Semiconductor device having a primary chip with bumps in joined registration with bumps of a plurality of secondary chips

[0002] 1. Field of the Invention

[0003] The present invention relates generally to a semiconductor device, and more particularly, to a semiconductor device having a structure in which a plurality of semiconductor chips are bonded to a surface of a solid (for example, a semiconductor chip or a wiring board) and a method of assembling the same.

[0004] 2. Description of Related Art

[0005] In order to increase a substantial integration degree of a semiconductor device, a semiconductor device having a chip-on-chip structure in which semiconductor chips in two layers are overlapped with each other has been paid attention to.

[0006] Specifically, a pad opening where internal wiring is partially exposed is provided on a device formation surface of each of semiconductor chips to be overlapped, for example. A projection electrode called a “bump” is provided in the pad opening. The semiconductor chips are overlapped with each other by so-called face-to-face bonding.

[0007] For example, one of the semiconductor chips is larger than the other semiconductor chip, and is taken as a mother chip or primary chip having an electrode for external connection. That is, an electrode is provided on a device formation surface of the primary chip, for example. The electrode is electrically connected to an electrode on an underlying substrate (a wiring board), and an external connecting electrode is provided on the underlying substrate. When the semiconductor device is mounted, the external connecting electrode on the underlying substrate is soldered to a printed board or a ceramic board.

[0008] There may, in some cases, be a plurality of semiconductor chips (daughter chip or secondary chips) to be overlapped with the primary chip. In this case, the secondary chips must be accurately aligned with and joined to the primary chip. The secondary chips are generally mounted on the primary chip one at a time. Accordingly, a mounting operation is repeatedly performed, and the number of times the operation is performed corresponds to the number of secondary chips. Therefore, a long time period is required for assembling. Since there are a plurality of secondary chips, the probability that the secondary chips are erroneously bonded or mounted (an erroneous mounting rate) is considerably high.

[0009] A first object of the present invention is to provide, in a semiconductor device having a structure in which a plurality of semiconductor chips are joined to a surface of a solid, a semiconductor device having a structure in which the production efficiency thereof can be improved.

[0010] A second object of the present invention is to provide, in a semiconductor device having a structure in which a plurality of semiconductor chips are joined to a surface of a solid, a semiconductor device having a structure in which the semiconductor chips can be prevented from being erroneously mounted.

[0011] A third object of the present invention is to provide a method of assembling a semiconductor device, in which a plurality of semiconductor chips can be efficiently joined to a surface of a solid and therefore, the productivity thereof can be improved.

[0012] A fourth object of the present invention is to provide a method of assembling a semiconductor device, in which a plurality of semiconductor chips can be accurately mounted on a surface of a solid, and the semiconductor chips can be prevented from being erroneously mounted.

[0013] A semiconductor device according to the present invention comprises a solid; a plurality of semiconductor chips respectively joined to predetermined positions of a surface of the solid; and a frame holding the plurality of semiconductor chips in a relative positional relationship corresponding to joint positions on the surface of the solid.

[0014] The solid may be another semiconductor chip or a wiring board.

[0015] The plurality of semiconductor chips may be bonded to a surface, opposite to the surface of the solid, of the frame.

[0016] The plurality of semiconductor chips may be respectively fitted in through holes formed in the frame.

[0017] The plurality of semiconductor chips and the frame may be together sealed with resin.

[0018] The frame may be also used as a lead frame. In this case, the solid and the frame may be electrically connected to each other by suitable means such as wire bonding.

[0019] According to the present invention, the plurality of semiconductor chips which are overlapped with and joined to the surface of the solid can be arranged in predetermined positions of the frame and can be overlapped with one another utilizing the frame. Consequently, the plurality of semiconductor chips can be overlapped with and joined to the surface of the solid by one mounting operation. If the plurality of semiconductor chips are arranged on the frame without error, erroneous mounting in the step of overlapping the frame with the surface of the solid is not a problem. Consequently, it is possible to realize a semiconductor device which is superior in production efficiency and in which semiconductor chips are hardly erroneously mounted.

