electric charges (electricity or signals) between the upper circuit element and the lower memory element, it may face downward as long as an electrical conduction state can be secured. It is preferable that at least the upper surface of the terminal portion used as the bonding site is exposed. For example, a signal terminal portion for a circuit element such as an upper ASIC to directly send and receive signals to the outside is often performed via a mounting terminal and a wire. It is desirable that the upper surface of the terminal portion of the circuit element is exposed.

  For this reason, it may be possible to form a wire bonding portion on the upper surface (back surface) of a circuit element such as an ASIC, but it is technically difficult to form a wiring pattern (terminal portion) on both surfaces of the circuit element such as an ASIC. is there.

  Similarly, it is also conceivable that wiring is drawn out from the terminal portion of the main surface of the circuit element by providing rewiring on the upper surface (main surface) of a memory element such as DRAM or SRAM. At this time, in order to provide the rewiring on the upper surface of the lower semiconductor chip, for example, the upper surface of the lower semiconductor chip is plated with metal, or the metal is deposited or formed into a predetermined shape. It is conceivable to form a rewiring layer by sticking a foil or the like.

  However, in this case, a general-purpose memory element cannot be used as the lower semiconductor chip, and a customized memory element subjected to special processing such as metal deposition must be used. In some cases, the lead time for designing and manufacturing a semiconductor device is prolonged and the manufacturing cost is increased.

  On the other hand, in the semiconductor device 100 shown in FIG. 1, it is not necessary to use a customized memory element or the like as the lower semiconductor chip 104, and a general-purpose memory element or the like can be used as it is. Therefore, the degree of freedom in designing the semiconductor device 1001 can be increased and the manufacturing cost can be reduced. In addition, since the general-purpose element can be used as it is, the development period and manufacturing lead time of the semiconductor device 1001 can be shortened.

  In addition, circuit elements such as the upper ASIC can directly send and receive signals to and from the outside by wire bonding connection, and high-speed data transmission and reception by the flip chip connection to the lower memory element, The degree of freedom in designing the semiconductor device 100 is improved, and the reliability and high speed of operation of the semiconductor device 100 are improved.

  FIG. 9 is a cross-sectional view when the ball grid array structure is applied to the semiconductor device having the multi-stage chip stack structure of the first embodiment.

  When the ball grid array structure is applied to the multi-layer stacked structure of this embodiment, the electrode pad 136 is provided on the back side of the substrate 102 and the solder ball 138 is provided on the electrode pad 136. The electrode pad 136 can be connected to the electrode pads 112a, 112b, 112c, and 112d in FIG.

  Further, the entire multi-layer laminated structure is sealed with a sealing resin layer 132. A gap between the lower semiconductor chip 104 and the silicon spacer 108 in FIG. 1 is sealed with an underfill resin layer 134a. The gap between the upper semiconductor chip 106 and the silicon spacer 108 is sealed with an underfill resin layer 134b.

  2 to 3 are process cross-sectional views illustrating the manufacturing method of the multistage chip stacked structure according to the first embodiment.

  In order to obtain the multi-stage chip stack structure shown in the first embodiment, first, electrode pads 112a, 112b, 112c, and 112d are formed on the substrate 102 as shown in FIG.

  Next, as shown in FIG. 2B, the lower semiconductor chip 104 is stacked on the upper surface of the substrate 102. On the upper surface (main surface) of the semiconductor chip 104, electrode pads 114c and 114d are provided in advance.

  Subsequently, as shown in FIG. 2C, a silicon spacer 108 is stacked on the semiconductor chip 104. In the silicon spacer 108, through electrodes 190a and 190b and rewirings 128a and 128b are formed by a method described later.

  At this time, the electrode pads 114c and 114d on the upper surface of the lower semiconductor chip 104 and the electrode pads 192c and 192d provided at the lower ends of the through electrodes 190a and 190b of the silicon spacer 108 are bump-connected, respectively.

  In addition, electrode pads 118a, 118b, 118c, 118d, 118e, 118f, 118g, and 118h are provided on the upper surface of the silicon spacer 108 in advance. The electrode pads 118c and 118d are provided at the upper ends of the through electrodes 190a and 190b, respectively. The electrode pads 118a and 118b are provided at the inner ends of the rewirings 128a and 128b, respectively. The electrode pads 118e and 118g are provided at the outer ends of the rewirings 128a and 128b, respectively. The electrode pads 118f and 118h are provided outside the electrode pads 118e and 118g, respectively.

