Patent Application
20190165091
The present invention relates to a semiconductor device using SiC and a method for manufacturing the same.
Conventionally, a problem of discharge in the atmosphere occurs when measuring electric characteristics of a high withstand voltage semiconductor device.
As a measure against this, for example, Patent Document 1 discloses a method for manufacturing a semiconductor device, including steps of forming base regions and emitter regions on a semiconductor wafer, patterning a base electrode and an emitter electrode, and then fitting and patterning a polyimide film onto the surface to cover regions except for a dicing region and other electrode bonding portions.
Patent Document 1: Japanese Patent Application Publication No. 60-50937
Patent Document 2: Japanese Patent Application Publication No. 54-45570
Patent Document 3: Japanese Patent Application Publication No. 2011-243837
Patent Document 4: Japanese Patent Application Publication No. 2001-176876
Patent Document 5: Japanese re-publication of PCT international application No. WO2009/101668
An object of the present invention is to provide a semiconductor device capable of improving a discharge starting voltage when measuring electric characteristics, and widening a pad area of a surface electrode to be larger than conventionally, and a method for manufacturing the same.
Another object of the present invention is to provide a semiconductor device capable of improving a discharge starting voltage when measuring electric characteristics, and increasing the number of semiconductor devices (number of chips) to be obtained from one wafer to be larger than conventionally, and a method for manufacturing the same.
A semiconductor device according to the present invention includes a first conductive-type SiC layer having a first surface, a second surface and end faces surrounding the first surface and the second surface, the SiC layer having a semiconductor device structure, a second conductive-type voltage relaxing layer formed in the SiC layer such that the voltage relaxing layer is exposed to an end portion of the first surface of the SiC layer, an insulating layer covering the voltage relaxing layer, and a surface electrode connected to the first surface of the SiC layer through the insulating layer, the surface electrode having a pad area selectively exposed (Claim 1).
This semiconductor device can be manufactured by a method for manufacturing a semiconductor device, including the steps of, in a first conductive-type SiC wafer which has a first surface and a second surface and a dicing region with a predetermined width set to segment a plurality of device regions, forming a semiconductor device structure in each of the device regions, forming a second conductive-type voltage relaxing layer with a width wider than that of the dicing region along the dicing region so as to be across the device regions adjacent to each other, forming an insulating layer on the SiC wafer so as to cover the voltage relaxing layer, forming a surface electrode in each of the device regions so that the surface electrode is connected to the first surface of the SiC wafer through the insulating layer and partially selectively exposed as a pad area, measuring electric characteristics of the semiconductor device structure of the device regions by applying a maximum applied voltage (BV) that generates a potential difference of 1000 V or more between the surface electrode and the SiC wafer in one of the device regions, and dividing the SiC wafer into individual pieces of the plurality of semiconductor devices by cutting the SiC wafer along the dicing region (Claim 20).
When measuring electric characteristics of the semiconductor device structure, in order to prevent discharge between the dicing region and the surface electrode, the discharge starting voltage V between the dicing region and the surface electrode must be set to a value higher than the maximum applied voltage (BV) (≥1000 V). The maximum applied voltage (BV) is a voltage (breakdown voltage: BV) when the semiconductor device structure causes avalanche breakdown, and between the dicing region and the surface electrode, a potential difference corresponding to this maximum applied voltage (BV) occurs. On the other hand, the discharge starting voltage V is a voltage when an insulator such as the atmosphere present between the dicing region and the surface electrode is broken down and an electric current starts to flow between the dicing region and the surface electrode. Specifically, as long as the relationship of the discharge starting voltage V>the maximum applied voltage (BV) is satisfied, an insulated condition between the dicing region and the surface electrode is maintained.
Therefore, according to the present invention, a voltage relaxing layer is formed along the dicing region, and further, the voltage relaxing layer is covered by an insulating layer. Therefore, when measuring electric characteristics of the semiconductor device structure, the maximum applied voltage (BV) can be relaxed in two stages of the insulating layer and the voltage relaxing layer. Accordingly, the burden of the voltage on the atmosphere between the dicing region and the surface electrode can be eased. In other words, the voltage applied between the dicing region and the surface electrode can be divided by the atmosphere, the insulating layer, and the voltage relaxing layer, so that even if the discharge starting voltage V in the atmosphere is set to be lower than conventionally, the relationship of the discharge starting voltage V>the maximum applied voltage (BV) can be maintained.
Here, based on Paschen's Law, the discharge starting voltage V between two electrodes is expressed by a function of the product of a gas pressure P and the interval between the electrodes (in the present invention, the distance X1 from the end of the pad area to the dicing region) (V=f (P·X1)). According to the present invention, the discharge starting voltage V in the atmosphere can be set to be lower than conventionally, so that based on Paschen's Law, the distance X1 from the end of the pad area to the dicing region (end face of the SiC layer) can be made shorter than conventionally.
