A channel defect detection method and a wafer detection method

Abstract

The present application belongs to the field of semiconductor technology and provides a channel defect detection method and a wafer detection method. The channel defect detection method applies a horizontal static magnetic field to the first channel layer to adjust the spin direction of the electrons in the defect energy level so that its spin direction is downward. After applying a radio frequency magnetic field, the spin direction of the electrons in the defect energy level is reversed and the direction is upward, so that the pairing behavior of the electrons from the conduction band that were originally paired with the defect energy level will be prohibited. At this time, the originally paired electrons from the conduction band will fall into the valence band and recombine with the holes in the valence band to form a first leakage current. Then, by detecting the channel layer of the power device after each epitaxial growth, the interface defect density of the channel layer during the epitaxial growth process can be detected early, thereby avoiding process waste and achieving accurate detection results.

Description

Technical Field

[0001] The present application belongs to the field of semiconductor technology, and in particular relates to a channel defect detection method and a wafer detection method. Background Art

[0002] Reliability assessment of power device performance is an important part of integrated circuit process development. However, reliability assessment of the gate dielectric layer (Gate Oxide) in power devices is a critical project in the front-end, primarily used to evaluate the performance of front-end dielectric materials. Many processes in the integrated circuit manufacturing process may affect the upper and lower interfaces of the gate dielectric layer, thereby affecting the TDDB (Time Dependent Dielectric Breakdown) of the gate dielectric layer, exhibiting polarity dependence. Therefore, it is extremely important to effectively characterize the defect density at the upper and lower interfaces of the gate dielectric layer.

[0003] After extensive research and development, the parallel conductance method for characterizing interface defect density is currently primarily used in Si-based MOS devices. However, this method often has strict application limits and requirements, such as requiring the device to have a very low first leakage current. Since AlGaN / GaN HEMTs exhibit a high first leakage current, simply applying this method to characterize their interface defect density will clearly yield inaccurate results. Furthermore, significant differences exist between Si and GaN materials in terms of conduction band effective state density, bandgap width, and interface defect density, raising doubts about the reliability of the results obtained using the parallel conductance method.

[0004] It can be seen that the existing channel defect detection method for power devices has the problems of limited application scope and inaccurate detection results. Summary of the Invention

[0005] In order to solve the above technical problems, the embodiments of the present application provide a channel defect detection method and a wafer detection method, which aim to solve the problems of small application scope and inaccurate detection results of existing channel defect detection methods for power devices.

[0006] A first aspect of an embodiment of the present application provides a channel defect detection method, comprising:

[0007] Step S10: forming a first channel layer, and sequentially forming a first source, a first drain, and a first gate on the first channel layer; wherein the first drain is connected to a power supply, the first source is grounded, and the first gate is left floating;

[0008] Step S20: applying a horizontal static magnetic field to the first channel layer; wherein the horizontal static magnetic field is used to adjust the spin direction of electrons located in the defect energy level;

[0009] Step S30: applying a radio frequency magnetic field above the first gate; wherein the radio frequency magnetic field is used to flip the spin direction of electrons in the defect energy level, so that electrons from the conduction band move into holes in the valence band, thereby forming a first leakage current;

[0010] Step S40: determining the interface defect density of the first channel layer according to the first leakage current, and removing the first source, the first gate, and the first drain;

[0011] Step S50: Repeat step S10, step S20, step S30 and step S40 until the thickness of the channel layer reaches a preset threshold thickness; wherein the channel layer is composed of multiple layers of the first channel layer.

[0012] In one embodiment, the step of increasing the thickness of the channel layer until the thickness reaches a preset threshold thickness includes:

[0013] The thickness of the channel layer is measured, and the thickness of the channel layer is compared with the preset threshold thickness. When the thickness of the channel layer is less than the preset threshold thickness, the steps S10, S20, S30, and S40 are repeatedly performed.

[0014] In one embodiment, sequentially forming a first source, a first drain, and a first gate on the first channel layer includes:

[0015] A first source electrode, a first drain electrode and a first gate electrode are sequentially formed on the first channel layer by photolithography.

[0016] In one embodiment, removing the first source, the first gate, and the first drain includes:

[0017] The first source electrode, the first gate electrode and the first drain electrode are washed away using a mixture of phosphoric acid, acetic acid and ammonia water.

[0018] In one embodiment, after determining the interface defect density of the first channel layer according to the first leakage current, the method further includes:

[0019] When the interface defect density of the first channel layer is greater than a preset interface defect density, a layer of the first channel layer is removed, and the steps S10, S20, S30 and S40 are repeatedly performed.

[0020] In one embodiment, before applying a radio frequency magnetic field above the first gate, the method further comprises:

[0021] A radio frequency coil is disposed above the first grid, and generates the radio frequency magnetic field when energized.

[0022] In one embodiment, the center of the radio frequency coil is located above the first grid.

