Semiconductor device and method for manufacturing the same
A dual-layer N buffer layer structure in semiconductor devices addresses the issues of turn-off controllability and leakage current by stabilizing breakdown voltage and reducing off-loss, ensuring reliable operation and compatibility with larger wafers.
Patent Information
- Application Number
- JP2020183522
- Authority / Receiving Office
- JP · JP
- Patent Type
- Patents
- Current Assignee / Owner
- Filing Date
- 2020-11-02
- Publication Date
- 2024-11-11
- Estimated Expiration
- 2040-11-02
AI Technical Summary
Conventional semiconductor devices with vertical structures face issues such as poor controllability of turn-off operation, decreased blocking ability, and increased leakage current due to crystal defects in the N-type buffer layer, leading to unstable breakdown voltage characteristics and high off-loss.
The semiconductor device incorporates a dual-layer N buffer layer structure comprising a first buffer layer with no lattice defects and a second buffer layer with controlled lattice defects, specifically Trap A and Trap B, to stabilize the breakdown voltage, reduce leakage current, and improve turn-off controllability.
The dual-layer N buffer layer structure enhances breakdown voltage stability, reduces off-loss, and improves the interrupting ability during turn-off operations, while being compatible with larger diameter wafers and reducing the risk of breakdown failures.
Smart Images

Figure 0007580245000001 
Figure 0007580245000002 
Figure 0007580245000003
Abstract
Description
[Technical field]
[0001] The present disclosure relates to a semiconductor device and a method for manufacturing a semiconductor device. [Background technology]
[0002] Conventional vertical semiconductor devices such as trench-gate IGBTs (Insulated Gate Bipolar Transistors) and PiN (P-intrinsic-N) diodes have a vertical structure. In an IGBT, a region including an N-type drift layer, an N-type buffer layer, and a P-type collector layer is a region of the vertical structure, and in a diode, a region including an N-type drift layer, an N-type buffer layer, and an N-type cathode layer is a region of the vertical structure. Patent Document 1 discloses an IGBT having a vertical structure. Patent Document 2 discloses a configuration in which an N-type buffer layer has a stacked structure consisting of multiple layers.
[0003] In conventional semiconductor devices such as IGBTs or diodes having a vertical structure, a wafer manufactured by the FZ method may be used as a Si wafer for forming the semiconductor device instead of a wafer manufactured by epitaxial growth. In this case, the vertical structure region of the wafer, for example, the N-type buffer layer in an IGBT, has a high impurity concentration, and the impurity profile has a steep gradient toward the junction with the N-type drift layer. [Prior art documents] [Patent documents]
[0004] [Patent Document 1] International Publication No. 2014 / 054121 [Patent Document 2] JP 2020-027921 A Summary of the Invention [Problem to be solved by the invention]
[0005] The above-mentioned impurity profile of the buffer layer in a semiconductor device having a vertical structure has various problems, such as poor controllability of the turn-off operation and a decrease in the cut-off ability at the time of turn-off.
[0006] The technology of Patent Document 2 aims to solve these problems. However, although Patent Document 2 describes the depth, impurity concentration, and impurity profile of the N-type buffer layer, it does not discuss the relationship between the crystal defects introduced into the Si wafer when the N-type buffer layer is formed and the device characteristics, making the feasibility of power semiconductor elements poor. As will be described later, the crystal defects introduced into the Si wafer are an important factor that affects the carrier plasma layer and the controllability of the electric field intensity, which are important during the dynamic operation of the power semiconductor element.
[0007] The present disclosure has been made to solve the problems described above, and aims to provide a semiconductor device having a vertical structure with stable voltage resistance characteristics, low off-loss due to reduced leakage current when off, improved controllability of turn-off operation, and improved blocking capability when turned off, while resolving problems caused by the effects of crystal defects introduced into the wafer when a buffer layer is formed. [Means for solving the problem]
[0008] A semiconductor device according to the present disclosure includes a semiconductor substrate having a first main surface and a second main surface, a drift layer of a first conductivity type formed in the semiconductor substrate, an impurity diffusion layer of a second conductivity type formed on the first main surface side of the drift layer, and a buffer layer of the first conductivity type formed on the second main surface side of the drift layer and having a peak impurity concentration higher than that of the drift layer, the buffer layer being a photoluminescent (PL) a first buffer layer in which no trap level derived from lattice defects is detected by the photoluminescence method; and a second buffer layer provided between the first buffer layer and the drift layer, the second buffer layer in which two types of trap levels derived from lattice defects are detected by the photoluminescence method. The two types of trap levels derived from lattice defects detected by the photoluminescence method in the second buffer layer are Trap A (photon energy: 1.0182 eV) and Trap B (photon energy: 1.0398 eV). . Effect of the Invention
[0009] According to the present disclosure, it is possible to solve problems caused by the effects of crystal defects introduced into the wafer during the formation of the N-type buffer layer, while achieving stable voltage resistance characteristics, low off-loss due to reduced leakage current during off-state, improved controllability of turn-off operation, and improved blocking capability during turn-off. [Brief description of the drawings]
[0010] [Figure 1] FIG. 1 is a plan view of a semiconductor device according to first to third embodiments. [Diagram 2] 1 is a cross-sectional view of an RFC diode according to a first embodiment. [Diagram 3] 1 is a cross-sectional view showing an example of the structure of the outer periphery of a semiconductor device including an RFC diode according to a first embodiment. [Figure 4] FIG. 1 is a diagram illustrating the targeted role of an N buffer layer. [Diagram 5] FIG. 1 is a diagram illustrating the targeted role of an N buffer layer. [Figure 6] FIG. 1 is a diagram illustrating the targeted role of an N buffer layer. [Figure 7] FIG. 13 shows PL spectrum results for the first buffer layer and the second buffer layer. [Figure 8] FIG. 13 is a diagram showing a change in the Trap B ratio from the lower surface side of the second buffer layer to the N-drift layer. [Figure 9] FIG. 13 is a diagram showing simulation results of recovery waveforms (1200 V class) of a conventional structure RFC diode and a new structure RFC diode. [Figure 10] FIG. 10 is an enlarged view of a portion of FIG. [Figure 11] 11 is a diagram showing a simulation result of the carrier concentration distribution and the electric field strength distribution between the first principal surface and the second principal surface at the analysis points in FIG. 10 in the PNP transistor region of the RFC diode of the conventional structure. [Figure 12]11 is a diagram showing a simulation result of the carrier concentration distribution and the electric field strength distribution between the first principal surface and the second principal surface at the analysis point shown in FIG. 10 in the PNP transistor region of the new structure RFC diode. [Figure 13] FIG. 13 is a diagram showing simulation results of recovery waveforms (4.5 kV class) during snappy recovery operation in a conventional structure RFC diode and a new structure RFC diode. [Figure 14] FIG. 14 is a diagram showing current density distributions in the conventional-structure RFC diode and the new-structure RFC diode at analysis points T1 to T7 in FIG. [Figure 15] FIG. 14 is a diagram showing the electron concentration distribution at analysis points T1 to T7 in FIG. 13 in the PiN diode region of the conventional RFC diode. [Figure 16] FIG. 14 is a diagram showing the electron concentration distribution at analysis points T1 to T7 in FIG. 13 in the PiN diode region of the new-structure RFC diode. [Figure 17] FIG. 14 is a diagram showing hole concentration distribution at analysis points T1 to T7 in FIG. 13 in the PiN diode region of the conventional RFC diode. [Figure 18] FIG. 14 is a diagram showing hole concentration distributions at analysis points T1 to T7 in FIG. 13 in the PiN diode region of the new-structure RFC diode. [Figure 19] FIG. 14 is a diagram showing the electric field intensity distribution at analysis points T1 to T7 in FIG. 13 in the PiN diode region of the conventional RFC diode. [Figure 20] FIG. 14 is a diagram showing the electric field intensity distribution at analysis points T1 to T7 in FIG. 13 in the PiN diode region of the new structure RFC diode. [Figure 21] FIG. 14 is a diagram showing the electron concentration distribution at analysis points T1 to T7 in FIG. 13 in the PNP transistor region of the RFC diode of the conventional structure. [Figure 22] FIG. 14 is a diagram showing the electron concentration distribution at analysis points T1 to T7 in FIG. 13 in the PNP transistor region of the new-structure RFC diode. [Diagram 23]FIG. 14 is a diagram showing hole concentration distribution at analysis points T1 to T7 in FIG. 13 in a PNP transistor region of a conventional RFC diode. [Figure 24] FIG. 14 is a diagram showing hole concentration distributions at analysis points T1 to T7 in FIG. 13 in a PNP transistor region of a new-structure RFC diode. [Diagram 25] FIG. 14 is a diagram showing the electric field intensity distribution at analysis points T1 to T7 in FIG. 13 in a PNP transistor region of a conventional RFC diode. [Figure 26] FIG. 14 is a diagram showing the electric field intensity distribution at analysis points T1 to T7 in FIG. 13 in the PNP transistor region of the new-structure RFC diode. [Figure 27] FIG. 13 is a graph showing the Trap B ratio dependence of JR-VR characteristics at 298 K for a prototype 1200 V class new structure RFC diode. [Figure 28] FIG. 13 is a graph showing the Trap B ratio dependence of the snappy recovery waveform at low temperatures in a prototype 1200V class new structure RFC diode. [Figure 29] This figure summarizes the relationship between the device performance of a 1200V class new structure RFC diode and the Trap B ratio. [Diagram 30] FIG. 13 is a diagram showing the relationship between the safe operating temperature range in snappy mode and the Trap B ratio for a 1200V class new structure RFC diode. [Diagram 31] FIG. 13 is a diagram showing the operating temperature dependence of the output characteristics of a prototype 1200V class new structure RFC diode. [Diagram 32] FIG. 13 is a diagram showing the operating temperature dependence of the JR-VR characteristics of a prototype 1200V class new structure RFC diode. [Diagram 33] FIG. 13 is a diagram showing recovery waveforms in a prototype 1200V class conventional structure RFC diode and a new structure RFC diode. [Diagram 34] FIG. 13 is a diagram showing recovery waveforms of the snappy mode at low temperatures in a prototype 1200V class RFC diode with a conventional structure and a new structure RFC diode. [Diagram 35] FIG. 13 is a diagram showing the relationship between the power supply voltage and the reverse recovery charge at 253 K for a prototype 1200 V class conventional structure RFC diode and a prototype new structure RFC diode. [Diagram 36] FIG. 13 is a diagram showing the relationship between the operating temperature and the reverse recovery charge at a power supply voltage of 1000 V for prototype 1200 V class conventional structure RFC diodes and new structure RFC diodes. [Figure 37] FIG. 13 is a diagram showing the results of a comparison of the safe operating temperature ranges in snappy mode at low temperatures in a prototype 1200V class RFC diode with a conventional structure and a prototype RFC diode with a new structure. [Figure 38] This figure shows the recovery waveform of a prototype 1200V class new structure RFC diode at high temperature (473K). [Figure 39] FIG. 13 is a diagram showing the VF-EREC trade-off characteristics of prototype 1200V class conventional structure RFC diodes and new structure RFC diodes. [Diagram 40] FIG. 11 is a cross-sectional view of a PiN diode according to a second embodiment. [Diagram 41] 11 is a cross-sectional view showing an example of the structure of the outer periphery of a semiconductor device including a PiN diode according to a second embodiment. FIG. [Diagram 42] FIG. 13 shows waveforms during snappy recovery operation in prototype 4.5 kV class PiN diodes of a conventional structure and a new structure. [Diagram 43] FIG. 13 is a graph showing the relationship between the power supply voltage and the snap-off voltage at 298 K for the prototype 4.5 kV class PiN diodes of the conventional structure and the new structure. [Diagram 44] FIG. 13 is a graph showing the relationship between the power supply voltage and the reverse recovery charge at 298 K for a prototype 4.5 kV class PiN diode with a conventional structure and a prototype PiN diode with a new structure. [Diagram 45] FIG. 11 is a cross-sectional view of an IGBT according to a third embodiment. [Figure 46] 11 is a cross-sectional view showing an example of the structure of the outer periphery of a semiconductor device including an IGBT according to a third embodiment. FIG. [Figure 47] 11 is a cross-sectional view showing an example of the structure of the outer periphery of a semiconductor device including an IGBT according to a third embodiment. FIG. [Figure 48] FIG. 13 is a diagram showing the simulation results of the electric field strength distribution inside the device in a short-circuit state for a conventional IGBT and a new IGBT. [Figure 49] FIG. 13 is a diagram showing the results of a simulation of the relationship between the electric field strength at the front and back surfaces of the device and the on-voltage in a short-circuit state for a conventional IGBT and a new IGBT. [Figure 50] FIG. 13 is a diagram showing the simulation results of the internal temperature distribution of a conventional IGBT and a new IGBT in a short-circuit state. [Figure 51] This is a short-circuit waveform obtained by simulation when analyzing the temperature distribution inside the device shown in FIG. [Figure 52] FIG. 13 shows the turn-off waveforms at 298K of a prototype 1200V class IGBT with a conventional structure and an IGBT with a new structure. [Figure 53] FIG. 13 shows short-circuit waveforms at 423 K for prototype 1200 V class IGBTs with a conventional structure and a new structure. [Figure 54] FIG. 1 is a diagram showing the relationship between the maximum energy density that can be interrupted during a short circuit and the device thickness in a prototype 1200V class IGBT with a conventional structure and an IGBT with a new structure. [Figure 55] FIG. 13 is a diagram showing the trade-off characteristics between the on-voltage and the turn-off loss in a prototype 1200V class IGBT with a conventional structure and an IGBT with a new structure. [Figure 56] FIG. 13 is a graph showing the relationship between the main junction leakage current and the collector-emitter voltage at 423 K for a prototype 6.5 kV class IGBT with a conventional structure and a prototype IGBT with a new structure. [Figure 57] FIG. 13 shows turn-off waveforms at 213 K for a prototype 6.5 kV class IGBT with a conventional structure and an IGBT with a new structure. [Figure 58] FIG. 1 is a process flow diagram of an IGBT. [Figure 59] FIG. 1 is a process flow diagram of an IGBT. [Figure 60] FIG. 1 is a process flow diagram of an IGBT. [Figure 61] FIG. 1 is a process flow diagram of an IGBT. [Figure 62] FIG. 1 is a process flow diagram of an IGBT. [Figure 63] FIG. 1 is a process flow diagram of an IGBT. [Figure 64] FIG. 1 is a process flow diagram of an IGBT. [Figure 65] FIG. 1 is a process flow diagram of an IGBT. [Figure 66] FIG. 1 is a process flow diagram of an IGBT. [Figure 67] FIG. 1 is a process flow diagram of an IGBT. [Figure 68] FIG. 1 is a process flow diagram of an IGBT. [Figure 69] FIG. 1 is a process flow diagram of an IGBT. [Figure 70] FIG. 1 is a process flow diagram of an IGBT. [Figure 71] FIG. 1 is a process flow diagram of an RFC diode. [Figure 72] FIG. 1 is a process flow diagram of an RFC diode. [Figure 73] FIG. 1 is a process flow diagram of an RFC diode. [Figure 74] FIG. 1 is a process flow diagram of an RFC diode. [Figure 75] FIG. 1 is a process flow diagram of an RFC diode. [Figure 76] FIG. 1 is a process flow diagram of an RFC diode. [Figure 77] FIG. 1 is a process flow diagram of an RFC diode. [Figure 78] FIG. 1 is a process flow diagram of an RFC diode. [Figure 79] FIG. 1 is a process flow diagram of an RFC diode. [Figure 80] FIG. 1 is a comparative diagram of process flows. [Figure 81]FIG. 13 is a diagram showing the relationship between the carrier lifetime and the thickness (tN−) of the N-drift layer by simulation. [Figure 82] This figure summarizes the performance of the 1200V class RFC diode prototyped using the process flows compared in Figure 80. [Figure 83] FIG. 13 is a graph showing the relationship between Trap B ratio and annealing temperature. [Figure 84] FIG. 1 shows the relationship between the PL intensity and the annealing time for Trap A and Trap B. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0011] In the following embodiments, IGBTs and freewheeling diodes (FWDs) are shown as representative examples of power semiconductor elements, and in the following, "diode" refers to FWDs. However, the technology according to the present disclosure also exerts the same effect on power semiconductor elements such as RC (Reverse Conducting)-IGBTs, RB (Reverse Blocking)-IGBTs, and MOSFETs (Metal-Oxide-Semiconductor Field Effect Transistors).
