Method for evaluating semiconductor wafer and method for manufacturing semiconductor wafer

Abstract

Provided is a new evaluation method capable of evaluating surface roughness of a semiconductor wafer with high precision. In one or more measurement regions of a surface of an evaluation target semiconductor wafer, a maximum height Sz and a roughness parameter other than Sz are obtained by performing measurement using a roughness measurement device; and for each measurement region, it is determined whether the value of the roughness parameter other than Sz obtained in the measurement region is a value that can be used in evaluation of the evaluation target semiconductor wafer, using the value of the maximum height Sz as an index.

Description

Technical Field

[0001] This invention relates to a method for evaluating semiconductor wafers and a method for manufacturing semiconductor wafers. Background Technology

[0002] In the manufacturing process of semiconductor wafers (hereinafter also referred to as "wafers"), it is common practice to conduct inspections for process management and to perform pre-shipment sampling inspections from batches for product quality management. Surface roughness is one such inspection item (see, for example, Patent Document 1).

[0003] Existing technical documents

[0004] Patent documents

[0005] Patent document 1: Japanese Patent Application Publication No. 2006-278972. Summary of the Invention

[0006] The problem that the invention aims to solve

[0007] Semiconductor wafers are generally manufactured by performing various processes, such as surface polishing, on wafers cut from ingots. By measuring the surface roughness of such manufactured semiconductor wafers, it is possible to evaluate, for example, whether the surface polishing process was performed appropriately, based on the measurement results. Moreover, by managing the surface polishing process based on the evaluation results and / or by distinguishing between qualified and unqualified batches, high-quality product wafers can be supplied to the market. However, the surface roughness measurement results may include various noises (see paragraph 0006 of Patent Document 1). To improve the accuracy of the evaluation based on the surface roughness measurement results, it is preferable to obtain surface roughness measurement results with reduced noise effects. In this regard, Patent Document 1 proposes a method that involves measuring the surface shape of a semiconductor wafer under at least two different measurement conditions, converting the data of each surface shape into a power spectrum, identifying peaks with inconsistent peak spatial frequencies in the power spectrum as noise components, removing the identified noise components from at least one power spectrum, and then performing an evaluation. However, the inventors believe that in such a method, the removable noise is limited to frequency-dependent noise (specifically, device-related noise), and cannot remove noise from external interference factors that are not frequency-dependent. However, from the viewpoint of evaluating the surface roughness of semiconductor wafers with higher accuracy, it is preferable to be able to reduce the influence of various types of noise.

[0008] In view of the above, one objective of the present invention is to provide a new evaluation method that can evaluate the surface roughness of semiconductor wafers with high precision.

[0009] Solution for solving the problem

[0010] One aspect of the present invention relates to a method for evaluating a semiconductor wafer (hereinafter also referred to as the "evaluation method"), comprising:

[0011] In one or more measurement areas on the surface of the semiconductor wafer being evaluated, roughness parameters, including the maximum height Sz and those other than Sz, are obtained by measurement using a roughness measuring device; and

[0012] For each measurement area, the value of the maximum height Sz is used as an indicator to determine whether the roughness parameter values ​​other than Sz obtained in that measurement area are applicable values ​​for evaluating the semiconductor wafer being evaluated.

[0013] In one embodiment, the evaluation method described above may include: when the value of the maximum height Sz is below or less than a preset threshold, using the value of a roughness parameter other than Sz obtained for one or more measurement areas where the maximum height Sz is obtained as the evaluation value for the semiconductor wafer being evaluated.

[0014] In one approach, in the aforementioned evaluation method, if the value of the maximum height Sz is above or exceeds a predetermined threshold, the roughness parameter value other than Sz obtained for the measurement area where the maximum height Sz is obtained is not used as the evaluation value for the semiconductor wafer being evaluated.

[0015] In one embodiment, the evaluation method may further include: if the value of the maximum height Sz is above or exceeds a preset threshold, then for the measurement area where the maximum height Sz has been obtained, the roughness measurement device is used again for measurement.

[0016] In one embodiment, the evaluation method may further include: when the value of the maximum height Sz is above or exceeds a preset threshold, performing a roughness measurement in a region outside the measurement area where the maximum height Sz of that value is obtained.

[0017] In one embodiment, the evaluation method may further include: recalculating the values ​​of one or more roughness parameters obtained for the measurement area where the maximum height Sz is obtained, if the value of the maximum height Sz is above or exceeds a preset threshold.

[0018] In one approach, in the aforementioned evaluation method, the recalculation can be performed by excluding the portion of the roughness parameter calculation object that is estimated to be the cause of outliers in the maximum height Sz.

[0019] In one approach, the roughness parameter other than Sz can be one or more roughness parameters selected from the group consisting of the arithmetic mean height Sa and the root mean square height Sq.

[0020] In one embodiment, the roughness measuring device described above can be an optical interference microscope.

[0021] In one embodiment, the aforementioned optical interference microscope can be a white interference microscope.