Therefore, when the size of the semiconductor device (chip size) is set to be equal to the conventional one, the outer edge of the pad area can be extended to the end face side of the SiC layer, so that the pad area of the surface electrode can be made wider than conventionally. On the other hand, when the size of the pad area of the surface electrode is set to be equal to the conventional one, the end faces of the SiC layer can be shortened to the pad area side, so that the number of semiconductor devices (number of chips) to be obtained from one SiC wafer can be increased to be larger than conventionally.
Furthermore, even if a defect (such as a hole, etc., caused by a process failure) occurs in one of the insulating layer and the voltage relaxing layer, the defect can be covered by the other. Therefore, discharge between the dicing region and the surface electrode can be effectively prevented from occurring.
In the semiconductor device according to the present invention, it is preferable that the maximum applied voltage (BV) Y (≥1000 V) of the semiconductor device structure and the distance X1 from the end of the pad area to the end face of the SiC layer satisfy the following relational expression (1) (Claim 2). Regarding the manufacturing method, the step of forming the surface electrode preferably includes a step of setting the size of the pad area so that the maximum applied voltage (BV) Y (≥1000 V) of the semiconductor device structure and the distance X1 from the end of the pad area to the dicing region satisfy the following relational expression (1) (Claim 21).
As described above, by providing the insulating layer and the voltage relaxing layer, in addition to the atmosphere and the first conductive-type SiC, at least a plurality of layers are interposed between the dicing region and the surface electrode. The interposition of these layers effectively prevents discharge between the dicing region and the surface electrode.
On the other hand, between the surface electrode of one device region and the surface electrode of a device region adjacent to the device region mentioned above, pad areas individually exposed are linked to each other via only the atmosphere. Therefore, when measuring the electric characteristics of the semiconductor device structure, if the maximum applied voltage (BV) (≥1000 V) exceeds the discharge starting voltage V in the atmosphere, discharge may occur between the surface electrodes adjacent to each other.
According to the function V=f (P·X1) of Paschen's Law, as X1 decreases, the discharge starting voltage V also lowers. That is, as a result of the present invention, the distance X1 from the end of the pad area to the dicing region (end face of the SiC layer) can be made shorter, however, according to this, the discharge starting voltage V between the surface electrodes linked to each other via only the atmosphere also lowers. Therefore, discharge between the surface electrodes is required to be prevented by preventing the maximum applied voltage (BV) from exceeding the discharge starting voltage V in the atmosphere while the distance X1 is kept as short as possible.
Therefore, in this arrangement, by satisfying the relational expression (1) described above, discharge between surface electrodes adjacent to each other can be reliably prevented while the distance X1 from the end of the pad area to the dicing region (end face of the SiC layer) can be made shorter than conventionally.
In the semiconductor device according to the present invention, it is preferable that the distance X2 from the end of the connected portion connected to the SiC layer of the surface electrode to the end face of the SiC layer is longer than the width of a depletion layer spreading in the horizontal direction along the first surface of the SiC layer from the connected portion when the maximum applied voltage (BV) is applied to the semiconductor device structure (Claim 3). Regarding the manufacturing method, the step of forming the surface electrode preferably includes a step of setting a relative position of the connected portion to the dicing region so that the distance X2 from the end of the connected portion connected to the SiC wafer of the surface electrode to the dicing region becomes longer than the width of a depletion layer spreading in the horizontal direction along the first surface of the SiC wafer from the connected portion when the maximum applied voltage (BV) is applied to the semiconductor device structure (Claim 22).
With this arrangement, in each of the semiconductor devices divided into individual pieces, the depletion layer spreading in the horizontal direction from the connected portion connected to the SiC layer of the surface electrode is prevented from reaching the end face of the SiC layer.
The semiconductor device according to the present invention preferably further includes a second conductive-type termination structure formed to be annular along the periphery of the surface electrode in the SiC layer (Claim 4).
With this arrangement, the degree of extension of the depletion layer from the connected portion connected to the SiC layer of the surface electrode can be adjusted. Furthermore, by adjusting the impurity concentration of the termination structure, the maximum applied voltage (BV) of the semiconductor device structure can also be adjusted. In this case, a plurality of the termination structures may be formed concentrically (Claim 5).
In the semiconductor device according to the present invention, the insulating layer preferably has a structure consisting of a plurality of layers including a first layer and a second layer laminated in order from the SiC layer (Claim 6). With this arrangement, according to the magnitude of the maximum applied voltage (BV) required for the semiconductor device structure, the kind of insulating layer can be variously changed.
