Defect evaluation method for single-crystal silicon wafers

Abstract

This invention provides a defect evaluation method for single-crystal silicon wafers that allows for the simple evaluation of processing-induced defects without revealing crystal defects. [Solution] A method for evaluating defects in a single-crystal silicon wafer, comprising: a first step of applying an isotropic chemical dry etching treatment to the surface of a mirror-finished single-crystal silicon wafer; and a second step of evaluating defects on the surface of the single-crystal silicon wafer that has undergone the isotropic chemical dry etching treatment.

Description

Technical Field

[0001] The present invention relates to a method for evaluating defects in a single crystal silicon wafer.

Background Art

[0002] Many single crystal silicon wafers are produced by slicing a single crystal silicon ingot grown by the Czochralski method (CZ method). The CZ method is a method of growing a single crystal silicon ingot by immersing a seed crystal in a silicon melt in a quartz crucible and slowly pulling up the seed crystal while rotating it.

[0003] In the single crystal silicon ingot grown by this method, since oxygen eluted from the quartz crucible is taken in supersaturation, minute oxygen precipitates (SiO2) are formed during the cooling process of the crystal.

[0004] Also, vacancies are taken in when the silicon melt crystallizes, and they aggregate during the cooling process of the crystal to form minute voids called COP. Although COP is a cavity, a silicon oxide film is formed on its inner wall during the cooling process of the crystal after COP formation, so COP is also a defect containing SiO2.

[0005] In addition to such crystal defects, the single crystal silicon wafer contains processing-induced defects introduced in the wafer processing steps.

[0006] The single crystal silicon wafer is completed through a slicing step of slicing the single crystal silicon ingot, a chamfering step of chamfering the outer peripheral portion of the wafer, a lapping step of flattening the wafer, an etching step of removing mechanical processing strain, a polishing step of mirror-finishing the surface, a cleaning step of removing foreign substances and cleaning liquid, etc., and is shipped as a single crystal silicon wafer that can be used as a substrate for semiconductor devices.

[0007] The processing-induced defects that occur in such wafer processing steps are extremely small and difficult to detect. However, in order to manufacture higher quality single-crystal silicon wafers, it is desirable to accurately evaluate processing-induced defects and feed this information back into the polishing process, thereby constructing a wafer processing process with fewer processing-induced defects.

[0008] Regarding the evaluation of defects caused by processing, Patent Document 1 discloses a method for evaluating processing-induced defects by dry etching the surface of a mirror-finished single-crystal silicon wafer under conditions where the etching rate of Si is greater than that of SiO2, thereby making processing-induced defects visible as pits (also called indentations), and evaluating the number of these pits.

[0009] Here, there are two methods of dry etching: reactive ion etching (RIE) and chemical dry etching (CDE).

[0010] Furthermore, there are two terms used to describe the properties of etching: isotropic etching and anisotropic etching. Isotropic etching means that etching occurs from all directions—up, down, left, and right. Anisotropic etching means that etching occurs from only one direction.

[0011] In the RIE method, the sample to be etched is placed between parallel plate electrodes, and the electrode on the material side is strongly negatively charged. This accelerates reactive ions generated in the plasma in a direction perpendicular to the sample surface. In other words, etching proceeds through collisions with a certain direction, involving both chemical and physical effects, and is therefore classified as anisotropic etching.

[0012] In contrast to the RIE method, the CDE method does not apply voltage to the sample. Furthermore, the discharge chamber and etching chamber are separated, and long-lived reactive species generated in the discharge chamber are transported to the etching chamber where etching proceeds. Therefore, reactive ions contribute very little to the etching process. In other words, etching proceeds through a purely chemical reaction rather than a physical action, and is therefore classified as isotropic etching.

[0013] Here, Patent Document 1 discloses an anisotropic RIE etching method, in which crystal defects containing SiO2 are not etched and instead become apparent as protrusions. Therefore, an identification step is required to distinguish between pits resulting from processing-induced defects and protrusions resulting from crystal defects based on differences in light scattering characteristics.

[0014] Therefore, if it is possible to reveal only processing-induced defects without revealing crystal defects, the identification process becomes unnecessary, and the number of defects obtained by the particle counter can be considered as the number of processing-induced defects, thus allowing for a simpler evaluation of processing-induced defects.

[0015] Patent Document 2 discloses a method for evaluating defects caused by contact with external objects at the edge surface by performing a dry etching process on a silicon wafer and observing the visual appearance.

