Semiconductor device

1. Field of the Invention

The present invention relates to a semiconductor device, and more particularly to a semiconductor device having bonding pads on two sides of a semiconductor chip.

2. Description of the Background Art

In a semiconductor device of a conventional BGA (Ball Grid Array) type, a semiconductor element (semiconductor chip) is fixed and adhered to a substrate made of glass epoxy resin, for example, with its active surface facing up. A connection pad (bonding pad) on the semiconductor element is connected to a bonding finger on the substrate by a wire made of gold, for example. The bonding finger is connected to a ball land by a copper interconnection. The ball land is provided for attachment of a solder ball provided for external connection. Such a BGA-type semiconductor device is disclosed, for example, in Japanese Patent Laying-Open Nos. 2000-68404 and 11-312705.

In the semiconductor element such as a memory, the connection pads are usually arranged along two sides of the semiconductor element, and the bonding fingers are generally arranged outside the side along which the connection pads are arranged. The wire for wire bonding is drawn out from the connection pad in a direction toward the outside of the side along which the connection pads are arranged.

An electric signal from the semiconductor element is taken out through a route from the connection pad through the bonding (wire), a copper interconnection on the substrate (including the bonding finger and an interconnection in a via) and the ball land to the solder ball. Such a routing is two-dimensional except for routing in the via, and intersection of interconnections is not allowed on the identical surface of the substrate. Accordingly, as the order or arrangement of the connection pads of the semiconductor element and pin arrangement of the solder ball are restricted, the order or the arrangement of the connection pads should be considered in advance at the time of designing the semiconductor element. In addition, when a product such as an existing semiconductor element of which design has already been completed and which can no longer accept consideration is employed, interconnections intersect two-dimensionally, resulting in failure in routing (failure in substrate design).

In general, a pitch between connection pads on the semiconductor element is smaller than that between pads of the bonding fingers. Therefore, in the conventional semiconductor device and structure, a package size necessarily tends to be restricted by a minimal dimension of the bonding finger.

Though the number of layers of the substrate may be increased in order to enable routing, this will cause cost increase as well as larger thickness of the substrate itself. Accordingly, the total height of the semiconductor device is increased, which will cause incompatibility with conventional products.

The present invention was made to solve the above-described problems. An object of the present invention is to improve a degree of freedom in routing without lowering a quality as well as to efficiently reduce an outer dimension of the semiconductor device.

A semiconductor device according to the present invention includes a substrate, a plurality of bonded portions, and a semiconductor element. The plurality of bonded portions are formed on a surface of the substrate. The semiconductor element is arranged above the surface of the substrate and has a plurality of connection pads on a surface opposite to a surface facing the substrate. The plurality of connection pads have a connection pad group aligned in the vicinity of one side of the semiconductor element along the same. The plurality of bonded portions have an adjacent side bonded portion arranged outside a side adjacent to the side of the semiconductor element along which the connection pad group is arranged. The connection pad group has a connection pad electrically connected to the adjacent side bonded portion by wire bonding.

According to the semiconductor device of the present invention, the bonded portion is arranged outside the side adjacent to the side along which the connection pad group is arranged, and the connection pad is connected to the bonded portion by wire bonding. Therefore, intersection of the interconnections arranged on the substrate or three-dimensional intersection of the bonding wires can be prevented, leading to smaller outer size of the semiconductor device. In this manner, the degree of freedom in routing can be improved without lowering the quality and the outer dimension of the semiconductor device can efficiently be reduced.

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.

FIG. 1 is a plan view schematically showing a configuration of a semiconductor device in a first embodiment of the present invention, without showing a mold resin.

FIG. 2 is a schematic cross-sectional view along the line II—II in FIG. 1 .

FIG. 3 is a diagram for illustrating a definition of “a side adjacent to a side of a semiconductor element along which a connection pad group is arranged” in the present specification.

FIG. 4 illustrates a distance between the side and positions of connection pads of the semiconductor element.

