Abstract: In a vertical power device with trenched insulated gates, there is an npnp layered structure. The vertical gates turn on the device with a suitable gate bias to conduct a current between a top electrode and a bottom electrode. In an example, implanted n+ source regions are formed in the top surface within a p-well. Between some gates, the overlying dielectric is opened up, by etching, to expose distributed p-type contact regions for the p-well. The dielectric is also opened up to expose areas of the n+ source regions. The top electrode metal directly contacts the exposed p-type contact regions and the n+ source regions to provide distributed emitter-to-base short across the cellular array to improve device performance in the presence of transients. The p-contact regions are isolated from the n+ source regions, prior to the deposition of the metal electrode, due to the p-type contact regions not abutting the n+ source regions.

Abstract: Trenches having a gate oxide layer are formed in the surface of a silicon wafer for vertical gates. Conductive doped polysilicon is then deposited in the trenches to form a relatively thin layer of doped polysilicon along the sidewalls. Thus, there is a central cavity surrounded by polysilicon. Next, the cavity is filled in with a much higher conductivity material, such as aluminum, copper, a metal silicide, or other conductor to greatly reduce the overall resistivity of the trenched gates. The thin polysilicon forms an excellent barrier to protect the gate oxide from diffusion from the inner conductor atoms. The inner conductor and the polysilicon conduct the gate voltage in parallel to lower the resistance of the gates, which increases the switching speed of the device. In another embodiment, a metal silicide is used as the first layer, and a metal fills the cavity.

Abstract: Trenches having a gate oxide layer are formed in the surface of a silicon wafer for vertical gates. Conductive doped polysilicon is then deposited in the trenches to form a relatively thin layer of doped polysilicon along the sidewalls. Thus, there is a central cavity surrounded by polysilicon. Next, the cavity is filled in with a much higher conductivity material, such as aluminum, copper, a metal silicide, or other conductor to greatly reduce the overall resistivity of the trenched gates. The thin polysilicon forms an excellent barrier to protect the gate oxide from diffusion from the inner conductor atoms. The inner conductor and the polysilicon conduct the gate voltage in parallel to lower the resistance of the gates, which increases the switching speed of the device. In another embodiment, a metal silicide is used as the first layer, and a metal fills the cavity.

Abstract: The present application teaches, among other innovations, power semiconductor devices in which breakdown initiation regions, on BOTH sides of a die, are located inside the emitter/collector regions, but laterally spaced away from insulated trenches which surround the emitter/collector regions. Preferably this is part of a symmetrically-bidirectional power device of the “B-TRAN” type. In one advantageous group of embodiments (but not all), the breakdown initiation regions are defined by dopant introduction through the bottom of trench portions which lie within the emitter/collector region. In one group of embodiments (but not all), these can advantageously be separated trench portions which are not continuous with the trench(es) surrounding the emitter/collector region(s).

Abstract: A sacrificial substrate wafer is provided. A low resistivity etch stop layer is formed on or in the top surface of the wafer. The etch stop layer may be a highly doped, p+ type epitaxially grown layer, or an implanted p+ type boron layer, or an epitaxially grown p+ type SiGe layer. Various epitaxial layers, such as an n? type drift layer, and doped regions are then formed over the etch stop layer to form a vertical power device. The starting wafer is then removed by a combination of mechanical grinding/polishing to leave a thinner layer of the starting wafer. A chemical or plasma etch is then used to remove the remainder of the starting wafer, using the etch stop layer to automatically stop the etching. A bottom metal electrode is then formed on the etch stop layer. The etch stop layer injects hole carriers into the drift layer.

Abstract: Methods and systems for power semiconductor devices integrating multiple quasi-vertical transistors on a single chip. Multiple power transistors (or active regions) are paralleled, but one transistor has a lower threshold voltage. This reduces the voltage drop when the transistor is forward-biased. In an alternative embodiment, the power device with lower threshold voltage is simply connected as a depletion diode, to thereby shunt the body diodes of the active transistors, without affecting turn-on and ON-state behavior.

Abstract: In an insulated trench gate device, polysilicon in the trench is etched below the top surface of the trench, leaving a thin gate oxide layer exposed near the top of the trench. An angled implant is conducted that implants dopants through the exposed gate oxide and into the side of the trench. If the implanted dopants are n-type, this technique may be used to extend an n+ source region to be below the top of the polysilicon in the trench. If the implanted dopants are p?type, the dopants may be used to form a p-MOS device that turns on when the polysilicon is biased with a negative voltage. P-MOS and n-MOS devices can be formed in a single cell using this technique, where turning on the n-MOS device turns on a vertical power switch, and turning on the p-MOS device turns off the power switch.

