Inspection method of ingot
The ingot inspection method addresses inaccuracies in identifying impurity concentrations by analyzing fluorescence photon count changes over time, enabling precise classification and reducing kerf loss in SiC ingot processing.
Patent Information
- Application Number
- JP2023080002
- Authority / Receiving Office
- JP · JP
- Patent Type
- Applications
- Current Assignee / Owner
- Filing Date
- 2023-05-15
- Publication Date
- 2024-11-27
AI Technical Summary
Existing ingot inspection methods for identifying regions with different impurity concentrations in single-crystal SiC ingots are inaccurate due to variations in fluorescence intensity caused by lattice defects and manufacturing inconsistencies, leading to non-uniform peeling layer formation and increased kerf loss.
An ingot inspection method that analyzes the temporal change in fluorescence photon count after excitation, utilizing the peak and subsequent decrease in photon count to accurately classify regions based on their impurity concentrations, independent of lattice defects.
Accurately identifies regions with high or low impurity concentrations, reducing kerf loss and improving the efficiency of wafer production from single-crystal SiC ingots.
Smart Images

Figure 2024164493000001_ABST
Abstract
Description
[Technical field]
[0001] The present invention relates to a method for inspecting an ingot to identify regions of the ingot that have different impurity concentrations. [Background technology]
[0002] Semiconductor device chips are generally manufactured using disk-shaped wafers. These wafers are manufactured, for example, by cutting out cylindrical semiconductor ingots using a wire saw. However, when manufacturing wafers in this manner, a large part of the ingot is lost as kerf loss, which is uneconomical.
[0003] Furthermore, the single crystal SiC (silicon carbide) used as a material for power devices is very hard, so when cutting wafers from a single crystal SiC ingot using a wire saw, it takes a long time to cut, resulting in low productivity.
[0004] In order to solve these problems, a method has been proposed in which a laser beam is used to cut out wafers from an ingot without using a wire saw (see, for example, Patent Document 1). In this method, a laser beam is irradiated onto an ingot so that the focal point of the laser beam, which has a wavelength that transmits through the ingot, is positioned inside the ingot.
[0005] As a result, a delamination layer including a modified portion and cracks extending from the modified portion is formed inside the ingot. When an external force is applied to the ingot with the delamination layer formed therein by ultrasonic vibration or the like, the ingot is separated at the delamination layer to produce a wafer.
[0006] Incidentally, single crystal SiC ingots are generally doped with impurities such as nitrogen to impart electrical conductivity, but single crystal SiC ingots are not doped with such impurities uniformly and may contain multiple regions with different impurity concentrations.
[0007] For example, the impurity concentration in atomically flat regions called facet regions formed during the growth process of single-crystal SiC is higher than in other regions (non-facet regions). Regions with high impurity concentrations, such as facet regions, have a higher refractive index and a higher energy absorption rate than non-facet regions.
[0008] Therefore, when a peeling layer is formed on a single crystal SiC ingot including a facet region by the above-mentioned method, there is a problem that the position (height) at which the peeling layer is formed becomes non-uniform, resulting in large kerf loss.
[0009] In view of this, an ingot inspection method has been proposed for identifying regions of a single crystal SiC ingot with different impurity concentrations (see, for example, Patent Document 2). Specifically, in this method, if the intensity of fluorescence generated by irradiating a specific region of a single crystal SiC ingot with excitation light is equal to or lower than a predetermined value, this specific region is determined to be a region with a high impurity concentration, i.e., a facet region. [Prior art documents] [Patent documents]
[0010] [Patent Document 1] JP 2016-111143 A [Patent Document 2] JP 2020-77783 A Summary of the Invention [Problem to be solved by the invention]
[0011] The intensity of the fluorescence generated by irradiating a specific region of an ingot with excitation light, i.e., the total number of photons, may vary depending not only on the impurity concentration of the specific region but also on individual differences between ingots, specifically, the amount of lattice defects caused by differences in the apparatus and / or method used in the manufacture of the ingot. Therefore, when regions of an ingot with different impurity concentrations are identified by the above-mentioned method, the accuracy may be low.
[0012] In view of this, an object of the present invention is to provide an ingot inspection method capable of accurately identifying regions with different impurity concentrations. [Means for solving the problem]
[0013] When an ingot is irradiated with excitation light, the number of photons of the fluorescence generated from the ingot changes over time. Specifically, the number of photons of this fluorescence reaches a peak immediately after the ingot is irradiated with excitation light, and then gradually decreases. By carefully observing such changes over time in the number of photons of the fluorescence, the inventors arrived at the present invention.