FIG. 5 shows that conventional bonding fingers are not arranged “outside the side adjacent to the side of the semiconductor element along which the connection pad group is arranged” in the present specification.

FIGS. 6A to 6C illustrate an effect achieved by the semiconductor device in the first embodiment of the present invention.

FIG. 7 shows that a plurality of bonding wires are drawn toward “outside the side adjacent to the side of the semiconductor element along which the connection pad group is arranged.”

FIG. 8 is a cross-sectional view showing a first preferred example of the bonding wire.

FIG. 9 is a cross-sectional view showing a second preferred example of the bonding wire.

FIGS. 10A and 10B are first plan views showing that an outer dimension of the semiconductor device can be reduced in a second embodiment of the present invention.

FIG. 11 is a second plan view showing that the outer dimension of the semiconductor device can be reduced in the second embodiment of the present invention.

FIGS. 12A and 12B are third plan views showing that the outer dimension of the semiconductor device can be reduced in the second embodiment of the present invention.

FIG. 13 is a plan view showing a first example of a configuration obtained by stacking semiconductor elements.

FIG. 14 is a plan view showing a second example of a configuration obtained by stacking semiconductor elements.

FIG. 15 is a plan view showing a third example of a configuration obtained by stacking semiconductor elements.

FIG. 16 is a plan view showing a fourth example of a configuration obtained by stacking semiconductor elements.

FIG. 17 is a plan view showing a fifth example of a configuration obtained by stacking semiconductor elements.

FIG. 18 is a cross-sectional view showing a preferred example of the bonding wire when the semiconductor elements are stacked.

In the following, embodiments of the present invention will be described with reference to the figures.

(First Embodiment)

Referring to FIGS. 1 and 2 , a semiconductor device 20 in the present embodiment mainly includes a substrate 1 , a bonding finger (bonded portion) 2 , copper interconnections 3 , ball lands 4 , solder balls 5 , a semiconductor element 6 , bonding wires 7 , and a mold resin 8 .

Substrate 1 is made of a glass epoxy resin, for example. On one surface of substrate 1 , bonding finger 2 and copper interconnection 3 electrically connected to bonding finger 2 are formed. On the other surface of substrate 1 , ball land 4 is formed. Bonding finger 2 and copper interconnection 3 on one surface are electrically connected to ball land 4 on the other surface through a conductive layer in a via (through hole) 1 a formed in substrate 1 . Solder ball 5 for external connection is formed on ball land 4 .

Semiconductor element 6 is a semiconductor chip, for example. Semiconductor element 6 has a plurality of connection pads (bonding pads) 10 on its surface. Each of the plurality of connection pads 10 is electrically connected to a circuit in semiconductor element 6 . The plurality of connection pads 10 have respective connection pad groups 110 A, 10 B aligned in the vicinity of sides 6 a , 6 b of semiconductor element 6 respectively. In a memory such as a DRAM (Dynamic Random Access Memory), in general, connection pad groups 10 A and 10 B are arranged along two sides 6 a and 6 b of semiconductor element 6 respectively.

Semiconductor element 6 is arranged above the surface of substrate 1 . In such an arrangement, the surface of semiconductor element 6 where connection pads 10 are not present is connected to one surface of substrate 1 with an adhesive or the like.

The plurality of connection pads 10 of semiconductor element 6 are electrically connected to the plurality of bonding fingers 2 respectively by bonding wires 7 implemented by a gold wire, for example. Mold resin 8 is molded on one surface of substrate 1 so as to cover bonding finger 2 , copper interconnection 3 , semiconductor element 6 , bonding wire 7 , and the like.

In the present embodiment, the plurality of bonding fingers 2 have a bonding finger (adjacent side bonded portion) 2 a and a bonding finger 2 b . Bonding finger 2 a is arranged outside sides 6 c and 6 d that are adjacent to sides 6 a , 6 b of semiconductor element 6 along which connection pad groups 10 A and 10 B are arranged respectively. Bonding finger 2 b is arranged outside sides 6 a and 6 b.

