Method for measuring the resistivity of a silicon wafer and a quartz jig used therefor.

Abstract

This invention provides a silicon wafer resistivity measurement method and a quartz jig for use therein, which can accurately measure the true resistivity of only the dopant while suppressing the effects of thermal donor regeneration after donor killer heat treatment. [Solution] The resistivity measurement method for a silicon wafer according to the present invention comprises: a sample preparation step S11 in which a silicon wafer is cut radially to prepare a sample; a sample introduction step S12 in which the sample is placed on a quartz jig and introduced into a heat treatment furnace so that the cut line is horizontal and upward; a donor killer heat treatment step S13 in which the sample is heat-treated in the heat treatment furnace to eliminate thermal donors; a cooling step S14 in which the quartz jig on which the sample is placed is removed from the heat treatment furnace and the sample is cooled by blowing air from above; and a resistivity measurement step S15 in which the resistivity of the cooled sample is measured by the four-probe method.

Description

[Technical Field]

[0001] This invention relates to a method for measuring the resistivity of a silicon wafer and a quartz jig used therefor. [Background technology]

[0002] Silicon wafers are widely used as substrate materials for semiconductor devices. In the manufacturing of silicon wafers, a silicon single crystal ingot grown by the CZ method is ground to adjust its diameter, then the top and tail portions are cut off, and the cylindrical ingot is further cut at regular intervals to form silicon blocks of a predetermined length. At this time, sample wafers (slag) for quality inspection are also cut from both ends of the silicon block, and the quality of the silicon block is inspected for resistivity, oxygen concentration, carrier recombination lifetime, presence or absence of crystal defects, etc., to determine whether the silicon block is acceptable or not.

[0003] If the silicon block passes the quality inspection, the silicon block is processed into finished products. In the processing of the silicon block, multiple silicon wafers are cut out at once by slicing the silicon block using a wire saw. After that, wafer products are completed through processes such as surface grinding, etching, surface polishing, and cleaning.

[0004] Regarding evaluation techniques for silicon single crystal ingots, for example, Patent Document 1 describes a method of cutting a silicon single crystal ingot into a block shape using a band saw or the like, cutting out sample wafers from both ends of the silicon block, and evaluating the resistivity, oxygen concentration, crystal defects, etc., to determine whether the silicon block is acceptable or not.

[0005] Patent Document 2 also describes a method for manufacturing a silicon wafer, comprising: a first outer circumference grinding step of grinding the outer circumference of a cylindrical ingot to a diameter larger than the diameter of an ingot block for wafer manufacturing; a block cutting step of cutting the cylindrical ingot after the first grinding step into a plurality of ingot blocks; a sample cutting step of cutting out silicon inspection samples from the plurality of ingot blocks; a quality evaluation step of performing quality evaluation using the cut inspection samples; a second outer circumference grinding step of grinding the outer circumference of an ingot block to a diameter for the wafer manufacturing step; a notch forming step of forming a notch on the outer circumference of an ingot block after the second outer circumference grinding step; and a wafer manufacturing step of cutting out a silicon wafer from the ingot block with the notch formed.

[0006] Patent Document 3 describes a method for measuring the resistivity of a high-resistivity silicon wafer with a resistivity of 2000 Ωcm or more. In this resistivity measurement method, after heat treatment of the silicon wafer with a donor killer, the oxide film on the surface to be measured is removed by a non-aqueous treatment such as buff polishing after at least two hours have passed. Then, the resistivity is measured by bringing an electrode needle into contact with the surface to be measured. This method prevents the deactivation of dopants due to contact between the surface to be measured and hydrogen ions, and allows for accurate measurement of the resistivity of the wafer.

[0007] Patent Document 4 describes a method for measuring the resistivity of a silicon wafer, comprising the steps of: measuring the resistivity and oxygen concentration of a silicon wafer subjected to donor killer heat treatment; determining the carrier concentration including dopants and thermal donors from the measured resistivity; determining the amount of thermal donors from the measured oxygen concentration; determining the amount of dopant by subtracting the determined amount of thermal donors from the determined carrier concentration; and converting the determined amount of dopant into resistivity. [Prior art documents] [Patent Documents]

[0008] [Patent Document 1] Japanese Patent Application Laid-Open No. 2014-201458 [Patent Document 2] Japanese Patent No. 6332422 [Patent Document 3] Japanese Patent Application Laid-Open No. 2015-26755 [Patent Document 4] Japanese Patent Application Laid-Open No. 2019-029387 [Summary of the Invention] [Problems to be Solved by the Invention]

[0009] As described above, when measuring the resistivity of a silicon wafer, by performing a donor killer heat treatment, the influence of thermal donors can be suppressed and the resistivity of the silicon wafer can be correctly measured.

[0010] However, in the conventional resistivity measurement method, even after performing a donor killer heat treatment, thermal donors may not disappear, and there are cases where the true resistivity cannot be correctly measured, and improvement is required.