[0016] Non-patent document 1 discloses an etching technique for Si using a gas mixture of CF4 and O2. Specifically, it describes how adding O2 to CF4 changes the etching rate of Si. [Prior art documents] [Patent Documents]

[0017] [Patent Document 1] Japanese Patent Publication No. 2011-249479 [Patent Document 2] Japanese Patent Publication No. 2011-228478 [Non-patent literature]

[0018] [Non-Patent Document 1] Dry etching technology for LSI manufacturing (Metal Surface Technology 30(5), 256-268, 1979) [Overview of the project] [Problems that the invention aims to solve]

[0019] Thus, in the conventional method for evaluating defects in a single crystal silicon wafer using an anisotropic RIE method, crystal defects are also revealed in addition to process-induced defects, so a discrimination process for discriminating between crystal defects and process-induced defects was necessary. Also, there was a problem that only process-induced defects could not be revealed.

[0020] The present invention has been made to solve the above problems, and an object thereof is to provide a method for evaluating defects in a single crystal silicon wafer that can simply evaluate process-induced defects without revealing crystal defects.

Means for Solving the Problems

[0021] In order to solve the above problems, a method for evaluating defects in a single crystal silicon wafer according to the present invention includes a first step of performing isotropic chemical dry etching treatment on the surface of a mirror-polished single crystal silicon wafer, and a second step of evaluating defects on the surface of the single crystal silicon wafer that has been subjected to the isotropic chemical dry etching treatment. A method for evaluating defects in a single crystal silicon wafer is provided, which is characterized by comprising these steps.

[0022] With such a method for evaluating defects in a single crystal silicon wafer, by performing isotropic etching, crystal defects containing SiO2 as seen in conventional anisotropic etching do not appear as protrusions. Therefore, etching treatment can be performed without revealing crystal defects. As a result, it is possible to simply evaluate process-induced defects on the surface without being affected by crystal defects.

[0023] At this time, in the first step, it is preferable to reveal process-induced defects among the defects on the surface of the mirror-polished single crystal silicon wafer.

[0024] By revealing process-induced defects, it is possible to more simply and accurately evaluate process-induced defects.

[0025] At this time, in the first step, it is preferable to make the processing-induced defect appear as a pit cluster composed of a plurality of pits.

[0026] If it is a pit cluster composed of a plurality of pits, since its size is larger than that of a single minute pit, it is easy to detect, and the processing-induced defect can be evaluated more simply and accurately.

[0027] At this time, in the first step, using the haze value (Haze) obtained by the particle counter as an index, it is preferable to perform the isotropic chemical dry etching treatment so that (Haze after etching) / (Haze before etching) is 50 or more.

[0028] By performing such treatment, the haze after etching can be enlarged to 50 times or more of the haze before etching, so that the processing-induced defect can be surely made apparent.

[0029] At this time, for the isotropic chemical dry etching treatment in the first step, it is preferable to use a gas in which at least a gas containing fluorine atoms and oxygen gas are mixed.

[0030] By using a gas in which a gas containing fluorine atoms and oxygen gas are mixed, the processing-induced defect can be suitably made apparent.

[0031] At this time, as the gas containing fluorine atoms, it is preferable to use a perfluorocarbon-based gas.

[0032] By using a perfluorocarbon-based gas, the processing-induced defect can be made apparent more suitably.

[0033] At this time, as the gas containing fluorine atoms, it is preferable to use CF4.

[0034] By using CF4, the processing-induced defect can be made apparent more suitably.

[0035] In this case, it is preferable to use a mixed gas of CF4 and O2 with an O2 content of 10 to 30 vol% for the isotropic chemical dry etching treatment in the first step.

[0036] By using this ratio, processing-induced defects can be more effectively and reliably identified.

[0037] At this time, it is preferable to measure the number of defects using a particle counter in the second step and to use the measured number of defects as the number of defects caused by processing.

[0038] As described above, crystal defects do not become apparent during the isotropic chemical dry etching process in the first step. Therefore, the defects measured by the particle counter in the second step do not include crystal defects, but only processing-induced defects. For this reason, there is no need to distinguish between crystal defects and processing-induced defects when measuring with the particle counter, and the number of measured defects can be used directly as the number of processing-induced defects. Thus, processing-induced defects can be evaluated simply. [Effects of the Invention]

[0039] The defect evaluation method for single-crystal silicon wafers of the present invention, by applying isotropic etching, prevents crystal defects containing SiO2 from becoming apparent as protrusions, as seen in conventional anisotropic etching. Therefore, etching can be performed without making crystal defects apparent. This allows for the simple evaluation of surface processing-induced defects without being affected by crystal defects. [Brief explanation of the drawing]

