Semiconductor Device in Which Electrodes Are Formed in a Self-Aligned Manner

The present invention relates to a semiconductor device and to a method of fabricating the same, and more specifically to the one which is adapted to forming a polycrystalline semiconductor layer having a highly precisely defined junction area on the side walls of a convex semiconductor layer or mesa.

The present invention relates to the structure of a semiconductor device and a method of fabricating the same, and particularly to the structure that is finely constructed and that is adapted to operating at high speeds and to a method of fabricating the same.

In the conventional semiconductor devices as disclosed in Japanese Patent Laid-Open No. 73156/1983, the base electrode of a bipolar transistor is taken out through a polycrystalline semiconductor layer sandwiched between the insulating films in order to realize high-speed operation by maintaining the parasitic capacitance small. The steps for fabricating the polycrystalline semiconductor layer that is a base electrode of the transistor, are shown in FIGS. 1A to 1E. The manufacturing steps will now be described. With reference to FIG. 1A, an n.sup.+ -type buried layer 2 is formed on a p-type silicon substrate 1, an n-type silicon epitaxial layer 3 is grown thereon, a silicon dioxide film 10 is deposited on the whole surface thereof, an insulating film other than the silicon dioxide film, such as a silicon nitride film (Si.sub.3 N.sub.4) 50 is deposited thereon, and another silicon dioxide film 20 is further deposited thereon, followed by patterning leaving three layers 10, 50 and 20 on the active region of the transistor. The silicon epitaxial layer is then etched using the three insulating layers 10, 50 and 20 as a mask, so that the active region assumes a convex shape. In this case, the silicon layer inwardly enters from the ends of the masks 10, 50, 20 due to etching.

With reference to FIG. 1B, an oxide film 30 is formed by the thermal oxidation, a silicon nitride film (Si.sub.3 N.sub.4) 60 is deposited on the whole surface, and the silicon nitride film 60 is left on the side surfaces of the convex silicon layer by selective etching.

With reference to FIG. 1C, an oxide film 40 is formed by the thermal oxidation. The silicon dioxide film grows very little on the side surfaces of the convex silicon layer but grows on the regions other than the convex or mesa area to form a thick oxide film 40.

With reference to FIG. 1D, the silicon nitride film 60 is removed and then the photoresist is applied followed by patterning (100), and a part of the silicon dioxide film 40 is then removed by etching resulting in an opening 200. In this case, the opening 200 is formed at an upper portion of the convex silicon layer.

With reference to FIG. 1E, a polycrystalline silicon layer is formed on the whole surface followed by patterning (300), so that only the side surfaces of the convex portion of the epitaxial layer will come into contact with the, polycrystalline silicon layer.

Through the above-mentioned steps, the structure can be realized in which the polycrystalline silicon layer comes into contact with the side surfaces of the convex silicon layer or mesa. However, the silicon dioxide film 40 that is formed by the thermal oxidation in FIG. 1C, also grows on the convex silicon layer in a manner as shown in FIG. 1C. When the silicon dioxide film in FIG. 1D is subjected to the etching, therefore, the area of the opening 200 on the side surface of the convex silicon layer tends to vary. Therefore, when p-type impurities are diffused from the polycrystalline silicon layer 300 in a subsequent step, it becomes highly difficult to precisely control the junction area between the p-type diffused region and the silicon epitaxial layer. Hence, the voltage V.sub.BE varies greatly across the base and the emitter in the inverse mode operation, making it difficult to adapt the device to an integrated circuit.

According to the present invention, a region where the base region comes into contact with the polycrystalline silicon is finely formed thereby maintaining high precision.

In the present invention, during the fabrication process a silicon nitride film is formed twice on the side surface of the convex silicon layer thereby forming a silicon opening which is maintained with high precision thereby decreasing the variance in the junction area relative to the polycrystalline silicon layer.

The semiconductor device according to the present invention comprises a semiconductor substrate of a first type of conductivity, a second region such as a buried 1 which is provided in the surface region of said substrate and which has a second type of conductivity opposite to said first type of conductivity, a first insulating film which is provided on the surface of said substrate and which has an opening on said third region, a first region formed as a mesa (island) which is provided in said opening and which consists of a first single crystalline semiconductor layer, a second insulating film which is provided on the side surface of said first mesa region, a second single crystalline or polycrystalline semiconductor layer which is formed on said first insulating film, which is located adjacent to said second insulating film, and which is in contact with a fourth region formed in the first region over a predetermined region inside the second insulating film, an insulating layer which is formed adjacent to said second single crystalline or polycrystalline semiconductor layer, a fifth region of the first type of conductivity provided in said first region, a sixth region of the second type of conductivity provided in said fifth region, and a third region of the first type of conductivity provided in said second single crystalline or polycrystalline semiconductor layer.

