A SiC trench structure with adaptive floating potential

Abstract

This invention discloses a SiC trench structure with adaptive floating potential. The trench is divided into upper and lower electrodes: the upper electrode is connected to the gate potential, and the lower electrode is connected to the generated internal potential. A thick oxide layer separates the two electrodes, while the common external surfaces of the two electrodes, namely the sidewalls and bottom of the trench, are covered by a thin oxide layer. The invention also includes a structure for generating the internal potential, comprising a drain potential readout section and a potential comparison section. The drain potential readout section generates a voltage lower than the drain voltage but higher than the source potential as the drain voltage changes. The potential comparison section compares the voltage generated by the drain potential readout section with the gate potential. As the generated voltage and gate potential dynamically change, the potential of the lower electrode follows the higher potential. This trench structure avoids the JFET effect introduced by the trench bottom protection of traditional SiC trench MOSFET structures and introduces an accumulation effect, thereby reducing the on-resistance of the device.

Description

Technical Field

[0001] This invention belongs to the field of SiC power device technology, specifically, it relates to a SiC trench structure with adaptive floating potential. Background Technology

[0002] Silicon carbide (SiC) materials possess advantages in physical properties such as a large bandgap, high breakdown electric field, fast electron mobility, and high thermal conductivity. These characteristics make SiC highly suitable for high-temperature, high-pressure, high-frequency, and radiation-resistant environments. SiC power MOSFETs are unipolar voltage-controlled devices primarily used in power supplies and power processing systems, playing a role in controlling power conversion. Compared to traditional Si-based power devices, SiC devices are easier to achieve in terms of high voltage, low loss, and high power density, thus gradually becoming the mainstream in the market. Currently, a key factor limiting the reduction of MOSFET device costs and further parameter improvements is how to reduce its on-resistance (on-resistance per unit chip area), which is composed of multiple distributed resistors connected in series. A common solution to reduce device on-resistance is to use a trench structure design, transforming the original transverse channel into a longitudinal channel, and increasing the channel density by increasing the processing density, thereby reducing the channel resistance.

[0003] In traditional Si materials, trench structures can significantly increase trench density, thereby reducing on-resistance. This is because increasing trench density significantly reduces channel resistance without increasing JFET resistance (no JFET effect), while introducing an accumulation region effect, increasing the current spread rate, and thus reducing overall on-resistance. However, for SiC devices, the change in the material's bandgap width increases the maximum electric field within the material by about 10 times, making the oxide layer at the bottom of the trench a weak point, prone to breakdown. Therefore, various commercial devices have unanimously designed P+ structures at various trench bottom positions to avoid localized oxide layer breakdown. For example... Figure 1 (a) and (b) represent two typical market structures. Figure 1 (a) adopts a semi-enclosed structure, while Figure 1 (b) A typical trench bottom injection structure is adopted. Since the reduction of the P+ injection spacing in both structures is synchronized with the increase of the channel density, while reducing the channel resistance, it will also greatly enhance the JFET effect in the adjacent P+ region, thereby increasing the channel resistance and limiting the reduction of the total resistance. Moreover, the protection of the trench bottom in these typical structures covers the original location of the accumulation area, so that the device no longer has an accumulation effect, which is also an important factor restricting the reduction of the on-resistance of current SiC trench technology. Summary of the Invention

[0004] To address the shortcomings of existing technologies, this invention proposes a SiC trench structure with adaptive floating potential. This invention divides the traditional trench internal electrode into two electrodes, one of which serves as the driving gate, controlling the device's on / off state. The other electrode is determined by the potential generation structure within the device provided by this invention, creating a floating potential. This raises the potential within the oxide layer, protecting the oxide layer's electric field, eliminating the JFET effect, and simultaneously enhancing the accumulation region effect at the trench bottom.

[0005] The technical solution of the present invention is described in detail below.