[0040] [Figure 1] This flowchart shows an example of a defect evaluation method for single-crystal silicon wafers according to the present invention. [Figure 2] This is a comparison of SEM images of processing-induced defects before and after the isotropic chemical dry etching treatment of the present invention. [Figure 3]This is a comparison of SEM images of crystal defects containing SiO2 before and after the isotropic chemical dry etching treatment of the present invention. [Figure 4] This graph shows the number of defects before and after isotropic chemical dry etching using a mixed gas of CF4 and O2, with varying O2 ratios. [Figure 5] This graph shows the number and percentage of defect pit clusters after isotropic chemical dry etching treatment using a mixed gas of CF4 and O2, with varying O2 ratios. [Figure 6] This graph shows the Si etching rate and Si / SiO2 selectivity when isotropic chemical dry etching is performed using a mixed gas of CF4 and O2, with varying O2 ratios. [Figure 7] This graph shows the haze before and after isotropic chemical dry etching treatment using a mixed gas of CF4 and O2, with varying O2 ratios, as well as the ratio of haze after treatment to haze before treatment. [Figure 8] This figure shows the number of defects and the map before and after treatment in Example 1. [Figure 9] This figure shows the number of defects and the map before and after treatment in Example 2. [Modes for carrying out the invention]

[0041] The present invention will be described in detail below, but the present invention is not limited to these descriptions.

[0042] As mentioned above, there was a need for a simple defect evaluation method for single-crystal silicon wafers that could evaluate processing-induced defects without revealing crystal defects.

[0043] The inventors of this invention diligently investigated the above-mentioned problems and focused on isotropic chemical dry etching, rather than anisotropic, and conducted research. They then discovered that by performing isotropic chemical dry etching, it is possible to evaluate processing-induced defects without revealing crystal defects, and thus completed the present invention.

[0044] In other words, the present invention provides a defect evaluation method for a single-crystal silicon wafer, comprising: a first step of applying an isotropic chemical dry etching treatment to the surface of a mirror-finished single-crystal silicon wafer; and a second step of evaluating defects on the surface of the single-crystal silicon wafer that has undergone the isotropic chemical dry etching treatment.

[0045] With this method of evaluating defects in single-crystal silicon wafers, isotropic etching prevents crystal defects containing SiO2 from becoming apparent as protrusions, as seen in conventional anisotropic etching. Therefore, etching can be performed without making crystal defects apparent. This allows for the simple evaluation of surface processing-induced defects without being affected by crystal defects.

[0046] Embodiments of the present invention will be described below with reference to the drawings, but the present invention is not limited thereto.

[0047] Figure 1 is a flowchart illustrating an example of a defect evaluation method for a single-crystal silicon wafer according to a preferred embodiment of the present invention.

[0048] Here, prior to the first step of performing isotropic chemical dry etching, which is essential for the present invention, and the second step of evaluating defects, there is a step of preparing a single-crystal silicon wafer, although this is not particularly limited.

[0049] [Process S1] Step S1 in Figure 1 is the step of preparing a single-crystal silicon wafer. Here, the method of manufacturing the single-crystal silicon wafer is not particularly limited. It may be a single-crystal silicon wafer manufactured by the CZ method or a single-crystal silicon wafer manufactured by the FZ method. Alternatively, an epitaxial wafer may be used, in which single-crystal silicon has been epitaxially grown on a single-crystal silicon wafer manufactured by the CZ method or the FZ method. Here, it is sufficient to prepare a mirror-finished single-crystal silicon wafer for which we want to evaluate processing-induced defects. A mirror-finished single-crystal silicon wafer is expected to contain processing-induced defects.

[0050] [Process S2] In step S2 of Figure 1, isotropic chemical dry etching is performed on the surface of the mirror-finished single-crystal silicon wafer prepared in step S1 (i.e., this is the first step essential to the present invention). Here, the conventional RIE method is undesirable because, due to anisotropic etching, crystal defects consisting of oxygen precipitates (SiO2) are not etched and become apparent as protrusions. Therefore, it is preferable to perform the etching in an isotropic chemical dry etching apparatus in which, for example, the discharge tube and etching chamber are separated.

[0051] Figure 2 shows scanning electron microscope (SEM) images before and after applying the isotropic chemical dry etching treatment of the present invention to processing-induced defects.

[0052] Here, the defect size is larger after isotropic chemical dry etching than before the treatment. In other words, isotropic chemical dry etching makes processing-induced defects more apparent.

[0053] While not particularly limited, making machining-induced defects visible allows for simpler and more accurate evaluation of these defects.

[0054] Furthermore, at this time, pit clusters composed of multiple pits are observed. For example, the top machining-induced defect in Figure 2 is a pit cluster in which multiple pits are arranged vertically. While not particularly limited, it is preferable to manifest machining-induced defects as pit clusters composed of multiple pits in this manner.

[0055] A pit cluster, composed of multiple pits, is larger in size than a single minute pit, making it easier to detect and allowing for simpler and more accurate evaluation of machining-induced defects.

[0056] On the other hand, Figure 3 shows scanning electron microscope (SEM) images of crystal defects consisting of oxygen precipitates (SiO2) before and after the isotropic chemical dry etching treatment of the present invention. The defect size did not change before and after the treatment, or some of the defects disappeared.