Abstract: Layout to reduce current crowding at endpoints. At least one example is a semiconductor device comprising: an emitter region defining an inner boundary in the shape of an obround with parallel sides, and the obround having hemispherical ends each having a radius; a base region having a first end, a second end opposite the first end, and base length, the base region disposed within the obround with the base length parallel to and centered between the parallel sides, the first end spaced apart from the first hemispherical end by a first gap greater than the radius, and the second end spaced apart from the second hemispherical ends by a second gap greater than the radius.

Abstract: The present application teaches, among other innovations, power semiconductor devices in which breakdown initiation regions, on BOTH sides of a die, are located inside the emitter/collector regions, but laterally spaced away from insulated trenches which surround the emitter/collector regions. Preferably this is part of a symmetrically-bidirectional power device of the “B-TRAN” type. In one advantageous group of embodiments (but not all), the breakdown initiation regions are defined by dopant introduction through the bottom of trench portions which lie within the emitter/collector region. In one group of embodiments (but not all), these can advantageously be separated trench portions which are not continuous with the trench(es) surrounding the emitter/collector region(s).

Abstract: After the various regions of a vertical power device are formed in or on the top surface of an n-type wafer, the wafer is thinned, such as by grinding. A drift layer may be n-type, and various n-type regions and p-type regions in the top surface contact a top metal electrode. A blanket dopant implant through the bottom surface of the thinned wafer is performed to form an n? buffer layer and a bottom p+ emitter layer. Energetic particles are injected through the bottom surface to intentionally damage the crystalline structure. A wet etch is performed, which etches the damaged crystal at a much greater rate, so some areas of the n? buffer layer are exposed. The bottom surface is metallized. The areas where the metal contacts the n? buffer layer form cathodes of an anti-parallel diode for conducting reverse voltages, such as voltage spikes from inductive loads.

Abstract: The present application teaches, among other innovations, power semiconductor devices in which breakdown initiation regions, on BOTH sides of a die, are located inside the emitter/collector regions, but laterally spaced away from insulated trenches which surround the emitter/collector regions. Preferably this is part of a symmetrically-bidirectional power device of the “B-TRAN” type. In one advantageous group of embodiments (but not all), the breakdown initiation regions are defined by dopant introduction through the bottom of trench portions which lie within the emitter/collector region. In one group of embodiments (but not all), these can advantageously be separated trench portions which are not continuous with the trench(es) surrounding the emitter/collector region(s).

Abstract: Methods and systems for double-sided semiconductor device fabrication. Devices having multiple leads on each surface can be fabricated using a high-temperature-resistant handle wafer and a medium-temperature-resistant handle wafer. Dopants can be introduced on both sides shortly before a single long high-temperature diffusion step diffuses all dopants to approximately equal depths on both sides. All high-temperature processing occurs with no handle wafer or with a high-temperature handle wafer attached. Once a medium-temperature handle wafer is attached, no high-temperature processing steps occur. High temperatures can be considered to be those which can result in damage to the device in the presence of aluminum-based metallizations.

Abstract: An npnp layered switch is modified to have a composite anode structure. Instead of the continuous p-type bottom anode layer of a typical npnp IGTO device, thyristor, or IGBT, the composite anode is formed of a segmented p-type layer with gaps containing n-type semiconductor material. The n-type material forms a majority carrier path between the bottom anode electrode and the n-type collector of the vertical npn bipolar transistor. When a trenched gate is biased high, the majority carrier path is created between the bottom anode electrode and the top cathode electrode. Such a current path operates at very low operating voltages, starting at slightly above 0 volts. Above operating voltages of about 1.0 volts, the npnp layered switch operates normally and uses regenerative bipolar transistor action to conduct a vast majority of the current. The two current paths conduct in parallel.

Abstract: In a vertical power device with trenched insulated gates, there is an npnp layered structure. The vertical gates turn on the device with a suitable gate bias to conduct a current between a top electrode and a bottom electrode. In an example, implanted n+ source regions are formed in the top surface within a p-well. Between some gates, the overlying dielectric is opened up, by etching, to expose distributed p-type contact regions for the p-well. The dielectric is also opened up to expose areas of the n+ source regions. The top electrode metal directly contacts the exposed p-type contact regions and the n+ source regions to provide distributed emitter-to-base short across the cellular array to improve device performance in the presence of transients. The p-contact regions are isolated from the n+ source regions, prior to the deposition of the metal electrode, due to the p-type contact regions not abutting the n+ source regions.

