Semiconductor device having a layout that reduces current congestion at endpoints - Patents.com
The obround-shaped emitter and base regions with strategic gaps and setbacks, along with dielectric trenches, address current congestion and electric field issues in B-TRANs, improving breakdown voltage and reducing leakage current.
Patent Information
- Application Number
- JP2023530294
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Priority Date
- 2020-11-19
- Filing Date
- 2021-11-11
- Publication Date
- 2024-11-26
- Estimated Expiration
- 2041-11-11
AI Technical Summary
Bidirectional bipolar junction transistors (B-TRANs) experience current congestion and reduced reverse bias breakdown voltage due to uniform spacing between base and emitter regions, leading to increased leakage current and lower than expected breakdown voltages.
The layout of the semiconductor device incorporates obround-shaped emitter regions with hemispherical ends and base regions spaced with gaps greater than the radius of curvature, along with base contacts set back from the ends by a distance equal to or greater than the radius, and trenches of dielectric material to reduce electric field strength and enhance reverse bias breakdown voltage.
The proposed layout effectively reduces current congestion and electric field strength at the base region edges, resulting in improved reverse bias breakdown voltage and decreased leakage current, enhancing the overall performance of B-TRAN devices.
Smart Images

Figure 0007589814000001 
Figure 0007589814000002 
Figure 0007589814000003
Abstract
Description
[Technical field]
[0001] This application relates to semiconductor devices, and more particularly to transistor layouts that reduce current crowding at the endpoints. [Background technology]
[0002] A bidirectional bipolar junction transistor (B-TRAN) is a junction transistor with a base and collector-emitter on a first side of a bulk region and a separate base and collector-emitter on a second side of the bulk region opposite the first side. When properly configured by an external driver, current can selectively flow through the B-TRAN in either direction, so a B-TRAN device is considered a bidirectional device. Based on the bidirectional nature, whether the collector-emitter is considered a collector (i.e., current flow into the B-TRAN) or an emitter (i.e., current flow out of the B-TRAN) depends on the applied external voltage and therefore on the direction of current flow through the B-TRAN. Summary of the Invention
[0003] For NPN type devices, the collector-emitter regions on each side of the semiconductor substrate can be considered to form a PN junction with the bulk region. One of the manufacturing performance tests is to measure the reverse bias breakdown voltage of the PN junctions on each side of the semiconductor substrate. If the reverse bias breakdown voltage of the PN junctions is too low, it can degrade the overall performance of the B-TRAN.
[0004] One example is a semiconductor device comprising: an emitter region defining an interior boundary in the shape of an obround having parallel sides, the obround having a first hemispherical end and a second hemispherical end, each having a radius; a base region having a first end, a second end opposite the first end, and a base length, the base length being parallel to the parallel sides and centered between the parallel sides and disposed within the obround; the first end is spaced from the first hemispherical end by a first gap, the first gap being larger than the radius by an amount that is greater than a manufacturing tolerance; and the second end is spaced from the second hemispherical end by a second gap, the second gap being larger than the radius by an amount that is greater than a manufacturing tolerance.
[0005] In an exemplary semiconductor device, the first gap may be at least 50% longer than the radius. In an exemplary semiconductor device, the first gap may be at least 100% longer than the radius.
[0006] The exemplary semiconductor device may further include a base contact electrically coupled to the base region through the base window by a dielectric. The base contact defines a length parallel to the base length and a first end that is an end closest to the first end, the first end being spaced from the first end by an amount of a setback distance at least equal to the radius. The setback distance may be at least 50% longer than the radius. The setback distance may be at least 100% longer than the radius.
[0007] The exemplary semiconductor device may further include a bottom side having an emitter region and a base region. The semiconductor device may define a bidirectional double-base bipolar junction transistor.
[0008] An exemplary semiconductor device may have a base region that is P-type and an emitter region that is N-type.
[0009] The exemplary semiconductor device may further include a trench of dielectric material surrounding the base region. The trench may have a depth between 10 μm and 15 μm, inclusive, and a width between 3 μm and 5 μm, inclusive.