[0057] This indicates that crystal defects do not become apparent in isotropic chemical dry etching. This is thought to be because isotropic chemical dry etching is isotropic and does not have the anisotropy of the conventional RIE method. In the conventional RIE method, due to the anisotropy, crystal defects containing SiO2 are not etched and become apparent as protrusions, but in isotropic chemical dry etching, crystal defects are etched from all directions, so they do not become apparent as protrusions.

[0058] On the other hand, processing-induced defects consist of small distortions or scratches present locally, and in these areas, etching proceeds more rapidly than in the Si substrate, resulting in the formation of pit clusters composed of multiple pits.

[0059] Next, we will describe in detail the conditions for isotropic chemical dry etching. In isotropic chemical dry etching, a fluorine (F)-containing gas is activated on a single-crystal silicon wafer by plasma discharge. The active species of F then act on Si to generate volatile SiF4, which leads to the etching of Si (Si + 4F → SiF4). Examples of fluorine-containing gases include fluorocarbon-based gases such as CF4, C2F6, C3F8, and C4F8, as well as inorganic fluorine-based gases such as SF6 and NF3; there are no particular restrictions. As long as fluorine is present, chlorine may also be present.

[0060] This section describes in more detail the isotropic chemical dry etching process using CF4 as a fluorine-containing gas. Polished wafers (mirror-finished single-crystal silicon wafers) for which processing-induced defects were to be evaluated were prepared, and surface defects were evaluated using a KLA SP5 particle counter with an inspection sensitivity of 19 nm Up. Next, isotropic chemical dry etching was performed for 60 seconds using CF4 alone or CF4 with O2 added at various concentrations (etching depth 200 nm), and after cleaning, surface defects were evaluated again using the SP5 with an inspection sensitivity of 60 nm Up.

[0061] Figure 4 shows the results. The number of defects (LLS) before treatment was generally low, ranging from 19 to 29 pcs, with little difference depending on the conditions. However, the number of defects after treatment ranged from 5 to 135 pcs, and increased significantly, especially at O2 ratios of 15 and 20 vol%, with the number of defects exceeding 130 pcs.

[0062] These defects were observed using SEM, and the number of pit clusters and the proportion of pit clusters, which were composed of multiple pits, were compiled and shown in Figure 5. Many pit clusters were observed at O2 ratios of 15% and 20% vol%. In other words, it can be said that processing-induced defects are more likely to become apparent at O2 ratios of 15% and 20% vol%.

[0063] Furthermore, minute processing-related defects with an inspection sensitivity of 15nm or 19nm before isotropic chemical dry etching, or even smaller processing-related defects that could not be detected even with 15nm or 19nm improvements, can now be evaluated as larger defects with an inspection sensitivity of 60nm after isotropic chemical dry etching.

[0064] Focusing on the O2 ratio, we can see that no pit clusters are observed at O2 ratios of 50% or 80% vol%. In other words, there are more favorable conditions for making processing-induced defects apparent.

[0065] Figure 6 shows the etching rate of Si and the etching selectivity ratio of Si to SiO2 (Si / SiO2 selectivity ratio). Specifically, the etching rate for Si was calculated by etching a single-crystal silicon wafer for a predetermined time and subtracting the wafer thickness after etching from the wafer thickness before etching. For SiO2, a single-crystal silicon wafer with a thermal oxide film grown by thermal oxidation was etched for a predetermined time and subtracted the thermal oxide film thickness after etching from the thermal oxide film thickness before etching. The etching selectivity ratio of Si to SiO2 was calculated as (etching rate of Si / etching rate of SiO2).

[0066] Figure 6 shows that the etching selectivity ratio of Si to SiO2 is higher than 1 under all conditions. Furthermore, this selectivity ratio increases as the O2 ratio decreases, and it is not the case that the O2 ratios of 15% and 20 vol% that showed the most pronounced manifestation effect were high. In other words, the etching selectivity ratio of Si to SiO2 is not the dominant factor in the manifestation of processing-induced defects in the present invention, and this selectivity ratio cannot be used as an indicator to reliably cause manifestation.

[0067] Focusing on the etching rate of Si, it can be seen that it is maximum at O2 ratios of 15 and 20 vol%. As mentioned above, processing-induced defects become apparent as pit clusters because etching proceeds more rapidly than on the Si substrate (defect-free areas). Therefore, it is thought that a higher etching rate of Si leads to a higher etching selectivity ratio of the defective areas compared to the Si substrate (defect-free areas), resulting in more pronounced manifestation of defects.

[0068] A general-purpose indicator that reflects this rapid etching behavior of the Si substrate is the haze value obtained from a particle counter. A higher haze value indicates a rougher surface.

[0069] Figure 7 shows the haze before and after treatment, as well as the ratio of haze after treatment to haze before treatment (after / before). It can be seen that the haze after treatment increases as the O2 ratio decreases, and this is particularly pronounced when the O2 ratio is 20 vol% or less. This is due to the rapid etching of the Si substrate.