[0023] In one embodiment, applying a radio frequency magnetic field above the first gate further comprises:

[0024] A magnetic core unit is provided at the center of the radio frequency coil, and the magnetic core unit is used to increase the intensity of the radio frequency magnetic field.

[0025] In one embodiment, applying a horizontal static magnetic field to the first channel layer includes:

[0026] A first magnet and a second magnet are respectively arranged on both sides of the first source and the first drain, and the first magnet and the second magnet are used to generate the horizontal static magnetic field.

[0027] A second aspect of an embodiment of the present application provides a method for detecting a wafer, wherein the wafer includes multiple power devices and at least one power device to be tested, and the at least one power device to be tested is prepared using a channel defect detection method as described in any one of the above items.

[0028] The beneficial effects of the embodiments of the present application compared with the prior art are: by applying a horizontal static magnetic field to the first channel layer, the spin direction of the electrons in the defect energy level is changed, and its spin direction is downward. After applying a radio frequency magnetic field, the spin direction of the electrons in the defect energy level is flipped, and the direction is upward, so that the pairing behavior of the electrons from the conduction band that were originally paired with the defect energy level will be prohibited. At this time, the originally paired electrons from the conduction band will fall into the valence band and recombine with the holes in the valence band to form a first leakage current. Then, the interface defect density is detected by detecting the magnitude of the first leakage current. The present application detects the interface defect density of the channel layer during the epitaxial process early by detecting each epitaxial growth of the channel layer of the power device, thereby avoiding process waste and achieving accurate detection results. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a schematic diagram of the steps of a channel defect detection method provided by an embodiment of the present application;

[0030] Figure 2 This is a schematic diagram of the specific structure of a channel defect detection method provided by an embodiment of the present application. Figure 1 ;

[0031] Figure 3 This is a schematic diagram of the specific structure of a channel defect detection method provided by an embodiment of the present application. Figure 2 ;

[0032] Figure 4 This is a schematic diagram of the structure of a channel defect detection method provided by an embodiment of the present application. Figure 1 ;

[0033] Figure 5 This is a schematic diagram of the structure of a channel defect detection method provided by an embodiment of the present application. Figure 2 ;

[0034] Figure 6 1 is a schematic top view of a radio frequency coil provided in one embodiment of the present application;

[0035] Figure 7 It is a structural diagram of a wafer detection method provided by an embodiment of the present application. DETAILED DESCRIPTION

[0036] In order to make the technical problems, technical solutions and beneficial effects to be solved by this application more clearly understood, this application is further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0037] It should be noted that when an element is referred to as being “fixed on” or “disposed on” another element, it may be directly on the other element or indirectly on the other element. When an element is referred to as being “connected to” another element, it may be directly connected to the other element or indirectly connected to the other element.

[0038] It should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc., indicating the orientation or position relationship, are based on the orientation or position relationship shown in the accompanying drawings, and are only for the convenience of describing this application and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on this application.

[0039] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of the technical features being referred to. Thus, a feature specified as "first" or "second" may explicitly or implicitly include one or more of the features. Throughout the description of this application, "plurality" means one or more, unless otherwise specifically defined.

[0040] References to "one embodiment," "some embodiments," or "an embodiment" in this specification mean that a particular feature, structure, or characteristic described in conjunction with that embodiment is included in one or more embodiments of the present application. Thus, phrases such as "in one embodiment," "in some embodiments," "in some other embodiments," "in some other embodiments," "in a specific embodiment," and "in a specific application" appearing in various places in this specification do not necessarily refer to the same embodiment, but rather mean "one or more but not all embodiments," unless otherwise specifically emphasized. Furthermore, in one or more embodiments, particular features, structures, or characteristics may be combined in any suitable manner.

[0041] Reliability assessment of power device performance is an important part of integrated circuit process development. However, reliability assessment of the gate dielectric layer (Gate Oxide) in power devices is a critical project in the front-end, primarily used to evaluate the performance of front-end dielectric materials. Many processes in the integrated circuit manufacturing process may affect the upper and lower interfaces of the gate dielectric layer, thereby affecting the TDDB (Time Dependent Dielectric Breakdown) of the gate dielectric layer, exhibiting polarity dependence. Therefore, it is extremely important to effectively characterize the defect density at the upper and lower interfaces of the gate dielectric layer.

[0042] After extensive research and development, the parallel conductance method for characterizing interface defect density is currently primarily used in Si-based MOS devices. However, this method itself often has relatively strict scopes and conditions, such as requiring the device to have a very low first leakage current. Since AlGaN / GaN HEMTs exhibit a relatively high first leakage current, simply applying the above method to characterize their interface defect density will clearly yield inaccurate results. Furthermore, SiC power devices have a high interface state density, which in turn leads to significant leakage current, making the detection of interface defect density particularly important. Furthermore, Si and GaN materials differ significantly in terms of conduction band effective state density, bandgap width, and interface defect density, raising doubts about the reliability of the results obtained using the parallel conductance method.