  Thereafter, as shown in FIG. 3D, the upper semiconductor chip 106 is stacked on the silicon spacer 108. On the lower surface (main surface) of the upper semiconductor chip 106, electrode pads 116a, 116b, 116c, and 116d are provided in advance.

  At this time, the electrode pads 116a, 116b, 116c, and 116d on the lower surface of the upper semiconductor chip 106 are bump-connected to the electrode pads 118a, 118b, 118c, and 118d on the silicon spacer 108, respectively.

  Next, wire bonding is performed as shown in FIG. Specifically, the electrode pads 118e, 118f, 118g, and 118h on the upper (outer) surface of the silicon spacer 108 are connected to the electrode pads 112a, 112b, 112d, and 112c on the substrate 102 by wire bonding, respectively.

  According to this method, since the through electrodes 190a and 190b are provided in the silicon spacer 108, the lower semiconductor chip 104 and the upper semiconductor chip 106 are connected to each other through the through electrodes 190a and 190b with a short path with good conductivity. A stable connection is possible. Therefore, the reliability and transmission speed of signal transmission between the lower semiconductor chip 104 and the upper semiconductor chip 106 can be stably improved.

  Further, according to this method, since the rewiring 128a, 128b for connecting the electrode pads 118a, 118b and the electrode pads 118e, 118g to the silicon spacer 108 is provided, the lower semiconductor chip 104 and the upper semiconductor Regardless of the chip size combination with the chip 106, the rewiring 128a, 128b connected to the electrode pads 116a, 116b on the surface of the upper semiconductor chip 106 on the silicon spacer 108 side through the electrode pads 118a, 118b is connected to the upper semiconductor. The electrode pads 118e and 118g can be pulled out outward from the outer peripheral edge of the chip 106.

  Then, by connecting the electrode pads 118e and 118g to the electrode pads 112a and 112d on the upper surface of the substrate 102 by the wires 120a and 120d, the electrode pads 116a and 116b are also connected to the electrode pads 118a and 118b, the rewiring 128a and 128b, and the electrode pads, respectively. It can be connected to electrode pads 112a and 112d on the upper surface of the substrate 102 via 118e and 118g. For this reason, the freedom degree of selection of the chip size of the lower semiconductor chip 104 and the upper semiconductor chip 106 and wiring connection by the wires 120a and 120d is improved.

  Further, according to this configuration, when the wiring connection is made to the electrode pads 118e and 118g of the overhanging portion of the silicon spacer 108 by a connection method such as wire bonding, the silicon spacer 108 has a rigidity, so that the connection is stable. Can do. For this reason, wiring connection by wire bonding etc. with a high degree of freedom can be formed stably.

  FIG. 8 is a process cross-sectional view illustrating the method for manufacturing the through electrode according to the first embodiment. 8 indicates the upper side of the silicon spacer 108 (the upper side in FIG. 1).

  In order to provide a through electrode in the silicon spacer shown in Embodiment Mode 1, first, a resist film (not shown) is provided on the silicon spacer 1108, and the silicon spacer 1108 is selectively etched using the resist film as a mask. A recess is formed on the upper surface of the substrate. Next, as shown in FIG. 8A, conductive members 1128c and 1128d are embedded in the recesses.

  These conductive members may be members made of a metal material containing, for example, Al or Cu. Further, these conductive members can be formed by, for example, a plating method. In addition, these conductive members may have a barrier metal film made of TiN or the like on the bottom and side surfaces, for example.

  Next, as illustrated in FIG. 8B, resist films 1132 a, 1132 b, and 1132 c are formed on the back surface of the silicon spacer 1108. These resist films are formed so as to cover the upper surfaces of the conductive members 1128c and 1128d and regions adjacent to the upper surfaces. Further, these resist films are formed so as to have an opening at a site where a rewiring is to be formed. Subsequently, conductive members 1138a and 1138b made of a metal material containing, for example, Al or Cu are formed in these openings by, for example, sputtering.

  Subsequently, as illustrated in FIG. 8C, the resist films 1132 a, 1132 b, and 1132 c are peeled from the back surface of the silicon spacer 1108. Then, the silicon spacer 1108 is polished (background) from the lower surface side to reduce the thickness of the silicon spacer 1108 to about 50 to 100 micrometers.