[0012] In addition, although each embodiment uses Si as the semiconductor material, the same effect can be obtained with semiconductor devices using wide band gap materials such as SiC, GaN, etc. In addition, in each embodiment, a Si wafer (FZ wafer) manufactured by the FZ (Floating Zone) method is shown as a representative Si material, but the residual oxygen concentration in the Si material is higher than that of the FZ wafer, and the residual oxygen concentration is 10 17 cm -3 Alternatively, a silicon wafer manufactured by the MCZ (Magnetic field applied Czochralski) method, which has an oxygen content of the order of 1000 V, may be used. In addition, although the embodiments show examples of semiconductor devices with a withstand voltage class of 1200 V and 4.5 kV, similar effects can be obtained regardless of the withstand voltage class.
[0013] <Embodiment 1> In the first embodiment, a relaxed field of cathode (RFC) diode having a structure in which a cathode layer is partially replaced with a P-type region is shown as a semiconductor element included in the power semiconductor device.
[0014] Fig. 1 is a plan view of a semiconductor device according to a first embodiment. As shown in Fig. 1, a plurality of active cell regions R1 are provided in the center of the semiconductor device, a front gate wiring portion R11 is provided between two adjacent active cell regions R1, and a gate pad portion R12 is provided in a part of the center. An intermediate region R2 is provided so as to surround the active cell region R1, the gate pad portion R12, and the front gate wiring portion R11, and an edge termination region R3 is provided so as to surround the intermediate region R2. In a diode, the gate pad portion R12 and the front gate wiring portion R11 may not exist.
[0015] The active cell region R1 is an element formation region that guarantees the basic performance of the power semiconductor chip. The intermediate region R2 and the edge termination region R3 are regions for maintaining the breakdown voltage. Among them, the intermediate region R2 is a region where the active cell region R1 and the edge termination region R3 are joined, and guarantees the breakdown resistance during dynamic operation of the semiconductor device, and supports the original performance of the semiconductor element formed in the active cell region R1. In addition, the edge termination region R3 is responsible for maintaining the breakdown voltage in the static state of the semiconductor device, guaranteeing the stability and reliability of the breakdown voltage characteristics, and suppressing defects in the breakdown resistance during dynamic operation of the semiconductor device, and supports the original performance of the semiconductor element formed in the active cell region R1.
[0016] Fig. 2 is a cross-sectional view of an RFC diode, which is a semiconductor element according to the first embodiment, and corresponds to the cross-section along the A1-A2 line in Fig. 1. Fig. 3 is a cross-sectional view showing an example of the structure of the outer periphery of a semiconductor device including an RFC diode according to the first embodiment, and corresponds to the cross-section along the A3-A4 line in Fig. 1. Here, the upper surface of the semiconductor substrate in the cross-sectional view is defined as the "first main surface" and the lower surface is defined as the "second main surface". The first main surface may also be expressed as the "upper surface" and the second main surface as the "lower surface" or "rear surface".
[0017] First, the structure of the RFC diode formed in the active cell region R1 will be described with reference to Fig. 2. The semiconductor substrate constituting the RFC diode has an N-type (first conductivity type) - The drift layer 14 is formed. - Under the drift layer 14, N - An N-type N buffer layer 15 having a higher peak impurity concentration than the drift layer 14 is formed. In this embodiment, the N buffer layer 15 has a two-layer structure consisting of a lower first buffer layer 15-1 and an upper second buffer layer 15-2. That is, the second buffer layer 15-2 is a layer structure consisting of the first buffer layer 15-1 and the N - The N buffer layer 15 is disposed between the drift layer 14. Although not shown, the N buffer layer 15 in Fig. 3 also includes a first buffer layer 15-1 and a second buffer layer 15-2. The N buffer layer 15 will be described in detail later.
[0018] Under the N buffer layer 15 in the semiconductor substrate, there is an N type N + The cathode layer 17 and the P cathode layer 18 of the P type (second conductivity type) are formed adjacent to each other. + The cathode layer 17 and the P cathode layer 18 are formed on the surface layer of the back surface side (second main surface side) of the semiconductor substrate, and a cathode electrode 19 is formed on the back surface of the semiconductor device. + It is formed so as to be in contact with the cathode layer 17 and the P cathode layer 18 .
[0019] N in semiconductor substrate -A P-type P anode layer 10 is formed on the drift layer 14. In the RFC diode, the P anode layer 10 and the N - The junction with the drift layer 14 becomes a main junction. The P anode layer 10 is formed in a surface layer portion on the upper surface (first main surface) side of the semiconductor substrate, and an anode electrode 5 is formed on the upper surface of the semiconductor substrate so as to be in contact with the P anode layer 10.
[0020] As shown in FIG. - The drift layer 14, the N buffer layer 15, and the N + The cathode layer 17 constitutes the PiN diode region R21, and the P anode layer 10, N - The drift layer 14, the N buffer layer 15, and the P cathode layer 18 constitute a PNP transistor region R22. In other words, when expressed as an equivalent circuit, the RFC diode has a configuration in which a PiN diode and a PNP transistor are connected in parallel. - When the RFC diode is turned on, the drift layer 14 becomes a variable resistance region because the carrier concentration changes due to conductivity modulation.
[0021] Compared to PiN diodes, RFC diodes exhibit distinctive effects in terms of diode performance, such as the electric field relaxation phenomenon that relaxes the electric field strength on the cathode side. In particular, the injection of holes from the P cathode layer 18 is promoted in the latter half of the recovery operation, which relaxes the electric field strength on the cathode side, suppressing the snap-off phenomenon (voltage jump phenomenon) at the end of the recovery operation and the subsequent oscillation phenomenon, improving the breakdown resistance during the recovery operation. These effects are explained, for example, in Patent No. 5256357 and Patent No. 6301776. N + In order to ensure the above-mentioned effects, the cathode layer 17 and the P cathode layer 18 are disposed so as to satisfy the relationship shown in Japanese Patent No. 5256357 or Japanese Patent No. 6301776 (US8686469).
[0022] Next, the structure of the intermediate region R2 and edge termination region R3 of the RFC diode will be described with reference to FIG.
[0023] N - The drift layer 14 is formed across the active cell region R1, the intermediate region R2 and the edge termination region R3. - A P-type guard ring 22 is formed in the surface layer of the drift layer 14, and is formed deeper than the P anode layer 10. The guard ring 22 extends to the active cell region R1 side and is connected to the P anode layer 10. In addition, the N - A P-type field limiting ring 23 is selectively formed on the surface layer of the drift layer 14.
[0024] In addition, on the outside of the field limiting ring 23, - The surface layer of the drift layer 14 is doped with N + A channel stopper layer 24 of N type is selectively formed. The channel stopper layer 24 is connected to the guard ring 22 and the field limiting ring 23. - The field limiting rings 23 are provided for the purpose of stopping the extension of the depletion layer extending from the junction with the drift layer 14. As the number of field limiting rings 23 increases, the withstand voltage class of the RFC diode becomes higher.
[0025] An insulating film 25 is formed on the upper surface of the semiconductor substrate in the intermediate region R2 and the edge termination region R3, and an interlayer insulating film 26 is formed on the insulating film 25. An FLR electrode 27 connected to the field limiting ring 23 through a contact hole, and a channel stopper electrode 28 connected to the channel stopper layer 24 through a contact hole are formed on the interlayer insulating film 26. The FLR electrode 27 and the channel stopper electrode 28 can be formed simultaneously with the anode electrode 5 in the active cell region R1.
[0026] A passivation film 29, which is a protective film covering the FLR electrode 27 and the channel stopper electrode 28, is formed across the intermediate region R2 and the edge termination region R3, and a passivation film 30 is further formed on the passivation film 29.
[0027] As shown in FIG. 3, N - Drift layer 14, N buffer layer 15, N + There is a structure in which the cathode layer 17 and the P cathode layer 18 are laminated, and this structure is the vertical structure 35 of the RFC diode. The region in which the vertical structure 35 is formed is called the "vertical structure region."
[0028] The vertical structure 35 is an area that supports the basic performance of the semiconductor device by guaranteeing total loss performance (the sum of on-state loss, turn-on state loss, and turn-off state loss), voltage resistance in a static state, stability of voltage resistance characteristics, leakage characteristics (off loss) when voltage is held at high temperatures, reliability, controllability during dynamic operation, and breakdown resistance.
[0029] Here, the concept of the vertical structure region in the present disclosure will be described. In manufacturing a semiconductor device having a vertical structure 35 such as a vertical IGBT or diode, - The process of forming the vertical structure 35 at the bottom of the drift layer 14 (the process shown in FIGS. 69 and 70, or the process shown in FIGS. 78 and 79 described later) is performed after the process of forming aluminum wiring such as the anode electrode 5 and the process of forming the passivation films 29 and 30. + The annealing used in forming the cathode layer 17 and the P cathode layer 18) includes annealing at a low temperature below the melting point of aluminum (660°C) and annealing using a laser with a wavelength that creates a temperature gradient in the device depth direction and prevents the temperature on the top surface of the semiconductor substrate from exceeding the melting point of aluminum (laser annealing).
[0030] The N buffer layer 15 has a process feature in which the N layer profile reproduces the profile in the depth direction at the time of ion implantation to introduce impurities, and is formed using the above-mentioned annealing technique, so that diffusion in the depth direction and lateral direction is unlikely to occur. In forming an N-type diffusion layer, a technique of annealing for a long time at high temperature (≧1100° C.) is used as a technique for realizing an N-type diffusion layer having a deep and gentle concentration gradient. However, this technique cannot be used in a process in which a low-melting-point metal such as aluminum used for the electrode is present, and can be used in an early process of the manufacturing process (for example, the process shown in FIG. 58 or FIG. 71). However, in that case, the thickness of the wafer needs to be adjusted to a desired thickness (40 μm to 700 μm) either before or after the high-temperature annealing, and in the subsequent process, each process device needs to be modified so that a wafer of that thickness can be processed, which is unrealistic because it incurs huge costs. In addition, high-temperature, long-time annealing is a process technique that does not match the large diameter of Si wafers.
[0031] In addition, IGBTs and diodes having the N buffer layer 15 have the following three performance problems. The first problem is that, in terms of voltage holding ability (hereinafter sometimes referred to as voltage resistance characteristic), which is an important performance characteristic for a power semiconductor, the device can go out of control due to increased off loss caused by increased leakage current when holding voltage at high temperatures, or due to thermal runaway caused by heat generation by the device itself at high temperatures, resulting in a state in which operation at high temperatures cannot be guaranteed.
[0032] The second problem is that during dynamic operations such as the turn-off of IGBTs and diodes, the N - The carrier plasma layer near the junction between the drift layer 14 and the N buffer layer 15 is depleted, and N -The electric field strength at the junction between the drift layer 14 and the N buffer layer 15 rises, or a snap-off phenomenon occurs, which is a voltage jump at the end of the turn-off operation on the turn-off waveform, causing an oscillation phenomenon due to the snap-off phenomenon, or the snap-off phenomenon causes the voltage to exceed the withstand voltage that can be maintained, resulting in device destruction. The carrier plasma layer is a neutral layer with a high carrier concentration where the electron and hole concentrations are almost the same, N ≒ P, and the carrier density is 10 16 cm -3 Higher, N - This layer has a carrier concentration two to three orders of magnitude higher than that of the drift layer 14 .
[0033] As a result, conventional IGBTs and diodes have poor controllability in the turn-off operation and a decrease in the cut-off capability at turn-off. In inverter systems that include power modules equipped with IGBTs or diodes that experience snap-off and subsequent oscillation, these can become a cause of noise generation, leading to malfunctions.
[0034] The third problem is that, due to the characteristics of the formation of the above-mentioned N buffer layer 15, scratches and foreign matter on the surface on which the N buffer layer 15 is formed, which occur during the wafer process in forming the vertical structure 35, tend to cause a phenomenon in which the N buffer layer 15 is partially not formed. This makes the IGBTs and diodes sensitive to the phenomenon of poor withstand voltage, leading to an increase in the failure rate of the withstand voltage characteristics of the IGBT and diode chips.