In the semiconductor device according to the present invention, it is possible that the first layer is formed on the entire first surface of the SiC layer, the second layer has an outer peripheral edge retreated to the inner side of the SiC layer with respect to the end faces of the SiC layer such that the first layer is partially exposed on the end portion of the the first surface of the SiC layer (Claim 7). In this case, the voltage relaxing layer may be formed to overlap the outer peripheral edge of the second layer in the thickness direction of the SiC layer (Claim 8), or may be formed so as not to overlap the outer peripheral edge of the second layer (Claim 9).
In the semiconductor device according to the present invention, the second layer preferably includes a convex portion that selectively penetrates the first layer and reaches the first surface of the SiC layer (Claim 10). In this case, a plurality of the convex portions of the second layer may be formed (Claim 11).
With this arrangement, even if the first layer peels from the end face of the SiC layer, this peeling can be stopped by the convex portion of the second layer. Therefore, adhesiveness of the insulating layer to the SiC layer can be improved.
In the semiconductor device according to the present invention, both of the first layer and the second layer may be formed on the entire first surface of the SiC layer (Claim 12).
In the semiconductor device according to the present invention, the first layer may be composed of any material of silicon oxide (SiO2) with a thickness of 1 μm or more, polyimide with a thickness of 0.2 μm or more, and silicon nitride (SiN) with a thickness of 1 μm or more (Claims 13 to 15).
In the semiconductor device according to the present invention, the insulating layer may have a structure consisting of a single layer (Claim 16).
In the semiconductor device according to the present invention, the semiconductor device structure may include a Schottky barrier diode structure formed by composing the surface electrode of a material that forms a Schottky barrier between the surface electrode and the SiC layer (Claim 17). Furthermore, the semiconductor device structure may include an MIS transistor structure including second conductive-type channel regions selectively formed in the SiC layer, first conductive-type source regions formed in contact with the channel regions, a gate insulating film formed on the channel regions, and a gate electrode opposed to the channel regions via the gate insulating film (Claim 18).
In the semiconductor device according to the present invention, the SiC layer may include an SiC substrate and an SiC epitaxial layer formed on the SiC substrate, and the SiC epitaxial layer may have an impurity concentration of 1×1016 cm−3 or less and a thickness of 5 μm or more.
Furthermore, in the method for manufacturing the semiconductor device according to the present invention, in the step of measuring a withstand voltage of the semiconductor device structure, the measurement is preferably performed at a pressure of 720 to 1520 Torr (Claim 23).
According to the function V=f (P·X1) of Paschen's Law, as the gas pressure P increases, the discharge starting voltage V also becomes higher. Therefore, by setting the gas pressure P for measurement of the electric characteristics in the above-described range, the distance X1 from the end of the pad area to the dicing region (end face of the SiC layer) can be further shortened.
Hereinafter, preferred embodiments of the present invention are described in detail with reference to the accompanying drawings.
The semiconductor device 1 is a device using silicon carbide (SiC). A number of semiconductor devices 1 are regularly aligned and formed on one SiC wafer 2. On the SiC wafer 2 (hereinafter, also referred to as “SiC layer 2”), a dicing region 4 with a predetermined width α (for example, 30 μm to 80 μm) that segments a plurality of device regions 3 is set. In the present preferred embodiment, the dicing region 4 is formed in a grid manner, and the plurality of device regions 3 are aligned in matrix as a whole. The semiconductor device 1 is formed in each device region 3, and the semiconductor devices are divided into individual pieces by cutting the SiC wafer 2 along the dicing region 4. The semiconductor device 1 according to the first preferred embodiment is a Schottky barrier diode.
The SiC layer 2 of each semiconductor device cut out by dividing into individual pieces has a first surface 2A and a second surface 2B, and end faces 2C surrounding the first surface 2A and the second surface 2B. The end faces 2C are SiC layer cutting surfaces (side surfaces) appearing by dividing into individual pieces, and segment the outer peripheries of the first surface 2A and the second surface 2B. In the present preferred embodiment, each semiconductor device 1 has a chip shape square in a plan view. As the size of the chip, the lengths in the up-down direction and the left-right direction of the sheet of
The SiC layer 2 includes a substrate 5 composed of n+-type SiC and an epitaxial layer 6 composed of n−-type SiC formed on the substrate 5. The thickness of the substrate 5 may be 50 μm to 1000 μm, and the thickness of the epitaxial layer 6 may be 5 μm or more (preferably 6 μm to 20 μm). As an n-type dopant contained in the substrate 5 and the epitaxial layer 6, for example, N (nitrogen), P (phosphorus), As (arsenic), etc., can be used (the same applies to the description given below). As a relationship between the dopant concentration of the substrate 5 and the dopant concentration of the epitaxial layer 6, the dopant concentration of the substrate 5 is relatively high, and the dopant concentration of the epitaxial layer 6 is relatively lower than that of the substrate 5. In detail, the dopant concentration of the substrate 5 may be 1×1017 to 1×1022 cm−3, and the dopant concentration of the epitaxial layer 6 may be 1×1016 cm−3 or less (preferably 1×1015 to 9×1015 cm−3).