[0006] A SiC trench structure with adaptive floating potential, wherein the trench is divided into upper and lower electrodes: electrode G1 Electrode G1 is the driving gate, connected to the gate potential for switching control; electrode G2 is the auxiliary gate, connected to the auxiliary gate potential generated by the internal potential generation structure for potential floating, thus providing trench bottom protection, reducing JFET effect, and enhancing accumulation effect. A first oxide layer separates the two electrodes to prevent breakdown between them. The common outer surfaces of the two electrodes, i.e., the sidewalls and bottom of the trench, are covered by a second oxide layer, used for channel opening and accumulation effect formation, respectively. The thickness of the first oxide layer is greater than that of the second oxide layer. The potential generation structure includes a drain-following potential reading section and a potential comparison section. The drain-following potential reading section generates an internal floating potential lower than the drain voltage but higher than the source potential as the drain voltage changes. The potential comparison section compares the internal floating potential with the gate potential as the internal floating potential and gate potential dynamically change, ensuring that the potential of electrode G2 follows the high potential.

[0007] In this invention, the thickness of the first oxide layer is between 100 nm and 600 nm, and the thickness of the second oxide layer is between 30 and 100 nm.

[0008] In this invention, the drain follower potential reading section includes a floating P-well structure for reading the floating potential, and a ground source P-well surrounding the floating P-well structure for protecting the breakdown voltage. The floating P-well structure and the ground source P-well are directly separated by a lightly doped N-type structure, or multiple floating rings surrounding the floating P-well structure are inserted.

[0009] In this invention, the drain-following potential reading section is implemented by directly connecting the metal to the SiC position at the bottom of the trench, and the SiC material at the corresponding position is P-type doped to form a PN junction for isolation, thereby ensuring the device breakdown voltage; the drain-following potential reading section reads the distributed potential generated by the P-type region at the bottom of the trench when the drain is at a high potential.

[0010] In this invention, the potential comparison section is formed by two PN junctions in an N-well within a P-well and an N-type ohmic contact electrode. The two PN junctions are interconnected by metal and connected to the gate potential and drain follower potential readout sections, respectively. The N-type ohmic contact electrode is connected to electrode G2 at the bottom of the trench structure. When the drain follower potential is higher than the gate potential, the PN junction connected to the drain follower potential is forward biased, making the potential in the N-well close to the drain follower potential, and the PN junction connected to the gate potential is reverse biased. At this time, electrode G2 follows this potential. Conversely, the potential of electrode G2 follows the gate potential. Therefore, with the dynamic changes of the internal floating potential and the gate potential, the two PN junctions are in forward bias and reverse bias states, respectively, so that the potential of electrode G2 follows the high potential.

[0011] Compared with the prior art, the beneficial effects of the present invention are as follows: This invention avoids the JFET effect introduced by the trench bottom protection of traditional SiC trench MOSFET structures and introduces a cumulative effect, increasing the current spread rate and thus reducing the on-resistance of the device. It also facilitates further increases in trench density. Instead of using a P-type network to pull down the trench bottom potential for protection, this invention raises the internal potential of the trench, solving a series of problems associated with existing trench structures. Attached Figure Description

[0012] Figure 1 It is the mainstream trench structure of SiC; (a) semi-enclosed structure, (b) trench bottom injection structure.

[0013] Figure 2 This is the main structure of the invention. Figure 1 - Cellular part.

[0014] Figure 3 This is the main structure of the present invention. Figure 2 - Potential generation section.

[0015] Figure 4 This is a structural diagram for reading the potential at the bottom of the tank.

[0016] Figure 5 The diagram shows the potential generation structure layout; (a) top view of the surface without metal, (b) top view after adding metal. Detailed Implementation

[0017] The technical solution of the present invention will now be described in detail with reference to the accompanying drawings and embodiments.