[0010] A second exemplary semiconductor device includes an emitter region defining an interior boundary in the shape of an obround having parallel sides, the obround having a first hemispherical end and a second hemispherical end, each having a radius; a base region having a first end, a second end opposite the first end, and a base length, the base length being parallel to the parallel sides and centered between the parallel sides and disposed within the obround; the first end is spaced from the first hemispherical end by a first gap; the second end is spaced from the second hemispherical end by a second gap; and a base contact electrically coupled to the base region; the base contact defines a length parallel to the base length and a first end that is an end closest to the first end; the first end is spaced from the first end by an amount of a setback distance at least equal to the radius.
[0011] In the second exemplary semiconductor device, the setback distance can be at least 50% longer than the radius. In the second exemplary semiconductor device, the setback distance can be at least 100% longer than the radius.
[0012] In the second exemplary semiconductor device, the first gap may be larger than the radius by an amount greater than manufacturing tolerances. The second gap may be larger than the radius by an amount greater than manufacturing tolerances. The first gap may be at least 50% longer than the radius. Alternatively, the first gap may be at least 100% longer than the radius.
[0013] The second exemplary semiconductor device may further include a bottom side having an emitter region and a base region. The semiconductor device may define a bidirectional double-base bipolar junction transistor.
[0014] A second exemplary semiconductor device may have a base region that is P-type and an emitter region that is N-type.
[0015] The second exemplary semiconductor device may further include a trench of dielectric material surrounding the base region, the trench may have a depth between 10 μm and 15 μm, inclusive, and a width between 3 μm and 5 μm, inclusive. [Brief description of the drawings]
[0016] For a detailed description of the exemplary embodiments, reference will now be made to the accompanying drawings, listed below.
[0017] [Figure 1] 1 illustrates a partially cross-sectional elevation view of a B-TRAN according to at least some embodiments. [Diagram 2] 1A-1D show top-side views of a semiconductor substrate during intermediate stages in the construction of a related art device. [Diagram 3] 1A-1C show top views of a top side of a semiconductor substrate during intermediate stages of construction according to at least some embodiments. [Figure 4] 1A-1D show top-side views of a semiconductor substrate during intermediate stages in the construction of a related art device. [Diagram 5] 1A-1C show top views of a top side of a semiconductor substrate during intermediate stages of construction according to at least some embodiments. [Figure 6] 6 illustrates a cross-sectional view taken substantially along line 6-6 of FIG. 5, according to at least some embodiments. [Figure 7] 1 illustrates a partially cross-sectional elevation view of a B-TRAN according to at least some embodiments. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0018] [Definition] Various terms are used to refer to specific system components. Different companies may call a component by different names. This specification does not intend to distinguish between components that have different names but not different functions. In the following description and claims, the terms "including" and "comprising" are used in an open format and therefore should be interpreted to mean "including, but not limited to." Additionally, the terms "couple" or "coupled" are intended to mean an indirect or direct connection. Thus, when a first device couples with a second device, the connection may be a direct connection or an indirect connection via other devices or connections.
[0019] "About" with respect to a stated parameter shall mean plus or minus ten percent (±10%) of the stated parameter.
[0020] "Bidirectional double base bipolar junction transistor" shall mean a junction transistor having a base and a collector-emitter on a first face or side of a bulk region and a base and a collector-emitter on a second face or side of the bulk region. The base and collector-emitter on the first side are different from the base and collector-emitter on the second side. The normal vector to the first side, which is outward, points in the opposite direction to the normal vector to the second side, which is outward.
[0021] "Upper base" means the base of a bidirectional double base bipolar junction transistor that is on the first side of the bulk region of the transistor and shall not be read to mean the position of the base relative to gravity.
[0022] "Lower base" means the base of a bidirectional double base bipolar junction transistor that is on a second side opposite the first side of the bulk region of the transistor and shall not be read to mean the position of the base with respect to gravity.