In the surficial portion of the epitaxial layer 6, a p-type voltage relaxing layer 7 is formed to be exposed to the end portion of the first surface 2A. In the present preferred embodiment, the voltage relaxing layer 7 is formed to be annular along the outer peripheral edge of the SiC layer 2 so as to be exposed to the corner portions on the first surface 2A side of the SiC layer 2 defined by the first surface 2A and the end faces 2C of the SiC layer 2. Accordingly, the voltage relaxing layer 7 is exposed to both of the first surface 2A and the end faces 2C, and the regions exposed to the faces 2A and 2C are integrated at the corner portion of the SiC layer 2. As a p-type dopant contained in the voltage relaxing layer 7, for example, B (boron), Al (aluminum), etc., can be used. The voltage relaxing layer 7 is formed so that the bottom portion thereof is positioned on the first surface 2A side with respect to the boundary between the substrate 5 and the epitaxial layer 6. In detail, the depth of the voltage relaxing layer 7 may be, for example, 1000 to 10000 angstroms.
On the SiC layer 2, an insulating layer 8 and an anode electrode 9 as a surface electrode are formed. The insulating layer 8 is formed to cover the voltage relaxing layer 7, and has a structure consisting of a plurality of layers including a first layer 81 and a second layer 82 laminated in order from the SiC layer 2. When the insulating layer 8 has a structure consisting of a plurality of layers, according to the magnitude of the maximum applied voltage (BV) required for the Schottky barrier diode, the kind of insulating layer 8 can be variously changed.
In the present preferred embodiment, the first layer 81 is formed on the entire first surface 2A of the SiC layer 2. On the other hand, the second layer 82 is formed so as to expose the portions on the corners of the first surface 2A of the SiC layer 2 in the first layer 81, and has an outer peripheral edge 83 retreated to the inner side of the SiC layer 2 with respect to the end faces 2C. In detail, the voltage relaxing layer 7 covered by the insulating layer 8 is formed to overlap the outer peripheral edge 83 of the second layer 82 in the thickness direction of the SiC layer 2. Specifically, the inner peripheral edge 71 of the voltage relaxing layer 7 is positioned on the inner side of the SiC layer 2 as compared with the outer peripheral edge 83 of the second layer 82. In the first layer 81, a contact hole 84 that selectively exposes the first surface 2A of the SiC layer 2 is formed.
The anode electrode 9 includes a lower end portion 91 embedded in the contact hole 84 and an upper end portion 92 protruding upward from the first layer 81, and the bottom surface of the lower end portion 91 is connected as a connected portion 93 to the first surface 2A of the SiC layer 2 inside the contact hole 84. The upper end portion 92 of the anode electrode 9 further includes a drawn-out portion 94 as a peripheral edge portion uniformly drawn-out in the horizontal direction (direction along the first surface 2A of the SiC layer 2) from the outer peripheral edge of the contact hole 84. Accordingly, the anode electrode 9 is formed to have a size larger than the opening diameter of the contact hole 84 in a plan view.
The peripheral edge portion of the upper end portion 92 of the anode electrode 9 is covered by the second layer 82 higher than the anode electrode. That is, in the second layer 82, a contact hole 85 that selectively exposes the central portion of the anode electrode 9 as a pad area 95 is formed. The contact hole 85 is formed so that the relative position of the outer peripheral edge to the end faces 2C of the SiC layer 2 is further inside than the outer peripheral edge of the contact hole 84. The detailed size of the contact hole 85 is, for example, 500 μm×300 μm when a bonding wire with a 125 μm diameter is connected to the pad area 95.
In the insulating layer 8, in the present preferred embodiment, the first layer 81 is composed of silicon oxide (SiO2) with a thickness of 1 μm or more, and the second layer 82 is composed of polyimide with a thickness of 0.2 μm or more. However, the materials of the insulating layer 8 are not limited to these. For example, the first layer 81 may be composed of polyimide with a thickness of 0.2 μm or more or silicon nitride (SiN) with a thickness of 1 μm or more. Among these materials, in terms of adhesiveness to the SiC layer 2, SiO2 is most preferable.
The anode electrode 9 can be composed of a material that forms a Schottky barrier or hetero junction with respect to the n-type SiC layer 2, specifically, Mo (molybdenum), Ti (titanium), Ni (nickel), or Al (aluminum) as an example of the former material, or polysilicon as an example of the latter material, or the like. That is, the anode electrode 9 forming a Schottky barrier in this semiconductor device 1 may be either of a metal electrode that forms a Schottky barrier between the electrode and the SiC layer 2, and a semiconductor electrode that is composed of a semiconductor having a band gap different from that of the SiC layer 2, and makes a hetero junction to the SiC layer 2 (junction forming a potential barrier between the electrode and the SiC layer 2 by utilizing a band gap difference).