In other words, according to the structure of the present invention, a region of the first type of conductivity does not exist on the side surface of the convex single crystalline semiconductor layer or mesa since there exists an insulating film, and a region of the first type of conductivity is provided only along the periphery of the surface of the convex single crystalline semiconductor layer, the region of the first type of conductivity being electrically connected to the single crystalline or polycrystalline semiconductor layer.

The object of the present invention is to provide a method of fabricating semiconductor devices, which improves the aforementioned defects inherent in the conventional method of fabricating semiconductor devices, and in which an opening is formed in the side surface of the convex silicon layer or mesa maintaining high precision in order to decrease variance in the junction area relative to the polycrystalline silicon layer.

Another object of the present invention is to provide a semiconductor device consisting of a bipolar transistor which operates at high speeds, which features a high breakdown voltage, and which enables the size to be reduced in the vertical direction of the element and a method of fabricating the same, overcoming the aforementioned problems inherent in the conventional semiconductor devices.

These and other objects and many of the attendant advantages of this invention will be readily appreciated as the same becomes better understood by reference to the following detailed description when considered in connection with the accompanying drawings.

FIGS. 1A to 1F are section views illustrating conventional fabrication steps;

FIGS. 2A to 2F are section views illustrating a first embodiment of the present invention;

FIG. 3 is a section view illustrating a second embodiment of the present invention;

FIG. 4 is a section view illustrating a third embodiment of the present invention;

FIGS. 5 and 6A to 6E are section views illustrating a fourth embodiment of the present invention;

FIGS. 7A to 7C are section views illustrating a fifth embodiment of the present invention;

FIG. 8 is a section view illustrating a sixth embodiment of the present invention;

FIG. 9 is a section view illustrating a seventh embodiment of the present invention;

FIG. 10 is a section view illustrating an eighth embodiment of the present invention;

FIG. 11 is a section view illustrating a ninth embodiment of the present invention;

FIG. 12 to 14 are section views illustrating a tenth embodiment of the present invention;

FIG. 15 is a section view illustrating an eleventh embodiment of the present invention;

FIG. 16 is a plan view showing fourth, sixth, seventh, eighth, ninth, tenth and eleventh embodiments of the present invention;

FIG. 17 is a section view illustrating a twelfth embodiment of the present invention; and

FIG. 18 is a section view illustrating a thirteenth embodiment of the present invention.

The invention will now be described in detail by way of embodiments.

Embodiment 1:

FIGS. 2A to 2F illustrate a method of fabricating a semiconductor device according to an embodiment of the of the present invention. With reference to FIG. 2A, an n.sup.+ -type buried layer 2 is formed on a p-type silicon substrate 1, an n-type silicon epitaxial layer 3 is grown thereon, a silicon dioxide film 10 is deposited on the whole surface thereof, an insulating film other than the silicon dioxide film, such as a silicon nitride film (Si.sub.3 N.sub.4) 50 is deposited thereon, and another silicon dioxide film 20 is further deposited thereon, followed by patterning leaving three layers 10, 50 and 20 on the active region of the transistor. The silicon epitaxial layer is then etched using the three insulating layers 10, 50 and 20 as a mask, so that the active region assumes a convex shape. The junction area relative to the polycrystalline silicon on the side surface of the convex silicon layer is determined depending upon the depth of etching.

With reference to FIG. 2B, an oxide film 31 is formed by thermal oxidation, and a silicon nitride film is deposited on the whole surface, followed by selective etching so that the silicon nitride film 70 is left on the side surface only of the convex silicon layer.

With reference to FIG. 2C, the silicon oxide film 31 is subjected to the etching so that the silicon epitaxial layer is exposed on the portions other than the convex portion. The silicon epitaxial layer is then subjected to the etching and is thermally oxidized to form an oxide film 32. Thereafter, silicon nitride is deposited on the whole surface, and a silicon nitride film 80 is left on the side surface only of the convex silicon layer by selective etching. The silicon nitride films 70 and 80 are formed in double-layer structure on the side surface of the silicon oxide film 31, and the silicon nitride film 80 is formed in a single-layer structure on the side surface of the silicon dioxide film 32.