[0023] "Upper collector-emitter" means the collector-emitter of a bidirectional double-base bipolar junction transistor on the first side of the bulk region of the transistor and shall not be read to mean the location of the base relative to gravity.
[0024] "Lower collector-emitter" means that the collector-emitter of a bidirectional double-base bipolar junction transistor is on a second side of the bulk region of the transistor opposite the first side, and shall not be read to mean the location of the base with respect to gravity.
[0025] [Detailed Description] This application claims the benefit of U.S. Provisional Application No. 63 / 116,078, filed November 19, 2020, entitled "Layout Design to Mitigate Current Crowding at Device Endpoints," which is hereby incorporated by reference as if fully set forth below.
[0026] The following description is directed to various embodiments of the present invention. Although one or more of these embodiments may be preferred, the disclosed embodiments shall not be interpreted as limiting the scope of the disclosure, including the claims, or used otherwise. In addition, those skilled in the art will understand that the following description has a broad application, and the description of any embodiment is intended to be only an example of that embodiment, and is not intended to mean that the scope of the disclosure, including the claims, is limited to that embodiment.
[0027] Various examples are directed to layouts that reduce current crowding at the ends of structures in semiconductor devices such as bidirectional double-base bipolar junction transistors (B-TRANs). In particular, in an exemplary system, each emitter region is in the shape of an obround having straight sides and hemispherical ends to define an emitter region having an internal boundary. A base region is disposed within the obround. The base regions have a base length and are disposed at opposing ends. In an exemplary system, each end of the base region has an offset or gap from its respective hemispherical end by a distance greater than the radius of the hemispherical end. The gap reduces current crowding at the ends of the base region, and this reduction in current crowding results in a lower electric field at the ends of the base region, which in turn results in a higher reverse bias breakdown voltage of the junction formed between the base and emitter. An example B-TRAN device is provided first as a guide for interpretation.
[0028] FIG. 1 illustrates a partial cross-sectional elevation of an exemplary B-TRAN. In particular, FIG. 1 illustrates a B-TRAN 100 having an upper surface or upper side 102 and a lower surface or lower side 104. The designations "upper" and "lower" are arbitrary and are used merely for convenience of description. The upper side 102 faces in the opposite direction to the lower side 104. In other words, an outward vector perpendicular to the average height of the upper side 102 (not specifically shown) faces in the opposite direction to an outward vector perpendicular to the average height of the lower side 104 (not specifically shown).
[0029] The top side 102 includes collector-emitter regions 106 that form a junction with a drift region or bulk substrate 108. The top side 102 further defines a base region 110 disposed between the collector-emitter regions 106. The collector-emitter regions 106 are electrically coupled to collector-emitter contacts 112, such as a metallic material applied through a window in an insulating material (not specifically shown) covering the top side 102. Similarly, the base region 110 is electrically coupled to a base contact 114, such as a metallic material applied through a window in an insulating material (not specifically shown) covering the top side 102. Although only two collector-emitter contacts 112 and only one base contact 114 are shown in the embodiment seen in FIG. 1, more than one collector-emitter contact 112 may be implemented and more than one base contact 114 may be implemented in an exemplary system, as described in further detail below. In the exemplary system, the collector-emitter contacts 112 are coupled together to form an upper collector-emitter 116. The base contacts 114 are coupled together to form an upper base 118.
[0030] Similarly, the lower side 104 includes a collector-emitter region 120 that forms a junction with the bulk substrate 108 and a collector-emitter contact 122 that is electrically coupled to the collector-emitter region 120. The collector-emitter contact 122 is coupled to form a lower collector-emitter 124. The lower side 104 includes a base region 126 and a base contact 128 that is electrically coupled to the base region 126. The base contacts 128 are coupled together to form a lower base 130.