In this semiconductor device 1, when it turns into a forward bias state where a positive voltage is applied to the anode electrode 9 and a negative voltage is applied to the cathode electrode (not illustrated), electrons (carriers) move from the cathode electrode to the anode electrode 9 through the SiC layer 9 and an electric current flows. Accordingly, the semiconductor device 1 (Schottky barrier diode) operates.
Next, a method for manufacturing the semiconductor device 1 is described with reference to
First, an SiC wafer 2 formed of the substrate 5 and the epitaxial layer 6 is prepared. In this SiC wafer 2, as described above, device regions 3 in which a plurality of semiconductor devices 1 are formed, and a dicing region 4 that segments the device regions 3 and defines the end faces 2C of the plurality of semiconductor devices 1 to be finally divided into individual pieces and separated from each other, are set in a grid manner.
Next, for example, by selectively applying ion implantation and annealing to the first surface 2A of the SiC layer 2, the voltage relaxing layer 7 is formed (Step S1). The voltage relaxing layer 7 is formed along the dicing region 4 so as to have a width β that is wider than the width α of the dicing region 4 and across the device regions 3 adjacent to each other in the middle of manufacturing. That is, in the present preferred embodiment, the voltage relaxing layer 7 is formed in a grid manner with a width β wider than the width α of the dicing region 4 (refer to the alternate long and short dash line in
Next, according to a known film formation method such as CVD, the first layer 81 is formed on the entire first surface 2A of the SiC layer 2 (Step S2).
Next, the contact holes 84 that selectively expose the first surface 2A of the SiC layer 2 are formed in the first layer 81, and then, for example, by sputtering, the material of the anode electrodes 9 is deposited and patterned to define the anode electrodes 9 (Step S3). The anode electrodes 9 are connected to the SiC layer 2 (epitaxial layer 6) through the contact holes 84 of the first layer 81.
Next, by a known film formation method such as CVD, the second layer 82 is formed on the first layer 81 so as to cover the entire anode electrodes 9 (Step S4). Next, by patterning the second layer 82, the contact holes 85 that selectively expose the central portions of the anode electrodes 9 as pad areas 95 are formed (Step S5). Simultaneously, the grid-shaped portion along the dicing region 4 of the second layer 82 is selectively removed, and the outer peripheral edge 83 of the second layer 82 is retreated to the inner side of each device region 3 with respect to the lines being the end faces 2C. Through the above-described steps, in each device region 3, a Schottky barrier diode (semiconductor device structure) is defined by junction between the SiC layer 2 and the anode electrode 9.
The next step is measurement of the electric characteristics of each Schottky barrier diode (Step S6). In detail, the anode electrode 9 of one device region 3 is set to 0 V, and the second surface of the SiC wafer 2 is set to 1000 V or more (for example, 1700 V). Accordingly, a maximum applied voltage (BV) that generates a potential difference of 1000 V or more is applied between the anode electrode 9 and the SiC wafer 2, and a withstand voltage of each Schottky barrier diode is measured.
At this time, the n-type portion of the SiC wafer 2 including a part (except for the voltage relaxing layer 7) of the dicing region 4 is fixed to a potential of 1000 V or more, so that between the dicing region 4 and the anode electrode 9, a potential difference of 1000 V or more is generated. Even in this case, according to the present preferred embodiment, the voltage relaxing layer 7 is formed along the dicing region 4, and the voltage relaxing layer 7 is further covered by the insulating layer 8. Therefore, the maximum applied voltage (BV) of 1000 V or more applied between the dicing region 4 and the anode electrode 9 can be relaxed in two stages of the insulating layer 8 and the voltage relaxing layer 7. Accordingly, the burden of the voltage on the atmosphere between the dicing region 4 and the anode electrode 9 can be eased. In other words, as shown in
Here, based on Paschen's Law, the discharge starting voltage V between two electrodes is expressed by a function of the product of a gas pressure P and the interval between the electrodes (in the present preferred embodiment, the distance X1 from the end of the pad area 95 (outer peripheral edge of the contact hole 85) to the dicing region 4) (V=f (P·X1)). With this semiconductor device 1, the discharge starting voltage V of the atmosphere section 10 can be made lower than conventionally, so that according to Paschen's Law, the distance X1 can be made shorter than conventionally.
Therefore, when the size of the semiconductor device 1 (chip size) is made equal to the conventional one, the outer edge of the pad area 95 can be widened to the end face 2C side of the SiC layer 2, so that the pad area 95 can be made wider than conventionally. On the other hand, when the size of the pad area 95 is made equal to the conventional size, the end faces 2C of the SiC layer 2 can be shortened to the pad area 95 side, so that the number of semiconductor devices 1 (number of chips) to be obtained from one SiC wafer 2 can be increased to be larger than conventionally.