[0022] One aspect of the present invention relates to a method for manufacturing a semiconductor wafer (hereinafter referred to as "manufacturing method 1"), comprising:

[0023] Manufacturing evaluation semiconductor wafers under test manufacturing conditions;

[0024] The evaluation method described above is used to evaluate the semiconductor wafer manufactured above.

[0025] Based on the results of the above evaluation, the manufacturing conditions for which changes were imposed on the aforementioned test manufacturing conditions will be determined as the actual manufacturing conditions, or the aforementioned test manufacturing conditions will be determined as the actual manufacturing conditions; and

[0026] Semiconductor wafers are manufactured under the actual manufacturing conditions determined above.

[0027] In one approach, the manufacturing conditions for which the above-described changes are applied can include surface polishing conditions for semiconductor wafers.

[0028] One aspect of the present invention relates to a method for manufacturing a semiconductor wafer (hereinafter referred to as "manufacturing method 2"), comprising:

[0029] Manufacturing a batch of semiconductor wafers that comprises multiple semiconductor wafers;

[0030] One or more semiconductor wafers are extracted from the aforementioned batch of semiconductor wafers;

[0031] The extracted semiconductor wafers were evaluated using the aforementioned evaluation method; and

[0032] We plan to ship semiconductor wafers from the same batch as those evaluated as qualified products as described above.

[0033] Invention Effects

[0034] According to one aspect of the present invention, a new evaluation method is provided that can evaluate the surface roughness of semiconductor wafers with high precision. Attached Figure Description

[0035] Figure 1 This is an explanatory diagram illustrating the measurement principle of a white interference microscope;

[0036] Figure 2 An example of the change in interference light during fringe scanning in measurements using a white interference microscope is shown;

[0037] Figure 3 This is an example of the measurement results of the surface of a silicon wafer using a white interference microscope image. Figure 3 (a) shows a white interference microscope image. Figure 3 (b) Shows a high-level profile;

[0038] Figure 4 This is a white interference microscope image of a silicon wafer surface containing a foreign object with a height of about 200 nm within the field of view. Figure 4 (a) is a white interference microscope image of the whole. Figure 4 (b) is a magnified image of the foreign object attachment site. Figure 4 (c) is a high-resolution profile of the foreign object. Figure 4 The table shown in (d) records the data from... Figure 4 Roughness parameters calculated from the overall image shown in (a);

[0039] Figure 5 The measurement procedure in the embodiments is shown;

[0040] Figure 6 An example of the evaluation results obtained in the embodiments is shown;

[0041] Figure 7 A graph is shown showing the arithmetic mean of the values ​​of Sq obtained for each of the 25 silicon wafers evaluated in the embodiment. Detailed Implementation

[0042] Evaluation methods for semiconductor wafers

[0043] One aspect of the present invention relates to a method for evaluating a semiconductor wafer, comprising: obtaining a maximum height Sz and roughness parameters other than Sz by measuring in one or more measurement areas on the surface of the semiconductor wafer to be evaluated using a roughness measuring device; and determining, for each measurement area, whether the value of the roughness parameter other than Sz obtained in that measurement area is an applicable value for evaluating the semiconductor wafer to be evaluated (hereinafter also referred to as "applicability determination"), using the value of the maximum height Sz as an indicator.

[0044] One of the two surfaces of a semiconductor wafer is the surface that forms the device (front side), and the other is the surface opposite the front side (back side). In this invention and this specification, the semiconductor wafer surface refers to either or both of the front and back sides.

[0045] In this invention and this specification, various roughness parameters can be referred to ISO 25178-2:2012 and the corresponding Japanese Industrial Standard JIS B0681-2:2018. The maximum height Sz, one of the roughness parameters, is expressed as "Sz = Sp + Sv". Sp is the maximum peak height (the maximum height from the average surface), and Sv is the maximum valley depth (the absolute value of the minimum height from the average surface). Sz, calculated as the sum of these, can be described as the distance from the highest point to the lowest point of the surface of the object being measured.

[0046] In the above evaluation method, in one or more measurement areas on the surface of the semiconductor wafer being evaluated, two or more roughness parameters, including the maximum height Sz, are obtained for each measurement area by using a roughness measuring device. Then, using the value of the maximum height Sz as an index, it is determined whether the values ​​of the roughness parameters other than Sz obtained in each measurement area are values ​​that can be used in the evaluation of the semiconductor wafer being evaluated. Regarding this, the inventors speculate as follows. However, the present invention is not limited to the inventors' speculation.

