JPS6329408B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a semiconductor device, and particularly to an electrode structure of a pressure contact type semiconductor device.

For example, in a three-terminal semiconductor device such as a transistor, gate turn-off type, or field effect thyristor, two different types of semiconductor regions (one main electrode region and one control electrode region) are exposed on one main surface of a semiconductor substrate. Generally, electrode films (one main electrode and one control electrode) are formed individually in the semiconductor region. In order to uniformly and effectively control each part of one of the main electrode regions, one of the main electrode regions is subdivided into a large number of rectangles when viewed in plan, and the control electrode region is divided into one of the main electrode regions. It is often formed to surround the Furthermore, one main electrode is formed by making one main electrode surface higher than the control electrode surface and bringing one main electrode plate into pressure contact with these many one main electrode surfaces simultaneously. This is particularly practiced in power semiconductor devices.

An example of such a semiconductor device will be described with reference to the drawings, taking a gate turn-off thyristor as an example. FIG. 1 shows an example of a conventional structure of this type of gate turn-off thyristor. In the figure, a Si semiconductor substrate 1 includes a p-type emitter layer 14 and an n-type base layer 1.
3. p-type base layer 12 and n-type emitter layer 11
It has a laminated structure. The n-type emitter layer 11 corresponds to one of the aforementioned main electrode regions, the p-type base layer 12 corresponds to the control electrode region, and the p-type emitter layer 14 corresponds to the other main electrode region. Note that 101 is a SiO ₂ surface protective film.

Further, one main electrode film 111 is placed on the exposed surface part of the n-type emitter layer 11, a control electrode film 121 is placed on the exposed surface part of the p-type base layer 12, and the other main electrode film 121 is placed on the exposed surface part of the p-type emitter layer 14. Main electrode films 141 are formed respectively. Here, the electrode formation surfaces of the n-type emitter layer 11 and the p-type base layer 12 have a step in the stacking direction of each semiconductor layer. That is, the n-type emitter layer 11 is formed at the top of the mesa that protrudes from the electrode formation surface of the p-type base layer 12. A cathode electrode film 111 is provided on the exposed portion of the n-type emitter layer 11, and a gate electrode film 1 is provided on the p-type base layer 12.
21 are formed with approximately the same thickness. Therefore, these two electrode films also have a step difference. For example, Al is used for these electrode films.

An anode electrode film 141 is formed on the exposed main surface of the p-type emitter layer 14.
41, the anode electrode plate 3 is a brazing material (not shown)
It is glued by.

The cathode electrode plate 2 is placed on the cathode electrode film 111 divided into many parts, and is pressed in the direction of the arrow in the figure to ensure good contact. The cathode electrode plate 2 and the cathode electrode film 111 are not soldered to each other in order to absorb thermal strain caused by heat generation during operation of the semiconductor device. For the cathode electrode plate and the anode electrode plate, materials such as W and Mo, whose coefficient of thermal expansion is close to that of Si, which is the material of the semiconductor substrate, are used.

Cathode electrode film 1 of the cathode electrode plate 2 described above
Normally, pressure is applied to 11 to reduce the electrical resistance between the two.
^{Approximately} 200Kg/cm2 can be added.

Although not shown, the gate electrode film 12
A gate lead for connecting the p-type base region 12 to the outside is connected to an appropriate portion of the p-type base region 1 . Further, external electrode plates made of, for example, Cu are connected to each of the cathode electrode plate 2 and the anode electrode plate 3, and the semiconductor substrate is sealed by connecting the external electrode plates with an insulating cylinder. The gate lead usually passes through the above-mentioned insulating tube or one of the external electrode plates and is guided to the outside.

The semiconductor device with the above structure has a current collection mechanism from the cathode electrode film, relaxation of thermal strain by bringing the cathode electrode plate and cathode electrode film into pressure contact, and equalization of turn-off operation by dispersing the n-type emitter layer. In this respect, it is particularly suitable for power equipment. However, there are still problems that need to be improved in the following points.

This is due to the fact that the cathode electrode film 111 is exposed to heat cycles under pressure. That is,
As mentioned above, this type of semiconductor device usually has a weight of 200Kg/cm ²
A pressure greater than that is required. During operation, the temperature rises to 100 to 150°C due to the heat generated by the semiconductor substrate, while the process of returning to room temperature during rest is repeated.
It has become clear that the cathode electrode film is extremely easily deformed under such conditions. In cases where the deformation of the cathode electrode film is significant, its width may be doubled. As a result, an accident occurs in which the cathode electrode film comes into contact with the gate electrode film and the semiconductor device becomes inoperable. FIG. 2 shows one mode of deformation of the cathode electrode, and in the figure, the cathode electrode film 111 is crushed by the heat cycle under pressure and is in contact with the gate electrode film 121.