  As a result, end portions on the lower surface side of the conductive members 1128c and 1128d are exposed, and a through electrode is formed. After polishing, the lower surface side of the silicon spacer 1108 is finish-polished to obtain a silicon spacer 1108 having through electrodes and rewiring.

  In the first embodiment, by using the silicon spacer 1108 having such a through electrode and rewiring as the silicon spacer, the electrode pad on the lower surface of the upper semiconductor chip can be connected to the upper side of the silicon spacer (via the rewiring ( Can be pulled out to the electrode pad on the outer side. In addition, the electrode pad on the upper surface of the lower semiconductor chip and the electrode pad on the lower surface of the upper semiconductor chip can be flip-chip connected via a through electrode.

  Therefore, when the silicon spacer having such a configuration is used, as shown in FIG. 1, the semiconductor device 100 is formed by stacking the lower semiconductor chip 104 and the upper semiconductor chip 106, and the lower semiconductor chip The semiconductor device 100 has a high degree of freedom in selecting the chip size and wiring connection of the chip 104 and the upper semiconductor chip 106, and is excellent in signal transmission reliability and high speed between the lower semiconductor chip 104 and the upper semiconductor chip 106. Is provided.

<Embodiment 2>
FIG. 4 is a cross-sectional view showing the multi-stage chip stack structure of the second embodiment.

  The multistage chip stack structure of the present embodiment has the same configuration as that of the first embodiment, but through electrodes 190c and 190d are formed in the silicon spacer 108.

  Electrode pads 118f and 118h are provided at the upper ends of the through electrodes 190c and 190d, respectively. In addition, electrode pads 192a and 192b are provided at the lower ends of the through electrodes 190c and 190d, respectively.

  The electrode pads 114a and 114b provided on the upper surface of the lower semiconductor chip 104 are bump-connected to the electrode pads 192a and 192b on the lower surface of the silicon spacer 108, respectively.

  Further, the electrode pads 118f and 118h on the upper surface of the silicon spacer 108 are connected to the electrode pads 112b and 112c provided on the upper surface of the substrate 102 by wire bonding via wires 120b and 120c, respectively.

  According to this configuration, the electrode pads 114a and 114b provided on the upper surface of the lower semiconductor chip 104 are made into silicon by the electrode pads 192a and 192b and the through electrodes 190c and 190d provided on the silicon spacer 108, respectively. It can be pulled out to the electrode pads 118 f and 118 h provided on the upper side (outside) of the spacer 108.

  At this time, the silicon spacer 108 may be larger or smaller in plan view than the lower semiconductor chip 104. That is, the silicon spacer 108 may be provided so as to narrow the space above the outer peripheral portion of the lower semiconductor chip 104.

  Therefore, wire bonding to the lower semiconductor chip 104 can be performed through the electrode pads 118f and 118h on the silicon spacer 108 without depending on the size combination of the lower semiconductor chip 104 and the silicon spacer 108. . Therefore, when the silicon spacer 108 having such a configuration is used, wire bonding connection is made to the lower semiconductor chip 104 without depending on the combination of the chip sizes of the lower semiconductor chip 104 and the upper semiconductor chip 106. be able to.

  In particular, in a semiconductor device having a structure such as COC, MCP, and three-dimensional SiP, wire bonding can be performed without depending on the combination of semiconductor chip sizes, so that a thin package in which semiconductor elements are highly integrated can be realized.

  Further, according to this configuration, the electrode pads 114 a and 114 b are provided on the upper surface of the outer peripheral portion of the semiconductor chip 104, and the electrode pads 192 a and 192 b are provided on the lower surface of the protruding portion of the silicon spacer 108. Therefore, the electrode pads 114a and 114b and the electrode pads 192a and 192b also function as support members for the silicon spacer 108. Therefore, wire bonding to the electrode pads 118e, 118f, 118g, and 118h on the upper surface of the silicon spacer 108 can be more stably performed.

  Therefore, according to this configuration, the semiconductor device 200 is formed by stacking the lower semiconductor chip 104 and the upper semiconductor chip 106, and the selection of the chip size of the lower semiconductor chip 104 and the upper semiconductor chip 106 and A semiconductor device 200 having a high degree of freedom in wiring connection and excellent in signal transmission reliability and high speed between the lower semiconductor chip 104 and the upper semiconductor chip 106 is provided.