[0070] Here, for the O2 ratio of 10 vol% with the highest post-treatment Haze / pre-treatment Haze ratio of 142.5 (Figure 7), when the treatment time was increased 1.5 times to 90 seconds (etching allowance of 300 nm), the number of defects after treatment was 75 pcs (more than 5 times more than the 14 pcs with a treatment time of 60 seconds (Figure 4)). When pit clusters were aggregated using SEM, the number was 65 pcs (more than 10 times more than the 5 pcs with a treatment time of 60 seconds (Figure 5)), indicating that processing-induced defects were significantly more apparent. In other words, for an O2 ratio of 10 vol%, a treatment time of 90 seconds (etching allowance of 300 nm) is more preferable than a treatment time of 60 seconds for a Si substrate (etching allowance of 200 nm).

[0071] In contrast, when the O2 ratio was 50 vol% with a low post-treatment Haze / pre-treatment Haze ratio of 6.5 (Figure 7), and the treatment time was tripled to 180 seconds, the number of defects after treatment was 32 pcs (more than three times the 10 pcs with a 60-second treatment time (Figure 4)). When pit clusters were aggregated using SEM, the number was 1 pc (an increase from 0 pcs with a 60-second treatment time (Figure 5)). Furthermore, the etching allowance of the Si substrate after this 180-second treatment time was 390 nm.

[0072] Therefore, comparing an O2 ratio of 10 vol% with a high post-treatment Haze / pre-treatment Haze ratio of 142.5 and an O2 ratio of 50 vol% with a low post-treatment Haze / pre-treatment Haze ratio of 6.5, the O2 ratio of 10 vol% is preferable because it results in more processing-induced defects that can be more clearly identified.

[0073] Therefore, it is preferable to perform isotropic chemical dry etching so that the haze value obtained by the particle counter is used as an indicator, and the ratio of (Haze after etching) / (Haze before etching) is, for example, 50 or more, although this is not particularly limited.

[0074] By processing under these conditions, the haze after etching can be expanded to more than 50 times the haze before etching, making it possible to more reliably reveal processing-induced defects.

[0075] Furthermore, while there are no particular limitations on the etching allowance, it is more preferable to have an allowance of, for example, 250 nm or more.

[0076] As described above, conditions that increase the ratio of (Haze after etching) to (Haze before etching) or the amount of etching removed result in more rapid etching, which allows for a more favorable manifestation of processing-induced defects.

[0077] The conditions for isotropic chemical dry etching have been explained above, but here we will summarize the gases used in the process.

[0078] While not particularly limited, it is preferable to use a gas mixture containing at least fluorine atoms and oxygen gas for isotropic chemical dry etching.

[0079] By using a gas mixture containing fluorine atoms and oxygen gas, processing-induced defects can be effectively made apparent.

[0080] Furthermore, although not particularly limited, it is preferable to use a perfluorocarbon gas as the gas containing fluorine atoms.

[0081] By using perfluorocarbon-based gases, processing-induced defects can be more effectively revealed.

[0082] Furthermore, although not particularly limited, it is preferable to use CF4 as the gas containing fluorine atoms.

[0083] By using CF4, processing-induced defects can be more effectively identified.

[0084] Furthermore, although not particularly limited, it is preferable to use a mixed gas of CF4 and O2 in which the proportion of O2 is 10 to 30 vol%.

[0085] By using this ratio, processing-induced defects can be more effectively and reliably identified.

[0086] [Process S3] Furthermore, in step S3 of Figure 1, surface defects of the single-crystal silicon wafer that was subjected to isotropic chemical dry etching in step S2 are evaluated (i.e., this is the second step essential to the present invention).

[0087] As mentioned above, since crystal defects consisting of oxygen precipitates (SiO2) do not become apparent in process S2, the defects acquired in process S3 are predominantly processing-induced defects that became apparent in process S2.

[0088] At this time, the method for evaluating defects is not particularly limited, but it is preferable to measure the number of defects using a particle counter, for example, and use the measured number of defects as the number of defects caused by processing.

[0089] The defects measured by the particle counter in step S3 do not include crystal defects, but only processing-induced defects. Therefore, there is no need to distinguish between crystal defects and processing-induced defects when measuring with the particle counter, and the measured number of defects can be used directly as the number of processing-induced defects. Thus, processing-induced defects can be evaluated simply.

[0090] With the defect evaluation method for single-crystal silicon wafers described above, isotropic etching prevents crystal defects containing SiO2 from becoming apparent as protrusions, as seen in conventional anisotropic etching. Therefore, etching can be performed without making crystal defects apparent. This allows for the simple evaluation of surface processing-induced defects without being affected by crystal defects.