Each of connection pad groups 10 A and 10 B has connection pads 10 electrically connected to bonding fingers 2 a respectively by wire bonding (that is, via bonding wire 7 ). In addition, each of connection pad groups 10 A and 10 B has connection pads 10 electrically connected to bonding fingers 2 b respectively by wire bonding.

It is noted that inner connection pad 10 other than that at an end portion of connection pad groups 10 A and 10 B arranged in a line respectively may electrically be connected to bonding finger 2 a by wire bonding.

Here, the phrase “the side adjacent to the side of the semiconductor element along which the connection pad group is arranged” refers to portions of sides 6 c , 6 d between a centerline C 1 —C 1 of connection pad group 10 A and a centerline C 2 —C 2 of connection pad group 10 B (portions shown with bold line), as shown in FIG. 3 . Therefore, the phrase “outside the side adjacent to the side of the semiconductor element along which the connection pad group is arranged” herein refers to surface areas (hatched areas) of substrate 1 outside the portions (portions shown with bold line) of sides 6 c , 6 d between centerline C 1 —C 1 and centerline C 2 —C 2 shown in FIG. 3 .

A direction in which bonding wire 7 is drawn out from connection pad 10 is generally set to a shorter Y direction out of X and Y directions shown in FIG. 4 . In the present embodiment, however, bonding wire 7 is positively drawn out in a longer X direction, so that the degree of freedom in routing can drastically be improved without lowering the quality due to intersection of bonding wires 7 as described later.

Here, if a bonding position of the semiconductor element is displaced or a space between bonding fingers 2 b becomes larger in a conventional example as shown in FIG. 5 , an X-direction component in the direction in which bonding wire 7 is drawn out may increase. Meanwhile, as the term in the present specification has been defined as described above, bonding finger 2 b shown in FIG. 5 is arranged in an area different from an area “outside the side adjacent to the side of the semiconductor element along which the connection pad group is arranged” (hatched area) in the present specification.

According to the present embodiment, connection pad 10 of connection pad groups 10 A, 10 B is electrically connected to bonding finger 2 a by wire bonding, whereby the following effect is obtained.

Referring to FIG. 6A , in an attempt to establish connection from connection pad 10 to bonding finger 2 b on substrate 1 by bonding wire 7 or the like, routing requiring intersection of copper interconnections 3 on substrate 1 (area P 1 ) may occur, resulting in failure in substrate design. In such a case, bonding wires 7 may three-dimensionally intersect by wiring as shown with an area P 2 in FIG. 6B . Bonding wires 7 , however, are too proximate to each other or they may contact with each other, which will result in lower quality.

According to the present embodiment, bonding finger 2 b is arranged outside side 6 c adjacent to side 6 a along which connection pad group 10 A is arranged, and bonding finger 2 b is connected to connection pad 10 by wire bonding, as shown in FIG. 6C . Therefore, intersection of copper interconnections 3 as shown in FIG. 6A and three-dimensional intersection of bonding wires 7 as shown in FIG. 6B can be prevented.

In addition, as shown in FIG. 7 , any bonding wire 7 may be selected and drawn out so as to facilitate routing of copper interconnections 3 on substrate 1 , and two or more bonding wires may be selected. In other words, two or more bonding fingers 2 a may be provided, and each of bonding fingers 2 a may be connected to connection pad 10 by wire bonding.

In wiring in the present embodiment, as shown with bonding wire 7 on the right in FIG. 8 , connection pad 10 and bonding finger 2 b are positioned apart from each other. That is, as bonding wire 7 is drawn out by a large length, the following disadvantages may occur.

(1) Bonding wire 7 may contact with an edge of semiconductor element 6 .

(2) Adjacent bonding wires 7 may contact with each other, depending on a direction in which bonding wire 7 is drawn out.