[0014] Specifically, the inventors of the present invention discovered that the number of lattice defects present in an ingot significantly affects the peak height of the number of photons in the fluorescence that changes over time, but does not significantly affect the number of photons in the fluorescence that occurs after the peak is exceeded. In other words, the inventors of the present invention discovered that the number of photons in the fluorescence that occurs after the peak is significantly affected by the impurity concentration contained in the region of the ingot irradiated with the excitation light, and thus arrived at the present invention.
[0015] For example, according to the present invention, there is provided an ingot inspection method for identifying regions of an ingot having different impurity concentrations, the ingot inspection method comprising: a light receiving step of receiving fluorescence, the number of photons of which changes over time, generated by irradiating multiple regions contained in the ingot with one pulse of excitation light of a predetermined wavelength from a light source that emits pulsed light in sequence; and a classification step of classifying each of the multiple regions as a region with high impurity concentration or a region with low impurity concentration by referring to the number of photons of the fluorescence generated during a predetermined period during which the number of photons of the fluorescence is gradually decreasing. Effect of the Invention
[0016] In the present invention, the region of the ingot irradiated with the excitation light is classified into a region with high impurity concentration or a region with low impurity concentration by referring to the number of fluorescent photons generated during a predetermined period during which the number of fluorescent photons is gradually decreasing.
[0017] In other words, this classification is performed by referring to the number of photons of the fluorescence, the amount of which is greatly affected by the impurity concentration in the region of the ingot irradiated with the excitation light, and therefore, in the present invention, it is possible to accurately identify the regions with different impurity concentrations. [Brief description of the drawings]
[0018] [Figure 1] FIG. 1(A) is a front view that shows a schematic diagram of an example of an ingot, and FIG. 1(B) is a top view that shows a schematic diagram of the ingot shown in FIG. 1(A). [Diagram 2] FIG. 2 is a diagram illustrating an example of an inspection device. [Diagram 3] FIG. 3 is a flow chart that illustrates an example of an ingot inspection method for identifying regions of the ingot with different impurity concentrations. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS
[0019] An embodiment of the present invention will be described with reference to the accompanying drawings. Fig. 1(A) is a front view showing an example of an ingot, and Fig. 1(B) is a top view showing the ingot shown in Fig. 1(A).
[0020] 1(A) and 1(B) is made of, for example, cylindrical single crystal SiC having a generally parallel upper surface (front surface) 11a and lower surface (back surface) 11b. This ingot 11 is produced by epitaxial growth such that the c-axis 11c of the single crystal SiC is slightly inclined with respect to a perpendicular line 11d to the front surface 11a and back surface 11b.
[0021] For example, the angle (off angle) α between the c-axis 11c and the perpendicular line 11d is 1° to 6° (typically 4°). In addition, two flat portions, i.e., a primary orientation flat 13 and a secondary orientation flat 15, are formed on the side surface of the ingot 11 to indicate specific crystal orientations of the single crystal SiC.
[0022] The primary orientation flat 13 is longer than the secondary orientation flat 15. In addition, the secondary orientation flat 15 is formed so as to be parallel to an intersection line where a plane parallel to the c-plane 11e of the single crystal SiC intersects with the front surface 11a or the back surface 11b.
[0023] Furthermore, ingot 11 is doped with impurities such as nitrogen to impart electrical conductivity. Ingot 11 also includes facet region 11f, which is a region that is flat at the atomic level, and non-facet region 11g, which is a region other than facet region 11f.
[0024] The impurity concentration in the facet region 11f is higher than the impurity concentration in the non-facet region 11g. Note that, in FIG. 1(B), the boundary between the facet region 11f and the non-facet region 11g is shown by a dotted line, but this boundary is an imaginary line and does not exist in the actual ingot 11.
[0025] The material of the ingot 11 is not limited to SiC, and may be LiTaO3 (lithium tantalate: LT) or GaN (gallium nitride). In addition, the side surface of the ingot 11 may not be provided with either or both of the primary orientation flat 13 and the secondary orientation flat 15.
[0026] Fig. 2 is a schematic diagram showing an example of an inspection device capable of identifying regions with different impurity concentrations in an ingot 11. Note that some of the components of the inspection device are shown in blocks in Fig. 2. The inspection device 2 shown in Fig. 2 has a disk-shaped chuck table 4 having a generally horizontal upper surface (holding surface) 4a on which the ingot 11 is placed.