[0091] Furthermore, when evaluating the number of processing-induced defects in a single-crystal silicon wafer by changing the processing conditions (for example, the conditions for polishing), by keeping the conditions for isotropic chemical dry etching in step S2 the same, it is possible to evaluate the difference in the number of processing-induced defects due solely to the difference in processing conditions with high accuracy, and this can be used to determine which processing conditions are better. [Examples]

[0092] The present invention will be described in detail below with reference to examples, but this is not intended to limit the present invention.

[0093] (Example 1) One mirror-finished single-crystal silicon wafer with a diameter of 300 mm, crystal plane orientation (100), conductivity type p-type, and resistance of 10 Ωcm was prepared. The prepared wafer was measured using a KLA SP7 particle counter with a sensitivity of 15 nm Up to obtain the initial (before isotropic chemical dry etching) surface defect count and surface roughness (Haze). The surface defect count was 33 pcs, and the Haze was 0.06 ppm.

[0094] Next, isotropic chemical dry etching was performed on the wafer using a single-wafer isotropic chemical dry etching apparatus. This apparatus is a discharge-separated type in which the discharge chamber and etching chamber are separated. CF4 and O2 are flowed into the discharge chamber as etching gases, and a 2.45 GHz microwave discharge is generated to produce active species such as fluorine radicals that contribute to etching. These active species are transported to the etching chamber via a transport tube connecting the discharge tube and the etching chamber. As a result, heat and plasma from the discharge tube do not reach the etching chamber, and isotropic etching can be performed mainly by chemical action without physical effects like those seen in RIE.

[0095] In this procedure, a mixed gas of 85 vol% CF4 and 15 vol% O2 was used with a total flow rate of 420 sccm, the chamber pressure was 40 Pa, the wafer support stage temperature was 25°C, and the processing time was 60 seconds. Afterwards, the wafers were cleaned using a washing machine.

[0096] Subsequently, using a KLA SP5 particle counter with a sensitivity of 60 nm, the number of surface defects and the haze (indicating surface roughness) after isotropic chemical dry etching were acquired. The number of surface defects was 106 pcs, and the haze was 7.21 ppm. Figure 8 shows the acquired defect map.

[0097] Here, using Haze as the indicator, the ratio of (Haze after etching) / (Haze before etching) was (7.21 / 0.06 =) 120, which was greater than 50.

[0098] The 106 defects after processing can be considered as the number of processing-induced defects. Furthermore, these 106 defects were more numerous than the 33 defects at 15nm Up before processing. This indicates that minute defects of 15nm Up or smaller became apparent and were large enough to be detected even at 60nm Up. As a result, processing-induced defects could be evaluated with higher sensitivity than before processing.

[0099] (Example 2) One mirror-finished single-crystal silicon wafer with a diameter of 300 mm, crystal orientation (100), conductivity type p-type, and resistance of 10 Ωcm was prepared. The prepared wafer was measured using a KLA SP7 particle counter with a sensitivity of 15 nm Up to obtain the initial (before isotropic chemical dry etching) surface defect count and surface roughness (Haze). The surface defect count was 11,320 pcs, and the Haze was 0.08 ppm.

[0100] Next, isotropic chemical dry etching and subsequent cleaning were performed under the same conditions as in Example 1.

[0101] Subsequently, using a KLA SP5 particle counter with a sensitivity of 60 nm, the number of surface defects and the haze (indicating surface roughness) after isotropic chemical dry etching were acquired. The number of surface defects was 157 pcs, and the haze was 6.89 ppm. Figure 9 shows the acquired defect map.

[0102] Here, using Haze as the indicator, the ratio of (Haze after etching) / (Haze before etching) was (6.89 / 0.08 =) 86, which was greater than 50.

[0103] The number of defects after processing, 157 pcs, can be considered to be the number of defects caused by the processing.

[0104] Looking at the defect map before processing, we see that the defects are distributed in a concentric pattern, and a large number of defects (11,320 pcs) are present. This is generally known to be a distribution characteristic of crystal defects containing SiO2. When a large number of crystal defects are present in this way, it has been difficult to evaluate only the small number of processing-induced defects using conventional methods. However, the isotropic chemical dry etching process of the present invention does not make the crystal defects visible, so they are not detected under conditions where the detection sensitivity is increased to 60 nm. By making only the processing-induced defects visible, even small processing-induced defects can be evaluated with high accuracy.

[0105] (Example 3) Seven single-crystal silicon wafers with a diameter of 300 mm, crystal orientation (100), conductivity type p-type, and resistance of 10 Ωcm were prepared, each with different polishing processes (e.g., CMP polishing) at different stages. Here, the processing stages are referred to as Processing Stage 1 and Processing Stage 2.

[0106] First, one wafer was selected from each of the seven wafers (designated as wafer #1), and measured using a KLA SP7 particle counter with a sensitivity of 15 nm Up to obtain the number of surface defects and the haze value, which indicates surface roughness. The wafer from processing time 1 had 28 surface defects and a haze value of 0.05 ppm. The wafer from processing time 2 had 25 surface defects and a haze value of 0.06 ppm.