[0047] The inventors believe that the maximum height Sz, as described above, is the sum of the maximum peak height and the maximum valley depth, and is therefore easily affected by noise from device factors, and further by noise from external interference factors such as foreign object attachment. For example, according to the inventors' research, when foreign objects larger than 100 nm are attached within the measurement field of view, the values ​​of other roughness parameters (e.g., Sa, Sq, described later) change only by a few to tens of nm compared to normal conditions, while the value of Sz changes significantly by more than 100 nm. Furthermore, when sudden device vibration occurs, a height difference of more than 200 nm may occur on the wafer surface, which is normally impossible. In this case, the value of Sz is around 200 nm or more. For example, in the above cases, by setting a threshold of, for example, 100 nm for the value of Sz, the presence or absence of noise influence can be determined based on this threshold. Specifically, it is possible to determine whether the measurement results of the roughness parameters obtained together with Sz include abnormal values ​​caused by noise, or whether the measurement results are normal values ​​that do not include noise influence or have minimal noise influence. For example, in this way, for each measurement area that has been measured using a roughness measuring device, the maximum height Sz can be used as an indicator to determine whether the roughness parameter values ​​other than Sz obtained in each measurement area are applicable values ​​for evaluating the semiconductor wafer being evaluated. As a result, roughness parameter values ​​that reduce the impact of noise can be used for evaluating semiconductor wafers.

[0048] The evaluation method described above will be explained in further detail below.

[0049] <Semiconductor wafers being evaluated>

[0050] The semiconductor wafers evaluated in the above evaluation method can include various semiconductor wafers such as silicon wafers (e.g., single-crystal silicon wafers). In this invention and specification, unless otherwise stated, a silicon wafer refers to a single-crystal silicon wafer. Furthermore, polished wafers can be cited as an example of semiconductor wafers to be evaluated. A polished wafer is a wafer whose one or two surfaces are polished surfaces, i.e., surfaces that have undergone surface polishing treatment. Polished wafers are typically manufactured by sequentially performing processes such as coarse polishing, etching, and mirror polishing (i.e., fine polishing), resulting in polished surfaces. The polished surfaces are typically mirror-like. The surface polishing treatment can be DSP (double-side polishing) or SSP (single-side polishing). Furthermore, the conductivity type of the semiconductor wafer being evaluated is not limited; it can be either n-type or p-type. Its dopant concentration (i.e., resistivity), oxygen concentration, etc., are also not limited. The diameter of the semiconductor wafer being evaluated can be, for example, 200 mm, 300 mm, or 450 mm, but is not particularly limited.

[0051] <Determination using a roughness measuring device>

[0052] In the above evaluation method, a roughness measuring device is used to measure roughness parameters in one or more measurement areas on the surface of the semiconductor wafer being evaluated. The surface of the semiconductor wafer being measured can be either the front or back side in one approach, or both in another approach. In the roughness measuring device measurement in the above evaluation method, the total number of measurement areas on one of the front or back surfaces is 1 or more; in one approach, it can be 1, and in another approach, it can be 2 or more. Furthermore, the total number can be, for example, 50 or less, 40 or less, 30 or less, or 20 or less, or it can be greater than the value exemplified here.

[0053] As a roughness measuring device, various roughness measuring devices capable of obtaining roughness parameters can be used. Specific examples of roughness measuring devices include optical interference microscopes, atomic force microscopes, laser microscopes, Fizzo interferometers, etc. As an example, an optical interference microscope will be further described below. However, the roughness measuring device used in the above evaluation method is not limited to an optical interference microscope.

[0054] An optical interferometer is a device that uses the interference of light to measure the surface roughness of an object. There are two main methods of optical interferometer measurement: vertical scanning white interferometry and phase-shifting interferometry. A vertical scanning white interferometer is a type of white interferometer. More specifically, a white interferometer is a device that uses white light as a light source and can perform three-dimensional measurements of the surface of an object with wide-area and high vertical resolution.

[0055] Figure 1 This is an explanatory diagram illustrating the measurement principle of a white interference microscope. (For example...) Figure 1 As shown in (a), a beam splitter divides the light from a white light source into two beams. One beam illuminates a reference mirror attached to the objective lens, while the other beam illuminates the surface of the object being measured. The two beams are reflected and coupled again. This creates a mechanism that mounts a vertical scanning scanner on the objective lens, as shown in [example image]. Figure 1 As shown in (b), height information can be obtained by measuring the height of the highest intensity of the generated interference fringes through scanning the objective lens in the vertical direction. Figure 1 In the example shown in (b), the difference in the peak positions of the interference intensity at locations A and B indicates the height difference. The interference of white light is formed by the superposition of interference from multiple wavelengths of light, and the peak position of the interference signal is the position where all phases are consistent. Therefore, according to white interference microscopy, high-precision height measurement can be performed regardless of the magnification of the observation.