[0035] As described as the second problem, IGBTs and diodes having an N buffer layer 15 have a problem that the N - The carrier plasma layer near the junction between the drift layer 14 and the N buffer layer 15 is easily depleted. - When the carrier plasma layer near the junction between the drift layer 14 and the N buffer layer 15 is depleted, the internal state of the device becomes N - This leads to an increase in the electric field strength at the junction between the drift layer 14 and the N buffer layer 15.
[0036] Regarding the waveform during turn-off operation, the IGBT at the end of the turn-off operation has a dj C / dt, dj for diodes A / dt value becomes large, and V=Ld i Due to the / dt relationship, a snap-off phenomenon and subsequent oscillation phenomenon occur in the voltage waveform, and the device may be destroyed by the snap-off phenomenon. As a result, the controllability of the turn-off operation of both IGBTs and diodes becomes poor, and the cut-off capability at turn-off decreases. The snap-off and oscillation phenomena also depend on the turn-off operation conditions. In other words, these phenomena occur when the N - A high power supply voltage (V CC ) conditions, the internal carrier density of the device is low at low current densities (J C ,J A ) conditions, or high stray inductance (L S ) and at low operating temperatures.
[0037] Furthermore, with regard to the voltage holding ability, which is an important performance of a power semiconductor device, the off loss increases due to an increase in leakage current when holding the withstand voltage at high temperatures, or the device itself heats up at high temperatures, causing thermal runaway, which makes it impossible to control and prevent operation at high temperatures. The heat generation components of a power semiconductor include the on-state loss, the switching loss that is the loss during switching operations (turning on and off), and the off-state loss. The above situation means that the off-loss, which is one of the three heat generation components of the power semiconductor itself, increases, which poses a problem in the thermal design of the power module that contains the power semiconductor.
[0038] To address the above issues, conventional IGBTs and diodes use - By increasing the thickness of the drift layer 14 or increasing the N - By increasing the impurity concentration of the drift layer 14 and reducing the variation in the impurity concentration, - The parameters of the drift layer 14 have been optimized.- Increasing the thickness of the drift layer 14 increases the ON voltage of both the IGBT and the diode, resulting in an increase in the total loss. - Reducing the variation in the impurity concentration of the drift layer 14 imposes restrictions on the Si wafer manufacturing technology and the Si wafers to be used, leading to a rise in the cost of the Si wafers. Thus, conventional IGBTs and diodes have a technical problem, which can be called a dilemma, in improving device performance.
[0039] As a solution to the second problem above, U.S. Patent Publication Nos. 6,482,681, 7,514,750, and 7,538,412 disclose a method for detecting protons (H + However, these techniques are not compatible with the trend to reduce the total loss of IGBTs or diodes by using N - Considering the thinning of the drift layer 14, it is necessary to increase the concentration of protons in order to maintain the breakdown voltage, which is a basic characteristic of power semiconductors. However, increasing the concentration of protons is accompanied by an increase in crystal defects when protons are introduced, or an increase in the defect density that becomes the center of carrier recombination due to crystal defects, which has the disadvantage of increasing the off-loss of the IGBT and diode (i.e., an increase in leakage current at high temperatures and high voltages) and reducing the breakdown resistance. The basic performance required of power semiconductors is to have the voltage holding ability and guarantee the breakdown resistance while reducing the total loss. Furthermore, if the off-loss increases, the amount of heat generated by the IGBT or diode itself increases, which poses problems for high-temperature operation or for the thermal design of the power module itself that contains the power semiconductor. In other words, the above technology is a step forward in the latest N - This technology does not satisfy the requirements for power semiconductors, which tend to have thinner drift layers 14. Furthermore, increasing the concentration of protons leads to a decrease in the processing capacity of the ion implantation equipment when introducing protons into Si, which is a disadvantage in terms of device manufacturing technology.
[0040] Therefore, an N buffer layer structure that solves the above problems is required in a wafer process that uses wafers manufactured by the FZ method or MCZ method and is also compatible with large-diameter Si wafers. It is also required to reduce sensitivity to the phenomenon of poor withstand voltage of IGBTs or diodes caused by partial non-formation of the N buffer layer 15 due to adverse effects during the wafer process.
[0041] In view of the above, the technology disclosed herein relates to a bipolar power semiconductor such as an IGBT and an FWD, which are key components of a power module (rated withstand voltage: ≧600V), (a) A vertical structure that increases the voltage blocking capability in the off state (static withstand voltage retention capability) and reduces the leakage current when the withstand voltage (voltage) is retained at high temperatures, achieving low off loss and high temperature operation. (b) A vertical structure that suppresses the snap-off phenomenon at the end of the turn-off operation and the oscillation phenomenon caused by the snap-off phenomenon. (c) Vertical structure that improves the cutoff capability (dynamic breakdown resistance) during turn-off operation and expands the safe operating temperature range (d) A vertical structure that can be incorporated into wafer process technology that can also handle larger diameter wafers (≧200 mm) used to manufacture semiconductors. The purpose is to achieve the following.
[0042] The concept of the vertical structure proposed in this disclosure will be described with reference to Figures 4 to 6. The meanings of the symbols in Figures 4 to 6 are as follows. Q pl :charge of plasma layer n pl :electron / hole concentration in plasma layer V: potential q:elementary charge(1.60218×10 -19 C) ε0:dielectric constant in vacuum(8.85418×10 -14 F / cm) ε r : relative dielectric constant (11.9 for Si) N + D :ionized donor concentration(cm -3 ) n,p:free electron / hole concentration (cm -3 ) J n ,J p :current density of electron / hole current(A / cm -3 ) ν sat(n) ,ν sat(p) :saturation drift velocity of electron / hole (cm / s) It is believed that the above technical issues caused by the problems with the vertical structures of conventional IGBTs and diodes can be resolved by realizing the following vertical structure, which is characterized in particular by the N buffer layer 15. The concepts shown below are common to IGBTs and diodes. The ideas regarding the N buffer layer 15 constituting the vertical structure proposed in this disclosure are as follows i, ii, and iii.
[0043] (i) Regarding the depletion phenomenon of the carrier plasma layer near the junction between the n drift layer 14 and the N buffer layer 15 during turn-off operation, the carrier plasma layer is made to remain as shown by the arrow circle 1 in Figure 4. In other words, the conductivity modulation phenomenon occurs even inside the N buffer layer 15 when the device is in the ON state, and the concentration of the N buffer layer 15 is reduced so that the carrier plasma layer exists. The concentration is 10 16 cm -3 Because of the high concentration of 10 or more, 15 cm -3 The charge amount of the carrier plasma layer at this time is expressed by the relational expression Qpl shown in FIG.
[0044] (ii) In a static state, the electric field strength is stopped inside the N buffer layer 15, and in dynamic operation, the concentration gradient near the junction between the n drift layer 14 and the N buffer layer 15 is made gentle so that the depletion layer extends gently inside the N buffer layer 15. Also, the depletion layer stops in the N buffer layer 15 due to the relationship with the remaining carrier plasma layer, as shown by the circled arrow 2 in Fig. 5. The gradient of the electric field strength distribution at this time is expressed by the relational expression dE / dx shown in Fig. 6.
[0045] (iii) By using a thick N layer with a low concentration and a concentration gradient, the basic performance of power semiconductors, namely, the voltage resistance characteristic is guaranteed, that is, the current amplification factor (α pnp ) to achieve low off-state loss by reducing leakage current when off.
[0046] In other words, the N buffer layer 15 plays a role of controlling the carrier plasma state inside the device during device operation, while ensuring the withstand voltage characteristics such as stable withstand voltage characteristics and low off-loss.
[0047] As a result, the snap-off phenomenon at the end of the turn-off operation and the oscillation phenomenon caused by the snap-off phenomenon can be suppressed while stabilizing the withstand voltage characteristics and ensuring low off-loss of the IGBT and diode, thereby realizing a power semiconductor with good controllability in switching operation and improving the breakdown resistance in dynamic conditions. Furthermore, the device is insensitive to the phenomenon of poor withstand voltage of the IGBT and diode caused by partial non-formation of the N buffer layer 15 due to adverse effects during the wafer process, and shows the effect of suppressing an increase in the defect rate of IGBT and diode chips.
[0048] Next, a description will be given of details of N buffer layer 15 of the semiconductor device according to embodiment 1. As described above, N buffer layer 15 of the semiconductor device according to embodiment 1 is composed of a lower first buffer layer 15-1 and an upper second buffer layer 15-2.
[0049] The first buffer layer 15-1 has a peak impurity concentration (C nb1,p) is 1.0 × 10 16 cm -3 Above 5.0×10 16 cm -3 The depth from the back surface of the semiconductor substrate (X j,nb1 ) is a region in which the particle size is 1.2 μm or more and 5.0 μm or less.
[0050] The second buffer layer 15-2 has a maximum peak impurity concentration ((C nb2,p ) max ) is the impurity concentration of the semiconductor substrate (C n- :1.0×10 12 cm -3 Above 5.0×10 14 cm -3 below) and 1.0×10 15 cm -3 The depth from the back surface of the semiconductor substrate (X j,nb2 ) is a region in which the thickness is 4.0 μm or more and 50 μm or less.
[0051] From the above relationship, the roles of the first buffer layer 15-1 and the second buffer layer 15-2 constituting the N buffer layer 15 are as follows, taking into consideration the roles of the targeted N buffer structure shown in FIGS.
[0052] The first buffer layer 15-1 plays the role of stopping the depletion layer extending from the main junction in a static state (the region indicated by the circled arrow 2 in Figure 5), and exhibits the effect of achieving stable voltage resistance characteristics and low off-loss due to low leakage current when off.
[0053] In the second buffer layer 15-2, when it is in the on state (when the rated main current is flowing), the impurity concentration increases from the doping profile at the time of forming the second buffer layer 15-2 in the wafer process due to a carrier plasma layer generated by the conductivity modulation phenomenon (region indicated by the circled arrow 1 in FIG. 4). Therefore, the second buffer layer 15-2 has the role of widening the base width of the PNP transistor, and the current amplification factor (α pnp) and shows the effect of reducing off-loss due to low leakage current during off-state. This carrier plasma layer acts as a residual carrier plasma layer in the dynamic state (area indicated by the arrow circle 1 in FIG. 6). In addition, the second buffer layer 15-2 slows down the growth speed of the depletion layer growing from the main junction in the static and dynamic states by N - This makes the movement slower than that within drift layer 14, and the presence of a residual carrier plasma layer from the on state plays a role in controlling the electric field strength distribution (the area indicated by arrow circle 2 in Figure 6), suppressing the snap-off phenomenon at the end of the turn-off operation and the oscillation phenomenon caused by the snap-off phenomenon, thereby improving the controllability of the switching operation and improving the breakdown resistance in the dynamic state.
[0054] In the following description, when simply referring to the "N buffer layer 15," it basically refers to the N buffer layer 15 consisting of the first buffer layer 15-1 and the second buffer layer 15-2 described above. However, when comparing and explaining the semiconductor device according to the present disclosure with a conventional semiconductor device, the N buffer layer 15 consisting of the first buffer layer 15-1 and the second buffer layer 15-2 may be referred to as the "new-structure N buffer layer 15," and the conventional single-layer N buffer layer 15 (consisting only of a layer corresponding to the first buffer layer 15-1) may be referred to as the "conventional-structure N buffer layer 15." In addition, the RFC diode of the first embodiment including the new-structure N buffer layer 15 may be referred to as the "new-structure RFC diode," and the conventional RFC diode including the conventional-structure N buffer layer 15 may be referred to as the "conventional-structure RCF diode."
[0055] The method for forming the new structure N buffer layer 15 will be described in the fourth embodiment in conjunction with the "Trap B ratio" described later.
[0056] In the RFC diode according to the first embodiment, the parameters of the diffusion layers other than the N buffer layer 15 are set as follows: - The drift layer 14 has an impurity concentration (C n- ) is 1.0 × 10 12 cm -3 Above 5.0×10 14 cm-3 The following is formed from a Si wafer (FZ wafer) fabricated by the FZ method. The final device thickness (t device The thickness of the P anode layer 10 is 40 μm or more and 700 μm or less. The surface impurity concentration of the P anode layer 10 is 1.0×10 16 cm -3 Peak impurity concentration is 2.0×10 or more 16 cm -3 Above 1.0×10 18 cm -3 Hereinafter, the depth is set to 2.0 μm or more and 10.0 μm or less. + The cathode layer 17 has a surface impurity concentration of 1.0×10 18 cm -3 Above 1.0×10 21 cm -3 The P cathode layer 18 has a surface impurity concentration of 1.0×10 16 cm -3 Above 1.0×10 20 cm -3 Hereinafter, the depth is set to 0.3 μm or more and 0.8 μm or less.
[0057] Figure 7 shows the spectrum results of the first buffer layer 15-1 and the second buffer layer 15-2 constituting the new-structure N-buffer layer 15, analyzed by the PL (Photoluminescence) method. The PL method is an analytical technique in which light is irradiated onto a semiconductor and the light emitted when electron-hole pairs recombine via defect levels is observed. The PL measurement was performed using a He-Ne laser (wavelength: 633 nm) at a temperature of 30 K, with a laser intensity of 4.5 mW hitting the sample surface, a laser diameter of 1.3 μm, and a laser intensity of 0.339 MW / cm at the sample surface. 2 This was done by setting
[0058] The vertical axis of FIG. 7 is the PL intensity normalized by the intensity at the band edge of each layer. The PL intensity on the vertical axis of FIG. 7 represents the defect density, and the stronger the PL intensity, the higher the defect density of the defects that are the source of the PL intensity. As shown in FIG. 7, nothing was detected in the first buffer layer 15-1, and the second buffer layer 15-2 had a PL spectrum with two characteristic peaks derived from lattice defects. These two characteristic peaks are the levels due to Trap A (photon energy: 1.0182 eV) and Trap B (photon energy: 1.0398 eV) present in the second buffer layer 15-2.
[0059] The detected Trap A and Trap B correspond to the photon energies of a lattice defect cluster (W-centre) that occurs when protons are introduced into silicon, and a lattice defect pair (X-centre) that occurs when protons diffuse and re-aggregate by annealing, respectively, from the following paper: R. Jones, TAG Eberlein, N. Pinho, BJ Coomer, JP Gross, PR Briddom and S. Oberg, “Self-Interstisial Clusters in Silicon,” Nucl. Instrum. Methods Phys. Res. B, Beam Interact. Mater. At., vol. 186, pp. 10.18, Jan. 2002, doi: 10.1016 / S0168-583X(01)00872-2.
