JP7622663B2 - Standard sample group for film thickness measuring device, its manufacturing method, and management method for film thickness measuring device using the standard sample group - Google Patents

Description

The present invention relates to a group of standard samples for a film thickness measuring device, a manufacturing method thereof, and a method for managing a film thickness measuring device using the group of standard samples.

In the manufacturing process of single crystal silicon wafers for semiconductor devices, their main surfaces are finished in a polishing process. In addition, a cleaning process is carried out to remove abrasives and metal impurities that have adhered to the silicon wafer surface during the polishing process. A cleaning method called RCA cleaning is used in this cleaning process.

This RCA cleaning is a cleaning method that combines SC1 (Standard Cleaning 1), SC2 (Standard Cleaning 2), and DHF (Diluted Hydrofluoric Acid) cleaning depending on the purpose. SC1 cleaning is a cleaning method that uses an alkaline cleaning solution made by mixing ammonia water and hydrogen peroxide water in any ratio, and lifts off attached particles by etching the silicon wafer surface, and removes particles while suppressing reattachment to the silicon wafer by utilizing the electrostatic repulsion between the silicon wafer and the particles. SC2 cleaning is a cleaning method that dissolves and removes metal impurities on the silicon wafer surface using a cleaning solution made by mixing hydrochloric acid and hydrogen peroxide water in any ratio. DHF cleaning is a cleaning method that removes natural oxide films on the silicon wafer surface using dilute hydrofluoric acid. In addition, ozone water cleaning, which has a strong oxidizing power, may also be used to remove organic matter adhering to the silicon wafer surface and to form a natural oxide film on the silicon wafer surface after DHF cleaning. The surface quality of the silicon wafer after cleaning, such as particle removal and surface roughness, is important, so these types of cleaning are combined depending on the purpose.

Semiconductor elements such as MOS (Metal Oxide Semiconductor) capacitors and transistors are formed on the surface of semiconductor silicon wafers. The insulating films formed on these semiconductor elements, such as the gate oxide film, are used under high electric field strength, and silicon oxide film, which is easy to form, is often used as this insulating film.

An ellipsometer is one method for evaluating the thickness of an oxide film on a silicon substrate. An ellipsometer is a device that irradiates a polarized light onto a substrate sample and measures the change in the polarization state between the incident light and the reflected light to determine the phase difference (Δ: delta) and amplitude ratio (Ψ: psi). Taking a silicon oxide film on a silicon substrate as an example, the polarization state of the incident light changes as it is reflected by the silicon oxide film on the top surface and the interface between the silicon oxide film and the silicon substrate. There are two types of ellipsometers: a single-wavelength type that uses a laser as a light source, and a spectral type that uses a white light source containing multiple wavelength components. The short-wavelength type is a method for measuring delta and psi for a specific wavelength (e.g., 633 nm). On the other hand, the spectral type can measure delta and psi for each wavelength, and it is known that the use of a spectral type with a large amount of information allows for more accurate evaluation of the film thickness (e.g., Patent Document 1).

As mentioned above, the information obtained by ellipsometer measurement is the phase difference and amplitude ratio, and it is not possible to directly calculate the film thickness. To calculate the film thickness, a model corresponding to the substrate sample is created, and the delta and psi theoretically obtained from this model are compared with the delta and psi obtained by ellipsometer measurement. The model is created by setting conditions according to the physical properties of the sample, and the set conditions include the material of the substrate and film, the film thickness of each film layer, and the optical constants of the substrate and film. In addition, a known reference corresponding to the sample and a required dispersion equation that shows the wavelength dependence of the dielectric constant and has multiple parameters are usually used to set each item.

Furthermore, a process is carried out to change the parameters of the dispersion equation and the film thickness of each film layer of the model so that the degree of difference between the two is minimized (also called fitting). The difference between the two is usually found by calculation using the least squares method, and when it is determined that the result obtained by the least squares method has become small to a certain extent through fitting, the refractive index and extinction coefficient of the film are found from the parameter values of the dispersion equation at that time, and the film thickness at that time is specified as the film thickness of the film possessed by the sample, thereby making it possible to find the film thickness. Note that model creation and fitting are generally performed manually or automatically using a computer based on the required program.

Patent Document 2 describes that the thickness of a native oxide film on a silicon wafer measured with an ellipsometer varies depending on the surface roughness. Specifically, the rougher the surface, the thicker the film thickness, and discloses a method for quantitatively evaluating surface roughness based on the correlation between roughness and the thickness of the native oxide film.

Atomic Force Microscope (AFM) is known as a method for evaluating the surface roughness of silicon substrates. Calculated average heights such as Ra and Sa are often used as roughness indices. Ra is the calculated average height over a reference length and is a two-dimensional roughness index, while Sa is a parameter that expands Ra to a surface and is a three-dimensional roughness index. The haze value obtained by a particle counter can also be used as an index of roughness. Haze is expressed as cloudiness and is widely used as an index of roughness on silicon surfaces, and a high haze level indicates that the wafer surface is rough.

A dense silicon oxide film with high insulating properties is produced by thermally oxidizing silicon wafers. Because silicon wafers shipped with a natural oxide film formed during cleaning have a natural oxide film present at the time of shipment to prevent particle adhesion, thermal oxidation is often carried out on silicon wafers on which a natural oxide film has already formed. In this case, it is known that the thickness of the thermal oxide film is affected by the quality (thickness and structure) of the natural oxide film before thermal oxidation.