[0031] The exemplary B-TRAN 100 is an NPN structure, meaning that the collector-emitter regions 106 and 120 are N-type, the base regions 110 and 126 are P-type, and the bulk substrate 108 is P-type. It is further noted that a PNP-type B-TRAN device is also contemplated, however, to avoid overly lengthening the description, a PNP-type B-TRAN device is not specifically illustrated.
[0032] In an exemplary case, the various structures and dopings associated with the top side 102 are intended to be mirror images of the various structures and dopings associated with the bottom side 104. However, in some cases, the various structures and dopings associated with the top side 102 are constructed at different times than the various structures and dopings associated with the bottom side 104, and thus slight differences in the structures and dopings may exist as between the two of the top and bottom sides. Thus, these differences may be due to variations in manufacturing tolerances, but do not adversely affect the operation of the device as a bidirectional double base bipolar junction transistor.
[0033] FIG. 2 shows an overhead view of the top side of a semiconductor substrate at an intermediate stage in the fabrication of a related art device. In particular, visible in FIG. 2 is emitter region 200. Emitter region 200 defines several interior regions that are not doped as emitter regions, such as interior region 202. Defined within exemplary interior region 202 is base region 204. In related art devices, the separation S between base region 204 and the surrounding emitter region 200 is uniform, as shown. In other words, the separation S between the ends of base region 204 is the same as the separation S along the long dimension of base region 204.
[0034] However, devices constructed with the layout shown in FIG. 2 have been found to have lower than expected reverse bias breakdown voltages, and higher than expected reverse bias leakage currents. As an example, consider that base region 204 is a P-type region and emitter region 200 is an N-type region. Thus, one may consider a PN junction (e.g., a diode) to be formed from base region 204 to emitter region 200, although it is part of a larger overall structure. At intermediate stages of fabrication, various characteristics of the device may be tested, such as the forward voltage drop and reverse bias breakdown voltage of the PN junction formed by base region 204 and emitter region 200. Testing of related art structures having the layout shown in FIG. 2 has shown them to have lower than expected reverse bias breakdown voltages, and a corresponding increased reverse bias leakage current.
[0035] The inventors herein have found that the lower than expected reverse bias breakdown voltage and corresponding increased reverse bias leakage current are at least partially due to the layout, such as between the emitter region 200 and the base region 204. In particular, the inventors herein have found that by implementing the same separation S at the ends of the base region 204 (such as at location 206) as is implemented along the long dimension of the base region 204, the electric field strength is increased, such as at location 208, as compared to straight sections. In other words, the uniform spacing S between the base region 204 and the emitter region 200 causes current bunching at the ends, such as at location 206. The increased electric field strength, which is due to the layout, results in breakdown at a lower reverse bias voltage than expected. That is, for a particular applied voltage at the base region 204, the electric field strength at location 206 is higher than the electric field strength along straight sections, such as at location 208. As a result, the reverse bias breakdown voltage is lower than desired (e.g., breakdown at 30-40V instead of the designed 60-90V).
[0036] FIG. 3 illustrates an overhead view of the top side 102 of a semiconductor substrate at an intermediate stage of construction, according to at least some embodiments. Again, illustrating the top side 102 is arbitrary, since in a symmetrical device the top side 102 and bottom side 104 (FIG. 1) would appear identical. For context, the cross-sectional view of FIG. 1 may be considered to be taken along line 1-1 in FIG. 3. However, it should be noted that FIG. 3 illustrates the top side 102 of the semiconductor substrate prior to metal deposition to form the collector-emitter contacts 112 (FIG. 1) and base contact 114 (FIG. 1).