[0027] The chuck table 4 is connected to a horizontal movement mechanism (not shown). This horizontal movement mechanism includes, for example, a ball screw and a motor. When the horizontal movement mechanism is operated, the chuck table 4 moves in a direction substantially parallel to the holding surface 4a (horizontal direction).
[0028] The chuck table 4 is connected to a rotation mechanism (not shown). The rotation mechanism includes, for example, a pulley and a motor. When the rotation mechanism is operated, the chuck table 4 rotates about a straight line L that is a rotation axis along a direction (vertical direction) substantially perpendicular to the holding surface 4a and that passes through the center of the holding surface 4a.
[0029] The chuck table 4 is also connected to a suction mechanism (not shown). This suction mechanism has, for example, an ejector. When the suction mechanism is operated, a suction force acts on the space near the holding surface 4a. Therefore, when the suction mechanism is operated with the ingot 11 placed on the holding surface 4a of the chuck table 4, the ingot 11 is sucked toward the chuck table 4 and held thereon.
[0030] An inspection unit 6 is provided above the chuck table 4. This inspection unit 6 has a light source 8 that emits pulsed light. This light source 8 has a light emitting diode (LED), and irradiates the mirror 10 with excitation light A of a predetermined wavelength (e.g., 375 nm) that is absorbed by the ingot 11.
[0031] Then, the excitation light A irradiated to the mirror 10 is reflected from the mirror 10 so as to head directly downward and reaches the condenser lens 12. Furthermore, the excitation light A that has reached the condenser lens 12 is condensed so as to form a focal point directly below the condenser lens 12.
[0032] The inspection unit 6 also has an annular elliptical mirror 14 having a reflecting surface 14a on the inside. Note that Fig. 2 shows a cross section of the elliptical mirror 14. This reflecting surface 14a corresponds to a part of the curved surface of a spheroid obtained by rotating an ellipse 14b, which has a major axis extending along the vertical direction and a minor axis extending along the horizontal direction, around the major axis.
[0033] The elliptical mirror 14 has two focal points F1 and F2, and focuses light generated from one of the focal points (e.g., the focal point F1) at the other focal point (e.g., the focal point F2). The focusing lens 12 is designed so that its focal point approximately coincides with the focal point F1. In other words, the focusing point at which the excitation light A is focused overlaps with the focal point F1.
[0034] The inspection unit 6 also has a light receiving section 16. The light receiving section 16 is provided so that the center of its light receiving surface 16a coincides with the focal point F2 of the elliptical mirror 14. The light receiving section 16 outputs an electrical signal indicating the change over time in the number of photons of the received light.
[0035] A bandpass filter 18 is provided in the optical path between focal points F1 and F2 of the elliptical mirror 14, which transmits light having a wavelength in a predetermined range longer than that of the excitation light A and blocks light having other wavelengths. The light generated at the focal point F1 and reflected by the elliptical mirror 14 passes through the bandpass filter 18 and reaches the focal point F2.
[0036] The bandpass filter 18 is a visible light filter that transmits visible light of, for example, 510 nm to 550 nm, typically about 530 nm, and blocks light of other wavelengths. Alternatively, the bandpass filter 18 may be an IR filter that transmits infrared (IR) light of, for example, 750 nm or more, and blocks light of other wavelengths.
[0037] Furthermore, the inspection unit 6 is connected to a vertical movement mechanism (not shown). This vertical movement mechanism has, for example, a ball screw and a motor. When the vertical movement mechanism is operated, the inspection unit 6 moves in the vertical direction.
[0038] In addition, the inspection device 2 includes a controller (not shown) for controlling the above-mentioned components. The controller has a processor and a memory. The processor includes, for example, a CPU (Central Processing Unit) and the like. The memory includes, for example, a volatile memory such as a DRAM (Dynamic Random Access Memory) or a SRAM (Static Random Access Memory), and a non-volatile memory such as an SSD (Solid State Drive) (NAND type flash memory) or an HDD (Hard Disk Drive) (magnetic storage device).
[0039] The memory stores various information (data, programs, etc.) used by the processor. For example, the memory stores data used to identify regions of the ingot 11 with different impurity concentrations. The processor reads out and executes various programs stored in the memory, and controls the components of the inspection device 2. For example, the processor controls the components of the inspection device 2 to perform an ingot inspection method for identifying regions of the ingot 11 with different impurity concentrations.
[0040] 3 is a flow chart showing a schematic example of this inspection method. In this inspection method, first, a plurality of regions included in the ingot 11 are sequentially irradiated with one pulse of excitation light A from the light source 8, and fluorescence B, the number of photons of which changes over time, is received (light receiving step S1).