[0011] Therefore, an object of the present invention is to provide a method for measuring the resistivity of a silicon wafer capable of correctly measuring the true resistivity after a donor killer heat treatment and a quartz jig used therefor. [Means for Solving the Problems]

[0012] As a result of the intensive research by the inventor of the present application, it has been clarified that the reason why the true resistivity of the wafer sample after the donor killer heat treatment cannot be correctly measured is that the sample has a tendency to cool slowly due to the residual heat from the quartz jig that holds the sample during the cooling process after the donor killer heat treatment, and thermal donors are regenerated. A sample of a silicon wafer is introduced into a heat treatment furnace while being placed on a quartz jig, and after the heat treatment, it is taken out of the heat treatment furnace together with the quartz jig and cooled. Here, in the region away from the quartz jig, the wafer is rapidly cooled and thermal donors are not regenerated, but in the region close to the quartz jig, the wafer is not rapidly cooled due to the influence of the residual heat of the quartz jig, and thermal donors are likely to be regenerated. Therefore, the true resistivity of the silicon wafer cannot be correctly measured.

[0013] The present invention is based on such technical findings. The method for measuring the resistivity of a silicon wafer according to the present invention comprises a sample preparation step of cutting the silicon wafer in the radial direction to prepare a sample, a sample introduction step of introducing the sample into a heat treatment furnace while placing the sample on a quartz jig such that the cutting line is horizontal and upward, a donor killer heat treatment step of heat-treating the sample in the heat treatment furnace to eliminate thermal donors, a cooling step of taking out the quartz jig on which the sample is placed from the heat treatment furnace and cooling it by blowing air from above the sample, and a resistivity measurement step of measuring the resistivity of the sample cooled in the cooling step by the four-probe method.

[0014] According to the present invention, a region where thermal donors are likely to be regenerated due to residual heat from the quartz jig can be excluded from the resistivity measurement region, and a region where thermal donors are unlikely to be regenerated can be used as the resistivity measurement region. That is, by distancing the resistivity measurement position as far as possible from the quartz jig, the resistivity based on the dopant concentration introduced during the growth of the silicon single crystal ingot can be correctly measured so that the resistivity of the silicon wafer falls within a predetermined range.

[0015] In the present invention, it is preferable that the resistivity measurement step measures the resistivity in a region above the midpoint of the height of the sample. This makes it possible to accurately measure the true resistivity of the silicon wafer without being affected by thermal donors regenerated during the cooling process after donor killer heat treatment.

[0016] The silicon wafer preferably has a diameter of 300 mm or more, and the resistivity measurement position is preferably 80 mm or more away from the contact point with the quartz jig. This allows for accurate measurement of the true resistivity of the silicon wafer without being affected by the regeneration of thermal donors.

[0017] The sample preferably has a quarter-circular shape. This allows multiple test samples to be prepared from a single silicon wafer, and enables the performance of multiple types of characteristic tests, including resistivity measurement.

[0018] The cutting line has a first cutting line and a second cutting line perpendicular to the first cutting line, and it is preferable that the sample is placed on the quartz jig such that the first cutting line is horizontal and upward and the second cutting line is vertical. When a semicircular sample is placed on the quartz jig with the cutting line horizontal and downward, the sample is susceptible to the regeneration of thermal donors near the right-angle corners, resulting in a large error in the true resistivity. Also, when a semicircular sample is placed on the quartz jig with the right-angle corners facing upward and the arc-shaped outer periphery facing downward, the resistivity measurement near the outer periphery of the sample is susceptible to the regeneration of thermal donors, resulting in a large error in the true resistivity. However, when a semicircular sample is placed on the quartz jig with the cutting line horizontal and upward, the radial resistivity distribution of the silicon wafer can be measured accurately and stably without being affected by the regeneration of thermal donors.

[0019] Preferably, the resistivity measurement points include a first measurement point set near the right-angle corner portion which is the intersection of the first cutting line and the second cutting line, a second measurement point set further outward than the first measurement point, and a third measurement point set further outward than the second measurement point and near the outer periphery. This makes it possible to accurately and stably measure the radial resistivity distribution of the silicon wafer without being affected by the regeneration of thermal donors.

[0020] Preferably, the first measurement point is at least 15 mm outside the right-angle corner, the second measurement point is at an intermediate radial position from the right-angle corner to the outermost circumference, and the third measurement point is at least 15 mm inside the outermost circumference. This improves the accuracy of resistivity measurement using the four-probe method.

[0021] The first measurement point is located on a first measurement line extending in a first radial direction from the right-angle corner, and the second and third measurement points are located on a second measurement line extending in a second radial direction different from the first radial direction from the right-angle corner. Preferably, the first angle formed by the first measurement line and the first cutting line is greater than the second angle formed by the second measurement line and the first cutting line. This allows for accurate and stable measurement of the radial resistivity distribution of the silicon wafer without being affected by the regeneration of thermal donors.

[0022] The sample has a shape slightly larger than a regular quarter circle, and the two mutually orthogonal cutting lines are set outside the line passing through the center of the wafer. The center of the wafer may be located inside the sample plane of the right-angle corner formed by the intersection of the two cutting lines. This allows the first measurement point set near the right-angle corner to coincide with the center of the wafer, and the first to third measurement points to be set on a single measurement line.