[0037] In particular, visible in FIG. 3 is the collector-emitter region 106. The collector-emitter region 106 defines several interior regions, such as interior region 300, that are not doped as collector-emitter regions. Each interior region 300 is defined by an interior boundary of the collector-emitter region 106, such as interior boundary 302. In some cases, interior boundary 302 is an interior boundary of a doped region that forms the collector-emitter region 106. In the exemplary case, interior boundary 302 defines a racetrack pattern or oblong that is comprised of a first linear side 304 that is parallel to and offset from a second linear side 306, as shown. Each of the exemplary linear sides 304 and 306 has a length L along its linear portion, as shown. S 3. Exemplary interior boundary 302 further defines a first hemispherical end 308 and a second hemispherical end 310 opposite first hemispherical end 308. Referring to hemispherical end 308 as representative, hemispherical end 308 defines a center of curvature 312 and a radius of curvature 314. That is, radius of curvature 314 sweeps out and defines the outer boundary of the circumference of hemispherical end 308. The total distance D between straight sides 304 and 306 (measured perpendicular to the straight sides) is twice radius of curvature 314. Again, hemispherical end 308 is representative of all hemispherical ends.
[0038] 3 further illustrates a respective base region disposed within each interior region. Referring to interior region 300 as representative, base region 110 is disposed within interior region 300. Exemplary base region 110 includes a first end 316 associated with hemispherical end 308, a second end 318 opposite first end 316 and associated with hemispherical end 310, and a base length L measured parallel to straight sides 304 and 306. B As shown, the exemplary base region 110 includes a base length L B 3. The base region 110 is disposed within the obround with straight sides 304 and 306 parallel to and centered between them. In some cases, the outer boundary 320 of the base region 110 is the outer boundary of the doped region that forms the base region 110. In the illustrated example, the outer boundary 320 of the base region 110 itself defines the obround. The obround has straight sides that run parallel to one another and ends 316 and 318 that are rounded or hemispherical.
[0039] According to various examples, first end 316 is spaced from hemispherical end 308 by a first gap G1 that is greater than radius of curvature 314. Similarly, second end 318 is spaced from hemispherical end 310 by a second gap G2 that is greater than radius of curvature 314. In some cases, gaps G1 and G2 are approximately the same. In other words, in the exemplary case, the layout of base region 110 is configured and constructed such that ends 316 and 318 have a clearance or gap that is greater than the radius of curvature that defines hemispherical ends 308 and 310, respectively, and the gap is greater than the radius of curvature by an amount greater than the manufacturing tolerance of the device (e.g., by an amount greater than 1 μm). Assuming that all radii of curvature are the same length (not strictly necessary), and further assuming that all gaps are the same length (not strictly necessary), the gaps can be at least 50% longer than the radii of curvature (e.g., gap = radius x 1.5), and in some cases can be 100% longer than the radii of curvature (e.g., gap = radius x 2.0).
[0040] By using a gap larger than the radius of curvature of the associated hemispherical end, the electric field strength can be lower than if the gap were approximately equal to the radius of curvature. Furthermore, a lower electric field strength at the end of the base region 110 (e.g., 1 μm beyond the boundary of the base region 110) can make the electric field strength approximately the same as the electric field strength along the straight sides of the outer boundary 320 of the base region 110. The lower electric field strength reduces the likelihood of breakdown initiation from both ends of the base region 110, reducing leakage current.
[0041] In some cases, addressing the reverse bias breakdown voltage using the gap alone may be sufficient, however, the inventors herein believe that additional factors also contribute to a lower than expected reverse bias breakdown voltage, such as the placement of metal contacts associated with the exemplary base region 110.
[0042] FIG. 4 illustrates an overhead view of the top side of a semiconductor substrate at an intermediate stage of construction of a related art device. In particular, visible in FIG. 4 is an emitter region 200, along with several interior regions that are not doped as emitter regions, such as interior region 202. Defined within exemplary interior region 202 is base region 204, which is only partially visible in FIG. 4 because base contact 400 partially obscures base region 204. Also visible in FIG. 4 are several emitter contacts 402. In the related art device, the separation S between base region 204 and the surrounding emitter region 200 is uniform, as shown. Additionally, the end of base contact 400 closest to the hemispherical portion of emitter region 200 has approximately the same separation S relative to the hemispherical portion of emitter region 200. In other words, the separation S between the edge of the base contact 400 and the associated hemispherical portion of the emitter region 200 is approximately the same as the separation S between the base contact 400 and the base region 204 along the long dimension.