Abstract: An npnp layered switch is modified to have a composite anode structure. Instead of the continuous p-type bottom anode layer of a typical npnp IGTO device, thyristor, or IGBT, the composite anode is formed of a segmented p-type layer with gaps containing n-type semiconductor material. The n-type material forms a majority carrier path between the bottom anode electrode and the n-type collector of the vertical npn bipolar transistor. When a trenched gate is biased high, the majority carrier path is created between the bottom anode electrode and the top cathode electrode. Such a current path operates at very low operating voltages, starting at slightly above 0 volts. Above operating voltages of about 1.0 volts, the npnp layered switch operates normally and uses regenerative bipolar transistor action to conduct a vast majority of the current. The two current paths conduct in parallel.

Abstract: A sacrificial substrate wafer is provided. A low resistivity etch stop layer is formed on or in the top surface of the wafer. The etch stop layer may be a highly doped, p+ type epitaxially grown layer, or an implanted p+ type boron layer, or an epitaxially grown p+ type SiGe layer. Various epitaxial layers, such as an n? type drift layer, and doped regions are then formed over the etch stop layer to form a vertical power device. The starting wafer is then removed by a combination of mechanical grinding/polishing to leave a thinner layer of the starting wafer. A chemical or plasma etch is then used to remove the remainder of the starting wafer, using the etch stop layer to automatically stop the etching. A bottom metal electrode is then formed on the etch stop layer. The etch stop layer injects hole carriers into the drift layer.

Abstract: After the various regions of a vertical power device are formed in or on the top surface of an n-type wafer, the wafer is thinned, such as by grinding. A drift layer may be n-type, and various n-type regions and p-type regions in the top surface contact a top metal electrode. A blanket dopant implant through the bottom surface of the thinned wafer is performed to form an n? buffer layer and a bottom p+ emitter layer. Energetic particles are injected through the bottom surface to intentionally damage the crystalline structure. A wet etch is performed, which etches the damaged crystal at a much greater rate, so some areas of the n? buffer layer are exposed. The bottom surface is metallized. The areas where the metal contacts the n? buffer layer form cathodes of an anti-parallel diode for conducting reverse voltages, such as voltage spikes from inductive loads.

Abstract: A design technique is disclosed that divides up a cellular power switch into different size segments. Each segment is driven by a different driver circuit. The selection of the combination of segments is made to minimize the combined conduction and switching losses of the power switch. For example, for very light loads, switching losses dominate so only a small segment is activated for driving the load. For medium and high load currents, conduction losses become more significant, so additional segments are activated to minimize the total losses. In one embodiment, the number of cells in the segments is binary weighted, such as 1×, 2×, and 4×, so that there are seven different combinations of segments. The drivers may be configured to achieve the same or different slew rates of the segments, such as to reduce transients. The segments may all be in the same die or a plurality of dies.

Abstract: Methods and systems for double-sided semiconductor device fabrication. Devices having multiple leads on each surface can be fabricated using a high-temperature-resistant handle wafer and a medium-temperature-resistant handle wafer. Dopants can be introduced on both sides shortly before a single long high-temperature diffusion step diffuses all dopants to approximately equal depths on both sides. All high-temperature processing occurs with no handle wafer or with a high-temperature handle wafer attached. Once a medium-temperature handle wafer is attached, no high-temperature processing steps occur. High temperatures can be considered to be those which can result in damage to the device in the presence of aluminum-based metallizations.

Abstract: A trenched, vertical MOS-gated switch is described that uses only three or four masking steps to fabricate. In one embodiment, one mask is used to form first trenches having a first depth, wherein the first trenches are filled with doped polysilicon to form gates to control the conduction of the switch. A second mask is used to form second trenches having a shallower second depth. The second trenches are filled with the same metal used to form the top source electrode and gate electrode. The metal filling the second trenches electrically contacts a top source layer and a body region. A third mask is used to etch the metal to define the source metal, the gate electrode, and floating rings in a termination region surrounding the active area of the switch. An additional mask may be used to form third trenches in the termination region that are deeper than the first trenches.