Furthermore, even if a defect (for example, a hole, etc., caused by a process failure) occurs in one of the insulating layer 8 and the voltage relaxing layer 7, the defect can be covered by the other. Therefore, discharge between the dicing region 4 and the anode electrode 9 can be effectively prevented from occurring.
Measurement of the electric characteristics of the Schottky barrier diode can be performed in a gas atmosphere of, for example, the atmosphere, nitrogen (N2), hydrogen (H2), argon (Ar), neon (Ne), helium (He), or the like. The gas pressure P at this time is preferably, for example, 720 Torr to 1520 Torr.
According to the function V=f (P·X1) of Paschen's Law, as the gas pressure P increases, the discharge starting voltage V also becomes higher. Therefore, by setting the gas pressure P for measurement of the electric characteristics in the above-described range, the distance X1 from the end of the pad area 95 to the dicing region 4 (end face 2C of the SiC layer 2) can be further shortened.
Thereafter, the electric characteristics of the Schottky barrier diodes in all device regions 3 are measured according to the same method. After the measurement, the SiC wafer 2 is cut into the individual semiconductor devices 1 by cutting along the dicing region 4. Thus, the semiconductor devices 1 having the structure shown in
Next, effects relating to the distance X1 from the end of the pad area 95 to the end face 2C of the SiC layer 2 are described with reference to
In the semiconductor device 1, it is preferable that the maximum applied voltage (BV) Y (≥1000 V) to be applied to the Schottky barrier diode and the distance X1 from the end of the pad area 95 to the end face 2C of the SiC layer 2 satisfy the following relational expression (1). In the manufacturing process, it is preferable that the size of the pad area 95 (the size of the contact hole 85) is set so that the maximum applied voltage (BV) Y (≥1000 V) to be applied to the Schottky barrier diode and the distance X1 from the end of the pad area 95 to the dicing region 4 satisfy the following relational expression (1) when exposing the pad area 95 by patterning the second layer 82 (Step S5).
As described above, by providing the insulating layer 8 and the voltage relaxing layer 7, in addition to the atmosphere section 10, the insulating layer section 11 and the voltage relaxing layer section 12 are interposed between the dicing region 4 and the anode electrode 9. The interposition of the insulating layer 8 and the voltage relaxing layer 7 effectively prevents discharge between the dicing region 4 and the anode electrode 9.
On the other hand, in the section (section 13) between the anode electrode 9 of one device region 3 and the anode electrode 9 of the device region 3 adjacent to the device region 3 mentioned above, pad areas 95 individually exposed are linked to each other via only the atmosphere. Therefore, at the time of measurement of the electric characteristics of the Schottky barrier diode (Step S6), if the maximum applied voltage (BV) (≥1000 V) exceeds the discharge starting voltage V in the atmosphere, discharge may occur between the anode electrodes 9 adjacent to each other.
According to the function V=f (P·X1) of Paschen's Law, as X1 decreases, the discharge starting voltage V also lowers. That is, as a result of the present invention, the distance X1 from the end of the pad area 95 to the dicing region 4 (end face 2C of the SiC layer 2) can be made shorter, however, according to this, the discharge starting voltage V in the section 13 linked via only the atmosphere also lowers. Therefore, discharge in the section 13 must be prevented by preventing the maximum applied voltage (BV) from exceeding the discharge starting voltage V in the atmosphere while the distance X1 is kept as short as possible.
Therefore, in this semiconductor device 1, by satisfying the above-described relational expression (1), discharge between anode electrodes 9 adjacent to each other can be reliably prevented while the distance X1 from the end of the pad area 95 to the dicing region 4 (end face 2C of the SiC layer 2) can be made shorter than conventionally.
In detail, as a result of investigation performed by the inventor of the present invention, the relationship between the discharge starting voltage V between two electrodes linked to each other via only the atmosphere and the discharge distance can be shown by the graph (Y=1.053E+03e5.846E-04X) of
The inventor of the present invention further examined the relationship between the maximum applied voltage (BV) Y and the distance X1 in the semiconductor device 1 based on
The graph of
However, even when the coordinates are included in the lower region, the distance X1 is preferably as short as possible to obtain the effects of widening the pad area 95 and increasing the number of semiconductor devices 1 to be obtained.
Therefore, in the present preferred embodiment, as described above, setting is made so that the maximum applied voltage (BV) Y and the distance X1 satisfy the following relational expression (1).
This relational expression (1) shows that the coordinates of the maximum applied voltage (BV) Y and the distance X1 are included in the region (hatched portion) surrounded by X1=855·In (Y/1053) and X1=855·In (Y/1053)+100. Accordingly, while the distance X1 from the end of the pad area 95 to the dicing region 4 (end face 2C of the SiC layer 2) can be made shorter than conventionally, discharge between the anode electrodes 9 adjacent to each other can be reliably prevented.