[0043] It can be seen that the existing channel defect detection method for power devices has the problems of limited application scope and inaccurate detection results.

[0044] In order to solve the above technical problems, refer to Figure 1 As shown, an embodiment of the present application provides a channel defect detection method, which includes steps S10 to S50.

[0045] In this embodiment, reference Figure 4 As shown, step S10 includes: forming a first channel layer 20, and sequentially forming a first source 50, a first drain 60 and a first gate 30 on the first channel layer 20; wherein the first drain 60 is connected to a power supply, the first source 50 is grounded, and the first gate 30 is suspended.

[0046] In this embodiment, taking the GaN power device as an example, the first channel layer 20 is a gallium nitride layer. The first channel layer 20 is formed on the semiconductor substrate 10 by epitaxial growth technology. Interface defects may exist when the first channel layer 20 is epitaxially grown. The gallium nitride layer is the electron channel of the GaN power device and is very important to the performance of the device. Interface defect detection can ensure the stability of the performance of the power device. At the same time, because the epitaxial process is very expensive, timely discovery of interface defects in the epitaxial process can be timely remedied to reduce costs.

[0047] In this embodiment, when step S10 is executed and it is confirmed that the first channel layer 20 is formed for the first time, the first channel layer 20 is formed on the semiconductor substrate 10, for example, Figure 4 As shown, a first source 50, a first drain 60 and a first gate 30 are then formed on the first channel layer 20. For example, the first source 50 and the first drain 60 are disposed on opposite sides of the first channel layer 20, and the first gate 30 is disposed at the center of the first channel layer 20.

[0048] In one embodiment, when step S10 is executed, it is found that the first channel layer 20 has been formed on the semiconductor substrate 10. Then, when step S10 is executed this time, the first channel layer 20 is directly formed again on the original first channel layer 20. At this time, two layers of first channel layers 20 are formed on the semiconductor substrate 10. Figure 5 As shown, the first source 50 , the first drain 60 and the first gate 30 are formed on the first channel layer 20 formed last, and the connection method is the same as above, which will not be repeated again.

[0049] In this embodiment, reference Figure 2 As shown, step S20 includes: applying a horizontal static magnetic field B1 to the first channel layer 20; wherein the horizontal static magnetic field B1 is used to adjust the spin direction of electrons located in the defect energy level.

[0050] In this embodiment, a horizontal static magnetic field B1 is applied to the first channel layer 20. It can be understood that the horizontal static magnetic field B1 is applied to the power device formed by the first channel layer. Specifically, the first channel layer 20 is generally obtained by epitaxy. However, interface defects may exist in the first channel layer 20 during the epitaxy process. The first channel layer 20 serves as a bridge for communication between the first source 50 and the first drain 60. Therefore, the interface defect density directly affects the performance of the power device. Therefore, it can be seen that the detection of interface defects in the first channel layer 20 of the power device is very important.

[0051] In this embodiment, continue to refer to Figure 2 As shown, when a horizontal static magnetic field B1 is applied to the first channel layer 20, the energy of the electrons in the defect energy level in the first channel layer 20 is consumed by the horizontal static magnetic field B1. After being consumed by the horizontal static magnetic field B1, the energy of the electrons in the defect energy level is reduced. At this time, the electrons in the defect energy level spin downward in the external horizontal static magnetic field B1 and are in the lowest energy state. For example, Figure 2 The dashed arrow line segment represents the trajectory of electrons in the conduction band, and the solid arrow line segment represents the spin direction of the corresponding electrons. It can be understood that at this time, the electrons in the defect energy level are vertically downward (for example, Figure 2 The arrow of the electron in the defect energy level is downward, and the energy level line in the figure means that the energy of electrons located on the same energy level line is the same), that is, perpendicular to the horizontal plane of the first channel layer 20. When the spin direction of the electron in the defect energy level is downward, the electron in the conduction band will pair with the electron in the defect energy level. Figure 2 As shown, electrons in the conduction band move and pair with electrons in defect levels.

[0052] In a specific embodiment, when a horizontal static magnetic field B1 is applied to the first channel layer 20 , the direction of the horizontal static magnetic field B1 is parallel to the upper surface of the first channel layer 20 , that is, the direction of the horizontal static magnetic field B1 is parallel to the line connecting the source and the drain.

[0053] In this embodiment, reference Figure 3 As shown, step S30 includes: applying a radio frequency magnetic field B2 above the first gate 30; wherein the radio frequency magnetic field B2 is used to flip the spin direction of electrons whose energy is located in the defect energy level, so that electrons from the conduction band move into holes in the valence band, thereby forming a first leakage current.