[0018] This invention provides a SiC trench structure with adaptive floating potential. Specifically, two electrodes are designed in the trench: the upper electrode serves as a conventional gate structure for switching control, and the lower electrode is used for potential floating, thereby providing trench bottom protection, reducing JFET effects, and enhancing accumulation region effects. When the device is in the blocking state, the lower electrode follows the high potential of the drain, reducing the electric field inside and outside the gate oxide layer, achieving trench bottom protection. When the device is in the turning state, the lower electrode follows the gate potential, making the trench bottom an accumulation state, overcoming the core defect of existing trench MOSFETs and reducing the on-resistance of the device. This invention provides a potential generation structure that is matched with the lower electrode. This structure is fabricated near the gate in the transition region. By comparing the internal floating potential (related to the drain) with the external gate potential, a new potential is generated to follow the higher of the two potentials. Example 1

[0019] The SiC trench structure of the present invention is as follows Figure 1 As shown, the trench is divided into two electrodes, G1 and G2, where G1 is the external gate potential and G2 is the generated internal potential. A first oxide layer (thick oxide layer) separates the two potentials to prevent breakdown between them. The common exterior of both, namely the sidewalls and bottom of the trench, is covered by a second oxide layer (thin oxide layer), used for channel opening and accumulation effect formation, respectively. The thickness of the first oxide layer is between 100 nm and 600 nm, and the thickness of the second oxide layer is between 30 and 100 nm.

[0020] It is important to note that while the active region of this structure shares similarities in cross-section with Si-based SGT (split-gate trench) devices, there are significant differences in their underlying principles. First, SGT devices have thick oxide layers on their sidewalls to achieve charge balance and reduce the distributed resistance in the epitaxial layer by increasing epitaxial doping. In contrast, the outer side of the G2 device in this invention has a thin oxide layer to form an accumulation region for rapid current expansion and to reduce the JFET effect. Second, the G2 potential in an SGT device is the source potential, i.e., a fixed ground potential, while the G2 potential in this invention is a floating potential generated internally. Therefore, there is a fundamental difference between the two. Figure 2 ).

[0021] Regarding how the G2 potential is formed Figure 3A simplified schematic diagram is provided: The potential generation structure is mainly used to form the G2 potential at the bottom of the trench, which is the auxiliary gate potential shown in the diagram. The left side is the drain-following potential readout section, which includes a floating P-well structure for reading the floating potential, and a grounded source P-well surrounding this structure. The function of this structure is to generate a voltage lower than the drain voltage but higher than the source potential through the voltage differentiation effect of space charge, as the drain voltage changes. Its relationship with the drain voltage is determined by its structure, doping, spacing with the P-well, and possible other auxiliary structures such as auxiliary rings, field plates, etc. There are various ways to obtain this voltage, which will not be listed here. Figure 4 This presents another method for reading the potential: reading it directly from the potential at the bottom of the tank. Figure 4 The right-hand side of the diagram is the potential comparison section, which is the calculation region for comparing the potential generated by the drain-following potential reading section with the gate potential. It is formed by two PN junctions within an N-well in a P-well. As the generated potential and gate potential change dynamically, the two PN junctions are in forward-biased and reverse-biased states, respectively, thus causing the G2 potential to follow the high potential.

[0022] Furthermore, the drain-following potential reading section and the potential comparison section in the aforementioned trench structure are quite complex in a two-dimensional layout, especially the metal interconnect section. However, in the actual three-dimensional design of the layout, they can be greatly simplified, such as... Figure 5 (a) and Figure 5 As shown in (b).

[0023] The overall working principle of the above structure is further described in a coordinated manner: When the external gate is at a high potential, electrode G1 in the active trench region is turned on. As the gate control current increases, the drain potential at the bottom of the chip gradually decreases. During the drain potential decrease, the drain potential read by the drain follower potential readout section will decrease synchronously. After decreasing to a certain level, the absolute value of the drain follower potential drops below the external gate potential, the PN junction on the left side of the potential comparison section will turn off, and the PN junction on the right side will turn on. At this time, the generated potential will be converted to follow the external gate potential, maintaining a high potential. At this time, an N-type accumulation region will be formed outside electrode G2 in the active trench region. When the device remains in the on-state, this accumulation effect will continue to exist to reduce the on-resistance.