[0020] A method of assembling a semiconductor device according to the present invention comprises the steps of holding a plurality of semiconductor chips in a frame in a relative positional relationship corresponding to joint positions on a surface of a solid; aligning and overlapping the frame with the surface of the solid; and joining the plurality of semiconductor chips held in the frame to the surface of the solid.

[0021] The frame may be removed from the plurality of semiconductor chips after the plurality of semiconductor chips are joined to the surface of the solid.

[0022] The foregoing and other objects, features, aspects and advantages of the present invention will become more apparent from the following detailed description of the present invention when taken in conjunction with the accompanying drawings.

[0023] FIG. 1 is a side view showing are embodiment of the present invention, in which a plurality of secondary chips mounted on a frame are overlapped with a primary chip by so-called face-down bonding;

[0024] FIG. 2 is a side view showing a state where a plurality of secondary chips are fixed to a frame having holes formed therein, and are overlapped with a primary chip; and

[0025] FIGS. 3A and 3B are side views showing another embodiment of the present invention, in which secondary chips are mounted, and a frame is separated from the secondary chips.

[0026] In embodiments of the present invention described below, it is presupposed that Si is used as the type of semiconductor. However, another semiconductor such as GaAs or Ge may be used.

[0027] FIG. 1 is a side view showing a state where a plurality of secondary chips 2aand 2bare overlapped with a primary chip 1 by so-called face-down bonding. Bumps 3 are provided on a device formation surface of the primary chip 1, and bumps 4 are also provided in corresponding positions on device formation surfaces of the secondary chips 2aand 2b. The corresponding bumps 3 and 4 are joined to each other, thereby having assembled a semiconductor device having a so-called chip-on-chip structure.

[0028] The secondary chips 2aand 2bare in a state where they are mounted on one common frame 5. A material for the frame 5 is not particularly limited. Examples are a resin plate, a glass plate, a metal plate, a ceramic plate, and the like, which are transparent or opaque. Further, the thickness of the frame 5 is not particularly limited.

[0029] The frame 5 and the secondary chips 2aand 2bare bonded to each other with adhesives. The secondary chips 2aand 2bare mounted on the frame 5 in a relative positional relationship corresponding to joint positions on the primary chip 1. If the frame 5 and the primary chip 1 are aligned with each other, therefore, alignment between the plurality of secondary chips 2aand 2band the primary chip 1 is collectively achieved.

[0030] A conventionally known method can be employed as a method of aligning the frame 5 and the primary chip 1. For example, a technique for aligning them by forming a through hole and seeing through the through hole a mark below the through hole (see Japanese Unexamined Patent Publication (KOKAI) No. 8-172111) may be applied. That is, they can be aligned by forming a through hole in the frame 5 and detecting a mark provided on the primary chip 1 through the through hole. Alternatively, a mark may be formed at an end of an upper surface of the frame 5 (an upper surface in FIG. 1), also forming a mark on an upper surface of the primary chip 1, and confirming a positional relationship between both the marks as viewed from above (see Japanese Unexamined Utility Model Publication (KOKAI) No. 5-87949). Further, if the frame 5 is transparent, they can be aligned by forming marks on a central region (a position avoiding the secondary chips 2aand 2b) of the frame 5 and an upper surface of the primary chip 1 and confirming a positional relationship between both the marks as viewed from above.

[0031] The frame 5 on which the secondary chips 2aand 2bare mounted is thus overlapped with the primary chip 1, thereby making it possible to align and join all the secondary chips 2aand 2bwith and to the primary chip 1 in one step.

[0032] Thereafter, the primary chip 1 is wired, via bonding wires 8, to a lead frame 7 for external connection, and the primary chip 1 and the secondary chips 2aand 2b, together with the frame 5, are molded with a sealing resin 9, whereby the semiconductor device having a chip-on-chip structure is completed.