<Embodiment 3>
FIG. 5 is a cross-sectional view showing the multi-stage chip stack structure of the third embodiment.

  The semiconductor device 300 having the multi-stage chip stack structure of the present embodiment has the same configuration as the multi-stage chip stack structure of the first embodiment, but electrode pads 114a and 114b are provided on the upper surface of the outer peripheral portion of the semiconductor chip 104. Structure.

  Further, the electrode pads 114a and 114b on the upper surface of the semiconductor chip 104 are connected to the electrode pads 112b and 112c provided on the upper surface of the substrate 102 by wire bonding via wires 120b and 120c, respectively.

  At this time, the silicon spacer 108 is preferably smaller in plan view than the lower semiconductor chip 104. That is, it is preferable that the silicon spacer 108 is not provided in the upper part of the outer peripheral portion of the lower semiconductor chip 104.

  According to this configuration, wire bonding to the lower semiconductor chip 104 can be performed directly from the electrode pads 112 b and 112 c on the upper surface of the substrate 102. Therefore, in the semiconductor device 300 having such a configuration, the degree of freedom of wiring connection to the lower semiconductor chip 104 can be further improved.

  In particular, in a semiconductor device having a structure such as COC, MCP, or three-dimensional SiP, it is possible to improve the degree of freedom of wiring connection by wire bonding, so that a thin package in which semiconductor elements are highly integrated can be freely designed. The improvement of the degree can be realized.

  Therefore, according to this configuration, the semiconductor device 300 is formed by stacking the lower semiconductor chip 104 and the upper semiconductor chip 106, and the selection of the chip size of the lower semiconductor chip 104 and the upper semiconductor chip 106 and A semiconductor device 300 having a high degree of freedom in wiring connection and excellent in signal transmission reliability and high speed between the lower semiconductor chip 104 and the upper semiconductor chip 106 is provided.

<Embodiment 4>
FIG. 6 is a cross-sectional view illustrating the multi-stage chip stack structure of the fourth embodiment.

  The semiconductor device 400 having the multi-stage chip stack structure of the present embodiment has the same configuration as that of the multi-stage chip stack structure of the first embodiment, except that the silicon spacer 108 is interposed between the lower semiconductor chip 104 and the silicon spacer 108. It is a structure provided with a reinforcing material for supporting.

  The reinforcing material provided between the lower semiconductor chip 104 and the silicon spacer 108 includes dummy bumps 194a and 194b provided on the upper surface of the outer peripheral portion of the lower semiconductor chip 104, and the protruding portion of the silicon spacer 108, respectively. And dummy bumps 196a and 196b provided on the lower surface of the substrate.

  At this time, the silicon spacer 108 may be larger or smaller in plan view than the lower semiconductor chip 104. That is, the silicon spacer 108 may be provided so as to narrow the space above the outer peripheral portion of the lower semiconductor chip 104.

  The dummy bumps 194a, 194b, 196a, and 196b may be made of a conductive material similar to a normal electrode pad, but may be made of a non-electric material. This is because the dummy bumps 194a, 194b, 196a, 196b do not need to function as electrode pads.

  As in this configuration, the reinforcing material that supports the protruding portion of the silicon spacer 108 is formed by the dummy bumps 194a, 194b, 196a, 196b, so that the electrode pads 118e, 118f, 118g, 118h on the silicon spacer 108 and the substrate are formed. Wire bonding can be more stably performed when the electrode pads 112a, 112b, 112d, and 112c on the upper surface of the wire 102 are connected by the wires 120a, 120b, 120d, and 120c, respectively.

  Therefore, according to this configuration, the semiconductor device 400 is formed by stacking the lower semiconductor chip 104 and the upper semiconductor chip 106, and the selection of the chip size of the lower semiconductor chip 104 and the upper semiconductor chip 106 and A semiconductor device 400 having a high degree of freedom in wiring connection and excellent in signal transmission reliability and high speed between the lower semiconductor chip 104 and the upper semiconductor chip 106 is provided.

<Embodiment 5>
FIG. 7 is a cross-sectional view illustrating the multi-stage chip stack structure of the fifth embodiment.