Considering these possibilities, a portion of bonding wire 7 may intentionally be located higher than the surface of substrate 1 . In this case, however, the following disadvantages may occur, as shown by bonding wires 7 on the left in FIG. 9 .

(1) The total height of the semiconductor device becomes larger.

(2) Displacement of bonding wire 7 tends to occur and short-circuit between bonding wires 7 takes place.

As described above, it is necessary to reduce the overall height of bonding wire 7 from the surface of substrate 1 as well as to prevent short-circuit between adjacent bonding wires 7 . Then, bonding wire 7 drawn “outside the side adjacent to the side of the semiconductor element along which the connection pad group is arranged” is shaped as shown with bonding wire 7 on the left in FIG. 8 or bonding wire 7 on the right in FIG. 9 , with a ball bonding 7 a provided on the side of bonding finger 2 a . In this manner, the disadvantages as above can be overcome and the quality of the product can be improved.

Though this configuration can be achieved in a leadframe using a 42 alloy, copper or the like, higher effect can be obtained in the BGA having the solder ball serving as an external connection terminal on a lower surface of the semiconductor device.

In an ASIC (Application Specific Integrated Circuit) or a microcomputer type semiconductor, a large number of input/output pins themselves are required. In addition, the ASIC or the microcomputer type semiconductor is custom-made in many cases. Therefore, connection pads are arranged uniformly on four sides of the semiconductor element at the time of design. Accordingly, few problems will be caused without measures as described above. On the other hand, in a memory type semiconductor, a demand for smaller semiconductor element size is great, and arrangement and an area for the connection pads are restricted. Accordingly, since the connection pads are arranged on two sides of the semiconductor element, the effect of the present embodiment will be more significant, as the number of connection pads is increased.

In order to avoid an interpretation of the side of the semiconductor element along which dummy connection pads are arranged as “the side of the semiconductor element along which the connection pad group is arranged” in a broader interpretation of “the side adjacent to the side of the semiconductor element along which the connection pad group is arranged,” “the side of the semiconductor element along which the connection pad group is arranged” is defined as the side along which the connection pads actually wired to the bonding fingers are arranged.

(Second Embodiment)

The configuration in the first embodiment attains not only the effect in routing but also an effect to suppress an outer size of the semiconductor element. When the number of input/output pins of the semiconductor element is large, the number of bonding fingers should correspondingly be large. As to the shape or arrangement of the bonding finger and the connection pad, a width of 150 μm of the bonding finger is necessary, a pitch between the bonding fingers should be set to approximately 200 μm, an opening diameter of the connection pad should be set to 60 μm to 80 μm, and a pitch between the connection pads should generally be set broader than in an example in which it is set to 100 μm to 120 μm. If the width of the bonding finger or the pitch between the bonding fingers is forcibly made smaller, disadvantages as below will occur, in which case it will be difficult to maintain a stable quality of the product.

(1) Substrate: Designing becomes impossible due to a problem in working accuracy of the substrate, or cost for working of the substrate increases.

(2) Wire bonding: A width for stitch bonding to the bonding finger is insufficient, resulting in poor connection.

(3) Quality: A distance between plated wires on an outer periphery of the package is made smaller, which can be a cause of a defect such as short-circuit at an end portion.

As described above, a minimal outer dimension of the semiconductor device employing the semiconductor element including a large number of input/output pins is determined by a width over which the bonding fingers can be arranged. On the other hand, the bonding finger is positively arranged “outside the side adjacent to the side of the semiconductor element along which the connection pad group is arranged,” so that the outer dimension of the semiconductor device can be made smaller by L 1 +L 2 each covered by one connection pad (bonding finger) in FIGS. 10A and 10B or by M 1 ×2 in FIG. 11 .

In other words, the outer dimension of the semiconductor device can be made smaller by L 1 +L 2 by arranging bonding fingers 2 a “outside the side adjacent to the side of the semiconductor element along which the connection pad group is arranged” as shown in FIG. 10A , as compared with the example in which all bonding fingers 2 b are arranged outside “the side of the semiconductor element along which the connection pad group is arranged” as shown in FIG. 10B .