[0023] The resistivity of the silicon wafer is 1000 Ωcm to 5000 Ωcm, and the oxygen concentration of the silicon wafer is 6 × 10⁻¹⁶. 17 atoms / cm 3 ~15×10 17 atoms / cm 3 Preferably, the resistivity is less than 1000 Ωcm, or the oxygen concentration is 6 × 10⁻⁶. 17 atoms / cm 3 For silicon wafers below a certain thickness, even when using conventional quartz jigs, the thermal donors regenerated during the cooling process after donor killer heat treatment did not affect the resistivity of the silicon wafer. However, when the resistivity is 1000 Ωcm or higher and the oxygen concentration is 6 × 10⁻⁶ 17 atoms / cm 3 When the resistivity of the silicon wafers described above was measured, the measurement results deviated from the target resistivity. According to the present invention, even with silicon wafers that have high resistivity and high oxygen concentration, the target resistivity based on the dopant concentration added during the growth of the silicon single crystal ingot can be accurately measured without being affected by the thermal donor.

[0024] Furthermore, the present invention relates to a quartz jig for holding a semicircular sample upright, obtained by cutting a silicon wafer radially, comprising: first and second sample holding portions that contact one end and the other end of the sample in the width direction, respectively; and third and fourth sample holding portions located below the first and second sample holding portions and that contact one end and the other end of the sample in the width direction, respectively, wherein the position of the third sample holding portion in the width direction is equal to that of the first sample holding portion, and the distance in the width direction between the first and second sample holding portions is wider than the distance in the width direction between the third and fourth sample holding portions.

[0025] According to the present invention, a quarter-circular sample can be placed on a quartz jig such that one of its cutting lines is horizontal and upward. This ensures that a region unaffected by thermal donor regeneration is maintained throughout the entire radial direction from the center to the outer edge of the wafer. Therefore, the true resistivity of the silicon wafer can be accurately measured without being affected by thermal donor regeneration.

[0026] In the present invention, it is preferable that the diameter of the silicon wafer is 300 mm or more, and the shortest distance in the height direction from the upper end of the sample placed on the quartz jig to the first and second sample holding parts is 80 mm or more. This makes it possible to accurately measure the true resistivity of the silicon wafer without being affected by the regeneration of thermal donors. [Effects of the Invention]

[0027] According to the present invention, it is possible to provide a method for measuring the resistivity of a silicon wafer that can accurately measure the true resistivity after donor killer heat treatment, and a quartz jig used therefor. [Brief explanation of the drawing]

[0028] [Figure 1] Figure 1 is a flowchart showing a method for measuring the resistivity of a silicon wafer according to an embodiment of the present invention. [Figure 2] Figure 2 is a schematic diagram showing the sample preparation process. [Figure 3] Figure 3 is a schematic diagram showing the sample introduction process. [Figure 4] Figure 4 shows the configuration of a quartz jig used during the heat treatment of a sample, where (a) is a front view and (b) is a side view. [Figure 5] Figure 5 is an explanatory diagram of the problems with donor killer heat treatment using conventional quartz jigs, where (a) and (b) are the conventional methods of holding samples using quartz jigs, and (c) is a graph showing the resistivity distribution of the sample. [Figure 6] Figure 6 is a schematic diagram showing the cooling process. [Figure 7] Figure 7 is a schematic diagram showing a modified sample for testing. [Figure 8] Figure 8 is a schematic diagram illustrating a method for evaluating the resistivity of a silicon wafer. [Figure 9] Figure 9 is a graph showing the measurement results of the resistivity of the first to third samples. [Figure 10] Figure 10 is a graph showing the relationship between the oxygen concentration of a silicon wafer sample and the degree of influence of the thermal donor, for each sample resistivity. [Figure 11] Figures 11(a) and (b) are graphs showing the relationship between the oxygen concentration of silicon wafer samples and the degree of influence of thermal donors for each sample resistivity. [Figure 12] Figure 12 is a graph showing the relationship between the distance from the sample holding part of the quartz jig to the resistivity measurement point and the degree of influence of the thermal donor. [Modes for carrying out the invention]

[0029] Preferred embodiments of the present invention will be described in detail below with reference to the attached drawings.

[0030] Figure 1 is a flowchart showing a method for measuring the resistivity of a silicon wafer according to an embodiment of the present invention.

[0031] As shown in Figure 1, the resistivity measurement method for a silicon wafer according to this embodiment comprises a sample preparation step S11 in which a silicon wafer is cut radially to prepare a sample for inspection, a sample introduction step S12 in which the sample is placed on a quartz jig and introduced into a heat treatment furnace so that the cut line is horizontal and facing upward, a donor killer heat treatment step S13 in which the sample is heat-treated in the heat treatment furnace to eliminate thermal donors, a cooling step S14 in which the sample is removed from the heat treatment furnace together with the quartz jig and cooled, and a resistivity measurement step S15 in which the resistivity of the sample cooled in the cooling step S14 is measured using the four-probe method.

[0032] Figure 2 is a schematic diagram showing the sample preparation process S11.