With reference to FIG. 2D, an oxide film 41 is formed by the thermal oxidation. In the portion where the silicon nitride films 70 and 80 are left in two layers on the side surface of the convex silicon layer, the oxide film grows less than that on the portion where the silicon nitride film 80 is left in one layer, and the thickness of the oxide film 31 remains almost unchanged.

With reference to FIG. 2E, the silicon nitride films 70 and 80 are removed and, then, a photoresist is applied followed by patterning (100), in order to subject the silicon oxide film 41 to the etching. In this case, an opening 201 is formed at an upper portion of the convex silicon layer. Here, though the silicon oxide film 41 is etched by using a photoresist, the patterning need not be particularly effected for the purpose of etching. In the step of forming the opening, the silicon oxide film 41 under the opening has a thickness greater than that of the silicon oxide film 31 due to the thermal oxidation of FIG. 2D, and only a portion of the silicon oxide film 31 can be selectively opened owing to the difference in the thickness.

With reference to FIG. 2F, the polycrystalline silicon layer is deposited on the whole surface followed by patterning (300), so that only the side surface of the convex epitaxial layer or mesa is brought into contact with the polycrystalline silicon layer.

Through the above-mentioned steps, it is possible to highly precisely control the area of the side surface of the convex silicon layer or mesa to which the polycrystalline silicon layer comes into contact. That is, the junction area is determined by the etching depth of the silicon layer in FIG. 2A. In the subsequent steps, the junction area is maintained covered with the silicon dioxide film and the silicon nitride film. Further, the silicon dioxide film can be selectively removed by etching from this portion.

Embodiment 2:

FIG. 3 shows a method of fabricating a semiconductor device according to the present invention, wherein the oxide film formed by the step of thermal oxidation of FIG. 2C has a thickness more than two times that of the oxide film 31, thereby to constitute an oxide film 33. Other steps are the same as those shown in FIGS. 2A to 2F. When the silicon dioxide film is removed by etching (FIG. 2E) to bring the polycrystalline silicon layer into contact with the side surface of the convex silicon layer in accordance with the method of this embodiment, the oxide film has a reduced thickness on a portion where the opening is to be formed, and increased degree of selectivity can be provided.

Embodiment 3:

FIG. 4 shows a method of fabricating a semiconductor device according to the present invention, wherein a lower portion of the convex silicon layer is tapered in the step of etching silicon of FIG. 2C. Other steps are the same as those shown in FIGS. 2A to 2F. According to the method of this embodiment, stress is not concentrated in the step of forming the thermally oxidized film 41 since the convex silicon layer is tapered, and defect is prevented from developing. Further, in selectively removing the silicon nitride film 80 by etching, the thickness of the silicon nitride film 80 decreases toward the lower end owing to the convex silicon layer that is tapered. Therefore, bird's beak of the oxide film reaches the lower end of the oxide film 31 due to the subsequent thermal oxidation and assumes a thickness greater than that of the oxide film 31. Therefore, the selectivity can be increased in the subsequent step (FIG. 2E) of removing the silicon dioxide film by etching.

The aforementioned embodiments 1 to 3 can be realized even when other semiconductors such as GaAs and the like are used. In the above embodiments, furthermore, the conductivity types, i.e., p-type conductivity and n-type conductivity, may be reversed as a matter of course.

According to the conventional art explained in conjunction with FIGS. 1A to 1F, the polycrystalline semiconductor layer 300 doped with p-type impurities is brought into contact with the side surface of the convex semiconductor or mesa, a p.sup.+ -type region 14 is isotropically diffused from the side surface of the n-type epitaxial layer 3 and is brought close to the n.sup.+ -type buried layer 2, causing the breakdown voltage to decrease between the base and the collector, and permitting the junction capacity to increase thereby impairing the high-speed operation. This problem becomes more conspicuous as the thickness of the n-type epitaxial layer 3 is further reduced.

The embodiments of the present invention that will be described hereinbelow are to solve the above-mentioned problem.

Embodiment 4:

FIG. 5 illustrates in cross section the structure of a semiconductor device according to a fourth embodiment of the present invention.