Furthermore, even if a short circuit accident as shown in Figure 2 does not occur immediately, some deformation may occur in some of the many dispersed cathode electrode films, resulting in uneven pressure between the cathode electrode films. In some cases, the degree of damage may progress and the cathode electrode plate may become tilted.
In such a case, a contact accident occurs between the base electrode film and the cathode electrode plate near the end of the semiconductor substrate.

Any type of contact accident is likely to occur after a large number of heat cycles, such as 10,000 times.
The reality is that this point has tended to be overlooked in the past.

Note that similar problems exist not only in gate turn-off thyristors, but also regardless of the material of the electrode film or electrode plate. Furthermore, if the pressure applied between the electrode plate and the electrode film is such that the contact resistance between the two is within a practical range, the above-mentioned problem will occur regardless of the pressure applied.

SUMMARY OF THE INVENTION An object of the present invention is to solve the problems with the conventional structure described above, and to provide a semiconductor device having an electrode structure in which short circuits between electrodes do not occur even when heat cycles under pressure are applied.

In order to achieve this object, the present invention is characterized by a semiconductor substrate in which different types of electrode films are arranged on one main surface side so that the top surfaces thereof have a step difference, and one main surface side of the semiconductor substrate. An electrode plate that is pressed into contact with the protruding electrode film on the surface side, and an electrode plate that is interposed between the protruding electrode film and the electrode plate and surrounds the top end of the protruding electrode film under pressure. The point is that it has a metal foil.

In the present invention, the metal foil described above is brought into contact with the top surface of the protruding electrode film under pressure to electrically connect the electrode film and the electrode plate. At the same time, just outside the top end of the electrode film, the metal foil is deformed into the shape of the cathode due to pressure, and intersects the direction of pressure at an acute angle at the side edge of the electrode film. This is because when the electrode film deforms and becomes thinner due to pressure, its apparent hardness increases due to the influence of the underlying hard semiconductor substrate.
This is because once the hardness of the metal foil is exceeded, the metal foil begins to deform due to the applied pressure. This deformation of the metal foil extends over the entire circumference of the top surface of the electrode film, and prevents deformation of the electrode film due to pressure from proceeding beyond a certain level. Therefore, deformation of the electrode film is limited to a level that causes no practical problems, and short circuits between the electrodes are prevented.

Embodiments of the present invention will be described below with reference to the drawings. FIG. 3 shows a cross section of a main part of a gate turn-off thyristor according to an embodiment of the present invention. In FIG. 3, the same parts as in FIG. 1 are designated by the same reference numerals as in FIG. In this embodiment, a foil 4 made of a Fe--Ni--Co alloy is interposed between the cathode electrode film 111 and the cathode electrode plate 2. Foil 4 is cathode electrode film 11
It has an area that completely covers 1. This foil 4
An alloy of 28% Fe, 16% Ni, and the balance Co was heated at 900℃ for 2 hours, cooled with water, rolled to a thickness of about 30μm, and annealed at 700℃ for 1 hour to remove strain. . The coefficient of thermal expansion is 4.5×10 ^-6 /°C.

The diameter of the semiconductor substrate is 30 mm, and two rows of 26 n-type emitter layers 11 each having a width of about 200 μm and a length of about 6 mm are formed on the main surface side. The cathode electrode film 111 and the gate electrode film 121 are made of Al and have a thickness of about 10 μm.
The cathode electrode plate is made of W and is pressed to the cathode electrode film through a metal foil 4 at a pressure of 200 kg/cm ² . In addition, the cathode electrode film 111 and the gate electrode film 121
The height difference between the top surfaces of each is approximately 20 μm.

Such a semiconductor device was subjected to heat cycles between 30°C and 120°C 10,000 times. As a result, it became clear that although the cathode electrode film 111 was slightly deformed as shown in FIG. 4, the deformation was prevented by the metal foil 4. on the other hand,
When the conventional example shown in FIG. 1 was subjected to a similar heat cycle test, the cathode electrode film was significantly crushed, and some of the cathode electrode film was in contact with the gate electrode film as shown in FIG. 2. in this way,
This embodiment is effective in preventing collapse of the electrode film in a semiconductor device of pressure contact type structure.

The same effect as this example can be obtained using Fe-Ni as the metal foil.
This can also be achieved using alloys. To give an example, the above-mentioned Fe-Ni-Co is added to an alloy of 35% Fe and 65% Ni.
The thickness obtained by applying the same treatment as in the case of alloys is approx.
Even when a 30 μm foil was used, the same results as Fe-Ni-Co alloy foil were obtained.

In the above embodiment, there was no problem that the metal foil other than the portion in contact with the cathode electrode film drooped down and came into contact with the gate electrode film during use of the semiconductor device. This is considered to be because a metal with relatively high hardness was used as the metal foil, and also because no pressing force was applied to the metal foil in areas other than those in contact with the cathode electrode film.