[0041] Specifically, in this light receiving step S1, first, the ingot 11 is placed on the holding surface 4a of the chuck table 4 so that the surface 11a of the ingot 11 faces upward. Next, the processor operates the suction mechanism so that the ingot 11 is sucked to the chuck table 4 and held thereon.
[0042] Next, the processor operates the vertical movement mechanism to move the inspection unit 6 so that the focal point F1 of the elliptical mirror 14 is positioned at the same height as the surface 11a of the ingot 11. Next, the processor operates the horizontal movement mechanism and / or the rotation mechanism to move and / or rotate the chuck table 4 so that the focal point F1 is positioned in any of a plurality of regions included in the ingot 11.
[0043] Next, the processor operates the light source 8 to irradiate one pulse of excitation light A from the light source 8. As a result, the excitation light A is irradiated onto one of a plurality of regions included in the ingot 11 via the mirror 10 and the condenser lens 12. When the excitation light A is irradiated onto this region, the excitation light A is absorbed in this region, and fluorescence B is generated.
[0044] If the wavelength of the excitation light A is 375 nm, the excitation light A penetrates from the surface 11a of the ingot 11 to a depth of about 10 μm. Then, fluorescence B is generated from the region of the ingot 11 into which the excitation light A penetrated. The number of photons of this fluorescence B peaks immediately after the ingot 11 is irradiated with the excitation light A, and then gradually decreases over time.
[0045] Furthermore, after being reflected by the reflecting surface of the elliptical mirror 14, the fluorescence B passes through the bandpass filter 18 and reaches the light-receiving unit 16. Therefore, light of the fluorescence B that is outside the predetermined range of wavelengths is blocked by the bandpass filter 18, and only light of the fluorescence B that is within the predetermined range of wavelengths is received by the light-receiving unit 16.
[0046] As a result, data indicating the change over time in the number of photons of light having a wavelength within a predetermined range of the fluorescence B (time-dependent change data) is output from the light-receiving unit 16. The time-dependent change data is then input to the processor, which stores the time-dependent change data in memory.
[0047] Furthermore, the above-mentioned operation is repeated until all of the multiple regions included in the ingot 11 are irradiated with the excitation light A. Specifically, movement and / or rotation of the chuck table 4 for positioning the focal point F1 at a region among the multiple regions of the ingot 11 that has not yet been irradiated with the excitation light A and irradiation of one pulse of the excitation light A from the light source 8 are alternately repeated.
[0048] As a result, a plurality of time-varying data indicating the change over time in the number of photons of light having a predetermined wavelength range among the fluorescence B generated from a plurality of regions included in the ingot 11 are sequentially output from the light receiving unit 16. The plurality of time-varying data are then input to a processor, which stores the plurality of time-varying data in memory. With the above, the light receiving step S1 is completed.
[0049] After the light receiving step S1, the number of photons of the fluorescence B generated during a predetermined period during which the number of photons of the fluorescence B is gradually decreasing is referenced to classify each of the multiple regions into a region with high impurity concentration or a region with low impurity concentration (classification step S2).
[0050] Specifically, in this classification step S2, the processor first identifies the period (not included period) from when the ingot 11 is irradiated with the excitation light A until the number of photons of the fluorescence B, which changes over time, gradually decreases from its peak to approximately n times the peak (n is, for example, a value that is 1 / 10 or less and 1 / 1000 or more, typically 1 / 100).
[0051] The value of the magnification factor used to specify the non-count period, i.e., the value of n, is stored in advance in the memory. The non-count period may be specified by the processor with reference to one piece of time-varying data, or may be specified by the processor as, for example, an average period with reference to multiple pieces of time-varying data.
[0052] Next, by referring to the plurality of pieces of time-dependent change data, the processor calculates the number of photons of light having a predetermined range of wavelengths among the fluorescence B generated from each of the plurality of regions included in the ingot 11 during a predetermined period from the lapse of the non-counting period. Note that this predetermined period is, for example, stored in advance in a memory.
[0053] Alternatively, the predetermined period may be specified by the processor with reference to one or more pieces of time-course change data. Specifically, the processor may specify the period from the non-counting period until the number of photons of the fluorescence B becomes about m (m is, for example, a value not larger than 1 / 100 and not smaller than 1 / 10000, typically 1 / 1000) times the peak number of photons of the fluorescence B as the predetermined period.