[0060] In this manner, the N buffer layer 15 is composed of the first buffer layer 15-1 that does not have lattice defects and the second buffer layer 15-2 that does have lattice defects. That is, the N buffer layer 15 is composed of multiple N layers with different carrier lifetimes, and the carrier lifetime of the first buffer layer 15-1 is longer than the carrier lifetime of the second buffer layer 15-2.
[0061] The carrier lifetime of the first buffer layer 15-1 is τ 15a , the carrier lifetime of the second buffer layer 15-2 is τ 15b, N - The carrier lifetime of the drift layer 14 is τ 14 , N + The carrier lifetime of the cathode layer 17 is τ 17 , the carrier lifetime of the P cathode layer 18 is τ 18 In the RFC diode of the first embodiment, τ 15b <τ 17 ≒ τ 18 ≒ τ 15a ≒ τ 14 (Details will be given in the fourth embodiment.)
[0062] When the above carrier lifetime relationship is satisfied, the efficiency of carrier injection from the bottom side in the on-state of the diode (electron injection efficiency γ n,active ), reducing the leakage current that occurs when a reverse bias is applied to the main junction and depletion occurs, and promoting carrier recombination by the second buffer layer 15-2. As a result, it is possible to expect effects such as stable on-operation of the diode (FIG. 31), low off-loss (FIG. 32), high-temperature operation based on thermally stable operation (FIGS. 32 and 38), and dynamic durability improvement (for example, FIG. 34 and FIG. 37).
[0063] Next, the effect of the new-structure N-buffer layer 15 on device performance is clarified in an RFC diode having the new-structure N-buffer layer 15, thereby demonstrating the effect of the technology according to the present disclosure in terms of device performance. In discussing the relationship between the two types of traps present in the second buffer layer 15-2 and device performance, the ratio of Trap B defined by the following formula (1) is set as an index representing the relationship between the two types of traps and device performance.
[0064] Trap B ratio (%) = {S PLB / (S PLA +S PLB )}×100...Equation (1) In formula (1), S PLB is the PL intensity of Trap B, S PLAis the PL intensity of Trap A.
[0065] FIG. 8 shows the change in the ratio of Trap B with respect to the depth direction of the second buffer layer 15-2. The values on the horizontal axis represent the device thickness (t device ) is the value obtained by normalizing the device thickness. j,nb1 and X j,nb2 are the junction between the first buffer layer 15-1 and the second buffer layer 15-2, and the junction between the second buffer layer 15-2 and the N - The junction with the drift layer 14. Trap A and Trap B are N - Since it is not detected from the drift layer 14, X j,nb2 Thereafter, the ratio of Trap B is 0. As will be described later, in order to obtain the effect of the new-structure N-buffer layer 15, it is necessary to control the ratio of Trap B within the range of 15% to 55%. From FIG. 8, it can be seen that the ratio of Trap B is constant within the above-mentioned allowable range in the second buffer layer 15-2. The ratio of Trap B in FIG. 27 to FIG. 30 described later is set to a value near the center of the second buffer layer 15-2 based on the behavior in FIG. 8.
[0066] The structures of the intermediate region R2 and edge termination region R3 of the prototype RFC diode are shown in FIG. 3. The structure of the N buffer layer 15 in the intermediate region R2 and edge termination region R3 is the same as that in the active cell region R1. That is, the N buffer layer 15 of the prototype new structure RFC diode includes the first buffer layer 15-1 and the second buffer layer 15-2 across the active cell region R1, intermediate region R2, and edge termination region R3. Also, the N buffer layer 15 of the prototype conventional structure RFC diode has a single-layer structure (only a layer corresponding to the first buffer layer 15-1) across the active cell region R1, intermediate region R2, and edge termination region R3.
[0067] FIG. 9 shows the simulation results for the recovery operation of the conventional structure RFC diode and the new structure RFC diode. FIG. 10 shows an enlarged view of the area surrounded by the dotted circle in FIG. 9. The thickness (t device ) and N - The impurity concentration of the drift layer 14 (C n- In each graph, "new diode-(b)" refers to the new structure RFC diode, and "con. diode-(b)" refers to the conventional structure RFC diode.
[0068] 11 and 12 show the results of simulation of the carrier concentration distribution and the electric field strength distribution between the first and second main surfaces in the PNP transistor region R22 of the RFC diode at the analysis point shown in FIG. 10. FIG. 11 shows the results for the conventional RFC diode, and FIG. 12 shows the results for the new RFC diode. From FIG. 11 and FIG. 12, it can be seen that the new RFC diode has a carrier plasma layer remaining on the cathode side during the recovery operation, unlike the conventional RFC diode. In the new RFC diode, the electric field strength on the cathode side in the latter half of the recovery operation is alleviated due to the presence of this residual carrier plasma layer. As a result, as shown in FIG. 9, the new RFC diode has a smaller dj / dt at the end of the recovery operation than the conventional RFC diode, and the snap-off phenomenon is suppressed.
[0069] Fig. 13 is a diagram showing the simulation results of recovery waveforms during snappy recovery operation in a 4.5 kV class conventional structure RFC diode and new structure RFC diode. Also, Fig. 14 is a diagram showing the device internal states (current density distribution) of the conventional structure RFC diode and the new structure RFC diode at device internal analysis points T1 to T7 shown in Fig. 13.
[0070] 15 to 26 are diagrams showing the internal states (electron and hole concentration distributions, electric field strength distributions) of the PiN diode region R21 and the PNP transistor region R22 of the RFC diode at the device internal analysis points T1 to T7 shown in FIG. 13. That is, FIG. 15 shows the electron concentration distribution in the PiN diode region R21 of the conventional structure RFC diode, and FIG. 16 shows the electron concentration distribution in the PiN diode region R21 of the new structure RFC diode. FIG. 17 shows the hole concentration distribution in the PiN diode region R21 of the conventional structure RFC diode, and FIG. 18 shows the hole concentration distribution in the PiN diode region R21 of the new structure RFC diode. FIG. 19 shows the electric field strength distribution in the PiN diode region R21 of the conventional structure RFC diode, and FIG. 20 shows the electric field strength distribution in the PiN diode region R21 of the new structure RFC diode. FIG. 21 shows the electron concentration distribution in the PNP transistor region R22 of the RFC diode of the conventional structure, and FIG. 22 shows the electron concentration distribution in the PNP transistor region R22 of the RFC diode of the new structure. FIG. 23 shows the hole concentration distribution in the PNP transistor region R22 of the RFC diode of the conventional structure, and FIG. 24 shows the hole concentration distribution in the PNP transistor region R22 of the RFC diode of the new structure. FIG. 25 shows the electric field strength distribution in the PNP transistor region R22 of the RFC diode of the conventional structure, and FIG. 26 shows the electric field strength distribution in the PNP transistor region R22 of the RFC diode of the new structure. The values on the horizontal axis of the graphs in each figure represent the device thickness (t device ) is the value obtained by normalizing the device thickness.
[0071] It can be seen from Fig. 13 that in the conventional RFC diode, a huge tail current is generated in the latter half of the recovery operation. Also, from Fig. 15 to Fig. 26, it is considered that in the conventional RFC diode, a tail current is generated in the latter half of the recovery operation in the following steps, which leads to device destruction.
[0072] First, in the first step, the time when the operation of the PiN diode region R21 converges is determined due to the difference in the rate-determining operation time between the PiN diode region R21 and the PNP transistor region R22. RR After passing the point (T2 in FIG. 13), the operation mode becomes one in which the operation of the PNP transistor region R22 is the rate-limiting factor. In the subsequent second step, the carrier concentration increases on the cathode side to promote hole injection, and the electric field on the cathode side is relaxed while the P anode layer 10 and the N - The electric field strength at the main junction with the drift layer 14 increases, and impact ionization is promoted. Then, in the third step, electrons generated by the impact ionization promoted at the main junction are transported to N - This is injected into the drift layer 14, causing an increase in the base current of the PNP transistor region R22, and generating a huge tail current on the recovery waveform. Finally, in the fourth step, a huge tail current is generated and at the same time the PNP transistor begins to operate, making it impossible to control the recovery operation, leading to device destruction.
[0073] In the new structure RFC diode having the N buffer layer 15 in which lattice defects exist in the second buffer layer 15-2, the above-mentioned third step mode does not occur, and the recovery operation can be controlled. In the new structure RFC diode, the carrier plasma layer on the cathode side disappears as the lattice defects existing in the second buffer layer 15-2 promote carrier recombination, and the P anode layer 10 and the N - This is because the increase in the electric field strength at the main junction with the drift layer 14 is suppressed, and the depletion layer extends toward the cathode side.
[0074] As a result, in the new structure RFC diode, as shown in the current density distribution in FIG. 14, the operation of the PNP transistor region R22 is suppressed to a minimum, and the period of occurrence of the current filament phenomenon that occurs at the device internal analysis point T2 and crosses the PiN diode region R21 and the PNP transistor region R22 is shortened. Therefore, the new structure RFC diode can complete the recovery operation while suppressing the tail current as shown in FIG. 13. As a result, the new structure RFC diode does not generate a huge tail current during the recovery operation in the snappy mode, and it is possible to expand the safe operating temperature range that guarantees the recovery operation. This effect is due to the presence of lattice defects in the second buffer layer 15-2 of the new structure N buffer layer 15.
[0075] Figure 27 shows the J at 298K of the prototype 1200V class new structure RFC diode with various Trap B ratios. R -V R It is a characteristic. R is the leakage current density (A / cm 2 ), V R is the reverse bias voltage (V). J of the diode R -V R The ratio of Trap B affects the characteristics, and if the ratio of Trap B is low, V R =Leakage current (J R ) increases, and J R -V R In other words, if the ratio of Trap B is low, the voltage retention capability, which is the basic performance of the power semiconductor, decreases.
[0076] Figure 28 shows snappy recovery waveforms at low temperatures for prototype 1200V class new structure RFC diodes with various Trap B ratios. The new structure RFC diode controls the lattice defects in the second buffer layer 15-2 to set the Trap B ratio to 40%, and therefore can increase the reverse recovery charge (Q RR It is possible to block the recovery action without increasing the
[0077] Fig. 29 is a table summarizing the relationship between the device performance and the ratio of Trap B of a 1200V class new-structure RFC diode. Fig. 29 shows that a diode having a new-structure N buffer layer 15 can improve the diode performance by controlling the lattice defects in the second buffer layer 15-2 in which the lattice defects constituting the N buffer layer 15 exist.
[0078] FIG. 30 is a diagram showing the relationship between the safe operating temperature range and the ratio of Trap B during snappy recovery operation of a 1200V class RFC diode with a new structure. From FIGS. 13 to 26, it can be inferred that the new structure RFC diode expands the safe operating temperature range that ensures snappy recovery operation. Furthermore, from FIG. 30, by controlling the Trap B in the lattice defects in the second buffer layer 15-2 to 15% to 55%, the Q in the latter half of the recovery operation can be reduced, as described later. RR As shown in FIG. 28, the tail current increase region does not exist due to the reduction, and the recovery operation can be cut off even at a low temperature of 213K, making it possible to expand the operating temperature range to the lower temperature side.
[0079] Below, we will show the results of demonstrating the device performance of a new structure RFC diode in which Trap B in the lattice defects in the second buffer layer 15-2 is controlled to 15% to 55%. Here too, the intermediate region R2 and edge termination region R3 of the prototype RFC diode have the structure shown in Figure 3. The structure of the N buffer layer 15 in the intermediate region R2 and edge termination region R3 is the same as that in the active cell region R1.
[0080] Fig. 31 shows the operating temperature dependence of the output characteristics of a prototype 1200V class new structure RFC diode. Although the new structure RFC diode has the second buffer layer 15-2 in which lattice defects exist in the N buffer layer 15, it satisfies the above-mentioned carrier lifetime relationship and therefore has a γ n,active is stabilized, and the snapback phenomenon does not occur even at a low temperature of 213K, so normal ON operation can be guaranteed.
[0081] In addition, power modules are required to operate many power semiconductor elements in parallel. The new RFC diode has a voltage (V F ) is positive in temperature dependence from 213K to 423K (with increasing temperature, V F This is an effective performance from the viewpoint of current sharing when multiple new structure RFC diodes are operated in parallel.
[0082] Figure 32 shows the J of the prototype 1200V class new structure RFC diode. R -V R The operating temperature dependence of the characteristics. Although the new structure RFC diode has the second buffer layer 15-2 in which lattice defects exist in the N buffer layer 15, by satisfying the above-mentioned carrier lifetime relationship, it does not go into thermal runaway even at a high temperature of 478K, and has both low off-loss performance and high voltage retention capability.
[0083] Figure 33 shows the recovery waveforms at 298K for the conventional and new structure RFC diodes of the 1200V class. device ) and N - The impurity concentration of the drift layer 14 (C n- ) is the same. Compared to RFC diodes with the conventional structure, the new structure RFC diode suppresses the snap-off phenomenon at the end of the recovery operation and the subsequent oscillation phenomenon, and exhibits excellent soft recovery behavior.
[0084] Figure 34 shows the recovery waveforms of the snappy mode at low temperature (253K) for the conventional structure RFC diode and the new structure RFC diode in the 1200V class. Compared to the conventional structure RFC diode, the new structure RFC diode suppresses the huge tail current even in the snappy mode and can cut off even under severe recovery conditions.
[0085] Figures 35 and 36 focus on the breakdown phenomenon caused by the huge tail current in Figure 34, and show the reverse recovery charge (QRR ) and power supply voltage (V CC ) and the operating temperature. Fig. 35 is a diagram showing the relationship between the power supply voltage and the reverse recovery charge at 253K for the prototype 1200V class conventional structure RFC diode and new structure RFC diode. Fig. 36 is a diagram showing the relationship between the operating temperature and the reverse recovery charge at a power supply voltage of 1000V for those conventional structure RFC diode and new structure RFC diode.
[0086] As can be seen from Figures 35 and 36, the new structure RFC diode has a higher Q RR V CC It can be seen that the dependence on operating temperature is small. RR This suppresses the operation of the PNP transistor region R22, which causes an increase in the resistance, and the durability of the RFC diode during dynamic operation is excellent. This behavior is due to the fact that the lattice defects in the second buffer layer 15-2 of the new structure N buffer layer 15 play a role in promoting the recombination of carriers in the latter half of the recovery operation.