This embodiment deals with a bipolar transistor, in which an insulating film 277 is provided on the side surface of a convex single crystalline semiconductor layer or mesa 23, and polycrystalline semiconductors 26 are connected only through a peripheral portion of the surface of the convex single crystalline semiconductor layer or mesa 23. Therefore, a p.sup.+ -type region 214 diffused from the polycrystalline semiconductor layer 26 which has been doped with impurities in advance can be formed near the surface only, and does not approach an n.sup.+ -type buried layer 22. Accordingly, the parasitic capacitance decreases and the speed of operation increases. The breakdown voltage increases, too. Further, since the side surfaces of the p.sup.+ -type region 214 are brought into contact with the insulating film 277, the junction exists on the bottom surface only, contributing to decreasing the capacitance. With this structure, furthermore, one time of photo step for determining the convex single crystalline semiconductor layer 23, helps determine the active region as well as diffused layers 24, 25, 214 inside thereof and an electrode 212, contributing to reducing the area of the element.

This structure makes it possible to broaden the gap between the buried layer of the second type of conductivity having a high concentration and the region of the first type of conductivity connected to the single crystalline or polycrystalline semiconductor layer, to decrease the capacitance between the base and the collector of the bipolar transistor, and to increase the speed of operation. The breakdown voltage can be increased, too. According to this embodiment, furthermore, the active region and the isolated regions can be determined by one time of photo step, contributing to increasing the degree of integration of the elements.

FIGS. 6A to 6E are section views illustrating the steps for fabricating the bipolar transistor of the fourth embodiment shown in FIG. 5. The fabricating steps will now be described in conjunction with the drawings.

With reference to FIG. 6A, an n.sup.+ -type buried layer 22 is formed on a p-type silicon substrate 2, an n-type silicon epitaxial layer 23 is grown maintaining a thickness of 1 .mu.m and a resistivity of about one ohm.multidot.cm. Then, a silicon dioxide film 2101, an insulating film other than the silicon dioxide film, such as a silicon nitride film (Si.sub.3 N.sub.4) 2102, and another silicon dioxide film 2103 are deposited on the whole surfaces, followed by patterning to leave the three layers 2101, 2102 and 2103 on only the active region A and the collector electrode take-out region B of the transistor.

With reference to FIG. 6B, the silicon epitaxial layer is etched to a depth of about 0.5 .mu.m by using the three insulating films 2101, 2102 and 2103 as a mask, so that the active region and the collector electrode take-out region will assume the convex shape. Thereafter, a silicon dioxide film 277 is formed by thermal oxidation, a silicon nitride film (Si.sub.3 N.sub.4) is deposited on the whole surface, and the silicon nitride film 2104 is left only on the side surface of the convex silicon layer by selective etching. Here, the oxide film 277 has a thickness two to three times as great as that of the oxide film 2101.

With reference to FIG. 6C, an oxide film 27 is formed by the thermal oxidation. Thereafter, the silicon nitride film 2104 is removed. At this moment, the sides of the silicon nitride film 2102 are removed by etching so that the silicon nitride film 2102 will enter inwardly from the ends of the surface of the convex silicon layer. Then, the oxide film is removed by etching by a thickness that corresponds to the oxide film 2101, to form an opening 2200 of silicon only at the ends of the surface of the convex silicon layer. Here, though a mask is used for the collector electrode take-out region to leave the oxide film 2101, the patterning need not be particularly effected for the purpose of etching.

With reference to FIG. 6D, a polycrystalline silicon layer is formed on the whole surface, followed by patterning so that only the end surface 2200 of the convex epitaxial layer or mesa is brought into contact with the polycrystalline silicon layer 250.

With reference to FIG. 6E, a silicon dioxide film 2110 and a silicon nitride film 2111 are formed on the whole surface, followed by patterning. Then, using the thus patterned two insulating films 2110 and 2111 as a mask, part of the polycrystalline silicon is thermally oxidized to form an oxide film 28. Then, n.sup.+ -type impurities are added at a high concentration to a collector electrode takeout semiconductor layer 233.

Thereafter, the silicon oxide film 2110 and the silicon nitride film 2111 are removed, p-type impurities are diffused into a polycrystalline silicon layer 250 to form a p-type diffused layer 214, followed by thermal oxidation to form an oxide film 29. Then, a base region 24 and an emitter region 25 of the transistor are formed by a customary method, contact holes are formed in the oxide film, and electrodes 211, 212 and 213 are formed by vaporization to form an element as shown in FIG. 5.