[0054] In this embodiment, reference Figure 3As shown, a radio frequency magnetic field B2 is applied above the first gate 30. Specifically, the direction of the radio frequency magnetic field B2 is perpendicular to the first gate 30, that is, the direction of the radio frequency magnetic field B2 is perpendicular to the first channel layer 20, and the direction of the radio frequency magnetic field B2 is perpendicular to the horizontal static magnetic field B1. When the radio frequency magnetic field B2 is applied above the first gate 30, the radio frequency magnetic field B2 can cause the spin direction of the electrons in the defect energy level to flip. For example, referring to Figure 2 、 Figure 3 As shown, after applying the horizontal static magnetic field B1, the spin direction of the electrons in the defect energy level is downward. After applying the radio frequency magnetic field B2, the spin direction of the electrons in the defect energy level is reversed, and its spin direction is upward. Under the constraint of the Pauli exclusion principle, the pairing behavior of the electrons from the conduction band that were originally paired with the defect energy level will be prohibited. At this time, the originally paired electrons from the conduction band will fall into the valence band and recombine with the holes in the valence band to form a recombination current (or the first leakage current).

[0055] In one embodiment, the Pauli exclusion principle states that electrons with identical motion states cannot reside in an atom.

[0056] In this embodiment, step S40 includes: determining the interface defect density of the first channel layer 20 according to the first leakage current, and removing the first source 50 , the first gate 30 , and the first drain 60 .

[0057] In this embodiment, when originally paired electrons from the conduction band fall (or move) into the valence band, forming a first leakage current, the density of interface defects in the first channel layer 20 can be detected by measuring the magnitude of the first leakage current. Specifically, the interface defect density is directly proportional to the magnitude of the first leakage current. That is, the greater the interface defect density, the more electrons fall into the valence band, and the greater the first leakage current formed by the electrons from the conduction band that fall into the valence band and combine with holes in the valence band. The interface defect density can be detected by measuring the magnitude of the first leakage current, thereby achieving accurate detection of interface defects in the first channel layer 20.

[0058] In one embodiment, the first leakage current of the first channel layer 20 may be detected by a current sensor.

[0059] In one embodiment, taking a SiC power device as an example, when manufacturing a SiC power device, interface defects may exist when the first channel layer 20 (i.e., the silicon carbide layer and the silicon dioxide layer) is epitaxially formed. The interface between the silicon carbide layer and the silicon dioxide layer is the electronic channel of the SiC power device and is very important to the performance of the device. Interface defect detection can play a role in ensuring the stability of the performance of the power device.

[0060] In one embodiment, taking a GaN power device as an example, the first channel layer 20 is a gallium nitride layer. When the first channel layer 20 (i.e., the gallium nitride layer) is formed by epitaxial growth, interface defects may exist. The gallium nitride layer is also the electron channel of the GaN power device and is very important to the performance of the device. Interface defect detection can ensure the stability of the performance of the power device.

[0061] In this embodiment, step S50 includes: repeatedly executing step S10 , step S20 , step S30 and step S40 until the thickness of the channel layer reaches a preset threshold thickness; wherein the channel layer is composed of multiple layers of the first channel layer 20 .

[0062] In this embodiment, it is understood that the thickness of the channel layer is detected after each execution of step S40. When the thickness of the channel layer reaches the preset threshold thickness, it indicates that the preparation of the channel layer of the power device is completed and the next preparation process can be performed. When the thickness of the channel layer does not reach the preset threshold thickness, step S10, step S20, step S30 and step S40 are performed again. In other words, it is understood that an interface defect detection is performed each time the first channel layer 20 is formed, so that the size of the interface defect can be discovered in a timely manner. When the interface defect density is unqualified, the subsequent process is stopped in time and corresponding remedial measures are taken, for example, the first channel layer 20 with a larger defect density is removed and re-epitaxially formed, etc., to reduce losses.

[0063] In one embodiment, when performing step S10 to form the first channel layer 20, if the first channel layer 20 has already been formed on the semiconductor substrate 10, epitaxial growth can be continued on the first channel layer 20 formed last time, referring to Figure 5 When the semiconductor substrate 10 does not have the first channel layer 20, the first channel layer 20 is initially formed on the semiconductor substrate 10. Figure 4 shown.

[0064] In one embodiment, until the thickness of the channel layer reaches a preset threshold thickness, the method includes: measuring the thickness of the channel layer, comparing the thickness of the channel layer with the preset threshold thickness, and when the thickness of the channel layer is less than the preset threshold thickness, repeating step S10, step S20, step S30 and step S40.

[0065] In one embodiment, each time step S40 is executed, the thickness of the channel of the power device is measured, and the thickness of the channel layer is compared with a preset threshold thickness. When the thickness of the channel layer is less than the preset threshold thickness, steps S10, S20, S30, and S40 are repeated. When the thickness of the channel layer is greater than or equal to the preset threshold thickness, the above steps are no longer executed.