[0033] An erroneous mounting rate will be discussed. If the secondary chips 2aand 2bare correctly mounted on the frame 5, the secondary chips 2aand 2bmay not be thereafter erroneously mounted on the primary chip 1 in the step of overlapping the frame 5 with the primary chip 1. Consequently, the erroneous mounting is a problem in the step of mounting the secondary chips 2aand 2bon the frame 5. For example, if the appearance of the frame 5 on which the secondary chips 2aand 2bare mounted can be examined, it is possible to improve the fabrication yield of the semiconductor device by previously finding out the erroneous mounting before the frame 5 is joined to the primary chip 1. The following items (1) and (2) are examples of a method of examining the appearance:

[0034] (1) If the frame 5 is transparent, identification indications including characters and other symbols are formed on rear surfaces of the secondary chips 2aand 2barranged on the frame 5 (surfaces opposite to the device forming surfaces, i.e., upper surfaces in FIG. 1). The identification indications can be pattern-recognized through the frame 5, thereby making it possible to confirm before assembling whether the secondary chips 2aand 2bare correctly arranged.

[0035] (2) If the frame 5 is opaque, identification indications including characters and other symbols are attached to bump mounting surfaces of the secondary chips 2aand 2b(the device formation surfaces, i.e., lower surfaces in FIG. 1). The identification indications can be pattern-recognized from a chip mounting surface of the frame 5, thereby making it possible to confirm before assembling whether the secondary chips 2aand 2bare correctly arranged.

[0036] FIG. 2 is a diagram showing a modified example of a frame. In the case shown in FIG. 2, a frame 51 having holes formed therein is used, and secondary hips 2aand 2bare respectively fitted in the holes. Stoppers (not shown) are respectively provided in predetermined portions on inner edges of the holes such that the inserted secondary chips 2aand 2bdo not slip off. In the frame 51, a plurality of holes corresponding to the secondary chips 2aand 2bare formed in a relative positional relationship corresponding to positions, to which the secondary chips 2aand 2bare to be joined, on the primary chip 1.

[0037] According to the present embodiment, rear surfaces of the secondary chips 2aand 2barranged in the frame 51 can be imaged irrespective of whether the frame 51 is transparent or opaque. If identification indications are formed on wiring of the secondary chips 2aand 2b, it is possible to confirm before assembling whether the secondary chips 2aand 2bare correctly arranged by pattern-recognizing the identification indications.

[0038] According to the present embodiment, the holes 25 corresponding to the sizes of the secondary chips 2aand 2bare formed. When the sizes of the secondary chips 2aand 2bdiffer from each other, the holes themselves are means for preventing the secondary chips 2aand 2bfrom being erroneously mounted.

[0039] The frame 5 or the frame 51 can be also used as a lead frame, for example. In this case, another lead frame need not be used, so that the number of members is saved. When the frame 5 or 51 is also used as a lead frame, an external connecting terminal of the primary chip 1 may be electrically connected to the frame by suitable means such as wire bonding. In FIG. 2, bonding wires 18 are shown which connect the frame 51 to the primary chip 1.

[0040] FIGS. 3A and 3B are side views showing another embodiment of the present invention. In the present embodiment, secondary chips 2aand 2bare mounted, and a frame 52 is then separated from the secondary chips 2aand 2b. The frame 52 is preferably made of a flexible material such as a resin film. The secondary chips 2aand 2bare previously mounted on the frame 52 in a relative positional relationship corresponding to positions, to which the secondary chips 2aand 2bare to be joined, on a primary chip 1.

[0041] FIG. 3A illustrates the step of joining the secondary chips 2aand 2bmounted on the frame 52 to the primary chip 1. FIG. 3B illustrates the step of joining bumps 3 and 4 to each other by heating or ultrasonic wave application and then stripping the frame 52 from the secondary chips 2aand 2b. When the frame 52 is stripped, the bumps 3 and 4 have already been firmly joined to each other. Therefore, the joining between the bumps 3 and 4 may not be released from an external force caused by an operation for stripping the frame 52.

[0042] The present invention is not limited to the above-mentioned embodiment. Although description has been made of a case where the number of secondary chips is two, for example, the present invention is not limited to the same. For example, the number may be an arbitrary number which is not less than two.

[0043] Although the present invention has been described and illustrated in detail, it is clearly understood that the same is by way of illustration and example only and is not to be taken by way of limitation, the spirit and scope of the present invention being limited only by the terms of the appended claims.

[0044] This application claims priority benefits under 35 USC §119 of Japanese Patent Application No. 11-45210 filed with the Japanese Patent Office on Feb. 23, 1999, the disclosure of which is incorporated hereinto by reference.