[0033] As shown in Figure 2, the silicon wafer 12 is made by slicing a single-crystal silicon ingot 10 grown by the CZ method. The single-crystal silicon ingot 10 has a top portion 10a in which the crystal diameter gradually increases, a straight body portion 10b in which the crystal diameter is constant, and a tail portion 10c in which the crystal diameter gradually decreases.

[0034] In processing the single-crystal silicon ingot 10, first, the straight body portion 10b of the ingot is cut radially to cut out the silicon wafer 12. Specifically, the top portion 10a and tail portion 10c of the ingot are cut off to process it into a cylindrical shape, and then the outer circumference is ground to make it slightly larger in diameter than the product wafer. Next, a block cutting process in which silicon blocks 11 are cut from the cylindrical ingot (straight body portion 10b) and a wafer cutting process in which silicon wafers 12 are cut alternately, thereby cutting out multiple silicon blocks 11 from the ingot and cutting out silicon wafers 12 from both ends of each silicon block 11. Cutting machines such as band saws, inner blades, and outer blades are used to cut out the silicon blocks 11 and silicon wafers 12. In the wafer cutting process, two or more silicon wafers 12 may be cut out in succession.

[0035] Next, each silicon wafer 12 is diced to produce quarter-size inspection samples 13. While four divisions are preferred for the silicon wafer 12, the number of divisions is not particularly limited and any number of divisions is acceptable. One of the samples 13 thus produced from the wafer is used to measure the resistivity, a quality indicator of the silicon wafer 12. The other samples 13 can be used to evaluate oxygen concentration, carrier recombination lifetime, crystal defects, etc.

[0036] The diameter of the silicon wafer 12 is preferably 300 mm or more. This is because in the case of a large-diameter silicon wafer with a diameter of 300 mm or more, in-plane variations in the temperature of the wafer are likely to occur during the cooling process after the donor killer heat treatment, and problems with the regeneration of thermal donors are likely to occur.

[0037] The resistivity of the silicon wafer 12 is preferably 1000 Ωcm or more and 5000 Ωcm or less, and the oxygen concentration is 6×10 17 atoms / cm 3 or more and 15×10 17 atoms / cm 3 or less. This is because a silicon wafer 12 with a high resistivity and a high oxygen concentration is easily affected by thermal donors, and measurement results deviated from the true resistivity are likely to be obtained.

[0038] In the sample preparation step S11, it is preferable to divide the silicon wafer 12 into four parts to produce a sample 13 having a semi-circular shape. The sample 13 having a semi-circular shape has a first cutting line E1 passing through the center of the wafer and a second cutting line E2 passing through the center of the wafer and orthogonal to the first cutting line E1. The intersection of the first cutting line E1 and the second cutting line E2 forms a right-angle corner portion C0. Further, the intersection of the first cutting line E1 and the arc E R forms a first outer peripheral corner portion C1, and the intersection of the second cutting line E2 and the arc E R forms a second outer peripheral corner portion C2.

[0039] It is preferable to grind the sample 13 after dicing the wafer to 1 / 4 size. The surface of the wafer cut with a band saw has a rough cut surface with deep processing scratches and irregularities. Therefore, when measuring the resistivity of sample 13, for example, using the four-probe method, the poor surface shape causes variability in the measured values. However, if sample 13 is flattened by grinding, the processing scratches and irregularities (undulations) on the surface can be removed, improving the accuracy of resistivity measurement. Grinding is a type of machining process in which a high-speed rotating grinding wheel is pressed against the workpiece surface to remove the surface layer and obtain a smooth surface. It is preferable to perform single-sided surface grinding, which grinds one side at a time, but it is not particularly limited as long as it is a machining process that can flatten the surface of the sample, and it may also be double-sided surface grinding, which grinds both sides simultaneously.

[0040] Figure 3 is a schematic diagram showing the sample introduction process S12.

[0041] As shown in Figure 3, the sample introduction step S12 involves introducing the sample 13 into the heat treatment furnace 30 while it is placed on the quartz jig 20. The quartz jig 20 can accommodate multiple samples 13, and the heat treatment furnace 30 is preferably a muffle furnace, allowing for simultaneous heat treatment (batch processing) of multiple samples 13.

[0042] Figure 4 shows the configuration of a quartz jig used during the heat treatment of a sample, where (a) is a front view and (b) is a side view.

[0043] As shown in Figure 4, the quartz jig 20 is a quartz holding member (quartz boat) that holds the upright position of the semicircular sample 13, and has a frame 21 consisting of a combination of legs and beams, and first to fourth sample holding parts 22a to 22d fixed to the frame 21.

[0044] The first to fourth sample holding sections 22a to 22d are substantially rod-shaped members that extend in a direction perpendicular to the main surface (Y direction) of the sample 13 placed on the quartz jig 20. The first and second sample holding sections 22a and 22b contact one end and the other end of the sample 13 in the width direction (X direction), respectively. The third and fourth sample holding sections 22c and 22d are located below the first and second sample holding sections 22a and 22b and also contact one end and the other end of the sample 13 in the width direction, respectively.