[0066] In a specific embodiment, the preset threshold thickness of the channel layer to be formed in a power device is 60 μm, and the thickness of the first channel layer 20 formed on the semiconductor substrate 10 when step S10 is performed for the first time can be 20 μm. Figure 4 As shown, step S10, step S20, step S30 and step S40 are then performed accordingly. When step S50 is performed, since step S10 is only performed once, the channel layer thickness of the power device is 20 μm less than 60 μm, and then step S10, step S20, step S30 and step S40 are repeated. It should be noted that when step S10 is performed for the second time, the first channel layer 20 is formed again on the original first channel layer 20, and the thickness is still 20 μm. Figure 5 As shown, step S10, step S20, step S30 and step S40 are then performed accordingly. When step S50 is performed, since step S10 is only performed twice, the channel layer thickness of the power device is 40 μm, which is less than 60 μm, and step S10, step S20, step S30 and step S40 are then repeated. After step S10 is performed for the third time, the first channel layer 20 formed for the third time is formed on the second channel layer. At this time, the thickness of the channel layer of the power device is 60 μm. After performing the corresponding steps S10, step S20, step S30 and step S40, it indicates that the channel layer of the power device has been prepared, and the above steps are no longer performed, and the preparation of other modules of the power device is started.

[0067] In one embodiment, reference Figure 4 、 Figure 5 As shown, the first source 50 , the first drain 60 and the first gate 30 are sequentially formed on the first channel layer 20 , including: sequentially forming the first source 50 , the first drain 60 and the first gate 30 on the first channel layer 20 by photolithography.

[0068] Specifically, a first source 50 groove, a first gate 30 groove and a first drain 60 groove are formed at preset positions on the first channel layer 20 by photolithography. The first source 50 groove and the first drain 60 groove are located at both ends of the upper surface of the first channel layer 20, and the first source 50 groove is located in the middle position of the upper surface of the first channel layer 20. Then, corresponding metal materials are deposited on the first source 50 groove, the first gate 30 groove and the first drain 60 groove to form the first source 50, the first drain 60 and the first gate 30.

[0069] In one embodiment, removing the first source electrode 50 , the first gate electrode 30 , and the first drain electrode 60 includes washing the first source electrode 50 , the first gate electrode 30 , and the first drain electrode 60 with a mixture of phosphoric acid, acetic acid, and ammonia.

[0070] In this embodiment, each time step S40 is performed, the first source electrode 50, the first gate electrode 30, and the first drain electrode 60 need to be washed with a mixture of phosphoric acid, acetic acid, and ammonia water to prevent contamination of the first channel layer 20 during the next epitaxial growth of the first channel layer 20. Specifically, washing the electrodes with a mixture of phosphoric acid, acetic acid, and ammonia water can prevent contamination of the first source electrode 50, the first gate electrode 30, and the first drain electrode 60 in subsequent processes.

[0071] In one embodiment, after determining the interface defect density of the first channel layer 20 according to the first leakage current, the method further includes: when the interface defect density of the first channel layer 20 is greater than a preset interface defect density, removing a layer of the first channel layer 20, and repeating steps S10, S20, S30, and S40.

[0072] In this embodiment, when it is detected that the interface defect density of the first channel layer 20 is relatively high, for example, the interface defect density of the first channel layer 20 is greater than a preset interface defect density, the first channel layer 20 formed in the most recent execution of step S10 is removed, and steps S10, S20, S30, and S40 are repeated. That is, the first channel layer 20 is formed again and the corresponding interface defect detection is performed until the thickness of the channel layer reaches a preset threshold thickness; wherein the channel layer is composed of multiple layers of the first channel layer 20.

[0073] In one embodiment, reference Figure 4 、 Figure 5 As shown, before applying the radio frequency magnetic field B2 above the first grid 30, the method further includes: disposing a radio frequency coil 70 above the first grid 30. The radio frequency coil 70 may be in contact with the first grid 30 or may be suspended above the first grid 30. The radio frequency coil 70 generates the radio frequency magnetic field B2 when powered on.

[0074] In this embodiment, reference Figure 6 As shown, Figure 6 FIG3 is a top view of the RF coil 70, which is disposed above the first grid 30. Specifically, after forming the first grid 30, the RF coil 70 can be fabricated on the plane of the first grid 30. A first end M1 and a second end M2 of the RF coil 70 are then reserved. The first and second ends of the RF coil 70 are used to connect to an RF power source, so that when the first and second ends M1, M2 of the RF coil 70 are connected to the RF power source and powered on, an RF magnetic field B2 is generated.

[0075] In a specific application, the first end M1 and the second end M2 of the RF coil 70 can be reserved by an air bridge so that the first end M1 and the second end M2 of the RF coil 70 can be connected to an RF power supply. The first end M1 and the second end M2 of the RF coil 70 are reserved by the air bridge because the air bridge has the advantages of low parasitic capacitance, easy preparation, and high reliability.

[0076] In one specific application, the air bridge method for preserving the first end M1 and second end M2 of the RF coil 70 is to use a composite adhesive structure composed of different photoresists and bake it to easily create an arched sacrificial layer approximately 60% thicker than the RF coil 70. This sacrificial layer provides strong protection for the metal underneath the bridge. Subsequently, the air bridge is thickened by electroplating using a non-toxic electroplating solution. This thickened air bridge has lower parasitic capacitance and higher reliability.