[0056] Figure 2 An example of the change in interference light during fringe scanning in measurements using a white interference microscope is shown. For example, as... Figure 1 As shown in (a), a vertical scanning scanner is mounted on the objective lens of the white interference microscope, such as... Figure 1 As shown in the example, height information is obtained through fringe scanning, which measures the height of the interference fringes (stripes) produced by scanning the objective lens along the vertical direction, where the intensity is highest. The number of interference fringes is determined by the tilt angle of the surface being measured, and the height difference between adjacent bright and dark areas of the fringes is approximately 140 nm, which is one-quarter of the center wavelength of the white light source. In the interference signal of the fringe scan, the point with the highest intensity of the interfering light is called the zero-order interference, and the second strongest point is called the first-order interference. If the influence of noise is excluded, for example, Figure 2 The interference waveform shown is... Figure 2 In the middle, A corresponds to Figure 1 (b) shows location A, and location B corresponds to... Figure 1Location B is shown in (b). In contrast, for example, when sudden device vibration occurs during fringe scanning, the interference waveform becomes disordered, and sometimes zero-order interference cannot be detected where it would normally be obtained. In such cases, first-order or second-order interference may become the highest interference intensity within that pixel, misidentifying height information. Furthermore, in cases where sudden device vibration occurs during measurement, causing interference waveform disorder and misidentifying first-order interference as zero-order interference, a height difference of approximately 280 nm, which is half the center wavelength λ (560 nm) of white light, may form at that location. Figure 3 This is an example of the measurement results of the surface of a silicon wafer using a white interference microscope image. Figure 3 (a) shows a white interference microscope image. Figure 3 (b) Shows a high-level profile. In Figure 3 It was confirmed that a height difference of approximately 260 nm existed within the measurement field of view due to sudden device vibration during the measurement. By using the maximum height Sz as an indicator, it was possible to determine whether the influence of device-related noise, as described above, was included in the measurement results obtained by the roughness measuring device. This is because, as previously mentioned, the maximum height Sz is the sum of the maximum peak depth and the maximum valley depth, and is therefore easily affected by noise from device-related factors. Furthermore, it was also considered that the maximum height Sz is easily affected by noise from external interference factors such as foreign matter adhesion. The following will... Figure 4 The following example further illustrates this point.

[0057] Figure 4 It is a white interference microscope image of a silicon wafer surface containing a foreign object with a height of about 200 nm within the field of view. Figure 4 (a) is a white interference microscope image of the whole. Figure 4 (b) is a magnified image of the foreign object attachment site. Figure 4 (c) is a high-resolution profile of the foreign object. Figure 4 The table shown in (d) records the data from... Figure 4 The roughness parameter calculated from the overall image shown in (a). When there are foreign objects, scratches, defects, etc. in the field of view of a white interference microscope, the value of the roughness parameter of the semiconductor wafer surface may sometimes be shifted due to the foreign objects, scratches, defects, etc. Figure 4 The Sa and Sq values ​​shown in table (d) are slightly higher than those of typical silicon wafers. Furthermore, Sa and Sq will be discussed in detail later. In contrast, Figure 4 The Sz value shown in (d) is affected by foreign objects in the field of view, showing values ​​above 200 nm. Therefore, by using Sz as an indicator, it is possible to determine whether the noise effect of external interference factors such as foreign object adhesion is included in the measurement results obtained by the roughness measuring device.

[0058] The above examples illustrate roughness measuring devices using a white interference microscope and semiconductor wafers using a silicon wafer. However, as mentioned earlier, the roughness measuring device used in the above evaluation method can also be a roughness measuring device other than a white interference microscope, and the semiconductor wafer being evaluated can also be a semiconductor wafer other than a silicon wafer.

[0059] <Determination of the Applicability of Roughness Parameters>

[0060] In the above evaluation method, as described above, measurements are performed using a roughness measuring device in one or more measurement areas, thereby obtaining the maximum height Sz and roughness parameters other than Sz. Then, for each measurement area, the value of the maximum height Sz obtained for that measurement area is used as an indicator to determine whether the value of the roughness parameter other than Sz obtained in that measurement area is a value that can be used in the evaluation of the semiconductor wafer being evaluated. The detailed reasons for using the maximum height Sz as such an indicator in the above evaluation method are as described above. Specific methods for determining whether it is acceptable can be, for example, the following methods. These various methods can also be combined arbitrarily.

[0061] (1) In one approach, when the value of the maximum height Sz is below or less than a predetermined threshold, the value of a roughness parameter other than Sz obtained for one or more measurement areas where the maximum height Sz is obtained is used as the evaluation value for the semiconductor wafer being evaluated. The case where the value of the maximum height Sz is below or less than the predetermined threshold is also recorded as "Sz is not abnormal". The threshold can be determined, for example, by considering the height of foreign matter that can adhere to the semiconductor wafer, the depth of defects or scratches that can be generated on the semiconductor wafer, the magnitude of noise from device factors that can be generated during measurement, and the required quality of the semiconductor wafer being evaluated. It is appropriate to determine whether to use a value below the threshold or a value less than the threshold. Furthermore, as a roughness parameter other than Sz, for example, one or more roughness parameters can be selected from the group consisting of the arithmetic mean height Sa and the root mean square height Sq. Only the arithmetic mean height Sa, only the root mean square height Sq, or both can be used as the evaluation value. Regarding the arithmetic mean height Sa and the root mean square height Sq, as mentioned above, refer to ISO 25178-2:2012.

[0062] The same applies to the various methods described below.