[0107] Next, each of the seven sheets was subjected to isotropic chemical dry etching and subsequent cleaning under the same conditions as in Example 1.

[0108] Subsequently, using a KLA SP5 particle counter, the number of surface defects and haze after isotropic chemical dry etching were obtained at inspection sensitivities of 60nm, 120nm, and 200nm.

[0109] For each wafer (wafer #1) that was measured with SP7 before processing, the number of surface defects and surface roughness with an inspection sensitivity of 60 nm Up were 84 pcs and 6.55 ppm for the wafer processed at processing time 1, and 161 pcs and 6.73 ppm for the wafer processed at processing time 2.

[0110] Here, using Haze as the indicator, the ratio of (Haze after etching) / (Haze before etching) was 131 (=6.55 / 0.05) for wafers processed at processing time 1, and 112 (=6.73 / 0.06) for wafers processed at processing time 2. In both cases, this ratio was 50 or higher.

[0111] The number of defects after processing, 84 pcs (processing period 1) and 161 pcs (processing period 2), can be considered as processing-induced defects, and it can be determined that processing-induced defects are evident in each of the 7 pieces processed this time.

[0112] Next, Table 1 shows the number of defects for each of the seven samples at each inspection sensitivity.

[0113] [Table 1]

[0114] The average values ​​for each inspection sensitivity (Ave on the far right of Table 1) were 90.0 pcs / wafer at 60 nm Up, 69.1 pcs / wafer at 120 nm Up, and 9.1 pcs / wafer at 200 nm Up for processing period 1. For processing period 2, the values ​​were 160.1 pcs / wafer at 60 nm Up, 144.0 pcs / wafer at 120 nm Up, and 67.6 pcs / wafer at 200 nm Up.

[0115] At all inspection sensitivity levels, the number of defects was higher at processing time 2 than at processing time 1. Since the defects counted in these defect counts can be considered processing-induced defects, it was determined that more processing-induced defects were introduced to the surface at processing time 2.

[0116] Furthermore, if there are clear differences in the polishing conditions for processing period 1 and processing period 2, we can propose using the polishing conditions for processing period 1 in the future based on these results. This is because using the polishing conditions for processing period 1 can reduce processing-related defects.

[0117] (Example 4) One mirror-finished single-crystal silicon wafer with a diameter of 300 mm, crystal plane orientation (100), conductivity type p-type, and resistance of 10 Ωcm was prepared. The prepared wafer was measured using a KLA SP7 particle counter with a sensitivity of 15 nm Up to obtain the initial (before isotropic chemical dry etching) surface defect count and surface roughness (Haze). The surface defect count was 31 pcs, and the Haze was 0.06 ppm.

[0118] Next, isotropic chemical dry etching was performed using the same apparatus as in Example 1, but with different etching conditions. Specifically, a mixed gas of 50 vol% CF4 and 50 vol% O2 was used with a total flow rate of 420 sccm, the chamber pressure was 40 Pa, the wafer support stage temperature was 25°C, and the processing time was 60 seconds. Subsequently, the same cleaning procedure as in Example 1 was performed.

[0119] Subsequently, using a KLA SP5 particle counter with a sensitivity of 60 nm, the number of surface defects and the haze (indicating surface roughness) after isotropic chemical dry etching were obtained. The number of surface defects was 8 pcs, and the haze was 0.62 ppm.

[0120] Here, using Haze as an indicator, the ratio of (Haze after etching) / (Haze before etching) was (0.62 / 0.06 =) 10, which was less than 50. However, since the isotropic chemical dry etching process increased the Haze after etching to 10 times that of the Haze before etching, it can be said that processing-induced defects were sufficiently revealed.

[0121] Furthermore, the number of surface defects decreased from 31 pcs to 8 pcs before and after isotropic chemical dry etching. This is the same phenomenon that occurred in Example 2, and it is thought that while crystal defects were the main component before the treatment, after the treatment the crystal defects did not become apparent, and disappeared under conditions where the detection sensitivity was increased to 60 nm, leaving only processing-induced defects. Therefore, the number of defects after the treatment of 8 pcs can be considered to be the number of processing-induced defects.

[0122] As described above, all of Examples 1-4 demonstrated that surface processing-induced defects can be evaluated without revealing crystal defects. Therefore, processing-induced defects occurring in single-crystal silicon wafers can be evaluated simply and with high accuracy.

[0123] Furthermore, comparing Examples 1-3 and 4, it can be said that Examples 1-3 more clearly reveals processing-induced defects.

[0124] Therefore, although not particularly limited, it is more preferable to perform isotropic chemical dry etching so that (Haze after etching) / (Haze before etching) is large, preferably 50 or more as in Examples 1-3.