[0077] In one embodiment, reference Figure 4 、 Figure 5 As shown, the center of the RF coil 70 is located above the first grid 30 .

[0078] In this embodiment, the primary function of the RF coil 70 is to generate an RF magnetic field B2, which flips the spin direction of electrons in the defect energy level. This causes electrons from the conduction band that were originally paired with electrons in the defect energy level to stop pairing, causing electrons from the conduction band to fall into the valence band and combine with holes in the valence band, forming a first leakage current. The magnitude of this first leakage current is then used to measure the interface defect density of the first channel layer 20. Because defects typically exist at the interface below the first gate 30 and in contact with the first channel layer 20, positioning the center of the RF coil 70 above the first gate 30 effectively changes the spin direction of electrons in the defect energy level. This causes electrons from the conduction band, unable to pair with electrons in the defect energy level, to fall into the valence band, forming the first leakage current. If the RF coil 70 is not positioned above the first gate 30, the spin direction of some electrons in the defect energy level will not change. Consequently, some electrons that have fallen into the conduction band will still combine with the electrons, reducing the number of electrons falling into the valence band and reducing the first leakage current, which can lead to inaccurate detection results. Therefore, by arranging the center of the RF coil 70 to be located above the first grid 30 , the detection result of the interface defect can be made more accurate.

[0079] In one embodiment, a radio frequency coil 70 is disposed above the first grid 30 , and further comprising: a magnetic core unit is disposed at the center of the radio frequency coil 70 , the magnetic core unit being used to increase the intensity of the radio frequency magnetic field B2 .

[0080] In this embodiment, the intensity of the RF magnetic field B2 can be increased by disposing a magnetic core unit at the center of the RF coil 70. The primary function of the RF coil 70 is to generate the RF magnetic field B2, thereby flipping the spin direction of electrons in the defect energy level. This causes electrons from the conduction band that were originally paired with electrons in the defect energy level to stop pairing and fall into the valence band, where they combine with holes in the valence band to form a first leakage current. By providing a higher intensity RF magnetic field B2, the spin direction of electrons in the defect energy level can be flipped more comprehensively, resulting in more accurate detection of the first leakage current and, consequently, more accurate detection of interface defects.

[0081] In a specific embodiment, the material of the magnetic core unit may be iron oxide.

[0082] In one embodiment, the RF coil 70 is helical. In this embodiment, by configuring the RF coil 70 to be helical, the RF coil 70 generates an RF magnetic field B2 when energized, thereby flipping the spin direction of electrons in defect energy levels, facilitating subsequent interface defect detection.

[0083] In a specific embodiment, referring to Figure 6 As shown, the RF coil 70 is in a square spiral shape, that is, each turn of the RF coil 70 is a square.

[0084] In one embodiment, reference Figure 6 As shown, the RF coil 70 includes at least two turns. Specifically, the number of turns of the RF coil 70 determines the strength of the RF magnetic field B2 generated by the RF coil 70. By configuring the RF coil 70 to include at least two turns, the strength of the RF magnetic field B2 can be increased, which can better change the spin direction of electrons in the defect energy level, thereby allowing electrons from the conduction band to fall into the valence band more easily, forming the first leakage current, and making the detection results more accurate.

[0085] In one embodiment, the diameter of the RF coil 70 is equal to the diameter of the first gate 30. In this embodiment, when the RF coil 70 is a square spiral, its diameter represents the maximum outer diameter of the RF coil 70. By setting the diameter of the RF coil 70 equal to the diameter of the first gate 30, the RF magnetic field B2 generated by the RF coil 70 can more comprehensively flip the spin direction of electrons in defect energy levels, avoiding the problem of the RF magnetic field B2 not fully affecting the electrons in the defect energy levels, resulting in the spin direction of some electrons in the defect energy levels not flipping, and leading to inaccurate detection results.

[0086] In one embodiment, the diameter of the RF coil 70 is 0.1-0.3 μm.

[0087] In one embodiment, applying the radio frequency magnetic field B2 above the first gate 30 includes adjusting the frequency of the radio frequency magnetic field B2 to flip the spin direction of electrons in the defect energy level.

[0088] In this embodiment, after applying the horizontal static magnetic field B1, the spin direction of the electrons in the defect energy level is adjusted, with their spin direction downward. The frequency of the radio frequency magnetic field B2 is then adjusted by sweeping the frequency, causing the spin direction of the electrons in the defect energy level to flip. Under the constraints of the Pauli exclusion principle, the pairing behavior of electrons from the conduction band that were originally paired with the defect energy level is prohibited. At this time, the originally paired electrons from the conduction band will fall into the valence band and recombine with holes in the valence band, forming a first leakage current. In this embodiment, adjusting the frequency of the radio frequency magnetic field B2 by sweeping the frequency can better flip the spin direction of the electrons in the defect energy level, preparing for the subsequent first leakage current detection.