In the foregoing were mentioned the fourth embodiment and the fabricating method according to the present invention. With the above fabricating method, the element isolating region through up to the emitter electrode can be formed all in a self-aligned manner through one time of the photo step. Further, the width of the external base region is determined by the amount of side etching of the silicon nitride film, and can be decreased to 0.2 .mu.m or smaller.

Embodiment 5:

In the foregoing was described a first fabricating method according to the present invention. Another method of fabricating the structure of the present invention are shown in FIGS. 7A to 7C.

With reference to FIG. 7A, three layers 2301, 2302 and 2303 are formed only on the active region A and on the collector electrode take-out region B of the transistor through the same steps as those of up to FIG. 6A, and a silicon oxide film 2304 is formed by the thermal oxidation. Here, the oxide film 2304 has a thickness which is one-half to one-third that of the oxide film 2301.

With reference to FIG. 7B, a silicon nitride film is deposited on the whole surface, and the silicon nitride film 2310 is left only on the side surfaces of the three layers 2301, 2302 and 2303 by selective etching. The oxide film 2304 is so removed by etching that it is left only on the lower portion of the silicon nitride film 2310 formed along the periphery of the three layers 2301, 2302 and 2303. Using the three layers 2301, 2302, 2303 and the silicon nitride film 2310 as masks, the silicon epitaxial layer is etched to a depth of about 0.5 .mu.m, so that the active region and the collector electrode take-out region assume the convex shape. Then, a silicon oxide film 2305 is formed by the thermal oxidation, a silicon nitride film is deposited on the whole surface, and the silicon nitride film 2311 is left only on the side surfaces of the convex silicon layer by selective etching. Here, the oxide film 2305 has a thickness the same as that of the oxide film 2301, which is about two to three times as great as the thickness of the oxide film 2304.

With reference to FIG. 7C, an oxide film 2307 is formed by the thermal oxidation, and the silicon nitride films 2310 and 2311 are removed. Thereafter, the oxide film is removed by etching by a thickness that corresponds to the oxide film 2304, and an opening 2200 is formed only at the ends of the surface of the convex silicon layer or mesa.

According to this fabricating method, the width of the opening 2200 is determined by the thickness of the silicon nitride film 2310, and can, hence, be controlled maintaining high precision.

After the step of FIG. 7C, the same steps as those of FIGS. 6D and 6E are effected to form the element shown in FIG. 5.

Embodiment 6:

FIG. 8 is a section view showing the structure of a lateral bipolar transistor to which the present invention is adapted. Electrodes of emitter and collector regions 214 are taken out through polycrystalline semiconductor layers 26 surrounded by insulating films 277, 27, 28 and 29, and wherein ends of the surface of the convex silicon layer or mesa are connected to single crystalline silicon of the polycrystalline semiconductor layers 26. This makes it possible to reduce the junction capacitance between the emitter and the collector, to reduce the thickness of the single crystalline semiconductor layer, and, hence, to realize a small transistor that operates at high speeds. In FIG. 8, a base electrode 233 is taken out via the n.sup.+ -type buried layer 22. As shown in FIG. 5, however, the n-type diffused layer 25 may be formed on the surface of the convex silicon layer or mesa to use it as a base electrode. Then, the n.sup.+ -type buried layer needs be formed only under the convex silicon layer, to reduce the junction capacitance between the n.sup.+ -type buried layer and the substrate and to realize the transistor in a small size.

Embodiment 7:

FIG. 9 shows in cross section the structure of a MOS to which the present invention is adapted. In this case, the source and drain regions 214 are formed maintaining a very small width and are self-aligned. In FIG. 9, an n-type epitaxial layer 23 is provided on an n-type silicon substrate 21000. However, the n-type epitaxial layer or mesa 23 is not particularly required, but the n-type silicon substrate 21000 only may be utilized.

Alternatively, the n.sup.+ -type buried layer may be provided on a p-type silicon substrate as in FIG. 5 of the drawings.

Embodiment 8:

With reference to FIG. 10, a metal or a metal compound (WSi.sub.2) 260 is formed on the polycrystalline silicon layer to decrease the wiring resistance, in order to form a transistor that operates at high speeds and that consumes a small amount of electric power.

Embodiment 9:

With reference to FIG. 11 which illustrates the structure of the device of the present invention, a polycrystalline silicon layer 270 is provided on the emitter to prevent metal atoms in the electrode 212 from infiltrating into the base region 24. This permits the depth of the emitter region 25 to be decreased (0.1 to 0.2 .mu.m), making it possible to form a transistor that operates at high speeds and that is small in size.