[0125] If the haze after etching is 50 times larger than the haze before etching, it can be more reliably determined that a processing-induced defect has become apparent.

[0126] Furthermore, although not particularly limited, it is preferable to use a mixed gas of CF4 and O2 with an O2 content of 10 to 30 vol%, as in Examples 1-3, for the isotropic chemical dry etching process.

[0127] By using this ratio, processing-induced defects can be more effectively and reliably identified.

[0128] The present invention encompasses the following aspects. [1]: A method for evaluating defects in single-crystal silicon wafers, The first step involves applying isotropic chemical dry etching to the surface of a mirror-finished single-crystal silicon wafer, A second step involves evaluating surface defects of the single-crystal silicon wafer that has undergone the isotropic chemical dry etching process, A method for evaluating defects in a single-crystal silicon wafer, characterized by comprising the following: [2]: The method for evaluating defects in a single-crystal silicon wafer according to [1] above, characterized in that, in the first step, processing-induced defects are made apparent among the defects on the surface of the mirror-finished single-crystal silicon wafer. [3]: The defect evaluation method for a single-crystal silicon wafer according to [2] above, characterized in that, in the first step, the processing-induced defects are made apparent as pit clusters composed of a plurality of pits. [4]: A method for evaluating defects in a single-crystal silicon wafer according to any one of the above [1] to [3], characterized in that, in the first step, the isotropic chemical dry etching process is performed such that (Haze after etching) / (Haze before etching) is 50 or more, using the haze value obtained by a particle counter as an indicator. [5]: The defect evaluation method for a single-crystal silicon wafer according to any one of the above [1] to [4], characterized in that the isotropic chemical dry etching process in the first step uses a gas obtained by mixing a gas containing at least fluorine atoms with an oxygen gas. [6]: The defect evaluation method for a single-crystal silicon wafer according to [5] above, characterized in that a perfluorocarbon gas is used as the gas containing fluorine atoms. [7]: The defect evaluation method for a single-crystal silicon wafer according to [5] or [6] above, characterized in that CF4 is used as the gas containing fluorine atoms. [8]: The defect evaluation method for a single-crystal silicon wafer according to any one of [5] to [7] above, characterized in that the isotropic chemical dry etching process in the first step uses a mixed gas of CF4 and O2 with an O2 content of 10 to 30 vol%. [9]: A defect evaluation method for a single-crystal silicon wafer according to any one of the above [1] to [8], characterized in that, in the second step, the number of defects is measured using a particle counter, and the measured number of defects is taken as the number of defects caused by processing.

[0129] It should be noted that the present invention is not limited to the embodiments described above. The embodiments described above are illustrative, and any configuration that has substantially the same technical idea as described in the claims of the present invention and achieves similar effects is included within the technical scope of the present invention. [Explanation of Symbols]

[0130] S1, S2, S3...process.

Claims

1. A method for evaluating defects in single-crystal silicon wafers, The first step involves applying isotropic chemical dry etching to the surface of a mirror-finished single-crystal silicon wafer, A second step involves evaluating surface defects of the single-crystal silicon wafer that has undergone the isotropic chemical dry etching process, A method for evaluating defects in a single-crystal silicon wafer, characterized by comprising the following:

2. The method for evaluating defects in a single-crystal silicon wafer according to claim 1, characterized in that, in the first step, processing-induced defects are made apparent among the defects on the surface of the mirror-finished single-crystal silicon wafer.

3. The defect evaluation method for a single-crystal silicon wafer according to claim 2, characterized in that, in the first step, the processing-induced defects are made apparent as pit clusters composed of a plurality of pits.

4. The defect evaluation method for a single crystal silicon wafer according to claim 1, characterized in that, in the first step, the isotropic chemical dry etching treatment is performed such that (Haze after etching) / (Haze before etching) is 50 or more, using the haze value obtained by a particle counter as an indicator.

5. The defect evaluation method for a single-crystal silicon wafer according to claim 1, characterized in that the isotropic chemical dry etching process in the first step uses a gas mixture of a gas containing at least fluorine atoms and oxygen gas.

6. The defect evaluation method for a single-crystal silicon wafer according to claim 5, characterized in that a perfluorocarbon gas is used as the gas containing fluorine atoms.

7. As the gas containing the aforementioned fluorine atoms, CF ４ The defect evaluation method for a single-crystal silicon wafer according to claim 5, characterized by using [the specified method].

8. The isotropic chemical dry etching process in the first step includes O ２ CF ４ and O ２ The defect evaluation method for a single-crystal silicon wafer according to claim 5, characterized by using a mixed gas.

9. The defect evaluation method for a single-crystal silicon wafer according to claim 1, characterized in that, in the second step, the number of defects is measured using a particle counter, and the measured number of defects is taken as the number of defects caused by processing.

Images