[0089] In one embodiment, applying the radio frequency magnetic field B2 above the first gate 30 includes adjusting the intensity of the radio frequency magnetic field B2 to flip the spin direction of electrons in the defect energy level.

[0090] In this embodiment, the main function of the RF coil 70 is to generate an RF magnetic field B2 to flip the spin direction of electrons in the defect energy level. Because the greater the intensity of the RF electric field, the better the effect of flipping the spin direction of electrons in the defect energy level. By adjusting the intensity of the RF magnetic field B2, the spin direction of electrons in the defect energy level can be flipped as much as possible, thereby stopping the pairing of electrons from the conduction band that were originally paired with electrons in the defect energy level, causing the electrons from the conduction band to fall into the valence band. The electrons from the conduction band then combine with holes in the valence band to form a first leakage current, thereby making the detection results more accurate.

[0091] In one embodiment, the intensity of the RF magnetic field B2 can be adjusted by adding a magnetic core unit made of different materials at the center of the RF coil 70 or by setting the number of turns of the RF coil 70 to be different.

[0092] In one embodiment, detecting the first leakage current and determining the interface defect size of the first channel layer 20 according to the first leakage current includes comparing the first leakage current with a preset mapping table and determining the interface defect density of the first channel layer 20 according to the comparison result.

[0093] In this embodiment, after the magnitude of the first leakage current is detected by the current detector, it is output to the control module. A mapping table corresponding to the first leakage currents is stored in the control module. For example, the sequence of the first leakage currents is: e1, e2, e3, e4, e5, and the corresponding sequence of the interface defect density is: f1, f2, f3, f4, f5. In this way, each first leakage current corresponds to an interface defect density result, making the detection result more accurate.

[0094] In one embodiment, reference Figure 4 As shown, a horizontal static magnetic field B1 is applied to the first channel layer 20 , including: arranging a first magnet 41 and a second magnet 42 on both sides of the first source 50 and the first drain 60 respectively, and the first magnet 41 and the second magnet 42 are used to generate the horizontal static magnetic field B1 .

[0095] In this embodiment, a first magnet 41 and a second magnet 42 are respectively arranged on both sides of the first source 50 and the first drain 60, wherein the magnetic north pole and the magnetic south pole of the first magnet 41 and the second magnet 42 are arranged opposite to each other, so that a horizontal static magnetic field B1 is generated between the first magnet 41 and the second magnet 42, and the spin direction of the electrons in the defect energy level is adjusted so that the electrons in the defect energy level spin downward after the horizontal static magnetic field B1 is applied.

[0096] In one embodiment, the heights of the first magnet 41 and the second magnet 42 are equal to the height of the first channel layer 20. In this embodiment, by setting the heights of the first magnet 41 and the second magnet 42 to be equal to the height of the first channel layer 20, the direction of the electrons in the defect energy level can be changed, and the impact on other parts of the power device can be minimized to avoid affecting the performance of the power device and causing inaccurate detection results.

[0097] In one embodiment, the first magnet 41 and the second magnet 42 are symmetrically arranged. For example, the first magnet 41 is arranged outside the first source 50, and the second magnet 42 is arranged outside the first drain 60, so that the first magnet 41 and the second magnet 42 can generate a horizontal static magnetic field B1. By arranging the first magnet 41 and the second magnet 42 symmetrically, the generated horizontal static magnetic field B1 can be made more uniform, and the spin direction of the electrons in the defect energy level can be better adjusted.

[0098] In one embodiment, applying a horizontal static magnetic field B1 to the first channel layer 20 includes placing the power device having the first channel layer 20 in a sealed cavity and placing a first magnet 41 and a second magnet 42 on both sides of the first source 50 and the first drain 60 , respectively.

[0099] In this embodiment, by placing the power device having the first channel layer 20 formed in a sealed cavity, the interference of external electromagnetic interference on the power device can be reduced, and the detection results can be made more accurate. A first magnet 41 and a second magnet 42 are respectively provided on both sides of the first source 50 and the first drain 60. The first magnet 41 and the second magnet 42 are used to generate a horizontal static magnetic field B1, which can adjust the spin direction of the electrons in the defect energy level. Then, a radio frequency coil 70 is provided on the first gate 30. When the radio frequency coil 70 is energized, a radio frequency magnetic field B2 is generated, which causes the spin direction of the electrons in the defect energy level to flip, thereby causing the electrons from the conduction band that were originally paired with the electrons in the defect energy level to stop pairing and fall into the valence band, combining with holes in the valence band to form a first leakage current. The first leakage current of the power device to be tested is detected, and the interface defect density of the first channel layer 20 is determined based on the magnitude of the first leakage current. In this embodiment, by placing the power device with the first channel layer 20 formed thereon in a sealed cavity, interference of external electromagnetic interference on the channel defect detection method can be reduced, and the detection result can be made more accurate.