[0063] In this method, when the total number of measurement areas is one, the roughness parameter value other than Sz obtained for that one measurement area is used as the evaluation value for the semiconductor wafer being evaluated. On the other hand, in this method, when the total number of measurement areas is two or more, and the total number of measurement areas with a maximum height Sz value below or below a preset threshold is two or more, the roughness parameter value other than Sz obtained for all of these two or more measurement areas can be used as the evaluation value for the semiconductor wafer being evaluated, or the roughness parameter value other than Sz obtained for a portion of the area can be used as the evaluation value for the semiconductor wafer being evaluated. When using the roughness parameter value other than Sz obtained for two or more measurement areas as the evaluation value, for example, the arithmetic mean, maximum value, minimum value, etc., obtained for multiple measurement areas can be used to determine whether the semiconductor wafer being evaluated is a semiconductor wafer with the desired quality (i.e., a qualified product) or a semiconductor wafer that does not meet the desired quality (i.e., a defective product). The threshold for determining whether it is a qualified or defective product can be determined by considering the required quality of the semiconductor wafer being evaluated.

[0064] The same applies to the various methods described below.

[0065] (2) In one method, if the value of the maximum height Sz is above or exceeds a preset threshold, the roughness parameter value other than Sz obtained for the measurement area where the maximum height Sz is obtained is not used as the evaluation value for the semiconductor wafer being evaluated. The case where the value of the maximum height Sz is above or exceeds the preset threshold is also recorded as "Sz is abnormal". According to this method, specifically, for a measurement area where the maximum height Sz is above or exceeds the preset threshold, it can be determined that the measurement results performed in this measurement area include abnormal values ​​affected by noise, and therefore the evaluation value for the semiconductor wafer being evaluated is not used.

[0066] In this method, in one aspect, the measurement area where the maximum height Sz value obtained is above or exceeds a preset threshold can be measured again using a roughness measuring device (hereinafter also referred to as "remeasurement"). If there are two or more measurement areas where the maximum height Sz value obtained is above or exceeds the preset threshold, remeasurement can be performed on some or all of them. For example, if the main reason for the maximum height Sz value being above or exceeding the preset threshold in a certain measurement area is device factors such as device vibration generated during measurement, the measurement result obtained by remeasurement of that measurement area may not include outliers. Whether the remeasurement result includes outliers can also be determined using the maximum height Sz as an indicator, as described above. If the maximum height Sz value obtained by remeasurement is below or less than the preset threshold, the roughness parameter value other than Sz obtained by remeasurement can be used as the evaluation value for the semiconductor wafer being evaluated. On the other hand, if even after re-measurement, a maximum height Sz value above or exceeding a predetermined threshold is obtained, it can be ultimately determined that the value of the roughness parameter other than Sz obtained for the measurement area is not used as the evaluation value for the semiconductor wafer being evaluated.

[0067] Furthermore, in one approach, if a maximum height Sz value exceeding or above a predetermined threshold is obtained, further measurements using a roughness measuring device can be performed in areas outside the measurement area where the maximum height Sz value was obtained. For example, if the measurement area in the first measurement includes a measurement area where a maximum height Sz value exceeding or above the predetermined threshold was obtained, then as a second measurement, measurements using a roughness measuring device can be performed on one or more areas not measured in the first measurement. If the main reason why the maximum height Sz obtained in the first measurement for a certain measurement area is above or above the threshold is due to foreign matter, scratches, defects, etc., present in that measurement area, it is possible to obtain measurement results that do not include these effects by changing the measurement area. In this case, if the maximum height Sz value obtained by changing the measurement area is below or below the predetermined threshold, then the roughness parameter value other than Sz obtained by the measurement can be used as the evaluation value for the semiconductor wafer being evaluated. On the other hand, if the measurement results obtained by changing the measurement area result in a maximum height Sz value that is above or exceeds a predetermined threshold, it can be determined that the value of the roughness parameter other than Sz obtained for that measurement area is not used as the evaluation value for the semiconductor wafer being evaluated, or the measurement area can be remeasured. Regarding the determination of the measurement results obtained through remeasurement, please refer to the previous description.

[0068] Furthermore, in one approach, when a maximum height Sz value above or exceeding a pre-set threshold is obtained, the values ​​of one or more roughness parameters obtained for the measurement area where the maximum height Sz value was obtained can be recalculated. The recalculation can exclude portions from the roughness parameter calculation objects that are presumed to be the cause of outliers in the maximum height Sz, i.e., the reasons why the maximum height Sz is above or exceeding the threshold. For example, if the reason why the maximum height Sz obtained for a certain measurement area is above or exceeding the threshold is due to foreign matter, the measurement results can be obtained by removing the portion containing foreign matter through image processing (e.g., mask processing) or data processing, and then the measurement results can be recalculated. This allows for the acquisition of measurement results that do not include or have reduced the influence of foreign matter. The object of recalculation can be one or more surface parameters other than the maximum height Sz. For example, the recalculated results obtained in this way can be used as evaluation values ​​for semiconductor wafers being evaluated. Alternatively, the maximum height Sz can be recalculated. If the recalculated maximum height Sz is below or less than the threshold, it can be determined that the values ​​of the surface parameters other than the maximum height Sz obtained through the recalculation are values ​​that can be used in the evaluation of the semiconductor wafer being evaluated.