The semiconductor device in the present embodiment is effective because a demand for products attaining high density and smaller size has been great in a market of PDA (Personal Digital Assistant), a portable telephone, a DSC (Digital Still Camera) or the like in recent days.

Moreover, in rectangular semiconductor element 6 in which connection pads 10 are arranged on a shorter side as shown in FIG. 12A , bonding wires 7 are drawn out from all connection pads 10 to bonding fingers 2 a arranged outside sides 6 c , 6 d adjacent to sides 6 a , 6 b along which connection pad groups 110 A, 10 B are arranged respectively, as shown in FIG. 12B . Accordingly, the outer dimension of the semiconductor device can be made smaller by M 1 ×2.

(Third Embodiment)

In a product such as an S-CSP (Stacked Chip Scale Package) obtained by stacking semiconductor elements, the number of connection pads is increased, routing of the copper interconnections is further complicated, and an extremely large number of bonding fingers are necessary unless connection among semiconductor elements is shared. In other words, it is more difficult to route the copper interconnections on the substrate as well as to reduce the outer dimension of the semiconductor device than in the product using a single semiconductor element as in the first or second embodiment.

Here, as shown in FIGS. 13 to 15 , in a configuration implemented by stacking two or more semiconductor elements (two semiconductor elements 6 A, 1 B in the present embodiment), bonding fingers 2 a are provided outside sides 6 c , 6 c adjacent to sides 6 a , 6 a of semiconductor elements 6 A, 6 B along which connection pad groups 10 A, 10 A are arranged respectively, and bonding fingers 2 a are electrically connected to connection pads 10 by bonding wires 7 respectively.

In the configuration shown in FIGS. 13 to 15 , semiconductor element 6 A and semiconductor element 6 B are stacked such that side 6 a of semiconductor element 6 A along which connection pad group 10 A is arranged and side 6 a of semiconductor element 6 B along which connection pad group 10 A is arranged are located respectively on phantom lines L 1 —L 1 , L 2 —L 2 parallel to each other.

Here, connection pad 10 of any one of semiconductor elements 6 A, 1 B should only be connected to bonding finger 2 a , and at least one connection pad 10 among the plurality of connection pads 10 should only be connected to bonding finger 2 a . Each of all connection pads in connection pad group 10 A of any one of semiconductor elements 6 A, 1 B may be connected to bonding finger 2 a.

In addition, as in the configuration shown in FIGS. 16 and 17 , semiconductor element 6 A and semiconductor element 6 B are stacked such that side 6 a of semiconductor element 6 A along which connection pad group 10 A is arranged and side 6 a of semiconductor element 6 B along which connection pad group 10 A is arranged are located respectively on phantom lines L 1 —L 1 , L 2 —L 2 intersecting (orthogonal, for example) with each other.

Here, connection pad 10 of semiconductor element 6 A is electrically connected by bonding wire 7 to bonding finger 2 a arranged outside side 6 c adjacent to side 6 a of semiconductor element 6 A along which connection pad group 10 A is arranged. In addition, connection pad 10 of semiconductor element 6 B is electrically connected by bonding wire 7 to bonding finger 2 b arranged outside side 6 d adjacent to side 6 a of semiconductor element 6 B along which connection pad group 10 A is arranged.

When the semiconductor elements are stacked, a distance between the position of connection pad 10 and the position of bonding finger 2 b further increases, as shown on the right in FIG. 18 . Accordingly, the disadvantage described in the first embodiment is further likely.

Then, ball bonding 7 a is provided on the side of bonding fingers 2 a , 2 b as shown on the left in FIG. 18 , as in the first embodiment. As such, contact of bonding wire 7 with semiconductor elements 6 A, 1 B or short-circuit between adjacent bonding wires 7 can be prevented.

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.