JPH01185952A - Flip-chip type semiconductor device - Google Patents

Abstract

PURPOSE:To prevent a chip from becoming electrically defective when a flaw is produced by a filler on the surface of the chip by a method wherein a gap between a solder bump electrode and a dummy electrode is set at a definite distance so as to limit the size of the filler. CONSTITUTION:A bump electrode 2a formed on a chip 1 and a dummy bump electrode 2b are provided; a distance 3 between the bump electrode 2a and the dummy bump electrode 2b is set to the size of 60% or less of the diameter of the bump electrode 2a. Accordingly, even when a semiconductor device is mounted on a substrate and is then coated with an epoxy-based molding material, a filler whose size exceeds 60% of the diameter of the dummy electrode 2b out of fillers contained in the molding material does not creep to the lower part of the semiconductor device. By this setup, it is possible to eliminate a wrong state that the filler produces a flaw on the surface of the semiconductor chip and that the semiconductor chip becomes electrically defective.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a flip-chip semiconductor device.

(Prior Art) Conventionally, semiconductor chips (free lag chips) equipped with pants electrodes have been used in order to improve packaging density.

Examples of methods for forming bump electrodes include electroplating,
Various methods have been proposed, such as a selective vapor deposition method, a solder dip method, and a solder dip method. Bump electrodes are suitable for boat fishing.
It is often formed with solder all around it by electroplating.

FIG. 2 is a cross-sectional view of a main part of a fly-knob chip schematically showing a conventional bump electrode connection structure, and shows a state in which bump electrodes are provided on a semiconductor chip. In the figure, the semiconductor chip fxo is made of a semiconductor substrate, for example, a P-type Si substrate.

In the illustrated example, the first layer 16 is an insulating film, the wiring electrode 18 is an Alx electrode, and the second layer 20 is a C'VD (Chemical) film.
ical Vapor Deposition)
A glass film (Hahishivation protective film) formed by the method is sequentially provided on the semiconductor chip.

The bump electrode 22 in the illustrated example includes a barrier layer 24 and a glass layer 2.
Consists of 6. In the illustrated example, the At-Ni alloy layer 28, the N
It consists of an i layer 30 and a Co layer 32, and further includes a barrier layer 24.
A solder layer is provided as a pad 26 on the Cu layer 32 . The At-Ni alloy layer 28 and the N1 layer 30 are formed, for example, by vapor deposition, and the Cu layer 32 and the pad 26 are formed, for example, by electroplating. - The GLAD 26 is formed into a spherical shape by, for example, treating a solder plating layer formed on the Co layer 32 at high temperature and utilizing surface tension. The method of forming solder bump electrodes by electroplating is well known (see, for example, Japanese Patent Application Laid-Open No. 160744-1/1983), and therefore detailed explanation will be omitted.

FIG. 3 is a side view schematically showing the structure of the flip chip. Flip chip type semiconductor device is semiconductor chip 7DI
A bump electrode 22 is provided at O.

In the illustrated example, the bump electrodes 22 are
Generally, 3 to 10 bump electrodes 22 are provided, and in some cases, nearly 100 bump electrodes 22 are provided.

FIG. 4 is a side view schematically showing the mounting state of the flip chip. In the figure, a mounted board (wiring board) 40
is equipped with a predetermined electric circuit pattern on its surface,
For example, a ceramic substrate is used as the substrate material of the mounted substrate 40. As shown in the figure, the flip chip 38
is directly mounted on the mounting board 40.

As a mounting method for a flip chip having two bump electrodes made of solder, there is, for example, a reflow one-type method. In this method, the flip chips f38'i are placed face-down at predetermined positions on the mounting board 40, and then these flip chips f3g and the mounting board 4
0 into a reflow oven and heated to 200 to 220°C.

As a result, the nine dots 26 (see FIG. 5) of the bump electrodes 22 are melted. Thereafter, the flip chip fss and the mounting board 40 are taken out of the reflow oven and cooled to room temperature. Therefore, the Rad 26 is also cooled.

Therefore, the flip chip f3g is connected to the electric circuit of the mounting board 40 via the wire 26.

Further, this substrate is coated with an epoxy molding material 42 to protect the chip surface from the outside, as shown in FIG. 5, in order to improve the reliability of the chip. Usually, this molding material filling is carried out while heating the substrate at 150°C to 200°C, and in order to improve the strength of the molding material, 10 to 30
% (weight ratio) of filler such as SiQ□ powder. The shape of this filler is generally spherical, with a diameter of 30 μm to 100 μmφ. However, in the method of connecting a substrate and a semiconductor chip using a solder bump having the above configuration, a spherical filler 44 is used as shown in FIG. When filling the molding material 42, the solder electrode 22 enters between the semiconductor chip f38 and the ceramic substrate 400 and solidifies the mold, and the solder electrode 22 expands at the curing temperature of 150°C, but contracts during the cooling process from the quenching temperature of 1500°C to room temperature. At this time, the semiconductor chip f38 and the substrate 4
0 is compressed by force in the direction shown by the arrow in the figure, and the filler 44 that has entered between them scratches the surface of the semiconductor chip 38, causing an electrical defect in the semiconductor chip 38. The height of the solder electrode required for normal connection is 70 to 100 μm, which is sufficient for a spherical filler to enter between the chip and the substrate, which affects the device due to shrinkage of the bump electrode.