[0069] In the above evaluation method, as described above, the semiconductor wafer to be evaluated can be evaluated based on the value of a roughness parameter other than Sz, which is determined to be usable based on the result of the determination of whether it is usable. Such evaluation can be based on one or more surface parameter values ​​other than Sz obtained for a single measurement area, or it can be based on the arithmetic mean, maximum value, minimum value, etc., of values ​​obtained for multiple measurement areas. In either case, if the value of the surface parameter is below or less than a predetermined threshold, the semiconductor wafer to be evaluated can be determined to be a semiconductor wafer with the desired quality (i.e., a qualified product); if the value is above or greater than the predetermined threshold, the semiconductor wafer to be evaluated can be determined to be a semiconductor wafer that does not meet the desired quality (i.e., a defective product). As mentioned earlier, the threshold used for determining whether a product is qualified or defective can be determined by considering the required quality of the semiconductor wafer to be evaluated. Furthermore, in one approach, the value of the maximum height Sz can also be used to determine the quality of the semiconductor wafer to be evaluated. Based on the maximum height Sz, the presence or absence of local defects on the surface (front or back) of the semiconductor wafer, and the roughness evaluation, etc., can be performed.

[0070] [Manufacturing Method 1]

[0071] Manufacturing method 1 is a method for manufacturing a semiconductor wafer, comprising:

[0072] Manufacturing evaluation semiconductor wafers under test manufacturing conditions;

[0073] The evaluation method described above is used to evaluate the semiconductor wafer manufactured above.

[0074] Based on the results of the above evaluation, the manufacturing conditions for which changes were imposed on the aforementioned test manufacturing conditions will be determined as the actual manufacturing conditions, or the aforementioned test manufacturing conditions will be determined as the actual manufacturing conditions; and

[0075] Semiconductor wafers are manufactured under the actual manufacturing conditions determined above.

[0076] In manufacturing method 1, as a preliminary stage for determining the actual manufacturing conditions, test manufacturing conditions are established, under which evaluation semiconductor wafers are manufactured. "Actual manufacturing conditions" refers to the manufacturing conditions of the product semiconductor wafer. The evaluation semiconductor wafer manufactured under the test manufacturing conditions is subjected to evaluation using the previously detailed evaluation method. There may be at least one, or more than two, evaluation semiconductor wafers; their number is not particularly limited. If the evaluation results for the surface parameters obtained from the evaluation semiconductor wafer (e.g., the previously described arithmetic mean height Sa and / or root mean square height Sq) are values ​​expected for the product semiconductor wafer, then by using these test manufacturing conditions as the actual manufacturing conditions, the product semiconductor wafer is manufactured and shipped, thereby enabling a stable supply of product semiconductor wafers with the desired surface properties to the market. On the other hand, if the evaluation results for the surface parameters obtained from the evaluation semiconductor wafer differ from the expected values ​​for the product semiconductor wafer, the manufacturing conditions under which the test manufacturing conditions were modified are determined as the actual manufacturing conditions. As an example of semiconductor wafer manufacturing processes, a polished wafer, such as a silicon wafer, can be manufactured through a process including: cutting (slicing), chamfering, rough grinding (e.g., polishing), etching, mirror polishing (fine polishing), and cleaning between or after the aforementioned processing steps, from a silicon wafer grown from a silicon single-crystal ingot using the CZ process or similar methods. For silicon wafers or semiconductor wafers other than silicon wafers, the manufacturing conditions to which the aforementioned modifications are applied are preferably manufacturing conditions considered to affect surface parameters. As an example of such manufacturing conditions, the polishing conditions for the surface (front and / or back) of a semiconductor wafer can be cited. Specific examples of such surface polishing conditions include rough polishing conditions and mirror polishing conditions; more specifically, the type of polishing slurry, the abrasive concentration of the polishing slurry, and the type of polishing pad (e.g., hardness). By determining the manufacturing conditions that modify the test manufacturing conditions as the actual manufacturing conditions, and manufacturing and shipping product semiconductor wafers under these actual manufacturing conditions, it is possible to reliably supply product semiconductor wafers with desired surface properties to the market. Alternatively, the evaluation semiconductor wafer can be remanufactured under different manufacturing conditions than those applied to the test manufacturing conditions. The evaluation method described above can then be used to evaluate the evaluation semiconductor wafer, repeating the process once or twice to determine whether the manufacturing conditions should be used as the actual manufacturing conditions or whether further changes should be applied. The threshold value can be appropriately set based on market requirements, quality, etc., and is used to determine whether the evaluation results of the surface parameters obtained for the evaluation semiconductor wafer are the values ​​expected for the product semiconductor wafer.

[0077] [Manufacturing Method 2]

[0078] Manufacturing method 2 is a method for manufacturing a semiconductor wafer, comprising:

[0079] Manufacturing a semiconductor wafer lot including a plurality of semiconductor wafers;

[0080] Extracting one or more semiconductor wafers from the above-mentioned semiconductor wafer lot;

[0081] Evaluating the extracted semiconductor wafers using the above evaluation method; and

[0082] Preparing to ship out as products semiconductor wafers of the same semiconductor wafer lot as the semiconductor wafers determined to be qualified in the above evaluation.