As in the above embodiment, Fe-Ni is used as the metal foil.
When using a material whose thermal expansion coefficient can be changed depending on the composition and processing method, such as -Co or Fe-Ni alloy, it is important to make the thermal expansion coefficient as close as possible to that of the semiconductor and cathode electrode plate. This is preferable in terms of preventing the king phenomenon. Further, these alloys are preferred because they can be easily formed into foils of 100 μm or less and are inexpensive. In addition, Fe−Ni
−For Co alloy, Ni is 27.5~31wt% and Co is 15~
If Ni deviates from the range of 20 wt%, or in the case of Fe-Ni alloys, if Ni deviates from the range of 30 to 42 wt%, the coefficient of thermal expansion will increase and the staking phenomenon with the cathode electrode plate is likely to occur, so this is preferable. do not have.

We also considered using Al, Ag, etc. as the metal foil, but Al softens significantly at high temperatures (approximately 80°C or higher) and risks coming into contact with the gate electrode plate.
It has become clear that since Ag has a significantly larger coefficient of thermal expansion than the cathode electrode plate, it causes a staking phenomenon with the cathode electrode plate, which is undesirable.

Although the present invention has been described above with reference to an embodiment of a gate turn-off thyristor, the present invention is not limited to this and can be widely applied.

First, the present invention can be applied to semiconductor devices such as transistors, high-frequency thyristors, and field-effect thyristors, regardless of the internal structure of their semiconductor substrates. For example, as shown in FIG. 5, the semiconductor substrate may have no level difference, but the electrode film may have a level difference. In addition, the collapse of the electrode film due to heat cycles under pressure is caused by the fact that the electrode film is made of aluminum.
Other metals, such as Al-Cu alloy, Cr, Ti, Ni,
Even if Mo or the like is used, or if these metals are appropriately laminated, this problem occurs to varying degrees. Therefore, the present invention is effective regardless of the type of electrode film.

The thickness of the metal foil does not matter as long as it surrounds the top end of the electrode film that protrudes under pressure. In other words, it is important that the direction of the pressurizing force and the metal foil intersect at an acute angle at the top edge of the protruding electrode film (for example, the cathode electrode film 111 in FIG. 3) mentioned above during pressurization. As long as this is satisfied, it is irrelevant to the thickness of the metal foil or the shape (both cross-section and plane) of the protruding electrode film.

In order to emphasize the effects of the present invention, it is desirable that the hardness of the metal foil be higher than the hardness of the protruding electrode film.

Furthermore, a plurality of metal foils may be stacked and used as long as the above-mentioned requirements are met. Alternatively, instead of covering the entire group of electrode films to be brought into contact with a single sheet of foil, it is also possible to use foils divided into a plurality of sheets.

Taking the above points into consideration, the material of the metal foil applied to the present invention is not limited to those of the embodiments, but can be selected from a wide range.

It goes without saying that all the electrode plates (indicated by reference numeral 2 in FIG. 3) that are normally used in semiconductor devices of this type can be used.

As explained in detail above, according to the present invention,
This is effective in preventing short circuits between electrodes in a semiconductor device having a pressure contact type structure.

[Brief explanation of the drawing]

Figure 1 shows a plan view (a) and a schematic cross-sectional view (b) of a conventional gate turn-off thyristor according to the present invention, and Figure 2 shows the view of the conventional example shown in Figure 1 (b) after being subjected to a heat cycle test under pressure. 3 is a sectional view of a gate turn-off thyristor according to an embodiment of the present invention, FIG. 4 is an enlarged sectional view of the main part of FIG. 3, and FIG. 5 is a diagram showing another embodiment of the present invention. be. 1... Semiconductor substrate, 2... Cathode electrode plate, 3
...Anode electrode plate, 4...Metal foil.

Claims (1)

[Claims] 1. A semiconductor substrate having a predetermined pn junction between a pair of main surfaces, a plurality of first electrode films formed on one main surface of the semiconductor substrate, and one main surface of the semiconductor substrate. a second electrode film formed on the surface to be insulated from the first electrode film and lower than the top surface of the first electrode film; and a second electrode film disposed astride the plurality of first electrode films. A metal foil having a thickness of 100 μm or less, which has higher hardness than the first electrode film and has a coefficient of thermal expansion substantially equal to that of the semiconductor substrate; and an electrode plate that is brought into pressure contact with the electrode film, and under pressure, the first
The electrode film is deformed by the applied pressure, its thickness is reduced, and the apparent hardness becomes greater than the hardness of the metal foil, and the metal foil is deformed by the applied pressure at the outside of the top end of the first electrode film. A semiconductor device characterized by being slightly curved so as to intersect at an acute angle with the direction of. 2. The semiconductor device according to claim 1, wherein the metal foil is made of Fe-Ni-Co or Fe-Ni alloy.