In order to solve the above problems, a flip chip type semiconductor device of the present invention has a bump electrode formed on a chip and a dummy panzo electrode, and the distance between the bump electrode and the dummy bump electrode is equal to the diameter of the bump electrode. The dimensions shall be 60 or less.

(Function) Since the flip-chip semiconductor device of the present invention has the above-described configuration, even if the semiconductor device described above is mounted on a mounting board and then the entire epoxy molding material is coated,
Of the fillers contained in this molding material, it is possible to prevent fillers whose size exceeds the diameter of the dummy electrode by 60c4''f! from entering the lower part of the semiconductor device.

(Embodiment) FIG. 1 is a diagram showing a second embodiment of the present invention, in which A is a sectional view and B is a top view. In this figure, 1 is a semiconductor chip and 2 is a solder bump electrode.

The size of this solder bump electrode is 70 μmφ.

Normally, only the number of solder bump electrodes 2a necessary for exchanging signals between the semiconductor chip l and the ceramic substrate are formed, but in the present invention, the gap 3 between the solder bump electrodes 2 on the chip is spherical. An appropriate number of dummy solder bump electrodes 2bf are added and placed on the chip at least all around the chip so that the distance is 40 μm or less that does not allow the filler to pass through. The distance between the bumps is 60% of the diameter of the bump electrode.

Next, a method for forming the above-mentioned dummy bump electrode 2b will be described. This dummy bump electrode Nzb may be formed by providing or forming a barrier layer on the lower layer as in the case of a normal bump electrode, or may simply be formed after forming a metal layer on the lower layer. However, while normal bump electrodes are connected to the substrate in order to exchange signals with the substrate, dummy bump electrodes are insulated from the substrate to prevent noise from entering.

In addition, metal pads are provided at positions corresponding to dummy bump electrodes on the mounting board on which the chip is mounted.
When mounted on a board, this metal/gradient and dummy bump electrode are well fused together. This metal wire is also used only for fusion bonding with the dummy bump electrode, and is insulated from others.

Figure 7 shows that when the size of the solder bump electrode is 70 μm, the distance between the bumps 3 is set at a level between 10 μm and 100 μm, and the semiconductor chip is mounted on the ceramic substrate.
A molding material having a weight ratio of 30 to the molding material having an average size of iO2-based spherical filler of 50 μm and a variation of 30 to 70 μm was filled at 150°C, and then heated to -30°C to 150°C. It is a graph showing the percentage of chips that became electrically defective due to scratches on the semiconductor chips when the temperature cycle test at °C was repeated for 100Ω.

From FIG. 7, it can be seen that if the solder gap 3 is less than about 60% of the size of the solder punch 70 μmφ, that is, 40 μm or less, the defective rate due to the filler scratching the surface of the chip is reduced. This means that the connection layer bump electrode shrinks due to the solder bump, connecting the tip and the board, but the percentage of this shrinkage is estimated to be about 60% at most, and the spherical filler, which is smaller than this shrinking hole, is used to connect the chip and the board. Even if it gets into the gap, it is considered that it will not cause any trouble such as scratching the surface of the chip.

Even when the size of the solder bump electrode is as large as 70 μmφ or as small as 70 μmφ, the distance between the solder bump electrodes must be at least 3 mm and the distance between the solder bump electrodes must be at least 6 mm.
A similar effect can be expected if it is less than 0.

(Effects of the Invention) As explained above, according to the present invention, the distance between the solder bump electrodes for signal extraction and the dummy electrodes is approximately 60% of the size of the solder bump electrodes all around the semiconductor chip. Due to this arrangement, the size of the spherical filler that enters the gap between the chip and the board during heating during mold filling of the connection layer between the chip and the board is limited to 60 cm or less, which is the size of the solder bump electrode. When the heating layer cools the chip to room temperature, problems such as the chip surface being scratched by the spherical filler existing in the gap between the chip and the substrate and causing the chip to become electrically defective do not occur.

[Brief explanation of the drawing]

FIG. 1 is a sectional view and a plan view of a semiconductor device according to the present invention. FIG. 2 is a sectional view showing the structure of a conventional bump electrode. FIG. 3 is a side view of a conventional flip chip. FIG. 4 is a side view showing a state where the device is mounted on a conventional flip-chip FF board. FIG. 5 is a side view showing a state in which a conventional free lag chip is mounted on the entire board and further coated with epoxy resin. FIG. 6 shows a state in which the epoxy resin is being cooled, and is a partially enlarged view of FIG. 5. FIG. 7 is a graph showing the distance between bump electrodes and defective rate 2. 1... Semiconductor chip, 2a... Solder bump electrode,
2b...Dummy solder bump electrode, 3゛...Distance between bumps. StepPβSehii>1aN A total of 14 accents, 1 β imm
Chapter 18 A: A long time ago, a chivalry A: Koji Kou question Figure 1 B) Demi 1! l! Product of 4 poems 2nd proverb free, 5 chip no ya) addition 3rd figure Q Pig

Claims (1)

[Claims] 1. A flip-chip semiconductor device comprising a bump electrode formed on a chip and a dummy bump electrode, the distance between the bump electrode and the dummy bump electrode being 60% or less of the diameter of the bump electrode.