[0024] When the external potential changes from high to low, the process is the reverse of the above: electrode G1 is turned off, the drain potential rises, and the drain potential read by the potential generation section will rise synchronously. After a certain point, the PN junction on the left side of the potential comparison section will conduct, while the PN junction on the right side will turn off. The potential of electrode G2 will continue to increase as the drain potential rises, thereby reducing the voltage on both sides of the oxide layer and forming protection for the bottom of the tank.

[0025] The SiC trench structure with adaptive floating potential of the present invention, when the device is in the blocking state, electrode G2 follows the high potential of the drain, reducing the electric field inside and outside the gate oxide layer and achieving trench bottom protection. This effectively avoids the JFET effect introduced by trench bottom protection in traditional SiC trench MOSFET structures. When the device is in the turning state, this potential follows the gate potential, making the trench bottom an accumulation state, overcoming the core defect of existing trench MOSFETs and reducing the on-resistance of the device.

Claims

1. A SiC trench structure with adaptive floating potential, characterized in that, The trench is divided into two electrodes: electrode G1 and electrode G2. Electrode G1 is the driving gate, connected to the gate potential for switching control. Electrode G2 is the auxiliary gate, connected to the auxiliary gate potential generated by the internal potential generation structure for potential floating, thus providing trench bottom protection, reducing JFET effect, and enhancing accumulation effect. A first oxide layer separates the two electrodes to prevent breakdown. The common outer surfaces of the two electrodes, i.e., the sidewalls and bottom of the trench, are covered by a second oxide layer, used for channel opening and accumulation effect formation, respectively. The thickness of the first oxide layer is greater than that of the second oxide layer. The potential generation structure includes a drain-following potential readout section and a potential comparison section. The drain-following potential readout section generates an internal floating potential lower than the drain voltage but higher than the source potential as the drain voltage changes. The potential comparison section compares the internal floating potential with the gate potential as the internal floating potential and gate potential dynamically change, ensuring that the potential of electrode G2 follows the high potential. The potential comparison section is formed by two PN junctions in an N-well within a P-well and an N-type ohmic contact electrode. The two PN junctions are interconnected by metal and connected to the gate potential and drain follower potential readout sections, respectively. The N-type ohmic contact electrode is connected to electrode G2 at the bottom of the trench structure. When the drain follower potential is higher than the gate potential, the PN junction connected to the drain follower potential is forward biased, making the potential in the N-well close to the drain follower potential, and the PN junction connected to the gate potential is reverse biased. At this time, electrode G2 follows the drain follower potential. Conversely, the potential of electrode G2 follows the gate potential. Therefore, with the dynamic changes of the internal floating potential and the gate potential, the two PN junctions are in forward bias and reverse bias states, respectively, so that the potential of electrode G2 follows the high potential.

2. The SiC trench structure with adaptive floating potential according to claim 1, characterized in that, The thickness of the first oxide layer is between 100 nm and 600 nm, and the thickness of the second oxide layer is between 30 and 100 nm.

3. The SiC trench structure with adaptive floating potential according to claim 1, characterized in that, The drain follower potential readout section includes a floating P-well structure for reading the floating potential, and a ground source P-well surrounding the floating P-well structure for protecting against breakdown voltage. The floating P-well structure and the ground source P-well are directly separated by a lightly doped N-type structure, or separated by inserting multiple floating rings surrounding the floating P-well structure.

4. The SiC trench structure with adaptive floating potential according to claim 1, characterized in that, The drain-following potential readout section is directly connected to the SiC position at the bottom of the trench via a metal, and the SiC material at this position is P-type doped to form a PN junction for isolation, thereby ensuring the device breakdown voltage. The drain-following potential readout section reads the distributed potential generated by the P-type region at the bottom of the trench when the drain is at a high potential.

Images