[0045] In this embodiment, the height of the first sample holding section 22a is equal to that of the second sample holding section 22b, and the height of the third sample holding section 22c is equal to that of the fourth sample holding section 22d. The widthwise position of the third sample holding section 22c is equal to that of the first sample holding section 22a, and the widthwise distance from the first sample holding section 22a to the second sample holding section 22b is wider than the widthwise distance from the third sample holding section 22c to the fourth sample holding section 22d.

[0046] Sample 13 is placed on the quartz jig 20 such that the first cutting line E1 is horizontal and upward, the first sample holder 22a and the third sample holder 22c are in contact with the second cutting line E2 of sample 13, and the second sample holder 22b and the fourth sample holder 22d are in contact with the arc E of sample 13 R It contacts the outer periphery. The sample 13 on the quartz jig 20 is maintained in such a position that the first cutting line E1 is horizontal and upward, and the second cutting line E2 is vertical.

[0047] Figure 5 is an explanatory diagram of the problems with donor killer heat treatment using conventional quartz jigs, where (a) and (b) show the conventional method of holding the sample using a quartz jig, and (c) is a graph showing the radial resistivity distribution of a sample in which the resistivity of the silicon wafer 12 is targeted to be 2000 Ωcm.

[0048] As shown in Figure 5(a), the right-angle corner C0 is pointing straight up and the arc E is facing upwards. RWhen sample 13 is placed on the quartz jig 20 with its outer edge facing directly downwards, the outer edge of sample 13 approaches the quartz jig 20, making it easier for thermal donors to regenerate near the outer edge of sample 13. As a result, as shown in Figure 5(c), the error in the true resistivity becomes larger near the outer edge of the wafer.

[0049] Furthermore, as shown in Figure 5(b), one cutting line E1 is vertical, and the other cutting line E2 is horizontal and downward, forming an arc E R When sample 13 is placed on the quartz jig 20 with its tip angled upwards, the right-angle corner C0 approaches the quartz jig 20, making it easier for thermal donors to regenerate near the right-angle corner C0. As a result, as shown in Figure 5(c), the error in the true resistivity becomes larger near the center of the wafer.

[0050] In contrast, as shown in Figure 4, when the sample 13 is placed on the quartz jig 20 such that the first cutting line E1 is horizontal and upward and the second cutting line E2 is vertical, the upper region of the sample 13 close to the first cutting line E1 is far from the quartz jig 20. Therefore, by setting measurement points P1, P2, P3, ... within this region, the radial resistivity distribution of the silicon wafer can be accurately and stably measured without being affected by thermal donors regenerated during the cooling process after donor killer heat treatment.

[0051] Single-crystal silicon grown by the CZ method contains supersaturated oxygen, and when heat-treated at a low temperature of around 450°C, several oxygen atoms condense to form oxygen clusters, which become thermal donors that release electrons, thus reducing the accuracy of oxygen concentration measurement. Therefore, a donor killer heat treatment process S13 is performed to eliminate the thermal donors by heat-treating sample 13 in a heat treatment furnace 30. The heating temperature of sample 13 is preferably 650-700°C, and the heating time is preferably 20-40 minutes.

[0052] Figure 6 is a schematic diagram showing the cooling process S14.

[0053] As shown in Figure 6, in the cooling step S14, after removing the sample 13, along with the quartz jig 20, from the heat treatment furnace, the sample 13 is rapidly cooled by blowing cold air from above. Although the thermal donors are eliminated by the above-mentioned donor killer heat treatment, if the sample stays in the 400-500°C temperature range for a long time during the cooling process, thermal donors will regenerate. Therefore, by rapidly cooling the sample and minimizing the stay in the 400-500°C temperature range, the regeneration of thermal donors can be prevented.

[0054] Even if sample 13 is rapidly cooled by applying cold air from above, the lower region close to the quartz jig 20 tends to cool slowly due to residual heat from the quartz jig 20, resulting in a longer stay in the 400-500°C temperature range and making it easier for thermal donors to regenerate.

[0055] However, in this embodiment, the resistivity measurement area is set in the upper region of the main surface of the sample 13, as far away from the quartz jig as possible, specifically, in the region above the midpoint in the height direction of the sample 13 placed in an upright position on the quartz jig 20. This allows for accurate measurement of the true resistivity without being affected by the regenerated thermal donor.

[0056] The resistivity measurement step S15 measures the radial resistivity distribution of the silicon wafer using the four-probe method. As shown in Figures 3 and 5, it is preferable that the resistivity measurement points include a first measurement point P1 located near the right-angle corner C0 of the sample 13, which corresponds to the center of the wafer; a second measurement point P2 located at the midpoint of the radial direction of the sample 13; and a third measurement point P3 located near the outer edge of the sample 13. This allows for the determination of the radial resistivity distribution of the silicon wafer. Since the resistivity tends to change near the outer edge of the wafer, it is preferable to measure the resistivity more precisely near the outer edge.

[0057] As described above, the resistivity measurement point is set as far away as possible from the contact point with the quartz jig 20. Specifically, the shortest distance from the contact point with the quartz jig 20 to the resistivity measurement point is preferably 80 mm or more. This allows for accurate measurement of the true resistivity of the silicon wafer without being affected by the thermal donor regenerated in the cooling process S14.