[0083] In Manufacturing Method 2, so-called sampling inspection is performed, and semiconductor wafers of the same lot as the semiconductor wafers determined to be qualified are prepared to be shipped out as product semiconductor wafers. The manufacturing of the semiconductor wafer lot in Manufacturing Method 2 can be carried out using the manufacturing process of general semiconductor wafers. As an example, regarding the manufacturing process of polished wafers, the previous description can be referred to.

[0084] The number of semiconductor wafers extracted from the manufactured semiconductor wafer lot and subjected to so-called sampling inspection is at least one, and can also be two or more, and its number is not particularly limited. The semiconductor wafers extracted from the semiconductor wafer lot are given an evaluation using the evaluation method detailed previously. Preparing to ship out as product semiconductor wafers semiconductor wafers of the same semiconductor wafer lot as the semiconductor wafers determined to be qualified in the evaluation. The criteria for being determined to be qualified, specifically the threshold values of surface parameters (such as the arithmetic mean height Sa and / or the root mean square height Sq described previously) are determined according to the quality required for the product semiconductor wafers. Regarding the preparation for shipping out as product semiconductor wafers, for example, packaging etc. can be cited. Thus, according to Manufacturing Method 2, product semiconductor wafers with a desired surface property can be stably supplied to the market.

[0085] [Examples]

[0086] Hereinafter, the present invention will be further described based on examples. However, the present invention is not limited to the embodiments shown.

[0087] [Semiconductor wafers to be evaluated]

[0088] As semiconductor wafers to be evaluated, 25 silicon wafers with a diameter of 300 mm were prepared. The above silicon wafers are polished wafers after double-sided grinding (DSP).

[0089] [Measurement using a roughness measuring device]

[0090] For the back surface of each of the above silicon wafers, measurements were carried out using the following measurement conditions.

[0091] <Measurement Conditions>

[0092] Roughness measuring device: white interference microscope

[0093] Measurement magnification: 10x (field of view: 1.6mm × 1.6mm)

[0094] Filter: Hi-pass filter 80μm (removes fluctuations in spatial wavelengths above 80μm)

[0095] Measurement areas: For each silicon wafer, Table 1 shows 9 areas within the wafer plane.

[0096] [Table 1]

[0097] Measurement area X-coordinate (mm) Y-coordinate (mm) Measurement area 1 0 0 Measurement area 2 75 0 Measurement area 3 0 75 Measurement area 4 -75 0 Measurement area 5 0 -75 Measurement area 6 140 0 Measurement area 7 0 140 Measurement area 8 -140 0 Measurement area 9 0 -140

[0098] *The origin is the center of the chip.

[0099] <Test Procedure>

[0100] Figure 5 The measurement procedure of this embodiment is shown in the figure. The details of the measurement procedure are described below.

[0101] The microscope image obtained by measuring the surface of a silicon wafer using a white interference microscope was subjected to image processing, including planarization. In the resolution section of the white microscope, the processed microscope image was analyzed, and roughness parameters Sa, Sq, and Sz were output for each of the nine points within the plane of each silicon wafer.

[0102] Regarding the threshold for Sz, in this embodiment, it is assumed that the height of the foreign object or the depth of the defect or scratch that causes noise is 50 nm or more, and the threshold for Sz is set to 50 nm. However, the threshold can be any value that does not obscure the surface shape of the wafer. For example, if the wafer surface is smooth, setting a value such as 10 nm or 20 nm as the threshold would also be appropriate.

[0103] In this embodiment, if the value of Sz is 50nm or higher, it is considered that "Sz is abnormal"; if it is less than 50nm, it is considered that "Sz is not abnormal". An Sz abnormality determination is performed. For each of the nine measurement regions of a silicon wafer, if Sz is not abnormal, it is determined that Sa and Sq in that measurement region can be used in wafer evaluation.

[0104] On the other hand, if Sz is abnormal, it is determined that further processing is required, which may involve one or more of the following: re-measurement, measurement in other measurement areas, and recalculation. Details of these processing are as described above.

[0105] Figure 6 An example of the evaluation results obtained in this embodiment is shown in the figure.

[0106] exist Figure 6 In the image, image A is the one where the value of Sz is determined to be below the threshold, indicating "no abnormality in Sz". No differences in level, foreign objects, defects, etc., that could potentially constitute noise in the roughness parameter measurement were identified in image A.

[0107] In contrast, images B, C, and D are those where the value of Sz is determined to be above the threshold, indicating an "abnormal Sz". Image B shows variations in the levels of what is presumed to be vibration factors in the device. Image C shows foreign matter attachment. Image D shows pitting-like defects.

[0108] The results above confirm that by using the value of Sz as an indicator, it is possible to determine whether the results of roughness parameter measurement include the influence of noise.

[0109] In this embodiment, for 25 silicon wafers, the number of measurement areas for each wafer was set to 9, thus obtaining a total of 225 images using a white interference microscope. Among these 225 images, the presence or absence of anomalies in Sz was determined: 216 images were judged as "no Sz abnormality," and 9 images were judged as "Sz abnormality present." Table 2 shows the arithmetic mean and standard deviation of the Sq values ​​for the 225 images, as well as the arithmetic mean and standard deviation of the Sq values ​​for the 216 images judged as having no Sz abnormality.