In the aforementioned embodiments 4 to 9, some or all of the combinations thereof may be employed. The device of the present invention can also be realized even by using other semiconductors such as GaAs and the like. In these embodiments, furthermore, the conductivity types, i.e., p-type and n-type may be reversed, as a matter of course.

According to the conventional art mentioned in conjunction with FIGS. 1A to 1F, the polycrystalline semiconductor layer 300 doped with p-type impurities contacts the side surfaces of the convex semiconductor layer or mesa, and the p.sup.+ -type region 14 is isotropically diffused from the side surface of the n.sup.- -type epitaxial layer 3 to approach the n.sup.+ -type buried layer 2. Therefore, the junction capacitance increases between the base and the collector to impair the high-speed operation. This problem becomes more conspicuous as the thickness of the n.sup.- -type epitaxial layer 3 decreases.

This embodiment is to provide a bipolar transistor which operates at high speeds, improving the aforementioned problems.

In order to achieve the above object, the following embodiment comprises a semiconductor substrate of a first type of conductivity, a first region which is provided on the surface region of said substrate and which has a second type of conductivity opposite to said first type of conductivity, a first insulating film which is provided on the surface of said substrate and which has an opening on said first region, a second region or mesa which is provided on said opening and which consists of a single crystalline semiconductor layer having the second type of conductivity, a second insulating film provided on the side surfaces of the second region, a second single crystalline or polycrystalline semiconductor layer which is provided on said first insulating film adjacent to said second insulating film and which is in contact with a corner portion, having top and side portions, near the surface of the second region over a region of a predetermined width, an insulating layer which is provided adjacent to said second single crystalline or polycrystalline semiconductor layer, a third region of the first type of conductivity which is provided in the upper corner portion of the second region that is electrically connected to said single crystalline or polycrystalline semiconductor layer, a fourth region of the first type of conductivity which is provided in the second region electrically connected to said third region, a fifth region of the second type of conductivity provided in said fourth region, and a sixth region of the first type of conductivity provided in said second single crystalline or polycrystalline semiconductor layer, wherein the impurity concentration in said third region is higher than that of said fourth region, and a position of junction between said third region and said second region is closer to the surface than a position of junction between said fourth region and said second region.

In other words, an insulating film is formed on the side surfaces of a convex single crystalline semiconductor layer or mesa, a single crystalline or polycrystalline conductive layer upper is connected to the corner portion on both the top and side surface near the surface of the convex single crystalline semiconductor layer or mesa maintaining a predetermined width, and the junction surface between the second region and the third region of the first type of conductivity which is electrically connected to the single crystalline or polycrystalline conductive layer and which has a high impurity concentration, is located closer to the surface than the junction surface between the second region and the fourth region.

Owing to the above-mentioned structure, the gap between the buried layer (first region) of the second type of conductivity and the third region of the first type of conductivity, which is connected to the second single crystalline or polycrystalline semiconductor layer and which has a high impurity concentration, can be selected to be greater than the gap between the first region and the fourth region of the first type of conductivity which has a low impurity concentration, and whereby the capacitance can be decreased between the base and the collector of the bipolar transistor to increase the speed of operation.

Embodiment 10:

FIG. 12 is a section view showing the structure of a semiconductor device according to a tenth embodiment of the present invention.