[0058] In the resistivity measurement step S15, it is preferable to set up multiple measurement lines extending radially from the right-angle corner C0 of the sample 13 and measure the resistivity along these measurement lines. Here, the measurement points on the multiple measurement lines are located within a measurement area unaffected by the thermal donor described above and are at least 80 mm away from the contact point with the quartz jig 20.

[0059] In this embodiment, the plurality of measurement lines include a first measurement line L1 extending in the first radial direction and a second measurement line L2 extending in the second radial direction, wherein the angle θ1 (first angle) of the first measurement line L1 with respect to the first cutting line E1 is greater than the angle θ2 (second angle) of the second measurement line L2 with respect to the first cutting line E1. Here, the first measurement point P1 is a point on the first measurement line L1, and the second measurement point P2 and the third measurement point P3 are points on the second measurement line L2. The first measurement point P1 is a point near the right-angle corner C0 (center of the wafer), at least 15 mm away from the right-angle corner C0. The second measurement point P2 is set on the second measurement line L2 and is a point located at an intermediate radial position from the right-angle corner C0 to the outermost circumference. The third measurement point P3 is set on the second measurement line L2 and is a point near the outer circumference, at least 15 mm away from the outermost circumference.

[0060] The four-probe method for measuring resistivity assumes that the measurement sample is an infinite plane. However, actual measurement samples are not infinite planes. Therefore, a "shape correction formula" is set according to the shape of the measurement sample to correct the measured resistivity. Thus, when dealing with a finite region, resistivity correction is necessary. However, by moving the resistivity measurement point slightly away from the edge of sample 13, the region can be brought closer to a finite region where resistivity correction is possible, and correcting to the correct resistivity is possible regardless of the measurement position within the sample surface.

[0061] Figure 7 is a schematic diagram showing a modified sample for testing.

[0062] As shown in Figure 7, this inspection sample 13 is not simply a circular wafer divided into four parts, but has a shape slightly larger than a regular quarter circle. The two mutually orthogonal cutting lines E1 and E2 are set slightly outside the line passing through the wafer center P0. Therefore, the lengths of the cutting lines E1 and E2 are greater than the radius of the wafer. Also, the intersection of the first cutting line E1 and the second cutting line E2 does not coincide with the wafer center P0, and the wafer center P0 is located inside the sample surface beyond the right-angle corner C0. This allows the first measurement point P1 to be set at the wafer center P0, and the first to third measurement points P1 to P3 can be set on a single measurement line (second measurement line L2). Furthermore, by keeping the wafer center P0 as far away as possible from the cutting lines E1 and E2, the resistivity of the wafer center P0 can be determined without using a shape correction formula.

[0063] If the resistivity measurement position is close to the contact point with the quartz jig 20, thermal donors are easily regenerated by residual heat from the quartz jig 20 during the cooling process S14 after the donor killer heat treatment. Measuring the resistivity affected by these thermal donors may result in a measurement value that deviates from the true resistivity. However, as in this embodiment, if the resistivity is measured at a position as far away as possible from the contact point with the quartz jig 20, the true resistivity, which is not affected by the regenerated thermal donors, can be determined, thereby improving the accuracy of resistivity measurement for high-resistivity silicon wafers.

[0064] As described above, the resistivity measurement method for a silicon wafer according to this embodiment involves placing the first cutting line E1 of a semicircular sample 13 cut from a silicon wafer 12 horizontally and upward on a quartz jig 20, performing donor killer heat treatment, and then measuring the resistivity in the upper region in the height direction of the sample 13. Therefore, the resistivity can be measured without being affected by thermal donors regenerated after the donor killer heat treatment. Consequently, the resistivity of only the dopants of a high-resistivity silicon wafer can be accurately measured.

[0065] Although preferred embodiments of the present invention have been described above, it goes without saying that the present invention is not limited to the above embodiments, and various modifications are possible without departing from the spirit of the invention, and these modifications are also included within the scope of the present invention. [Examples]

[0066] Resistivity is 5000 Ωcm, oxygen concentration is 14.8 × 10⁻⁶ 17 atoms / cm 3Ten p-type silicon wafers with a diameter of 300 mm were prepared under the following conditions, and each silicon wafer was divided into four sections to create quarter-circular samples. As shown in Figure 8, of the four quarter-circular samples obtained by dividing one wafer into four sections, the first to third samples were used for resistivity measurement, and the fourth sample was used for oxygen concentration measurement by FTIR (ASTM F121, 1979). As a result of the oxygen concentration measurement, the oxygen concentration of the silicon wafer was found to be 14.8 × 10⁻⁶. 17 atoms / cm 3 I confirmed that this was the case.

[0067] For the first sample, donor killer heat treatment and cooling were performed using the quartz jig according to the present invention shown in Figure 4. Specifically, the sample was subjected to donor killer heat treatment using a quartz jig capable of holding the sample with one cut line horizontal and facing upward. After that, the sample, along with the quartz jig, was removed from the heat treatment furnace and cooled by applying cold air from above. After the sample cooled, the resistivity at three locations—the center of the wafer, the R / 2 portion, and the outer edge—wafer was determined using the four-probe method. At this time, as shown in Figure 8, the resistivity measurement points were set in the vicinity of one cut line, which is less susceptible to the regeneration of thermal donors.