Next, effects relating to the distance X2 from the end of the connected portion 93 connected to the SiC layer 2 of the anode electrode 9 to the end face 2C are described.
In the semiconductor device 1, the distance X2 from the end of the connected portion 93 (outer peripheral edge of the contact hole 84) connected to the SiC layer 2 of the anode electrode 9 to the end face 2C is preferably longer than the width X3 of the depletion layer 14 spreading in the horizontal direction along the first surface 2A of the SiC layer 2 from the connected portion 93 when the maximum applied voltage (BV) is applied to the Schottky barrier diode. In the manufacturing process, the contact hole 84 is formed in the first layer 81, and when the anode electrode 9 is connected to the SiC layer 2 through the contact hole 84 (Step S3), the relative position of the connected portion 93 to the dicing region 4 is preferably set so that the distance X2 from the end of the connected portion 93 to the dicing region 4 becomes longer than the width X3 of the depletion layer 14.
Generally, it is said that the depletion layer in the semiconductor layer composed of SiC extends to two times in the orthogonal direction (horizontal direction) with respect to the thickness direction (vertical direction) of the semiconductor layer. If the distance X2 is shorter than the width X3 of the depletion layer 14 at the time of application of the maximum applied voltage (BV), when the maximum applied voltage (BV) is applied to each of the semiconductor devices 1 divided into individual pieces, the depletion layer 14 may extend to the end face 2C of the SiC layer 2. Therefore, in this semiconductor device 1, by setting distance X2>width X3, the depletion layer 14 can be prevented from reaching the end face 2C of the SiC layer 2.
In addition, in this semiconductor device 1, the outer peripheral edge of the contact hole 85 which becomes the starting point of the distance X1 is positioned on the inner side of the end faces 2C of the SiC layer 2 as compared with the outer peripheral edge of the contact hole 84 which becomes the starting point of the distance X2. Therefore, the distance X1 and the distance X2 satisfy X1>X2. Accordingly, by making the setting so that the distance X2 satisfies the relational expression (1) described above, the depletion layer 14 can be prevented from being exposed at the end face 2C, and simultaneously, discharge between the anode electrodes 9 adjacent to each other can be reliably prevented.
Here, an example of the distance X2 is described by using detailed numerical values. For example, when the impurity concentration of the epitaxial layer 6 is 7×1015 cm−3 and the thickness thereof is 7 μm, the maximum applied voltage (BV) becomes 1450V, theoretically. In this case, theoretically, the depletion layer 14 extends by 15.2 μm in the vertical direction of the epitaxial layer 6. Therefore, theoretically, the width X3 in the horizontal direction of the depletion layer 14 becomes 30.4 μm. The distance X2 is required to be longer than the width X3, so that under this condition, the distance becomes X2>30.4 μm.
On the other hand, when the maximum applied voltage (BV) is 1450 V, the discharge distance in the atmosphere becomes 550 μm with reference to
That is, to prevent only exposure of the depletion layer 14 at the end face 2C, the distance becomes X2>30.4 μm, and to prevent discharge between the anode electrodes 9 adjacent to each other as well, the distance X2>275 μm.
In the semiconductor device 102 shown in
In the first preferred embodiment, the voltage relaxing layer 7 is formed to overlap the outer peripheral edge 83 of the second layer 82 in the thickness direction of the SiC layer 2, however, the voltage relaxing layer 7 may be formed so as not to overlap the outer peripheral edge 83 of the second layer 82 as in the semiconductor device 104 (fourth preferred embodiment) shown in
In the first preferred embodiment, only the first layer 81 is formed on the entire first surface 2A of the SiC layer 2, however, both of the first layer 81 and the second layer 82 may be formed on the entire first surface 2A of the SiC layer 2 as in the semiconductor device 105 (fifth preferred embodiment) shown in
In the semiconductor devices 106 and 107 shown in
In the first preferred embodiment, the insulating layer 8 has a structure composed of a plurality of layers including the first layer 81 and the second layer 82, however, the insulating layer 8 may have a structure composed of a single layer as in the semiconductor device 108 (eighth preferred embodiment) shown in
In the first preferred embodiment described above, the semiconductor device structure formed in the SiC layer 2 is a Schottky barrier diode structure including the SiC layer 2 and the anode electrode 9 that forms a Schottky barrier between the anode electrode and the SiC layer 2, however, in the semiconductor device 109 shown in
The MIS transistor structure includes the SiC layer 2, the p-type channel regions 16, the n+-type source regions 17, the p+-type channel contact regions 18, the gate insulating film 19, and the gate electrode 20. The semiconductor device 109 further includes an interlayer insulating film 21 and a source electrode 22 as a surface electrode as components attached to the MIS transistor structure.