This embodiment deals with a bipolar transistor, wherein an insulating film 377 is provided on the side surface of a convex single crystalline semiconductor layer or mesa 33, a polycrystalline semiconductor layer 36 is connected to a corner portion near the surface of the convex single crystalline semiconductor relayer 33 over a region of a predetermined width which is as small as 0.1 .mu.m or less, and the thickness of the polycrystalline semiconductor layer is reduced near the connecting portion to restrain the diffusion from the polycrystalline semiconductor layer 36 which has been doped with impurities in advance. Further, the distance between a p.sup.+ -type region 314 and an n.sup.+ -type buried layer 32 is greater than the distance between the p.sup.+ -type region 314 and a p.sup.- -type region 34. The junction capacity varies in proportion to an inverse number of the width of a depletion layer under the condition where a reverse bias is applied. Therefore, the junction capacity decreases with the increase in the width of the depletion layer. Here, the width of the depletion layer is determined by the concentration of the diffused layer which forms the junction. In the junction between a high-concentration layer and a low-concentration layer, the depletion layer extends toward the side of the low-concentration layer. If a capacitance between the collector and the base is taken into consideration, the depletion layer extends toward the side of the n.sup.- -type epitaxial layer 33 in the junction between the p.sup.+ -type external base layer 314 of a high concentration and the n.sup.- -type epitaxial layer 33 of a low concentration. When the n.sup.+ -type buried layer 32 of the high concentration is reached, however, the depletion layer does not extend any more, and the thickness of the n.sup.- -type epitaxial layer 33 between the p.sup.+ -type external base layer 314 and the n.sup.+ -type buried layer 32 becomes nearly equal to the width of the depletion layer. Therefore, the junction capacitance between the p.sup.+ -type external base layer 314 and the n.sup.- -type epitaxial layer 33 is determined by the thickness of the n.sup.- -type epitaxial layer 33 under the p.sub.+ -type external base layer 314. To reduce the junction capacitance, the depth of diffusion of the p.sup.+ -type external base layer 314 must be decreased. In the junction between a p.sup.- -type intrinsic base layer 34 of a low concentration and the n.sup.- -type epitaxial layer 33, on the other hand, the depletion layer extends toward both sides. Compared with the case of the p.sup.+ -type external base layer 314, therefore, the junction capacitance is not much affected by the thickness of the n.sup.- -type epitaxial layer 33 under the p.sup.- -type intrinsic base layer 34. Therefore, the high-speed operation is not impaired even when the p.sup.- -type intrinsic base layer is formed having a thickness deeper than the p.sup.+ -type external base layer 314. Therefore, development of defective operation such as leakage of current between the emitter and the collector can be restrained, and high yields can be realized.

FIGS. 13A to 13E are section views illustrating the steps for fabricating a bipolar transistor of the tenth embodiment shown in FIG. 12. The fabricating steps will now be described in conjunction with the drawings.

With reference to FIG. 13A, an n.sup.+ -type buried layer 32 is formed on a p-type silicon substrate 31, an n-type silicon epitaxial layer 33 is grown maintaining a thickness of 1 .mu.m and a resistivity of 1 ohm.multidot.cm, and a silicon dioxide film 3101, an insulating film other than the silicon dioxide film, such as a silicon nitride film (Si.sub.3 N.sub.4) 3102 as well as another silicon dioxide film 3103 are deposited on the whole surface, followed by patterning to leave the three layers 3101, 3102 and 3103 on only the active region A and the collector electrode take-out portion B of the transistor.

With reference to FIG. 13B, the silicon epitaxial layer is etched to a depth of about 0.5 .mu.m using the three insulating films 3101, 3102 and 3103 as a mask, so that the active region and the collector electrode take-out region assume a convex shape. Thereafter, a silicon dioxide film 377 is formed by the thermal oxidation, a silicon nitride film (Si.sub.3 N.sub.4) is deposited on the whole surface, and the silicon nitride film 3104 is left only on the side surfaces of the convex silicon layer by selective etching. Here, the oxide film 377 has a thickness about two to three times as great as that of the oxide film 3101.

With reference to FIG. 13C, an oxide film 37 is formed by the thermal oxidation. Then, the silicon nitride film 3104 is removed. At this moment, the sides of the silicon nitride film 3102 are removed by etching so that the silicon nitride film 3102 will enter to the inside from the ends of the surface of the convex silicon layer or mesa. Then, the oxide film is removed by etching by a thickness corresponding to the oxide film 3101, and an opening 3200 of silicon is formed only at a corner portion near the surface of the convex silicon layer. Though the oxide film 3101 is left on the collector electrode take-out portion by using a mask, the patterning need not necessarily be effected for the purpose of etching.

With reference to FIG. 13D, a ploycrystalline silicon layer is formed on the whole surface, followed by patterning so that only the end surface 3200 of the convex epitaxial layer or mesa is brought into contact with the polycrystalline silicon layer 350.

With reference to FIG. 13E, a silicon dioxide film 3110 and a silicon nitride film 3111 are formed on the whole surface, and are subjected to the patterning. Using the thus patterned two insulating films 3110 and 3111 as a mask, part of the polycrystalline silicon layer is thermally oxidized to form an oxide film 38. Thereafter n.sup.+ -type impurities are added to the collector electrode take-out semiconductor layer 333 at a high concentration.