[0068] For the second sample, donor killer heat treatment and cooling were performed using the conventional quartz jig shown in Figure 5(a). Specifically, the sample was subjected to donor killer heat treatment using a conventional quartz jig capable of holding the sample with the right-angle corners facing directly upwards and the arc facing directly downwards. After that, the sample, along with the quartz jig, was removed from the heat treatment furnace and cooled by applying cold air from above. After the sample cooled, the resistivity at three locations—the center of the wafer, the R / 2 portion, and the outer edge—wafer was determined using the four-probe method. At this time, as shown in Figure 8, the resistivity along the measurement line extending directly downwards from the center of the wafer was measured.

[0069] For the third sample, in order to confirm the true resistivity of the silicon wafer, 20 x 20 mm square sample pieces were cut from three locations: the center, the R / 2 portion, and the outer edge of the semicircular sample. Processing scratches were removed by etching, and donor killer heat treatment was performed. After that, the sample pieces were immersed in pure water and forcibly rapidly cooled, resulting in a very short stay at 400-500°C. During the water cooling process, some of the sample pieces shattered as soon as they were immersed in pure water, so the resistivity of the unshattered sample pieces was determined using the four-probe method.

[0070] Figure 9 is a graph showing the resistivity measurement results for the first to third samples, with the horizontal axis representing the distance from the wafer center (mm) and the vertical axis representing the resistivity (Ωcm). The resistivity values ​​in the graph are the average of the resistivity measurements obtained from each of the 10 wafers.

[0071] As shown in Figure 9, assuming that the resistivity distribution of the third sample, which underwent water cooling, is closest to the true value (i.e., there is no change in resistivity due to the thermal donor), the resistivity distribution of the second sample deviates significantly from that of the third sample, indicating a change in resistivity due to the thermal donor. In contrast, the resistivity distribution of the first sample is almost identical to that of the third sample, and no change in resistivity due to the thermal donor was observed.

[0072] To confirm the temperature during sample cooling, the temperature at the resistivity measurement point of the sample after donor killer heat treatment using the quartz jig according to the present invention was measured with a contact thermometer. As a result, the residence time in the temperature range of 400°C to 500°C was approximately 5 seconds.

[0073] Based on the above results, the degree of influence of thermal donors on silicon wafers when donor killer heat treatment is performed using the quartz jig according to the present invention (degree of thermal donor influence) was determined by simulation. The degree of thermal donor influence is calculated as Δρ = {(ρ2-ρ1) / ρ1} × 100, where ρ1 is the target resistivity and ρ2 is the resistivity after donor killer heat treatment.

[0074] Figure 10 is a graph showing the relationship between the oxygen concentration of a silicon wafer sample and the degree of influence of thermal donors for each sample resistivity, with the horizontal axis representing oxygen concentration (×10 17 atoms / cm 3 The vertical axis shows the impact of thermal donors (%).

[0075] As shown in Figure 10, the influence of thermal donors increased with increasing oxygen concentration and also with increasing target resistivity. In particular, when donor killer heat treatment was performed using the quartz jig according to the present invention, the resistivity was 5000 Ωcm or less and the oxygen concentration was 15 × 10⁻⁶ 17 atom / cm 3 It was found that the influence of thermal donors was 1.5% or less under the following conditions. Considering the measurement accuracy of the resistivity meter, 1.5% or less is within an acceptable range, and it can be said that resistivity is not affected by thermal donors.

[0076] Next, we simulated the degree of influence that thermal donors have on silicon wafers when donor killer heat treatment is performed using a conventional quartz jig (the degree of influence of thermal donors).

[0077] Figures 11(a) and (b) are graphs showing the relationship between the oxygen concentration of silicon wafer samples and the degree of influence of thermal donors for each sample resistivity, with the horizontal axis representing oxygen concentration (×10). 17 atoms / cm 3 The vertical axis shows the degree of influence of the thermal donor (%). Figure 11(a) shows the oxygen concentration ranges from 0 to 20 × 10 17 atoms / cm 3 The range is broad, and Figure 11(b) shows the range where the oxygen concentration is 0 to 10 × 10 17 atoms / cm 3 It is shown only within that range.

[0078] As shown in Figures 11(a) and (b), the effect of thermal donors increases with increasing oxygen concentration in the silicon wafer, especially when the oxygen concentration is 6 × 10⁻⁶. 17atoms / cm 3 It was found that the effect of thermal donors exceeds 1.5% under the above conditions. Furthermore, it was found that the influence of thermal donors increases with higher resistivity of the silicon wafer, and particularly exceeds 1.5% when the resistivity is 1000 Ωcm or higher. Considering the measurement accuracy of the resistivity measuring instrument, if the influence of thermal donors exceeds 1.5%, it can be said that the resistivity is being affected by thermal donors, and therefore improvement is necessary.

[0079] Next, the degree to which the distance from the sample holding part of the quartz jig to the resistivity measurement point affects the accuracy of resistivity measurement was evaluated by simulation. The resistivity of the silicon wafer was 4600 Ωcm, and the oxygen concentration was 15 × 10⁻¹⁶. 17 atoms / cm 3 That's what I decided.