[0087] Figure 37 shows the operating temperature at low temperature where the diode can be cut off under snappy recovery operation and V CC 37 is a diagram showing the relationship between the SOA and the VCC. The inside of the data shown in FIG. 37 is the safe operating area (SOA). From FIG. 37, the new structure RFC diode has a high V CC This has the effect of expanding the safe operating area to the side.
[0088] Figure 38 shows waveforms indicating the breaking capability of a 1200V class new structure RFC diode during recovery operation at high current density and high temperature (473K). In order to satisfy the above-mentioned carrier lifetime relationship, the new structure RFC diode can withstand a high current density (3000A / cm2), which is about 10 times the rated current density, even at a high temperature of 473K. 2 ) and has sufficient blocking capacity.
[0089] Figure 39 shows the on-state voltage (V F ) and recovery loss (E REC The new RFC diode has a trade-off characteristic with the device thickness (t device ) is shrunk compared to the conventional RFC diode, and durability during dynamic operation is improved as shown in FIGS. 33 to 38, while achieving a reduction in total loss performance as shown in FIG.
[0090] From the above, the new-structure N-buffer layer 15 includes the second buffer layer 15-2 in which lattice defects exist, but since the Trap B in the lattice defects in the second buffer layer 15-2 is controlled to 15% to 55% and the above-mentioned carrier lifetime relationship is satisfied, the new-structure N-buffer layer 15 is thermally stable. Therefore, the new-structure N-buffer layer 15 can guarantee, in an RFC diode, normal on-operation, sufficient withstand voltage retention capability with low off-loss without thermal runaway even at high temperatures, high breaking capability under high temperatures, high voltages or large currents, low total loss performance, operation at high temperatures of 473K, and the like. In other words, the new-structure N-buffer layer 15 is a technology that realizes the above-mentioned objects (a) to (c).
[0091] <Embodiment 2> In the second embodiment, the new-structure N buffer layer 15 is applied to a PiN diode, and it is demonstrated that even in this case, the new-structure N buffer layer 15 exhibits the same behavior as in the first embodiment.
[0092] 40 is a cross-sectional view of a PiN diode, which is a semiconductor element according to the first embodiment, and corresponds to the cross section taken along line A1-A2 in FIG. 1. Also, FIG. 41 is a cross-sectional view showing an example of the structure of the outer periphery of a semiconductor device including a PiN diode according to the first embodiment, and corresponds to the cross section taken along line A3-A4 in FIG. 1. The configurations in FIGS. 40 and 41 are obtained by deleting the P cathode layer 18 from the configurations in FIGS. 2 and 3 shown in the first embodiment, and replacing the cathode layer with an N + Only the cathode layer 17 is included.
[0093] As shown in FIG. 41, N - The drift layer 14, the N buffer layer 15 and the N + There is a structure in which the cathode layer 17 is stacked, and this structure is the vertical structure 35 of the PiN diode.
[0094] 40, the N buffer layer 15 of the PiN diode according to the second embodiment is a new-structure N buffer layer 15 including a first buffer layer 15-1 and a second buffer layer 15-2. Hereinafter, the PiN diode according to the second embodiment including the new-structure N buffer layer 15 may be referred to as a "new-structure PiN diode," and the conventional PiN diode including the conventional-structure N buffer layer 15 may be referred to as a "conventional-structure PiN diode."
[0095] The carrier lifetime of the first buffer layer 15-1 is τ 15a , the carrier lifetime of the second buffer layer 15-2 is τ 15b , N - The carrier lifetime of the drift layer 14 is τ 14 , N + The carrier lifetime of the cathode layer 17 is τ 17 In the PiN diode of the second embodiment, τ 15b <τ 17 ≒ τ 15a ≒ τ 14 (Details will be given in the fourth embodiment.)
[0096] Hereinafter, it will be shown that the PiN diode according to the second embodiment also realizes suppression of the snap-off phenomenon during recovery operation and the subsequent oscillation phenomenon, and improves the breakdown resistance during recovery operation. Here, the demonstration results using a 4.5 kV class PiN diode are shown. In the new structure PiN diode described in Figs. 42 to 44, Trap B in the lattice defects in the second buffer layer 15-2 is controlled to 15% to 55%, as in the first embodiment. In addition, the intermediate region R2 and edge termination region R3 of the prototype PiN diode have the structure shown in Fig. 41. However, the N buffer layer 15 in the intermediate region R2 and edge termination region R3 has the same structure as the active cell region R1.
[0097] In the PiN diode according to the second embodiment, the parameters of the diffusion layers other than the N buffer layer 15 are set as follows: - The drift layer 14 has an impurity concentration (C n- ) is 1.0 × 10 12 cm -3 Above 5.0×10 14 cm -3 The following is formed from a Si wafer (FZ wafer) fabricated by the FZ method. The final device thickness (t device The thickness of the P anode layer 10 is 40 μm or more and 700 μm or less. The surface impurity concentration of the P anode layer 10 is 1.0×10 16 cm -3 Peak impurity concentration is 2.0×10 or more 16 cm -3 Above 1.0×10 18 cm -3 Hereinafter, the depth is set to 2.0 μm or more and 10.0 μm or less. + The cathode layer 17 has a surface impurity concentration of 1.0×10 18 cm -3 Above 1.0×10 21 cm -3 , the depth is set to 0.3 μm or more and 0.8 μm or less. The parameters of the N buffer layer 15 (the first buffer layer 15-1 and the second buffer layer 15-2) may be the same as those in the first embodiment.
[0098] Figure 42 shows the waveforms during snappy recovery operation at 298K for a 4.5kV class PiN diode with a conventional structure and a new structure PiN diode. Compared to an RFC diode, a PiN diode is more likely to have a depleted residual carrier plasma layer on the cathode side in the latter half of the recovery operation, so it has a smaller effect of suppressing the snap-off phenomenon during recovery operation. For this reason, in Figure 42, the conventional PiN diode experiences a snap-off phenomenon, which then leads to device destruction.
[0099] In contrast, the new PiN diode has a - Due to the influence of the residual carrier plasma layer near the junction between the drift layer 14 and the second buffer layer 15-2, the speed at which the depletion layer extends from the main junction during recovery operation slows down in the second buffer layer 15-2, and even if a snap-off phenomenon occurs, the snap-off voltage is kept low. In other words, in the new structure PiN diode, a carrier plasma layer exists in the second buffer layer 15-2 from the on state, as shown in the region indicated by the arrow circle 1 in Figures 4 and 6. Therefore, at the point where the snap-off phenomenon occurs in the conventional structure PiN diode, a residual carrier plasma layer exists in the second buffer layer 15-2 region, and this controls the electric field strength distribution, thereby delaying the point at which the snap-off phenomenon occurs and preventing device destruction.
[0100] Figure 43 shows the power supply voltage (V CC ) and snap-off voltage (V snap-off ) is shown. With the new structure PiN diode, the device does not break down even at voltages that would cause breakdown in a conventional structure PiN diode, and it can be seen that the breakdown tolerance during snap recovery operation is improved. Furthermore, with the new structure PiN diode, the snap-off voltage V CC The dependency is desensitized to high V CC It can be seen that this is the most effective way to increase the breakdown resistance on the
[0101] Figure 44 shows the power supply voltage (V CC ) and reverse recovery charge (Q RR ) is shown in the figure. The new structure PiN diode has a higher Q RR V CC This leads to the effect of improving the breakdown resistance during snappy recovery operation in the new structure PiN diode.
[0102] As described above, the effect of improving the breakdown voltage can be obtained also in a PiN diode by using the new structure N buffer layer 15 in which the Trap B in the lattice defects in the second buffer layer 15-2 is controlled to 15% to 55%.
[0103] <Embodiment 3> In the third embodiment, the new-structure N-buffer layer 15 is applied to an IGBT, and it is demonstrated that even in this case, the new-structure N-buffer layer 15 exhibits the same behavior as in the first embodiment.
[0104] Fig. 45 is a cross-sectional view of a trench-gate IGBT that is a semiconductor element according to the third embodiment, and corresponds to the cross section taken along the line A1-A2 in Fig. 1. Also, Fig. 46 is a cross-sectional view showing an example of the structure of the outer periphery of a semiconductor device including an IGBT according to the third embodiment, and corresponds to the cross section taken along the line A3-A4 in Fig. 1.
[0105] First, the structure of the IGBT formed in the active cell region R1 will be described with reference to Fig. 45. The semiconductor substrate constituting the IGBT has an N-type (first conductivity type) - The drift layer 14 is formed. - Under the drift layer 14, N -An N-type N buffer layer 15 having a higher peak impurity concentration than the drift layer 14 is formed. The N buffer layer 15 is a new-structure N buffer layer 15 made up of a lower first buffer layer 15-1 and an upper second buffer layer 15-2. Although not shown, the N buffer layer 15 in FIG. 46 also includes the first buffer layer 15-1 and the second buffer layer 15-2.
[0106] A P-type (second conductivity type) P collector layer 16 is formed below the N buffer layer 15 in the semiconductor substrate. The P collector layer 16 is formed in a surface layer portion on the back surface side (second main surface side) of the semiconductor substrate, and a collector electrode 20 is formed on the back surface of the semiconductor device so as to be in contact with the P collector layer 16.
[0107] N in semiconductor substrate - On the drift layer 14, N - An N layer 11 having a higher peak impurity concentration than the drift layer 14 is formed, and a P-type P base layer 9 is formed on the N layer 11. The P base layer 9 is formed in the surface layer portion on the upper surface side (first main surface side) of the semiconductor substrate. + The emitter layer 7 and the P-type P + Layer 8 is selectively formed on the IGBT. In the IGBT, the junction between P base layer 9 and N layer 11 serves as a main junction.
[0108] On the upper surface of the semiconductor substrate, a P base layer 9 and an N layer 11 are vertically penetrated, and an N - A trench is formed so as to reach the drift layer 14. A gate insulating film 12 is formed on the inner wall of the trench, and a gate electrode 13 is formed on the gate insulating film 12 so as to fill the trench. The gate electrode 13 is connected to the N - The drift layer 14, the N layer 11, the P base layer 9, and the N + The gate electrode 13 faces the emitter layer 7. + The emitter layer 7, the P base layer 9 and the N layer 11 form an insulated gate transistor structure (MOS transistor structure) in the IGBT.
[0109] An interlayer insulating film 6 is formed on the upper surface of the semiconductor substrate so as to cover the gate electrode 13, and an emitter electrode 4 is formed on the interlayer insulating film 6. The emitter electrode 4 is connected to an N + Emitter layer 7 and P + The gate electrode 13 formed on the outer periphery of the active cell region R1 (the right one of the two gate electrodes 13 shown in the active cell region R1 in FIG. 46) does not serve as an original gate electrode, but is a dummy electrode set to a dynamic potential with the emitter electrode 4. The purpose and effect of the dummy electrode are described in Japanese Patent No. 4205128, Japanese Patent No. 4785334, and Japanese Patent No. 5634318, and are to suppress the saturation current density of the IGBT, to suppress oscillation in a no-load short-circuit state by controlling the capacitance characteristics, thereby improving the short-circuit resistance, and to lower the on-voltage by improving the carrier concentration on the emitter side.
[0110] Next, with reference to FIG. 46, the structure of the intermediate region R2 and edge termination region R3 of the IGBT will be described.
[0111] N - The drift layer 14 is formed across the active cell region R1, the intermediate region R2 and the edge termination region R3. - A P-type guard ring 22 is formed in the surface layer of the drift layer 14, and is formed deeper than the P-base layer 9. The guard ring 22 extends to the active cell region R1 side, connects to the P-base layer 9, and contains the gate electrode 13, which is a dummy electrode (i.e., the guard ring 22 is formed deeper than the gate electrode 13, which is a dummy electrode). - A P-type field limiting ring 23 is selectively formed on the surface layer of the drift layer 14.
[0112] Further, outside the field limiting ring 23 in the edge termination region R3, a structure consisting of an N layer 11, a gate insulating film 12, and a gate electrode 13 similar to that in the active cell region R1 is provided. + A channel stopper layer 24 of N type is selectively formed. The channel stopper layer 24 is connected to the guard ring 22 and the field limiting ring 23. - The purpose of this is to stop the extension of the depletion layer from the junction with the drift layer 14 .
[0113] An insulating film 25 is formed on the upper surface of the semiconductor substrate in the intermediate region R2 and the edge termination region R3. A surface gate electrode 13a connected to the gate electrode 13 is formed on the insulating film 25 in the intermediate region R2, and an interlayer insulating film 6 is formed on the insulating film 25 and the surface gate electrode 13a.
[0114] An FLR electrode 27 connected to the field limiting ring 23 through a contact hole, a channel stopper electrode 28 connected to the channel stopper layer 24 through a contact hole, and a surface gate electrode 31 connected to the surface gate electrode 13a through a contact hole are formed on the interlayer insulating film 6. The FLR electrode 27, the channel stopper electrode 28, and the surface gate electrode 31 can be formed simultaneously with the emitter electrode 4 in the active cell region R1.
[0115] In addition, a passivation film 29, which is a protective film covering the FLR electrode 27, the channel stopper electrode 28 and the surface gate electrode 31, is formed across the intermediate region R2 and the edge termination region R3, and a passivation film 30 is further formed on the passivation film 29.
[0116] As shown in FIG. 46, N - There is a structure in which the drift layer 14, the N buffer layer 15 and the P collector layer 16 are laminated, and this structure is the vertical structure 35 of the IGBT.
[0117] 46 shows a structure in which the P collector layer 16 is formed also in the intermediate region R2 and the edge termination region R3, but as shown in FIG. 47, the P collector layer 16 does not have to be formed in the intermediate region R2 and the edge termination region R3. That is, the vertical structure 35 in the intermediate region R2 and the edge termination region R3 is an N - It may be composed of only drift layer 14 and N buffer layer 15. Compared with the IGBT of Fig. 46, the IGBT of Fig. 47 can suppress adverse effects on on-voltage, switching loss, and SOA in a short-circuit state, and can significantly improve the SOA during turn-off operation. Details of this technology are described in Japanese Patent No. 6165271.
[0118] As described above, the N buffer layer 15 of the IGBT according to the third embodiment is a new-structure N buffer layer 15 including the first buffer layer 15-1 and the second buffer layer 15-2. Hereinafter, the IGBT according to the third embodiment including the new-structure N buffer layer 15 may be referred to as a "new-structure IGBT," and the conventional IGBT including the conventional-structure N buffer layer 15 may be referred to as a "conventional-structure IGBT."