[0054] Then, the processor stores in memory a plurality of data (a plurality of fluorescence data) each of which is linked to one of a plurality of regions contained in ingot 11 and the number of photons of light having a predetermined wavelength range among the fluorescence B generated from one of the plurality of regions during a predetermined period after the non-counting period has elapsed.
[0055] Next, by referring to the plurality of pieces of fluorescence data, the processor classifies each of the plurality of regions included in the ingot 11 into a region with a high impurity concentration or a region with a low impurity concentration. For example, by referring to the plurality of pieces of fluorescence data, the processor classifies a region associated with a photon number less than a photon number previously stored in memory as a threshold value into a region with a high impurity concentration, and classifies a region associated with a photon number greater than the threshold value into a region with a low impurity concentration.
[0056] Alternatively, the processor may compare the number of photons contained in each of the multiple fluorescence data and classify areas associated with a relatively low number of photons as areas of high impurity concentration, and classify areas associated with a relatively high number of photons as areas of low impurity concentration.
[0057] In the ingot inspection method shown in FIG. 3, the region of the ingot 11 irradiated with the excitation light A is classified into a region with a high impurity concentration or a region with a low impurity concentration by referring to the number of photons of light of a predetermined range of wavelengths among the fluorescence B generated during a predetermined period while the number of photons of the fluorescence B is gradually decreasing.
[0058] That is, this classification is performed by referring to the number of photons of the fluorescence B, the amount of which is greatly affected by the impurity concentration in the region of the ingot 11 irradiated with the excitation light A. Therefore, with this method, it is possible to accurately identify regions with different impurity concentrations.
[0059] The above is one embodiment of the present invention, and the present invention is not limited to the above. For example, in the above-mentioned method, only light of a predetermined range of wavelengths of the fluorescence B is received by the light receiving unit 16, but in the method of the present invention, all of the fluorescence B may be received by the light receiving unit 16. In this case, the inspection device 2 does not need to be provided with the bandpass filter 18.
[0060] Furthermore, in the above-mentioned method, light of the fluorescence B within a predetermined range of wavelengths is received by the light-receiving unit 16 during both the not-included-in-count period and the predetermined period, but in the present invention, light of the fluorescence B within a predetermined range of wavelengths may be received by the light-receiving unit 16 only during the predetermined period. In this case, in the inspection device 2, a light-shielding filter that blocks the fluorescence B during the not-included-in-count period and transmits the fluorescence B during the predetermined period may be provided in the optical path between the focal points F1 and F2 of the elliptical mirror 14.
[0061] In the method of the present invention, the light receiving step S1 and the classification step S2 may be performed in parallel. That is, in the method of the present invention, the regions already irradiated with the excitation light A may be classified into high impurity regions or low impurity regions at any time before the irradiation of all of the regions included in the ingot 11 with the excitation light A is completed.
[0062] Furthermore, in the method of the present invention, each of the multiple regions included in the ingot 11 may be classified into three or more regions having different impurity concentrations. Such classification is performed, for example, by a processor comparing the number of photons included in each of the multiple fluorescence data with two or more photon number thresholds having different values that are stored in advance in a memory.
[0063] In addition, the structures and methods according to the above-described embodiments can be modified as appropriate without departing from the scope of the present invention. [Explanation of symbols]
[0064] 2: Inspection equipment 4: Chuck table 6: Inspection unit 8:Light source 10: Mirror 11: Ingot (11a: Top surface (front surface), 11b: Bottom surface (back surface), 11c: c axis) (11d: perpendicular line, 11e: c surface) (11f: faceted area, 11g: non-faceted area) 12: Condenser lens 13: Primary orientation flat 14: Elliptical mirror (14a: Reflecting surface) 15: Secondary orientation flat 16: Light receiving part (16a: Light receiving surface) 18: Bandpass filter
Claims
[Claim 1] 1. A method for inspecting an ingot to identify regions of the ingot having different impurity concentrations, comprising: a light receiving step of receiving fluorescence generated from one of a plurality of regions included in the ingot by irradiating the plurality of regions with one pulse of excitation light having a predetermined wavelength from a light source that emits pulsed light in sequence to the plurality of regions, and the number of photons of the fluorescence changes over time; a classification step of classifying each of the plurality of regions into a region with a high impurity concentration or a region with a low impurity concentration based on the number of photons of the fluorescence generated during a predetermined period during which the number of photons of the fluorescence is gradually decreasing; The ingot inspection method comprises:
Citation Information
Patent Citations
Generation method of wafer
JP2016111143A
Facet area detection method and detection apparatus, wafer generation method, and laser processing apparatus
JP2020077783A