The channel regions 16 are selectively formed in the surficial portion of the epitaxial layer 6, for example, in a plurality of regions discretely disposed periodically on the SiC layer 2. The channel regions 16 may be disposed in, for example, a matrix, zigzag, or striped manner.
The source region 17 is formed in the inside region of the channel region 16. In this region, the source region 17 is selectively formed in the surficial portion of the channel region 16. The source region 17 is formed inside the channel region 16 so as to be positioned at a predetermined distance inward from the interface between the channel region 16 and the epitaxial layer 6. Accordingly, in the surface layer region of the semiconductor layer including the epitaxial layer 6 and the channel region 16, etc., between the source region 17 and the epitaxial layer 6, the surficial portion of the channel region 16 is interposed, and this interposed surficial portion provides the channel portion 23.
The channel contact region 18 penetrates through the source region 17 and is connected to the channel region 18.
The gate insulating film 19 may be composed of, for example, a silicon oxide film, a silicon nitride film, a silicon oxynitride film, a hafnium oxide film, an alumina film, a tantalum oxide film, or the like. The gate insulating film 19 is formed to cover at least the surfaces of the channel regions 16 in the channel portions 23.
The gate electrode 20 is formed such that it is opposed to the channel portions 23 via the gate insulating film 19. The gate electrode 20 may be composed of polysilicon that is reduced in resistance by implanting impurities. In the present preferred embodiment, the gate electrode 20 is formed in almost the same pattern as that of the gate insulating film 19, and covers the surfaces of the gate insulating film 19. Accordingly, a planar gate structure is arranged.
The interlayer insulating film 21 can be formed as, for example, an extended portion obtained by extending the first layer 81 along the first surface 2A of the SiC layer 2. The interlayer insulating film 21 is formed in a pattern that covers the upper surfaces and side surfaces of the gate electrode 20 and has contact holes 24 in the central regions of the channel regions 16 and the inner edge regions of the source regions 17 linked to the central regions.
The source electrode 22 is composed of aluminum (Al) and other metals. The source electrode 22 is formed to cover the surfaces of the interlayer insulating film 21 and be embedded in the contact holes 24. Accordingly, the source electrode 22 forms an ohmic contact with the source regions 17. Examples of a planar profile of the source electrode 22 are shown in
This semiconductor device 109 may include a p-type annular region 29 surrounding the MIS transistor structure in the SiC layer 2, and a contact region 30 formed in the surficial portion of the annular region 29. The annular region 29 and the contact region 30 may be exposed from the contact hole 84 of the first layer 81. That is, when the source electrode 22 as a surface electrode is connected at a plurality of points of the SiC layer 2, the outermost connected portion (in the present preferred embodiment, the connected portion 222 connected to the contact region 30) corresponds to the “connected portion connected to the SiC layer of the surface layer” of the present invention.
In this ninth preferred embodiment, a planar gate structure is described as an example of the MIS transistor structure, however, the MIS transistor structure may be a trench-gate structure.
Preferred embodiments of the present invention are described above, however, the present invention may be carried out in other modes.
For example, an arrangement in which the conductive-types of each semiconductor portion of the semiconductor devices 1, 101 to 109 are inverted may be adopted. For example, in the semiconductor device 1, the p-type portion may be n-type, and the n-type portion may be p-type.
The voltage relaxing layer 7 must be a conductive-type (in the preferred embodiments described above, p-type) different from that of the SiC layer 2, however, in the inside region thereof, may have a portion the conductive-type of which is the same as that of the SiC layer 2. For example, the voltage relaxing layer 7 may have an n-type region positioned at a predetermined distance inward from the interface between the voltage relaxing layer 7 and the epitaxial layer 6. As the n-type region is formed, for example, in the arrangement shown in
The semiconductor device (semiconductor power device) according to the present invention can be installed in a power module that is used in an inverter circuit constituting a drive circuit for driving an electric motor to be used as a power source of, for example, an electric vehicle (including a hybrid vehicle), train, an industrial robot, or the like. The semiconductor device can also be installed in a power module to be used in an inverter circuit that converts electric power generated by a power generator (in particular, private electric generator) such as a solar cell, a wind power generator, or the like, so as to match electric power of a commercial power source.
The features understood from the disclosure of the above-described preferred embodiments may also be combined between different preferred embodiments. Furthermore, the components presented in the preferred embodiments may be combined within the scope of the present invention.
In addition, various design changes are possible within the scope of the matters described in the claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2012-054953 | Mar 2012 | JP | national |
| Number | Date | Country | |
|---|---|---|---|
| Parent | 15419565 | Jan 2017 | US |
| Child | 16242744 | US | |
| Parent | 14384598 | Sep 2014 | US |
| Child | 15419565 | US |