[0119] The carrier lifetime of the first buffer layer 15-1 is τ 15a , the carrier lifetime of the second buffer layer 15-2 is τ 15b , N - The carrier lifetime of the drift layer 14 is τ 14 , the carrier lifetime of the P collector layer 16 is τ 16 In the IGBT of the third embodiment, τ 15b <τ 16 ≒ τ 15a ≒ τ 14 (Details will be given in the fourth embodiment.)
[0120] By satisfying the above carrier lifetime relationship, the carrier injection efficiency from the bottom side in the IGBT on state (hole injection efficiency (γ p,active), reducing the leakage current that occurs when a reverse bias is applied to the main junction and depletion occurs, and promoting carrier recombination by the second buffer layer 15-2. As a result, it is possible to expect effects such as stable on-operation of the IGBT, low off-loss (Fig. 56), high-temperature operation based on thermally stable operation (Fig. 56), and dynamic durability improvement (Fig. 53).
[0121] In the IGBT according to the third embodiment, the parameters of the diffusion layers and trenches other than the N buffer layer 15 are set as follows. - The drift layer 14 has an impurity concentration (C n- ) is 1.0 × 10 12 cm -3 Above 5.0×10 14 cm -3 The following is formed from a Si wafer (FZ wafer) fabricated by the FZ method. The final device thickness (t device ) is greater than or equal to 40 μm and less than or equal to 700 μm.
[0122] The P base layer 9 has a peak impurity concentration of 1.0×10 16 cm -3 Above 1.0×10 18 cm -3 , depth N + It is set to be deeper than the emitter layer 7 and shallower than the N layer 11. The N layer 11 has a peak impurity concentration of 1.0×10 15 cm -3 Above 1.0×10 17 cm -3 The depth of the following layers is set to be 0.5 μm to 1.0 μm deeper than the P base layer 9. + The emitter layer 7 has a peak impurity concentration of 1.0×10 18 cm -3 Above 1.0×10 21 cm -3 Hereinafter, the depth is set to 0.2 μm or more and 1.0 μm or less. + Layer 8 has a surface impurity concentration of 1.0×10 18 cm -3 More than 1.010 21 cm -3 Below, the depth is N +The P collector layer 16 is set to be the same as or deeper than the emitter layer 7. The surface impurity concentration of the P collector layer 16 is 1.0×10 16 cm -3 Above 1.0×10 20 cm -3 Hereinafter, the depth is set to 0.3 μm or more and 0.8 μm or less. The trench in which the gate electrode 13 is buried has a depth (D trench ) is set to 2.0 μm or more and to a depth that reaches at least the N layer 11. The parameters of the N buffer layer 15 (first buffer layer 15-1 and second buffer layer 15-2) may be the same as those in the first embodiment.
[0123] The intermediate region R2 and edge termination region R3 of the prototype IGBT structure have the structure shown in Fig. 46 or 47. However, the N buffer layer 15 in the intermediate region R2 and edge termination region R3 has the same structure as that in the active cell region R1.
[0124] 48 to 50 show the results of verifying, by simulation, the contribution of the new-structure N-buffer layer 15 in a peculiar short-circuit state in the operation of an IGBT that cannot be seen in a diode.
[0125] Figure 48 shows the electric field distribution inside the IGBT in a short circuit state. Figure 49 shows the electric field strength at the top and bottom of the device and the on-state voltage (V CE (sat)) on the horizontal axis. CE (sat) is the hole injection efficiency from the collector side (γ p,active ) in the figure. "Front side" in the figure is the electric field strength of the main junction on the front surface of the device, and "Back side" is the maximum electric field strength in the N buffer layer 15 on the back surface of the device. CE This shows the internal temperature distribution above the device when the conventional and new structure IGBT devices are short-circuited with (sat) held constant. The temperature distribution shown in Fig. 50 was measured at the times shown in Fig. 51. Fig. 51 shows the short-circuit waveform in a simulation when the internal temperature distribution of the device shown in Fig. 50 was analyzed.
[0126] In conventional IGBTs, the hole injection efficiency from the collector side (γ p,active ) is high, the electric field strength at the main junction increases, and as a result, the temperature at the top of the device increases as shown in Figure 50, which is highly likely to lead to a decrease in the breaking ability in a short-circuit state. On the other hand, in the new-structure IGBT, the carrier plasma layer in the collector part in a short-circuit state is controlled by the effect of promoting carrier recombination due to lattice defects in the second buffer layer 15-2, which makes it easier for the depletion layer to extend to the back side of the device, resulting in a well-balanced distribution of electric field strength even in a short-circuit state, as shown in Figure 48. As a result, in the new-structure IGBT, as shown in Figure 50, the temperature rise at the top of the device is suppressed more than in the conventional-structure IGBT, and it is expected that the breaking ability in a short-circuit state will be improved.
[0127] FIG. 52 shows the turn-off waveforms at 298K for the conventional and new structure IGBTs of 1200V class. devic ) and N - The impurity concentration of the drift layer 14 (C n- ) is the same. In the new-structure IGBT, a carrier plasma layer remains on the collector side in the latter half of the turn-off operation, just like the new-structure RFC diode shown in Figure 12. Therefore, in the new-structure IGBT, the electric field strength on the collector side is mitigated, suppressing the snap-off phenomenon and subsequent oscillation phenomenon observed in the conventional-structure IGBT, and thus demonstrating turn-off operation with excellent controllability.
[0128] Figure 53 shows the short-circuit waveforms for a 1200V class IGBT with a conventional structure and a new structure. Figure 54 shows the maximum energy density (E SC ) and device thickness (t device 53 and 54 show that the new structure IGBT has a wider safe operating area (SOA) in a short circuit state than the conventional structure IGBT, and can reduce the device thickness (t device ) can be shrunk.
[0129] Figure 55 shows the on-voltage (V CE (sat)) and turn-off loss (E OFF The new IGBT structure shows a trade-off characteristic with the device thickness (t device ) can be shrunk compared to conventional IGBTs, and as shown in Figures 52 and 53, it has excellent controllability during dynamic operation and improved durability, while achieving a reduction in total loss as shown in Figure 55.
[0130] 56 and 57 are diagrams showing the effect of the new-structure N-buffer layer 15 in a high-voltage class (6.5 kV) IGBT. FIG. 56 shows the main junction leakage current (J CES ) and collector-emitter voltage (V CES 56 shows the relationship between the current amplification factor (α pnp ), which results in low off-state loss due to low leakage current when off.
[0131] Figure 57 shows the turn-off waveforms at 213K for a 6.5kV class IGBT with a conventional structure and a new structure IGBT. From the turn-off waveforms at low temperatures, it can be seen that in the IGBT with the new structure N buffer layer 15, a residual carrier plasma layer exists on the collector side in the latter half of the turn-off operation. Therefore, in the new structure IGBT, the electric field strength on the collector side is mitigated, suppressing the snap-off phenomenon and the subsequent oscillation phenomenon, and the device exhibits excellent turn-off operation even at low temperatures.
[0132] As described above, it is clear that the above objectives (a) to (c) can be achieved even in an IGBT using a new structure N-buffer layer 15 in which the ratio of Trap B among lattice defects in the second buffer layer 15-2 is controlled to 15% to 55%.
[0133] <Fourth embodiment> In the fourth embodiment, a method for manufacturing an IGBT and an RFC diode using a new structure N buffer layer 15 will be described.
[0134] First, a manufacturing method of a semiconductor device including an IGBT will be described with reference to Figures 58 to 70. These figures show the manufacturing method in the active cell region R1.
[0135] N - A semiconductor substrate, which is an FZ wafer on which a drift layer 14 is formed, is prepared, and ion implantation and annealing processes are performed to form an N layer 11 and a P base layer 9 in the surface layer portion of the semiconductor substrate, as shown in Fig. 58. In addition, a SiO2 film 101 is formed on the P base layer 9.
[0136] Furthermore, by subjecting the semiconductor substrate to ion implantation and annealing treatment, N is formed in the upper layer of the P base layer 9 as shown in FIG. + The emitter layer 7 is selectively formed.
[0137] Next, as shown in FIG. 60, an oxide film 102 is formed on the upper surface of the semiconductor substrate and patterned using photolithography. Then, reactive ion etching using plasma is performed on the portion exposed at the opening of the oxide film 102 to form a trench 103. After that, chemical dry etching and sacrificial oxidation are performed to remove crystal defects and plasma damage layers around the trench 103, round the bottom of the trench 103, and flatten the inner wall of the trench 103. The chemical dry etching and sacrificial oxidation are disclosed in, for example, Japanese Patent Application Laid-Open No. 7-263692. The appropriate depth of the trench 103 is disclosed in, for example, WO2009-122486.
[0138] 61, a gate insulating film 12 is formed on the inner wall of the trench 103 by thermal oxidation or CVD (Chemical Vapor Deposition) (for example, see JP-A-2001-085686). Then, a polysilicon layer 104 doped with phosphorus is formed in the trench 103 including the gate insulating film 12 to fill the trench 103. Note that an oxide film 105 is formed on the lower surface of the semiconductor substrate at the same time as the gate insulating film 12 is formed, and a polysilicon layer 106, which is a polysilicon layer doped with phosphorus, is formed on the oxide film 105 at the same time as the polysilicon layer 104 is formed.
[0139] Next, as shown in FIG. 62, the portion of the polysilicon layer 104 that is exposed outside the trench 103 is etched. At this time, the polysilicon layer 104 remaining in the trench 103 becomes the gate electrode 13. Thereafter, an oxide film 107 is formed on the exposed surface of the gate electrode 13 by a thermal oxidation method or a CVD method. Furthermore, a P + Layer 8 is then formed. Thereafter, oxide film 108 doped with boron or phosphorus and TEOS film 109 are formed by CVD on the upper surface of the semiconductor substrate, thereby forming interlayer insulating film 6. A TEOS film or silicate glass may be formed as oxide film 108. When oxide film 108 and TEOS film 109 are formed, TEOS film 110 is formed on the lower surface of the semiconductor substrate.
[0140] Next, as shown in FIG. 63, the TEOS film 110, the polysilicon layer 106, and the oxide film 105 on the lower surface of the semiconductor substrate are etched using a liquid containing hydrofluoric acid or a mixed acid (for example, a mixed liquid of hydrofluoric acid, nitric acid, and acetic acid) to obtain a N - The drift layer 14 is exposed.
[0141] Next, as shown in FIG. 64, a polysilicon layer 111 doped with impurities is deposited on the N exposed on the lower surface of the semiconductor substrate by LPCVD (Low-Pressure Chemical Vapor Deposition). - The polysilicon layer 111 is formed on the drift layer 14. The polysilicon layer 111 is doped with N+ The impurity concentration of the polysilicon layer 111 is 1×10 19 (cm -3 ) or more. The thickness of the polysilicon layer 111 is set to 500 (nm) or more. At the same time as the formation of the polysilicon layer 111, a polysilicon layer 112 is also formed on the upper surface of the semiconductor substrate.
[0142] Next, as shown in FIG. 65, the temperature of the semiconductor substrate is heated to about 900° C. to 1000° C. in a nitrogen atmosphere to convert the impurities in the polysilicon layer 111 into N. - The N - A high concentration of N + The gettering layer 113 is formed by the layer 113a and the high crystal defect density layer 113b. + The impurity concentration at the surface of the layer 113a is, for example, 1.0×10 19 cm -3 Above 1.0×10 22 is set to the range.
[0143] After that, the temperature of the semiconductor substrate is lowered to about 600 to 700°C at an arbitrary cooling rate and maintained at that temperature for 4 hours or more. This process is called the annealing process. In the annealing process, the semiconductor substrate is heated and the N - Metal impurities, contaminating atoms, and damage introduced into the drift layer 14 diffuse and are captured in the gettering layer 113 .
[0144] Next, as shown in Fig. 66, the polysilicon layer 112 on the upper surface of the semiconductor substrate is selectively removed using a liquid of hydrofluoric acid or a mixed acid (for example, a mixed liquid of hydrofluoric acid / nitric acid / acetic acid). The gettering process shown in Fig. 64 to Fig. 66 is disclosed in, for example, WO2014-054121.
[0145] Then, as shown in FIG. 67, the interlayer insulating film 6 is partially etched on the upper surface side of the semiconductor substrate to form a N + Emitter layer 7 and P +A contact hole is formed down to layer 8, and a trench exposed portion 114 is formed in part of active cell region R1, where the upper surface of gate electrode 13 embedded in the trench is exposed. The portion other than trench exposed portion 114 functions as a MOS transistor portion in the IGBT.
[0146] The purpose of forming the exposed trench portion 114 in a part of the region where the gate electrode 13 is formed is to reduce the effective gate width and adjust the capacitance by setting a part of the gate electrode 13 at the emitter potential. This makes it possible to suppress the saturation current density, suppress oscillation during a short circuit by controlling the capacitance, improve short circuit resistance (see WO2002-058160 and WO2002-061845 for details), and lower the on-voltage by improving the emitter side carrier concentration in the on-state.
[0147] 68, a metal for forming a silicide layer 115 and a barrier metal layer 116 is formed on the upper surface of the semiconductor substrate by sputtering. Furthermore, a metal wiring layer containing about 1 to 3% Si is formed by sputtering, and the metal wiring layer is patterned to form an emitter electrode 4. The material of this metal wiring layer is, for example, AlSi, AlSiCu, or AlCu. The emitter electrode 4 is electrically connected to the trench exposed portion 114.
[0148] Next, as shown in FIG. 69, the gettering layer 113 and the polysilicon layer 111 formed on the lower surface of the semiconductor substrate are removed by polishing or etching. The process of removing the gettering layer 113 and the like is called a removal process. - The portion of the drift layer 14 that is in contact with the gettering layer 113 may be removed by a desired thickness. - The thickness of the drift layer 14) can be set to correspond to the breakdown voltage class of the semiconductor device.
[0149] 70, an N buffer layer 15 is formed in a surface layer portion on the back surface side of the semiconductor substrate. Phosphorus, selenium, sulfur, or protons (hydrogen) are implanted from the back surface side of the semiconductor substrate, and the semiconductor substrate is annealed to activate the implanted ions, thereby forming the N buffer layer 15. Details of the process flow for the N buffer layer 15 of the fourth embodiment (i.e., the new structure N buffer layer 15) will be described later.