  The semiconductor device 500 including the multi-stage chip stack structure of the present embodiment has the same configuration as that of the multi-stage chip stack structure of the first embodiment, but for supporting the silicon spacer 108 between the substrate 102 and the silicon spacer 108. In this structure, the reinforcing members 130a and 130b are provided. The reinforcing members 130a and 130b can be formed by injecting resin with a dispenser or the like and curing the resin.

  At this time, the silicon spacer 108 is preferably larger in plan view than the lower semiconductor chip 104. That is, it is desirable that the silicon spacer 108 has a portion protruding outward from the outer peripheral portion of the lower semiconductor chip 104.

  According to this configuration, the electrode pads 118e, 118f, 118g, and 118h on the silicon spacer 108 and the electrode pads 112a, 112b, 112d, and 112c on the upper surface of the substrate 102 are respectively connected by the wires 120a, 120b, 120d, and 120c. When connecting, wire bonding can be more stably performed.

  Therefore, according to this configuration, the semiconductor device 500 is formed by stacking the lower semiconductor chip 104 and the upper semiconductor chip 106, and the selection of the chip size of the lower semiconductor chip 104 and the upper semiconductor chip 106 and A semiconductor device 500 having a high degree of freedom in wiring connection and excellent in signal transmission reliability and high speed between the lower semiconductor chip 104 and the upper semiconductor chip 106 is provided.

  As mentioned above, although the structure of this invention was demonstrated, what combined these structures arbitrarily is effective as an aspect of this invention. Moreover, what converted the expression of this invention into the other category is also effective as an aspect of this invention.

  For example, in the above embodiment, a spacer made of a single plate is used as the plate-like spacer, but there is no particular limitation. For example, a spacer in which two plates are joined together may be used, or a plurality of plate-like spacers arranged at regular intervals may be used.

  Even in such a configuration, the electrode pad on the lower surface of the upper semiconductor chip 106 can be drawn out from the outer peripheral edge of the upper semiconductor chip 106 by the wiring provided in the spacer.

  In the above-described embodiment, a silicon spacer is used as the plate-like spacer, but there is no particular limitation. For example, a plate-like spacer made of another semiconductor may be used, or a plate-like spacer made of a resin composition may be used.

  Any spacer made of any material can be used as long as it is a plate-shaped spacer with a certain rigidity, even when the outer end of the drawn wiring is connected to an arbitrary location by wire bonding. The occurrence of tilt is suppressed.

  However, compared to the insulating film used in the conventional chip-on-chip structure shown in FIG. 12, a plate-like spacer having higher rigidity can perform wire bonding more favorably, and manufacture of a multi-stage chip laminated structure. Increased stability.

  In the above embodiment, the connection by wire bonding is used for the connection between the electrode pad on the upper surface of the plate-like spacer and the electrode pad on the upper surface of the substrate, but there is no particular limitation. For example, the wiring provided on the lower surface of the plate-like spacer connected by the electrode pad on the upper surface of the plate-like spacer and the through electrode may be directly connected to the electrode pad on the substrate.

  Even in such a configuration, the electrode pad on the lower surface of the upper semiconductor chip 106 is drawn out from the outer peripheral edge of the upper semiconductor chip 106 by the wiring provided in the spacer and connected to the electrode pad on the substrate. can do.

1 is a cross-sectional view showing a multistage chip stack structure of Embodiment 1. FIG. FIG. 5 is a process cross-sectional view illustrating the manufacturing method of the multistage chip stack structure of Embodiment 1. FIG. 5 is a process cross-sectional view illustrating the manufacturing method of the multistage chip stack structure of Embodiment 1. FIG. 5 is a cross-sectional view showing a multistage chip stack structure of Embodiment 2. It is sectional drawing which shows the multistage chip laminated structure of Embodiment 3. It is process sectional drawing which shows the multistage chip laminated structure of Embodiment 4. FIG. 10 is a process cross-sectional view illustrating a multistage chip stack structure of Embodiment 5. FIG. 5 is a process cross-sectional view illustrating the method for manufacturing the through electrode according to the first embodiment. FIG. 3 is a cross-sectional view when a ball grid array structure is applied to the semiconductor device having the multi-stage chip stack structure of Embodiment 1. It is sectional drawing which shows the multistage chip | tip laminated structure using the connection method of the conventional flip chip format. It is an expanded sectional view which shows a part of multistage chip | tip laminated structure using the connection method of the conventional flip chip format. It is sectional drawing which shows the chip-on-chip structure using the conventional insulating film.