[0100] The present application also provides a wafer detection method, referring to Figure 7 As shown, the wafer 100 includes a plurality of power devices and at least one power device to be tested 200 . The at least one power device to be tested is prepared using any of the above-mentioned channel defect detection methods.

[0101] In this embodiment, when multiple power devices are provided on a wafer 100, and only one of the power devices needs to be tested, the power device to be tested is referred to as the power device to be tested 200. During the preparation of the power device to be tested 200, any of the above-described channel defect detection methods is employed. For example, steps S10 to S50 are performed once each time the first channel layer 20 is prepared. Specifically, a radio frequency coil 70 can be provided on the first gate 30 of the power device to be tested 200, the first drain 60 can be connected to a power source, the first source 50 can be grounded, and the first gate 30 can be left floating. Alternatively, etched lines on the wafer 100 can be selected to serve as the first source 50 and the first drain 60. The wafer 100 is then placed in a sealed chamber. Two magnets are set on opposite sides of the sealed chamber, and the wafer 100 is set in the middle of the two magnets. The horizontal static magnetic field B1 generated by the two magnets adjusts the spin direction of the electrons in the defect energy level. When the radio frequency coil 70 is energized, an radio frequency magnetic field B2 is generated, causing the spin direction of the electrons in the defect energy level to flip, thereby causing the electrons from the conduction band that were originally paired with the electrons in the defect energy level to stop pairing and fall into the valence band, combining with the holes in the valence band to form a first leakage current. The first leakage current of the power device 200 to be tested is detected, and the interface defect density of the channel layer of the power device 200 to be tested is determined based on the magnitude of the first leakage current. In this embodiment, multiple power devices 200 to be tested on the wafer 100 can also be tested simultaneously, which can improve the detection effect while ensuring the detection accuracy.

[0102] In the above embodiments, the description of each embodiment has its own focus. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant description of other embodiments.

[0103] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A channel defect detection method, characterized in that: include; Step S10: forming a first channel layer, and sequentially forming a first source, a first drain, and a first gate on the first channel layer; wherein the first drain is connected to a power supply, the first source is grounded, and the first gate is left floating; Step S20: applying a horizontal static magnetic field to the first channel layer; wherein the horizontal static magnetic field is used to adjust the spin direction of electrons located in the defect energy level; Step S30: applying a radio frequency magnetic field above the first gate; wherein a radio frequency coil is disposed above the first gate, with the center of the radio frequency coil located above the first gate. When energized, the radio frequency coil generates the radio frequency magnetic field, which is used to flip the spin direction of electrons in defect energy levels, so that electrons from the conduction band move into holes in the valence band, thereby forming a first leakage current; Step S40: determining the interface defect density of the first channel layer according to the first leakage current, and removing the first source, the first gate, and the first drain; Step S50: Repeat step S10, step S20, step S30 and step S40 until the thickness of the channel layer reaches a preset threshold thickness; wherein the channel layer is composed of multiple layers of the first channel layer.

2. The trench defect detection method according to claim 1, wherein: The step of increasing the thickness of the channel layer until the thickness of the channel layer reaches a preset threshold thickness includes: The thickness of the channel layer is measured, and the thickness of the channel layer is compared with the preset threshold thickness. When the thickness of the channel layer is less than the preset threshold thickness, the steps S10, S20, S30, and S40 are repeatedly performed.

3. The channel defect detection method according to claim 1, wherein: The step of sequentially forming a first source electrode, a first drain electrode, and a first gate electrode on the first channel layer includes: A first source electrode, a first drain electrode and a first gate electrode are sequentially formed on the first channel layer by photolithography.

4. The channel defect detection method according to claim 1, wherein: The removing of the first source, the first gate, and the first drain includes: The first source electrode, the first gate electrode and the first drain electrode are washed away using a mixture of phosphoric acid, acetic acid and ammonia water.

5. The trench defect detection method according to claim 1, wherein: After determining the interface defect density of the first channel layer according to the first leakage current, the method further includes: When the interface defect density of the first channel layer is greater than a preset interface defect density, a layer of the first channel layer is removed, and the steps S10, S20, S30 and S40 are repeatedly performed.

6. The trench defect detection method according to claim 1, wherein: The step of applying a radio frequency magnetic field above the first gate further comprises: A magnetic core unit is provided at the center of the radio frequency coil, and the magnetic core unit is used to increase the intensity of the radio frequency magnetic field.

7. The trench defect detection method according to any one of claims 1 to 6, wherein: The applying a horizontal static magnetic field to the first channel layer comprises: A first magnet is disposed on the left side of the first source electrode, and a second magnet is disposed on the right side of the first drain electrode. The first magnet and the second magnet are used to generate the horizontal static magnetic field.

8. A wafer detection method, characterized in that: The wafer includes a plurality of power devices and at least one power device to be tested, and the at least one power device to be tested is prepared using the channel defect detection method according to any one of claims 1 to 7.

Images