[0150] Thereafter, a P-type P collector layer 16 is formed on the surface layer portion on the back surface side of the N buffer layer 15. Furthermore, a collector electrode 20 is formed so as to be in contact with the P collector layer 16. The collector electrode 20 is a portion that is soldered to a semiconductor substrate or the like in a module when the semiconductor device is mounted on a module. For this reason, it is preferable to form the collector electrode 20 by stacking a plurality of metals to achieve low contact resistance.
[0151] Next, a method for manufacturing a semiconductor device provided with an RFC diode will be described with reference to FIGS.
[0152] In FIG. 71, an active cell region R1, a middle region R2 and an edge termination region R3 are shown. - A semiconductor substrate on which only the drift layer 14 is formed is prepared. Then, an insulating film 25 made of an oxide film is formed on the semiconductor substrate and patterned, ion implantation is performed using the insulating film 25 as a mask, and then an annealing process is performed on the semiconductor substrate to remove N ions in the intermediate region R2 and the edge termination region R3. - A guard ring 22, a field limiting ring 23, and a channel stopper layer 24 are formed on the surface layer of the drift layer 14. When the insulating film 25 is formed, an oxide film 121 is formed on the lower surface of the semiconductor substrate.
[0153] Next, as shown in FIG. 72, N - The surface portion of the drift layer 14 is subjected to ion implantation and annealing to form the P anode layer 10. In addition, the N -Ion implantation and annealing are performed on the surface layer of the drift layer 14 to form a channel stopper layer 24. Furthermore, an interlayer insulating film 26 made of a TEOS film is formed on the upper surface of the semiconductor substrate, and then a process is performed to expose the lower surface of the semiconductor substrate.
[0154] 73, the impurity-doped polysilicon layer 122 is then etched onto the N - It is formed so as to be in contact with the drift layer 14. At this time, a polysilicon layer 123 is also formed on the upper surface of the semiconductor substrate.
[0155] Next, as shown in FIG. 74, the semiconductor substrate is heated to convert the impurity of the polysilicon layer 122 into N. - The N - A high concentration of N + A gettering layer 124 consisting of a layer 124a and a high crystal defect density layer 124b is formed. This process is the same as the process of forming the gettering layer 113 in the IGBT manufacturing method shown in FIG. 65. Then, an annealing process is performed to remove N - Metal impurities, contaminating atoms, and damage in the drift layer 14 are captured by the gettering layer 124 .
[0156] 75, the polysilicon layer 123 on the upper surface of the semiconductor substrate is selectively removed using a liquid of hydrofluoric acid or a mixed acid (for example, a mixed liquid of hydrofluoric acid / nitric acid / acetic acid). This gettering process is the same as the gettering process of the IGBT described above.
[0157] 76, contact holes reaching the P anode layer 10, the guard ring 22, the field limiting ring 23, and the channel stopper layer 24 are formed in the insulating film 25 and the interlayer insulating film 26. Thereafter, an aluminum film containing about 1% to 3% Si is formed by sputtering on the upper surface of the semiconductor substrate and patterned to form the anode electrode 5, the FLR electrode 27, and the channel stopper electrode 28.
[0158] Subsequently, as shown in FIG. 77, passivation films 29 and 30, which are protective films, are formed on the upper surface of the semiconductor substrate.
[0159] Thereafter, as shown in FIG. 78, the gettering layer 124 and the polysilicon layer 122 on the lower surface side of the semiconductor substrate are removed by polishing or etching. - The thickness of the drift layer 14) is set to correspond to the breakdown voltage class of the semiconductor device.
[0160] And, as shown in Figure 79, N - An N buffer layer 15 is formed on the lower surface side of the drift layer 14. Then, a P cathode layer 18 is formed on the lower surface of the N buffer layer 15. Next, in the active cell region R1, an N + The cathode layer 17 is formed. The N buffer layer 15, + The cathode layer 17 and the P cathode layer 18 are diffusion layers formed by ion implantation and annealing treatment. Finally, a cathode electrode 19 is formed on the lower surface of the semiconductor substrate.
[0161] The substrate concentration (C n- ) is determined according to the voltage class of the semiconductor device being manufactured. For example, C n- is 1.0 × 10 12 cm -3 Above 5.0×10 14 cm -3 The process is as follows. The Si wafer is produced by the FZ method. Then, during the wafer process shown in FIG. 69 or FIG. 78, the thickness of the device is precisely adjusted according to the breakdown voltage class, and the vertical structure 35 is constructed during the wafer process shown in FIG. 70 or FIG. 79. The wafer process in which the vertical structure region is constructed during the wafer process using the FZ wafer in this way is becoming mainstream for the following reasons.
[0162] a) N as wafer -In the case of a wafer in which the drift layer 14 is formed by the epitaxial method, the cost of the Si wafer is very high because it depends on the Si thickness formed by the epitaxial method. - Only the concentration of the drift layer 14 is set to an appropriate value for each withstand voltage class, and at the start of the wafer process, the same thickness of N is used regardless of the withstand voltage class. - By using a Si wafer for the drift layer 14, it becomes possible to employ wafers with low unit prices.
[0163] b) In order to utilize wafers manufactured by the above FZ method, the thickness of the device can be controlled to a value required for the breakdown voltage class in the final stage of the wafer process shown in Figure 69 or Figure 78, and a vertical structure can be constructed, thereby adopting a wafer process that minimizes the modification of the process equipment. This makes it possible to handle a variety of wafer thicknesses, from 40 μm to 700 μm, even in the wafer process for large-diameter Si wafers with diameters of 200 mm or more.
[0164] c) Based on background b), the device structures such as MOS transistor structures, various diffusion layers, and wiring structures formed on the wafer surface for both IGBTs and diodes can be fabricated by directly utilizing the latest process equipment.
[0165] The impurity concentration of the n drift layer and the device thickness are device parameters that affect not only the withstand voltage characteristics of the IGBT and diode, but also the total loss, controllability during dynamic operation, and breakdown resistance, and therefore require high precision.
[0166] 80 is a diagram comparing two types of process flows for forming the second buffer layer 15-2 of the new-structure N-buffer layer 15. Process A is the process flow of the fourth embodiment. The prototype results shown in the first to third embodiments are the results of prototyping the second buffer layer 15-2 by Process A.
[0167] Details of the above-mentioned wafer process are also described in Japanese Patent Nos. 3,396,553, 5,622,814, 6,065,067, and 6,558,462.
[0168] That is, even when the new structure N buffer layer 15 consisting of the first buffer layer 15-1 and the second buffer layer 15-2 is formed, as described in Japanese Patent No. 6065067 and Japanese Patent No. 6558462, the IGBT and the diode are subjected to the N-type diffusion process during the wafer process so as to satisfy a value equal to or greater than the carrier lifetime calculated by the following formula (2). - In the step of restoring the carrier lifetime of drift layer 14 (the steps of FIGS. 64 to 65 or the steps of FIGS. 73 to 74), a high concentration (for example, a surface concentration of 1.0×10 20 cm -3 Above 1.0×10 22 cm -3 The following is a depth of 1.0 μm or more and 10 μm or less) + The gettering layer 113 or 124 is formed of the layer 113a or 124a and the high crystal defect density layer 113b or 124b. However, in this embodiment, the N buffer layer 15, the P collector layer 16, the N + Before forming the cathode layer 17, the P cathode layer 18, etc., the gettering layer 113 or 124 is removed by a polishing step and a subsequent wet etching step (step of FIG. 69 or step of FIG. 78).
[0169] τ t =1.5×10 -5 exp(5.4×10 3 t N- )...Equation (2) In equation (2), t N- is N - is the thickness (m) of the drift layer 14, and is shown in FIG. N- This is the device parameter that corresponds to τ t N - This is the carrier lifetime (sec) in the drift layer 14.
[0170] Equation (2) is derived from the following viewpoint. That is, the on-voltage of IGBT and FWD is N - The dependence on the carrier lifetime of the drift layer 14 is almost eliminated from a certain value. If the relationship between the on-voltage and the carrier lifetime can be controlled, the effect of the carrier lifetime on the switching loss can also be controlled. In addition, the off-loss is also affected by the carrier lifetime, and if the carrier lifetime can be set so that the effect of the carrier lifetime on the on-voltage is eliminated, it is effective in reducing the off-loss and suppressing thermal runaway.
[0171] Based on this idea, the vertical axis shows the N of the on-state voltage for the assumed voltage class (600V to 6500V class). - The carrier lifetime value at which the drift layer 14 is independent of the carrier lifetime and N - The thickness of the drift layer (t N- ) is plotted as shown in Figure 81. Figure 81 shows the relationship between carrier lifetime and N - The thickness of the drift layer (t N- 81) and is a result of examining the device structures shown in FIGS. 2, 40, and 45. The relationship of the straight lines in FIG. 81 corresponds to the relationship of formula (2).
[0172] In other words, the “device characteristic stabilization region” in Figure 81 (τ t By controlling the lifetime during the wafer process to the range above (lifetime region above), the effect of N on the final device characteristics can be - It is believed that the adverse effects of the carrier lifetime in the drift layer 14 can be suppressed, and the inherent performance of the device structure and the variation in electrical characteristics between multiple chips present on a wafer can be minimized. - The carrier lifetime value of the drift layer is given by τ t The manufacturing technique of this embodiment is to manufacture the device structure shown in embodiments 1 to 3 so as to satisfy the lifetime relationships of the elements constituting the device structure while controlling them within the above range (i.e., the "device characteristic stabilization region" in Figure 81).
[0173] In order to realize the formula (2) and the "device characteristic stabilization region" of FIG. 81, the gettering layer (113 or 124) used in the manufacturing method of this embodiment is formed in the following procedure. First, in order to expose the Si surface on the back surface of the semiconductor substrate (wafer), only the back surface of the wafer is selectively etched (FIGS. 63 and 72). The etching technique used here uses a liquid of hydrofluoric acid or a mixed acid (for example, a mixture of hydrofluoric acid / nitric acid / acetic acid).
[0174] Next, high concentration of N + As a source for forming the layer (113a or 124a) and the high crystal defect density layer (113a or 124a), + A polysilicon layer (111 or 122) doped with atoms that form the layer (hereinafter referred to as "d-polysilicon layer") is formed by LPCVD (Figures 64 and 73). + The atoms that form the layer are, for example, phosphorus, arsenic, antimony, etc., which diffuse into Si. + Select atoms that can form a layer. The d-polysilicon layer (111 or 122) is 1×10 19 cm -3 The film is doped with a high concentration of impurities and has a thickness of 500 nm or more. The reason why it is necessary to dope the film with a high concentration of impurities is that the high concentration impurities diffuse into the Si surface on the back side of the wafer during the subsequent annealing and become N + This is because high density dislocations and lattice defects, including the high crystal defect density layer (113a or 124a), are introduced during the formation of the layer (113a or 124a), and the effect of capturing heavy metals and contaminant atoms is utilized. At this time, the d-polysilicon layer (111 or 122) is in direct contact with the Si surface exposed on the backside of the wafer.
[0175] After the deposition of d-polysilicon, thermal annealing is performed at 900°C to 1000°C in a nitrogen atmosphere, and then the temperature is lowered from 900°C to 1000°C at an arbitrary rate to 600°C to 700°C. If an annealing technique is performed at 500°C to 700°C in a nitrogen atmosphere at a lower temperature than the previous annealing temperature, impurities diffuse from the d-polysilicon into the Si on the back surface of the wafer where the d-polysilicon and Si are in direct contact, and N + While forming the layer (113a or 124a), crystal defects (i.e., the high crystal defect density layer (113a or 124a)) are generated secondarily. + The layer (113a or 124a), the high crystal defect density layer (113a or 124a), and the d-polysilicon layer (111 or 122) are in direct contact with the exposed Si surface, and the materials with different thermal expansion coefficients are then heated at the interface (N + Strain occurs in the surface layer of the layer (113a or 124a), and N + layer (113a or 124a), high crystal defect density layer (113a or 124a) and N + The strained layer present in the surface portion of the layer (113a or 124a) becomes the getter site (FIG. 65, FIG. 74). As a result, when the above-mentioned low-temperature thermal annealing technique is performed, heavy metals and contaminant atoms that are taken into the wafer during the wafer process diffuse and move in the crystal lattice to the getter site.
[0176] This technology reduces the N - By restoring the τ of the drift layer, we can obtain a sufficiently long carrier lifetime N such that the carrier lifetime does not affect the electrical characteristics of IGBTs and FWDs of various voltage classes, as shown in equation (2). - The drift layer is realized. In addition, the low-temperature thermal annealing time in a nitrogen atmosphere at 600°C to 700°C is - As described later, there is an appropriate annealing time that is based on the carrier lifetime of the drift layer and the electrical characteristics of IGBTs and FWDs that are affected by the carrier lifetime.
[0177] In addition to the method using the d-polysilicon layer (111 or 122), a similar effect can be obtained by forming a high crystal defect density layer (113a or 124a) on the back surface of the wafer using laser annealing technology (rapid heating / rapid cooling and local annealing technology using a laser with a wavelength of 500°C to 1000 nm). In that case, the power density of the laser annealing is set to 4 J / cm. 2 By using the above-mentioned laser annealing and the subsequent annealing techniques (thermal annealing at 900°C to 1000°C in a nitrogen atmosphere and annealing at 500°C to 700°C in a nitrogen atmosphere), a high crystal defect density layer (113a or 124a) is formed. By forming this high crystal defect density layer (113a or 124a), the effect of improving and stabilizing the carrier lifetime is obtained.
[0178] The vertical structure (35) provided on the back surface of the semiconductor substrate is formed after the aluminum wiring process and the passivation film formation process (Figures 69 and 78). On the top surface of the semiconductor substrate (the surface on which the vertical structure (35) is not formed), for example, a MOS transistor structure is formed in the case of an IGBT, and aluminum wiring and a passivation film are present. Therefore, the diffusion layers (N buffer layer 15, P collector layer 16, N + When forming the cathode layer 17 and the P cathode layer 18), in order to maintain the upper surface of the semiconductor substrate at a temperature lower than the melting point of the metal used for the aluminum wiring (melting point of aluminum: 660°C), an annealing technique (laser annealing) is used that uses a laser with a wavelength that does not transmit heat to the upper surface of the semiconductor substrate when there is a temperature gradient in the device depth direction, or an annealing technique at a low temperature below the melting point of the metal is used.