[0110] [Table 2]

[0111] Number of images Arithmetic mean (nm) Standard deviation (nm) All images 225 0.432 1.895 Sz has no abnormalities. 216 0.202 0.022

[0112] exist Figure 7 The arithmetic mean of the Sq values ​​obtained from 25 silicon wafers is plotted (referred to as "Sq arithmetic mean").

[0113] Regarding the double-sided polishing performed on the silicon wafer evaluated in this embodiment, the normal value of Sq on the polished surface after double-sided polishing is about 0.2 nm.

[0114] Figure 7 In the middle, the Sq arithmetic mean of the nine in-plane measurement regions for 25 silicon wafers is plotted on the left. Figure 7 The plot on the left also includes plots of Sq values ​​with an arithmetic mean of 1 nm or more.

[0115] Figure 7 In the middle, the arithmetic mean of Sq obtained by excluding the images judged to be abnormal by Sz is drawn on the right. Figure 7 The graphs on the right side of the image all have an arithmetic mean of approximately 0.2 nm.

[0116] From Table 2 and Figure 7 The results show that by using the value of Sz as an indicator, the surface roughness of semiconductor wafers can be evaluated with high precision.

[0117] Industrial availability

[0118] One aspect of the present invention is useful in the manufacture of various semiconductor wafers, such as silicon wafers.

Claims

1. A method for evaluating semiconductor wafers, wherein, include: In one or more measurement areas on the surface of the semiconductor wafer being evaluated, roughness parameters other than the maximum height Sz are obtained by performing three-dimensional measurements using a roughness measuring device. as well as For each measurement area, the value of the maximum height Sz is used as an indicator to determine whether the roughness parameter values ​​other than Sz obtained in that measurement area are applicable values ​​for evaluating the semiconductor wafer being evaluated.

2. The evaluation method for semiconductor wafers according to claim 1, wherein, include: When the maximum height Sz is below a preset threshold, the roughness parameter other than Sz obtained from one or more measurement areas where the maximum height Sz is obtained is used as the evaluation value for the semiconductor wafer being evaluated.

3. The evaluation method for semiconductor wafers according to claim 1, wherein, If the maximum height Sz is above a preset threshold, the roughness parameter values ​​other than Sz obtained for the measurement area where the maximum height Sz is obtained are not used as the evaluation values ​​for the semiconductor wafer being evaluated.

4. The evaluation method for semiconductor wafers according to claim 3, wherein, Also includes: If the maximum height Sz is above a preset threshold, a three-dimensional roughness measurement is performed again on the measurement area where the maximum height Sz is obtained.

5. The evaluation method for semiconductor wafers according to claim 3, wherein, Also includes: When the maximum height Sz is above a preset threshold, a three-dimensional measurement is performed in the area outside the measurement area where the maximum height Sz is obtained, using a roughness measuring device.

6. The evaluation method for semiconductor wafers according to claim 3, wherein, Also includes: If the maximum height Sz is above a preset threshold, the values ​​of one or more roughness parameters obtained for the measurement area where the maximum height Sz is obtained are recalculated.

7. The evaluation method for semiconductor wafers according to claim 6, wherein, The recalculation is performed by excluding the portion of the roughness parameter calculation object that is estimated to be the cause of outliers in the maximum height Sz.

8. The evaluation method for semiconductor wafers according to claim 1, wherein, The roughness parameter other than Sz is one or more roughness parameters selected from the group consisting of the arithmetic mean height Sa and the root mean square height Sq.

9. The evaluation method for semiconductor wafers according to claim 1, wherein, The roughness measuring device is an optical interference microscope.

10. The evaluation method for a semiconductor wafer according to claim 9, wherein, The optical interference microscope is a white interference microscope.

11. A method for manufacturing a semiconductor wafer, wherein, include: Manufacturing evaluation semiconductor wafers under test manufacturing conditions; The manufactured evaluation semiconductor wafer is evaluated using the evaluation method according to any one of claims 1 to 10; Based on the evaluation results, the manufacturing conditions that have been modified under the test manufacturing conditions will be determined as the actual manufacturing conditions, or the test manufacturing conditions will be determined as the actual manufacturing conditions. as well as Semiconductor wafers are manufactured under the actual manufacturing conditions determined in the above decision.

12. The method for manufacturing a semiconductor wafer according to claim 11, wherein, The manufacturing conditions under which the changes are applied include the surface polishing conditions of the semiconductor wafer.

13. A method for manufacturing a semiconductor wafer, wherein, include: Manufacturing a batch of semiconductor wafers that comprises multiple semiconductor wafers; One or more semiconductor wafers are extracted from the batch of semiconductor wafers; The extracted semiconductor wafer is evaluated using the evaluation method according to any one of claims 1 to 10; and The semiconductor wafers from the same batch as those evaluated as qualified semiconductor wafers are prepared to be shipped out as products.

Images