[0043] The inventors herein believe that the close proximity, if not coextensive, of the ends of the base contact 400 to the ends of the base region 204 may contribute to a lower than expected reverse bias breakdown voltage. In particular, when the base contact 400 is electrically coupled to the base region 204, charge carriers (e.g., electrons) injected into the base region 204 do not experience an appreciable voltage drop as the charge carriers propagate to the ends of the base region 204. The higher the voltage at the ends of the base region 204, the greater the electric field associated with current crowding at the ends of the base region. Additionally, the charge carriers in the metal base contact 400 also create an electric field with respect to the hemispherical portion of the emitter region 200. Although those charge carriers may not directly cross the depletion region around the base region 204, the additional electric field may hasten breakdown in the depletion region between the base region 204 and the emitter region 200.
[0044] 5 shows an overhead view of the top side 102 of a semiconductor substrate at an intermediate stage of construction, according to at least some embodiments. Again, the top side 102 is shown arbitrarily, since in a symmetric device the top side 102 and bottom side 104 (FIG. 1) would appear identical. For context, the cross-section of FIG. 1 may be considered to be taken along line 1-1 in FIG. 5, including the base and collector-emitter contacts.
[0045] In particular, visible in FIG. 5 is the collector-emitter region 106. The collector-emitter region 106 defines several interior regions, such as interior region 300, that are not doped as collector-emitter regions. Each interior region 300 is defined by an interior boundary 302 of the collector-emitter region 106. As before, the interior boundary 302 may be an interior boundary of a doped region that forms the collector-emitter region 106. And, as before, the exemplary interior boundary 302 defines an obround shape that includes straight sides 304 and 306 and an exemplary hemispherical end 308. With reference to the hemispherical end 308 as a representative, the hemispherical end 308 defines a center of curvature 312 and a radius of curvature 314. That is, the radius of curvature 314 sweeps out and defines an outer boundary of the obround of the exemplary hemispherical end 308. As with the previous example, the first end 316 is spaced from the hemispherical end 308 by a first gap G1 that is greater than the radius of curvature 314. In other words, in the exemplary case, the layout of the base region 110 is configured and constructed such that the exemplary end 316 has a gap G1 that is greater than the radius of curvature that defines the exemplary hemispherical end 308.
[0046] Also visible in FIG. 5 are a number of base contacts, such as base contact 114. The exemplary base contact 114 is disposed over and electrically coupled to the underlying base region 110. In particular, during fabrication, the collector-emitter region 106 and base region 110 may be constructed, and then a dielectric layer (e.g., an oxide layer) may be grown or deposited over the entire top side 102, although the dielectric layer is not specifically shown in FIG. 5 so as not to obscure the underlying regions. Using various photolithographic and etching techniques, windows may be opened through the dielectric layer to expose the underlying collector-emitter region 106 and base region 110. A metal layer may then be sputtered or evaporated to cover the top side 102, with the metal electrically contacting the underlying regions. Again using various photolithographic and etching techniques, much of the metal may be removed, leaving only the collector-emitter contact 112 and base contact 114 (and other contacts not specifically labeled) shown in FIG. 5. The exemplary base contact 114 has a base length LB (FIG. 3) and defines a first termination 500. The first termination 500 is associated with the exemplary first end 316 of the base region 110 and thus with the exemplary hemispherical end 308. More specifically, the first termination 500 is the termination closest to the first end 316 of any base contact 114 coupled to a particular underlying base region 110. In the illustrated example, the base contacts 114 and associated base contacts along the shared base region 110 are periodically separated to reduce stresses associated with temperature-based differential expansion, such as between metal and semiconductor layers. However, in other cases, the base contacts may be continuous structures along and above the respective base regions 110.