Then, the silicon dioxide film 3110 and the silicon nitride film 3111 are removed, p-type impurities are diffused into the polycrystalline silicon layer 350 to form a p.sup.+ -type diffused layer 314, followed by the thermal oxidation to form an oxide film 39. Next, a base region 34 and an emitter region 35 of a transistor are formed by a custmary method, contact holes are formed in the oxide film, and electrodes 311, 312 and 313 are formed by vaporization thereby to form the element as shown in FIG. 12.

In the foregoing were described the tenth embodiment of the present invention and a fabricating method according thereto. According to such a fabricating method, the isolating region of an element through up to an emitter electrode can be formed in a self-aligned manner through one time of the photo step. Further, the outer base region is formed maintaining a constant width at the corner portion near the surface of the convex silicon layer or mesa.

In the foregoing were described major portions of the present invention. FIG. 14 shows a method of decreasing the diffusion of the p.sup.+ -type outer base layer in fabricating the structure according to the present invention. Namely, FIG. 14 shows a step between the step of FIG. 13C and the step of FIG. 13D, and wherein after a polycrystalline silicon layer is formed on the whole surface, an insulating layer is deposited on the whole surface, and the insulating layer 340 is left on the vertically rising portion of the polycrystalline silicon layer by selective etching. Immediately thereafter, p-type impurities (such as boron ions B.sup.+) are injected into the polycrystalline silicon layer 350. According to this method, an impurity-injected layer 351 is formed at a position separated away from the convex silicon layer by the thickness (distance C) of the polycrystalline silicon layer 350 and the insulating layer 340. Owing to this offset effect, the depth of the p.sup.+ -type diffused layer 314 can be reduced.

Embodiment 11:

FIG. 15 shows in cross section the structure of a semiconductor device according to an eleventh embodiment of the present invention. In this embodiment, an insulating film 370 is formed adjacent to the convex single crystalline semiconductor layer or mesa 33 nearly in flush with the surface of the convex single crystalline semiconductor layer or mesa, and a polycrystalline semiconductor layer 36 is connected to the corner near the surface of the convex single crystalline semiconductor layer 33 maintaining a constant width. According to this structure, there exists the insulating film 370 having a large thickness between the polycrystalline semiconductor layer 36 and the convex single crystalline semiconductor layer 33, and the MOS capacitance therebetween can be decreased compared with that of the tenth embodiment (FIG. 12).

FIG. 16 is a plan view of the semiconductor device of the present invention, wherein the cross section along the line X-X' corresponds to those of FIGS. 5, 8, 9, 10, 11, 12 and 15. There is provided the p.sup.+ -type diffused layer region 3140 of a constant width at the corner portion near the surface of the convex single crystalline semiconductor layer, and there is further provided the n.sup.+ -type diffused layer region 3500 inside thereof maintaining an equal distance from the p.sup.+ -type diffused layer region 3140.

Embodiment 12:

FIG. 17 shows the structure of a device according to the present invention, wherein a metal or a metal compound 366 is formed on the polycrystalline semiconductor layer 36 to decrease the wiring resistance, thereby to form a transistor that operates at high speeds.

Embodiment 13:

FIG. 18 illustrates the structure of a device according to the present invention, wherein a polycrystalline semiconductor layer 355 is provided on the emitter to prevent metal atoms in the electrode 312 from infiltrating into the base region 34. Owing to this structure, the thickness of the emitter region 35 can be reduced (0.1 to 0.2 .mu.m) to realize a transistor that operates at high speeds.

In the aforementioned embodiments 10 to 13, some or all of the combinations thereof may be employed. The device of the present invention can also be realized even by using other semiconductors such as GaAs and the like. In these embodiments, furthermore, the conductivity types, i.e., p-type and n-type may be reversed, as a matter of course.

According to the structure of the present invention in which a polycrystalline semiconductor layer sandwiched by the insulating films is brought into contact with the side wall of the convex semiconductor layer or mesa, the junction area can be controlled highly precisely contributing to markedly improving the precision of the semiconductor devices.

According to the present invention, furthermore, there are provided a transistor and an integrated circuit thereof that operates at high speeds, that can be highly densely integrated and that has a high breakdown voltage.

It is further understood by those skilled in the art that the foregoing description deals with preferred embodiments of the disclosed device and that various changes and modifications may be made in the invention without departing from the spirit and scope thereof.