Explanation of symbols

DESCRIPTION OF SYMBOLS 100 Semiconductor device 102 Substrate 104 Semiconductor chip 106 Semiconductor chip 108 Silicon spacer 112 Electrode pad 114 Electrode pad 116 Electrode pad 118 Electrode pad 120 Wire 128 Rewiring 130 Reinforcement material 132 Sealing resin layer 134 Underfill resin layer 136 Electrode pad 138 Solder ball 142 Rewiring layer 190 Through-electrode 192 Electrode pad 194 Dummy bump 196 Dummy bump 200 Semiconductor device 300 Semiconductor device 400 Semiconductor device 411 First semiconductor chip 412 First semiconductor chip surface 413 First semiconductor chip bump 414 Insulating film 415 Film 416 wiring pattern 417 second semiconductor chip 418 surface of second semiconductor chip 419 bump 420 of second semiconductor chip wiring pattern 423 connection portion 24 connecting portion 425 connecting portion 426 connecting portion 427 connecting portion 500 semiconductor device 1001 semiconductor device 1002 first semiconductor chip 1002a main surface 1003 second semiconductor chip 1003b lateral region 1004 die pad 1005 resin package 1006 adhesive 1007 anisotropic conductive Resin 1020 First terminal portion 1020a First terminal pad 1020b Bump 1030 Second terminal portion 1030a Second terminal pad 1030b Bump 1031 Signal terminal portion 1031a Signal terminal pad 1031b Bump 1032 Second terminal portion 1032a Second terminal pad 1032b Bump 1033 Wiring portion 1040 External connection terminal 1041 Internal lead 1042 External lead 1070 Resin component 1071 Conductive ball W Wire 1100 Semiconductor device 1102 Substrate 110 The semiconductor chip 1106 semiconductor chip 1108 silicon spacer 1112 electrode pad 1116 electrode pad 1120 wire 1128 conductive member 1132 resist film 1138 electrically conductive member

Claims (11)

A first semiconductor element;
A second semiconductor element;
A plate-like body that is provided between the first semiconductor element and the second semiconductor element and has a protruding portion that protrudes outward from the outer peripheral edge of the second semiconductor element;
The first semiconductor element has a first electrode pad on the surface on the plate-like body side,
The second semiconductor element has a second electrode pad and a third electrode pad on the surface on the plate-like body side,
The chip size of the first semiconductor element is larger than the chip size of the second semiconductor element,
The first semiconductor element is a memory element, and the second semiconductor element is a logic element;
The plate-like body is
A through electrode connecting the first electrode pad and the second electrode pad;
A fourth electrode pad provided on the surface of the projecting portion on the second semiconductor element side;
The third electrode pad and the fourth electrode pad are connected without passing through any electrode member on the first semiconductor element, and only through the second semiconductor element. And a wiring connected to the semiconductor element . The semiconductor device according to claim 1,
The semiconductor device, wherein the through electrode is bump-bonded to the first electrode pad and the second electrode pad. The semiconductor device according to claim 1 or 2,
The semiconductor device, wherein the plate-like body is a plate-like spacer. The semiconductor device according to claim 1,
The plate-like body is a silicon spacer. The semiconductor device according to claim 4,
The semiconductor device, wherein the first semiconductor element and the second semiconductor element are silicon-based semiconductor elements. The semiconductor device according to claim 1,
The semiconductor device, wherein the fourth electrode pad is provided outside an outer peripheral edge of the first semiconductor element. The semiconductor device according to claim 1,
The semiconductor device, wherein the fourth electrode pad is connected by wire bonding. The semiconductor device according to claim 1,
Further comprising a substrate,
The first semiconductor element is provided on an upper portion of the substrate,
The semiconductor device, wherein the second semiconductor element is provided on an upper part of the first semiconductor element. The semiconductor device according to claim 8,
A semiconductor device, further comprising: a reinforcing material that supports the plate-like body at the projecting portion on the substrate. The semiconductor device according to claim 8 or 9,
A semiconductor device, further comprising a reinforcing material that supports the plate-like body at the projecting portion on the first semiconductor element. The semiconductor device according to claim 8,
A fifth electrode pad is provided on the upper surface of the substrate;
The semiconductor device according to claim 4, wherein the fourth electrode pad is connected to the fifth electrode pad by wire bonding.

Images