[0047] In the illustrated layout, the first termination 500 has a setback distance S measured along the long dimension of the base region 110. D 3, and in some examples may be spaced apart from the first end 316 by a setback distance S D is at least equal to the radius of curvature 314. Assuming that all of the radii of curvature are the same length (not strictly necessary), and further assuming that all of the setback distances are the same length (not strictly necessary), the setback distance S of each base contact 114 is D is at least 50% greater than the radius of curvature (e.g., setback distance=radius×1.5), and in some cases can be 100% greater than the radius of curvature (e.g., setback distance=radius×2.0). In other words, the difference between the gap G1 and the setback distance S, measured parallel to the long dimension of the base region, is D The sum of may be at least 2.5 times the length of the radius of curvature 314.
[0048] A setback distance S that is greater than the radius of curvature of the associated hemispherical end D Using the setback distance S as shown, the electric field strength can be lower compared to when the base contact 114 is coextensive with the base region 110. In one possible explanation, although others are possible, the setback distance S as shown DBy having a base contact 114 having a setback distance S 1 , charge carriers (e.g., electrons) injected into the base region 110 via the base contact 114 experience a non-trivial voltage drop as they propagate through the base region 110 towards the first end 316. The non-trivial voltage drop therefore reduces the voltage at the first end 316 and, as a result, reduces the electric field strength at the exemplary first end 316 of the base region 110 (e.g., 1 μm beyond the boundary of the base region 110). D reduces any electric field contribution from charge carriers in the base contact 114 itself. The lower electric field strength reduces the likelihood of breakdown initiation across the base region 110, and also reduces leakage current.
[0049] Figure 6 illustrates an exemplary cross-sectional view taken substantially along line 6-6 of Figure 5. In particular, Figure 6 illustrates a portion of the collector-emitter region 106 of the top side 102, along with a portion of the collector-emitter region 120 of the bottom side 104. Figure 6 further illustrates the interior boundary 302 associated with the hemispherical end 308. For purposes of description, it will be assumed that the center of curvature 312 is as indicated, and thus a radius of curvature R extends between the center of curvature 312 and the interior boundary 302.
[0050] 6 further illustrates a portion of base region 110 and base contact 114 of top side 102 and a portion of base region 126 and base contact 128 of bottom side 104. FIG 6 illustrates that in various examples, a base region, such as base region 110, can have a gap G that is greater than radius of curvature R, and in the illustrated example, gap G is about twice radius of curvature R. Additionally, FIG 6 illustrates that in various examples, a base contact, such as base contact 114, can have a setback distance S that is at least equal to radius of curvature R. D This indicates that the
[0051] In the various embodiments described thus far, it has been assumed that the volume between the base region and the collector-emitter region on each side of the device contains only the P-type bulk substrate, but in other cases other structures may be present.
[0052] FIG. 7 illustrates a partial cross-sectional elevation view of an exemplary B-TRAN. In particular, FIG. 1(7) illustrates a B-TRAN 100 having a top side 102 and a bottom side 104. The top side 102 includes a collector-emitter region 106 that forms a junction with a drift region or bulk substrate 108. The top side 102 further defines a base region 110 disposed between the collector-emitter regions 106. As before, the collector-emitter region 106 is electrically coupled to a collector-emitter contact 112. Similarly, the base region 110 is electrically coupled to a base contact 114. Although only two collector-emitter contacts 112 and only one base contact 114 are shown in the embodiment seen in FIG. 7, more than one collector-emitter contact 112 may be implemented and more than one base contact 114 may be implemented in an exemplary system. The collector-emitter contacts 112 are coupled together to form an upper collector-emitter 116. The base contacts 114 are coupled together to form an upper base 118.
[0053] Similarly, the lower side 104 includes a collector-emitter region 120 that forms a junction with the bulk substrate 108 and a collector-emitter contact 122 that is electrically coupled to the collector-emitter region 120. The collector-emitter contact 122 is coupled to form a lower collector-emitter 124. The lower side 104 includes a base region 126 and a base contact 128 that is electrically coupled to the base region 126. The base contacts 128 are coupled together to form a lower base 130. As in FIG. 1, in FIG. 7, the exemplary B-TRAN 100 is an NPN structure, so that the collector-emitter regions 106 and 120 are N-type, the base regions 110 and 126 are P-type, and the bulk substrate 108 is P-type. It is further noted that PNP-type B-TRAN devices are also contemplated. However, to avoid overly lengthy description, PNP-type B-TRAN devices are not specifically illustrated.