[0179] The first buffer layer (15-1) and the second buffer layer (15-2) constituting the vertical structure (35) are formed to have a thickness (t device The gettering layer (113 or 124) is removed at this time, but the N - N -A necessary diffusion layer is formed in the drift layer 14.
[0180] In the process of the fourth embodiment (Process A in FIG. 80), the order in which the first buffer layer (15-1) and the second buffer layer (15-2) are formed and the setting of the peak position of the acceleration energy when the second buffer layer (15-2) is introduced are important. That is, after the first buffer layer (15-1) is formed by ion implantation and annealing technology (first annealing step), the second buffer layer (15-2) is formed by ion implantation and annealing technology (second annealing step). The reason is that the annealing technique used in forming the first buffer layer (15-1) is a higher temperature annealing technique than the annealing technique used in forming the second buffer layer (15-2), and performing the annealing technique used in forming the first buffer layer (15-1) later has an adverse effect on the impurity profile after activation of the second buffer layer (15-2) and on the type of lattice defects introduced in the second buffer layer (15-2) to form the second buffer layer (15-2), resulting in an adverse effect on carriers (electrons or holes) in the device on state.
[0181] Regarding the second buffer layer 15-2, ions are introduced into the Si after the activation annealing of the first buffer layer (15-1), and the second annealing process is performed on the P collector layer 16 and the N + This can be performed after forming the cathode layer 17 and the P cathode layer 18 using ion implantation and annealing techniques, or after forming metallization (cathode electrode 19 or collector electrode 20), to form the desired second buffer layer (15-2).
[0182] Here, the first buffer layer 15-1 and the second buffer layer 15-2 are formed so as to satisfy the following relationship. The peak position of the second buffer layer 15-2 is the junction (X j,nb1 ) above (Joint X j,nb2This makes it possible to form the second buffer layer 15-2 with high precision without causing interference between the first buffer layer 15-1 and the second buffer layer 15-2.
[0183] As ion species for forming the first buffer layer 15-1 and the second buffer layer 15-2, phosphorus is used for the first buffer layer 15-1, and selenium, sulfur, phosphorus, and protons (H + ) or helium. + ), when helium is used, a diffusion layer formation process technology is used to form an N layer by the donor phenomenon caused by the annealing conditions described below. + ), helium can be introduced into silicon by irradiation technology using a cyclotron other than ion implantation. + ) into Si, (a) Vacancies (V) generated during the introduction diffuse and react with impurities (hydrogen atoms (H), oxygen atoms (O) and carbon atoms (C)) or form complex defects (VOH, V2H, VO, CiOS (G-centre) and Cio (C-centre)) due to substitution reactions between carbon atoms and lattice defects. (b) Lattice defects that occur during the introduction diffuse, self-aggregate, and react with oxygen atoms during annealing to form complex defects (VO, VO). (c) A lattice defect pair (Trap B) is formed when a lattice defect agglomeration (Trap A) occurs during the introduction and diffuses and reaggregates upon annealing. exists.
[0184] Among these, the complex defects formed by the reaction of vacancies with impurities contain hydrogen, which acts as an electron source (donor). The donor concentration increases due to the increase in the complex defect density caused by annealing, and the donor concentration is increased by a mechanism that promotes the thermal donor phenomenon caused by the ion implantation / irradiation process. As a result, N - A donor-type layer having a higher impurity concentration than the drift layer 14 is formed, and contributes to the operation of the device as the second buffer layer 15-2. This technology utilizes complex defects formed in the second buffer layer 15-2 to improve device performance.
[0185] On the other hand, since the complex defects formed in the second buffer layer 15-2 include defects that become lifetime killers that shorten the carrier lifetime, the impurity dose when forming the second buffer layer 15-2 and the process flow when forming the second buffer layer 15-2 (as described above, the second buffer layer 15-2 is formed by ion implantation and annealing techniques after the formation of the first buffer layer 15-1 (Process A in FIG. 80)) and the annealing conditions for making the second buffer layer 15-2 a donor) are important. Also, the first buffer layer 15-1 and the second buffer layer 15-2 are activated using methods that use different annealing techniques.
[0186] Figure 82 shows a device performance table of a 1200V class new structure RFC diode prototyped using two types of process flows (Process A and Process B) for forming the second buffer layer 15-2 shown in Figure 80. The ratio of Trap B in the lattice defects in the second buffer layer 15-2 differs due to the difference in process flows, and in particular in Process B, in which the ion implantation step for forming the second buffer layer 15-2 occurs before the ion implantation step for forming the first buffer layer 15-1, the ratio of Trap B becomes 55% or more, resulting in deterioration of the device performance of the RFC diode.
[0187] FIG. 83 shows the results of examining the annealing temperature of the second annealing step for forming the second buffer layer 15-2 required for controlling the ratio of Trap B in the lattice defects in the second buffer layer 15-2 to the target 15% to 55% in Process A. Here, the annealing time is 120 minutes. As can be seen from FIG. 83, even when Process A is used, the setting of the annealing temperature for forming the second buffer layer 15-2 is important, and in order to control the ratio of Trap B to 15% to 55%, it is necessary to control the annealing temperature to 370° C. or more and 425° C. or less. By doing so, the lifetime shown in formula (2) is satisfied.
[0188] Fig. 84 is a diagram showing the relationship between the PL intensity of Trap A and Trap B and the annealing time. Fig. 84 shows the relationship between the PL intensity when Trap A and Trap B are analyzed by the PL method and the annealing time of the annealing step after controlling the ratio of Trap B in the lattice defects in the second buffer layer 15-2 to 15% to 55% in Process A. Note that points A and B in Fig. 84 are data at points A and B shown in Fig. 8.
[0189] Here, the annealing temperature is 400°C. Even if the ratio of Trap B is controlled to 15% to 55%, it is better to reduce the defect density of Trap A and Trap B in order to satisfy the lifetime shown in formula (2) from the viewpoint of device performance of IGBT.FWD. From FIG. 84, the annealing time dependency of the PL intensity expressing the defect density of Trap A and Trap B becomes insensitive when the annealing time is 90 minutes or more. In other words, in order to reduce the defect density of Trap A and Trap B while controlling the ratio of Trap B in the lattice defects in the second buffer layer 15-2 constituting the N buffer layer 15 in ProcssA to 15% to 55%, the annealing time needs to be 90 minutes or more.
[0190] It should be noted that the embodiments can be freely combined, and each embodiment can be modified or omitted as appropriate. [Explanation of symbols]
[0191] R1 active cell region, R2 middle region, R3 edge termination region, R11 surface gate wiring section, R12 gate pad section, R21 PiN diode region, R22 PNP transistor region, 4 emitter electrode, 5 anode electrode, 6 interlayer insulating film, 7 N + Emitter layer, 8 P + layer, 9 P base layer, 10 P anode layer, 11 N layer, 12 gate insulating film, 13 gate electrode, 13a surface gate electrode, 14 N -drift layer, 15 N buffer layer, 15-1 first buffer layer, 15-2 second buffer layer, 16 P collector layer, 17 N + cathode layer, 18 P cathode layer, 19 cathode electrode, 20 collector electrode, 22 guard ring, 23 field limiting ring, 24 channel stopper layer, 25 insulating film, 26 interlayer insulating film, 27 FLR electrode, 28 channel stopper electrode, 29 passivation film, 30 passivation film, 31 surface gate electrode, 35 vertical structure, 101 SiO2 film, 102 oxide film, 103 trench, 104 polysilicon layer, 105 oxide film, 106 polysilicon layer, 107, 108 oxide film, 109, 110 TEOS film, 111, 112 polysilicon layer, 113 gettering layer, 113a N + layer, 113b high crystal defect density layer, 114 exposed trench portion, 115 silicide layer, 116 barrier metal layer, 117 metal wiring layer, 121 oxide film, 122, 123 polysilicon layer, 124 gettering layer, 124a N + layer, 124b high crystal defect density layer.
Claims
1. a semiconductor substrate having a first main surface and a second main surface; a drift layer of a first conductivity type formed on the semiconductor substrate; an impurity diffusion layer of a second conductivity type formed on the first main surface side of the drift layer; a buffer layer of a first conductivity type formed on the second main surface side of the drift layer and having a peak impurity concentration higher than that of the drift layer; Equipped with The buffer layer is a first buffer layer in which no trap levels derived from lattice defects are detected by a photoluminescence (PL) method; a second buffer layer provided between the first buffer layer and the drift layer, the second buffer layer being a layer in which two types of trap levels derived from lattice defects are detected by the photoluminescence method; Including, The two types of trap levels derived from lattice defects detected by the photoluminescence method in the second buffer layer are Trap A (photon energy: 1.0182 eV) and Trap B (photon energy: 1.0398 eV). A semiconductor device comprising:
2. The PL intensity of Trap B is S PLB , the PL intensity of Trap A is S PLA When Trap B ratio = {S PLB / (S PLA +S PLB ) x 100 The ratio of Trap B in the second buffer layer represented by the formula (1) is 15% or more and 55% or less. The semiconductor device according to claim 1 .
3. an insulated gate transistor structure is formed on the first main surface side of the impurity diffusion layer; 3. The semiconductor device according to claim 1 or 2.
4. a collector layer of a second conductivity type is formed on the second main surface side of the buffer layer; The semiconductor device according to claim 3 .
5. An active cell area, an edge termination area provided outside the active cell area, and an intermediate area between the active cell area and the edge termination area are defined on the semiconductor substrate; the buffer layer and the collector layer are formed across the active cell region, the intermediate region and the edge termination region; The semiconductor device according to claim 4.
6. An active cell area, an edge termination area provided outside the active cell area, and an intermediate area between the active cell area and the edge termination area are defined on the semiconductor substrate; the buffer layer is formed across the active cell region, the intermediate region and the edge termination region; the collector layer is formed only in the active cell region; The semiconductor device according to claim 4.
7. the impurity diffusion layer functions as an anode of a diode; a cathode layer of a first conductivity type functioning as a cathode of the diode is formed on the second principal surface side of the buffer layer; 3. The semiconductor device according to claim 1 or 2.
8. An active cell area, an edge termination area provided outside the active cell area, and an intermediate area between the active cell area and the edge termination area are defined on the semiconductor substrate; the buffer layer and the cathode layer are formed across the active cell region, the intermediate region and the edge termination region; The semiconductor device according to claim 7.
9. the impurity diffusion layer functions as an anode of a diode; a first cathode layer of a first conductivity type and a second cathode layer of a second conductivity type functioning as a cathode of the diode are formed on the second main surface side of the buffer layer; 3. The semiconductor device according to claim 1 or 2.
10. An active cell region, an edge termination region provided outside the active cell region, and an intermediate region between the active cell region and the edge termination region are defined on the semiconductor substrate; the buffer layer and the first cathode layer are formed across the active cell region, the middle region and the edge termination region; the second cathode layer is formed only in the active cell region; The semiconductor device according to claim 9.
11. The distribution of the ratio of Trap B in the depth direction of the second buffer layer is constant within a range of 15% to 55%. The semiconductor device according to claim 2 .
12. a carrier lifetime of the second buffer layer is shorter than a carrier lifetime of each of the first buffer layer, the drift layer, and the collector layer; The semiconductor device according to claim 4 .
13. a carrier lifetime of the second buffer layer is shorter than a carrier lifetime of each of the first buffer layer, the drift layer, and the cathode layer; 9. The semiconductor device according to claim 7 or 8.
14. a carrier lifetime of the second buffer layer is shorter than a carrier lifetime of each of the first buffer layer, the drift layer, the first cathode layer, and the second cathode layer; The semiconductor device according to claim 9 or 10.
15. (a) forming a predetermined device structure on a semiconductor substrate having a first main surface and a second main surface and a drift layer formed thereon; (b) The thickness of the drift layer is t N- When the carrier lifetime in the drift layer is τ t = 1.5 x 10 -5 exp(5.4 x 10 3 t N- ) or more after the step (a), (c) after the step (b), forming an electrode and a protective film for protecting the electrode on the first main surface of the semiconductor substrate; (d) forming a buffer layer on the second main surface side of the drift layer, the buffer layer including a first buffer layer and a second buffer layer provided between the first buffer layer and the drift layer, after the step (c); Equipped with The step (b) (b-1) exposing the second main surface of the semiconductor substrate by etching; (b-2) forming polysilicon doped with a first conductivity type impurity on the second main surface of the semiconductor substrate; (b-3) after the step (b-2), performing a first annealing in a nitrogen atmosphere at a first temperature, lowering the temperature to a second temperature lower than the first temperature, and performing a second annealing in a nitrogen atmosphere at the second temperature, thereby forming a first conductivity type high concentration layer and a high crystal defect density layer on the second main surface of the semiconductor substrate; Including, The step (d) (d-1) removing the high concentration layer and the high crystal defect density layer from the second main surface of the semiconductor substrate by polishing or etching, and adjusting a thickness of the semiconductor substrate; (d-2) after the step (d-1), a first impurity introduction step of introducing an impurity for the first buffer layer into the semiconductor substrate from the second main surface; (d-3) a first annealing step of annealing the semiconductor substrate to form the first buffer layer after the step (d-2); (d-4) after the step (d-3), a second impurity introduction step of introducing an impurity for the second buffer layer into the semiconductor substrate from the second main surface; (d-5) a second annealing step of annealing the semiconductor substrate after the step (d-4) to form the second buffer layer and generate trap levels Trap A (photon energy: 1.0182 eV) and Trap B (photon energy: 1.0398 eV) derived from lattice defects detected by a photoluminescence (PL) method in the second buffer layer; Including, A method for manufacturing a semiconductor device.
16. The temperature of the second annealing step is 370° C. or more and 425° C. or less. The method for manufacturing a semiconductor device according to claim 15.
17. The duration of the second annealing step is 90 minutes or more. The method for manufacturing a semiconductor device according to claim 15.
Citation Information
Patent Citations
Semiconductor device and manufacturing method therefor
JP2009176892A
Semiconductor device, power conversion device, and manufacturing method of semiconductor device
JP2018107303A
Semiconductor device and manufacturing method thereof
JP2020027921A
Semiconductor device and method for producing semiconductor device
WO2013100155A1
Semiconductor device and method for manufacturing semiconductor device
WO2014054121A1