[0080] Figure 12 is a graph showing the relationship between the distance from the sample holding part of the quartz jig to the resistivity measurement point and the degree of influence of thermal donors. The horizontal axis represents the distance from the sample holding part of the quartz jig to the resistivity measurement point (mm), and the vertical axis represents the degree of influence of thermal donors (%).

[0081] As shown in Figure 12, it was found that the influence of thermal donors decreases as the resistivity measurement point moves further away from the sample holding part of the quartz jig. In particular, it was found that the influence of thermal donors could be reduced to 1.5% or less if the resistivity measurement point was 80 mm or more away from the sample holding part of the quartz jig. [Explanation of Symbols]

[0082] 10 Single-crystal silicon ingots 10a Top section 10b Straight body part 10c Tail section 11 Silicone Blocks 12. Silicon wafer (sample wafer) 13. Sample for testing 20 Quartz jigs 21 frames 22a First sample holding section 22b Second sample holding section 22c Third sample holding section 22d Fourth sample holding section 30 Heat treatment furnace C0 Right-angle corner C1 First Outer Corner Section C2 Second Outer Corner Section E1 First Cutting Line E2 Second Cutting Line E R Arc (outer circumference) L1 First Measurement Line L2 Second Measurement Line First measurement point of P1 resistivity Second measurement point of P2 resistivity Third measurement point of P3 resistivity S11 Sample preparation process S12 Sample Introduction Process S13 Donor Killer Heat Treatment Process S14 Cooling process S15 Resistivity measurement process

Claims

1. A sample preparation process involves cutting a silicon wafer radially to create a sample, A sample introduction step involves introducing the sample into a heat treatment furnace while it is placed on a quartz jig so that the cutting line is horizontal and facing upwards. A donor killer heat treatment step is performed to eliminate thermal donors by heat-treating the sample in the heat treatment furnace, A cooling step is performed in which the quartz jig on which the sample is placed is removed from the heat treatment furnace and the sample is cooled by blowing air from above, A method for measuring the resistivity of a silicon wafer, characterized by comprising a resistivity measurement step of measuring the resistivity of the sample cooled in the cooling step using a four-probe method.

2. The resistivity measurement step is to measure the resistivity in a region above the midpoint of the height of the sample, as described in claim 1.

3. The diameter of the aforementioned silicon wafer is 300 mm or more. The resistivity measurement position for a silicon wafer according to claim 1, wherein the resistivity measurement position is located at a distance of 80 mm or more from the contact point with the quartz jig.

4. The method for measuring the resistivity of a silicon wafer according to claim 1, wherein the sample has a semicircular shape.

5. The method for measuring the resistivity of a silicon wafer according to claim 4, wherein the cutting line has a first cutting line and a second cutting line perpendicular to the first cutting line, and the wafer is placed on the quartz jig such that the first cutting line is horizontal and upward and the second cutting line is vertical.

6. The resistivity measurement method for a silicon wafer according to claim 5, wherein the resistivity measurement points include a first measurement point set near a right-angle corner portion which is the intersection of the first cutting line and the second cutting line, a second measurement point set further outward than the first measurement point, and a third measurement point set further outward than the second measurement point and near the outer periphery.

7. The first measurement point is a point at least 15 mm outside the right-angle corner, The second measurement point is an intermediate position in the radial direction from the right-angle corner to the outermost circumference, The method for measuring the resistivity of a silicon wafer according to claim 6, wherein the third measurement point is a point at least 15 mm inward from the outermost edge.

8. The first measurement point is located on a first measurement line extending radially from the right-angle corner portion. The second and third measurement points are located on a second measurement line extending from the right-angle corner portion in a second radial direction different from the first radial direction. The method for measuring the resistivity of a silicon wafer according to claim 7, wherein the first angle formed by the first measurement line and the first cutting line is greater than the second angle formed by the second measurement line and the first cutting line.

9. The resistivity of the aforementioned silicon wafer is 1000 Ωcm or more and 5000 Ωcm or less. The oxygen concentration of the silicon wafer is 6 × 10 １７ atoms / cm ３ The above 15 x 10 １７ atoms / cm ３ The method for measuring the resistivity of a silicon wafer according to claim 1 is as follows:

10. A quartz jig for holding upright a semicircular sample obtained by cutting a silicon wafer radially, First and second sample holding portions that contact one end and the other end of the sample in the width direction, respectively, It has third and fourth sample holding portions located below the first and second sample holding portions, and which contact one end and the other end of the sample in the width direction, respectively. The position of the third sample holding portion in the width direction is equal to that of the first sample holding portion. A quartz jig characterized in that the distance in the width direction from the first sample holding portion to the second sample holding portion is wider than the distance in the width direction from the third sample holding portion to the fourth sample holding portion.

11. The diameter of the aforementioned silicon wafer is 300 mm or more. The quartz jig according to claim 10, wherein the shortest distance in the height direction from the upper end of the sample placed on the quartz jig to the first and second sample holding parts is 80 mm or more.

Images