[0054] 7, the exemplary B-TRAN 100 further includes additional structures between the base region and the collector-emitter region. Referring first to the top side 102, there is a trench 700 between the base region 110 and the surrounding collector-emitter region 106. In an exemplary case, the trench may have a depth between 10 μm and 15 μm, inclusive, although other trench depths are contemplated having sufficient depth to be "below" the bottom of the respective regions. In an exemplary case, the trench 700 further includes a width between 3 μm and 5 μm, inclusive, although larger and smaller widths are also contemplated. In an exemplary case, the trench includes a dielectric material (e.g., oxide) that electrically insulates the base region 110 from the collector-emitter region 106. The trench 700 may be used in conjunction with the gaps and setbacks described above to improve or increase the reverse bias breakdown voltage and reduce leakage current of the exemplary B-TRAN 100.
[0055] Similarly for the bottom side 104, the exemplary B-TRAN 100 includes additional structures between the base region and the collector-emitter region. In this example, a trench 702 exists between the base region 126 and the surrounding collector-emitter region 120. As with the top side 102, the trench 702 may have a depth between 10 μm and 15 μm, inclusive, and a width between 3 μm and 5 μm, inclusive. The exemplary trench is filled with a dielectric material (e.g., oxide) that electrically insulates the base region 126 from the collector-emitter region 120. The trench 702 may be used in conjunction with the gaps and setbacks described above to improve or increase the reverse bias breakdown voltage and reduce leakage current of the exemplary B-TRAN 100.
[0056] In the exemplary cross-sectional view of Figure 7, trenches 700 and 702 span the volume between each base region and the surrounding collector-emitter region on that same side. However, in other cases, trenches 700 and 702 may span only a portion of the volume. In one exemplary case, the trenches surround the base region and abut or closely abut the base region. This may allow additional depletion regions to extend from the outer surface of each trench into the surrounding collector-emitter region.
[0057] The above description is intended to illustrate the principles and various embodiments of the present invention. Numerous variations and modifications will become apparent to those skilled in the art once the above disclosure is fully appreciated. For example, the various structures may be implemented in any semiconductor device having structures intercombined with one another. It is intended that the following claims be interpreted to include all such variations and modifications.
Claims
1. an emitter region defining an interior boundary in the shape of an obround having parallel sides, the obround having first and second hemispherical ends, each end having a radius; a base region having a first end, a second end opposite the first end, and a base length, the base length being parallel to the parallel sides and centered between the parallel sides and disposed within the obround; the first end is spaced from the first hemispherical end by a first gap, the first gap being greater than the radius by an amount greater than a manufacturing tolerance; the second end is spaced from the second hemispherical end by a second gap, the second gap being greater than the radius by an amount greater than the manufacturing tolerance.
2. The semiconductor device of claim 1 , wherein the first gap is at least 50% longer than the radius.
3. a base contact electrically coupled to the base region through the base window by a dielectric; the base contact defines a length parallel to the base length and a first termination that is an end closest to the first end; The semiconductor device of claim 1 , wherein said first termination is spaced from said first end by an amount of a setback distance at least equal to said radius.
4. further comprising a lower side having an emitter region and a base region; 10. The semiconductor device of claim 1, wherein the semiconductor device defines a bidirectional double-base bipolar junction transistor.
5. The semiconductor device of claim 1 , further comprising a trench of dielectric material surrounding said base region.
Citation Information
Patent Citations
Semiconductor device
JP2000315695A
High voltage drain-extended transistor
US20050145930A1
Device parameter normalization about a periphery of a non-circular emitter of a lateral bipolar transistor
US20160380086A1
Bipolar layout for improved performance
US5352924A