In recent years, with the miniaturization and multi-layering of semiconductor integrated circuits, there is a demand for further thinning of various films, including insulating films constituting elements. This thinning requires the formation of an extremely thin insulating film, i.e., a silicon oxide film, uniformly and reproducibly within a surface or between substrates. For this purpose, it is required to control the quality of the natural oxide film at the time of shipment of the silicon wafer, particularly the film thickness, which affects the quality of the silicon oxide film. Generally, if the natural oxide film is thick, the thickness of the thermal oxide film is also thick. If it is desired to make the thermal oxide film thin, it is better that the natural oxide film is thin, and if it is desired to make the thermal oxide film thick, it is better that the natural oxide film is thick. Therefore, in recent years, there has been a particular demand for controlling the thickness of the natural oxide film with good reproducibility within a certain range. Patent Document 3 describes the relationship between silicon wafers cleaned under various conditions and the oxide film thickness after thermal oxidation. Specifically, it is disclosed that if the NH ₄ OH concentration of the SC1 cleaning solution is made high, the amount of OH groups contained in the natural oxide film increases, resulting in a thicker film thickness after thermal oxidation, and a method of controlling the film thickness after thermal oxidation by using the correlation between the structure (film quality) of the natural oxide film and the film thickness after thermal oxidation.

Thus, in recent years, it has become clear that the thickness of a natural oxide film or a film thickness after thermal oxidation changes slightly depending on cleaning conditions. For example, FIG. 6 of Patent Document 3 discloses a thermal oxide film measured by spectroscopic ellipsometry for NH ₄ OH concentration, with the low concentration side being about 5.10 nm and the high concentration side being about 5.15 nm, the difference being 0.05 nm. That is, there is an increasing demand for accurate evaluation of slight film thickness differences in such thin samples. With this demand, the management of film thickness evaluation devices (film thickness measurement devices) represented by ellipsometers is also becoming more important.

Generally, for example, silicon substrates with an oxide film are used to manage film thickness measurement equipment, and the thickness of the oxide film is at least about 25 nm, and may be even thicker. Therefore, in general equipment management, the film thickness of a silicon substrate with an oxide film of at least 25 nm or more is obtained, and if it is compared with a film thickness value obtained previously and is equivalent, it can be determined that a valid film thickness evaluation has been performed. If necessary, the validity can be evaluated in more detail by using statistical methods.

In this way, when the film thickness is thick, 25 nm or more, the equipment can be managed by the conventional method. In other words, when measuring a natural oxide film with a thickness of about 1 nm or an extremely thin thermal oxide film with a thickness of about 5 nm, even if a film thickness value of 25 nm or more is found to be valid, it cannot be determined that the film thickness of an extremely thin oxide film with a thickness of about 1 nm or 5 nm is valid. Furthermore, the guaranteed lower limit of the film thickness value of commercially available film thickness measurement devices is thicker than 1 nm or 5 nm, for example, a film thickness value of about 25 nm, which shows that it is difficult to accurately measure an extremely thin thermal oxide film with a thickness of about 1 nm or 5 nm and to manage the equipment, and the current situation is that equipment management is only possible when the film thickness is thick.

JP 2005-283502 A Japanese Patent Application Publication No. 6-163662 Patent No. 6791453 Patent No. 4192410

As mentioned above, when measuring the thickness of a natural oxide film with a thickness of about 1 nm or an extremely thin thermal oxide film with a thickness of about 5 nm, there is a demand for an equipment management method that can verify the validity of the obtained thickness value. This is especially true when the thickness is thinner than 1 nm. As such an equipment management method, it is easy to imagine a method in which a Si substrate with a natural oxide film with a thickness of about 1 nm is prepared, and the thickness of the prepared Si substrate is obtained with a thickness measurement device each time a thickness measurement is performed, and compared with the thickness value obtained in the past. However, when the inventors performed this method, the thickness value tends to increase with the length of time left unattended, especially when the thickness is as thin as about 1 nm, and even when comparing with the past thickness values, it was not possible to obtain confirmation of the validity of the measurement device. The reason for this is presumably that the natural oxide film with a thickness of about 1 nm grew due to being left unattended in the air, causing the thickness to increase. It is thought that moisture and oxygen in the air became oxidizing species, diffused through the natural oxide film, reached the interface between the oxide film and the substrate, and the oxidation reaction proceeded. As such, thin oxide films, especially natural oxide films with a thickness of about 1 nm, become thicker the longer they are left exposed, so it is not possible to perform valid device management using a single standard sample.

Patent Document 4 also discloses the effect of exposure time when the gate oxide film is about 10 nm (100 Å). In this case, it is described that the oxide film appears thicker due to deposits on the oxide film. However, it is very time-consuming to obtain the exposure time dependency of the film thickness value. Furthermore, Figure 6 of Patent Document 4 shows the corrected film thickness values of the gate oxide film for many samples with different exposure times, but the values range from about 9.7 nm (97 Å) to about 10 nm (100 Å), a difference of about 0.3 nm. This 0.3 nm is much larger than the above-mentioned 0.05 nm, and it is unlikely that a slight difference in film thickness between cleaning levels can be detected.

The present invention has been made to solve the above problems, and aims to provide a standard sample group that can stably and accurately measure oxide films and manage film thickness measuring devices even when the oxide film is very thin, a method for manufacturing the standard sample group, and a method for managing film thickness measuring devices using the standard sample group.

In order to achieve the above object, the inventors conducted extensive research into the relationship between native oxide film-coated Si substrates with various surface roughnesses and the thickness of the native oxide film measured by an ellipsometer. They discovered that when the surface roughness formed by SC1 cleaning is greater than a certain level, the oxide film becomes slightly thicker, and that the native oxide film formed by ozone water cleaning continues to oxidize uniformly even after being left for a certain period of time, thus completing the present invention.

That is, we provide a standard sample group for a film thickness measuring device, which includes a plurality of silicon substrates having different surface roughness, each of which has an ozone oxide film of a different thickness on its surface.

Such a standard sample group for a film thickness measuring device is capable of stably and accurately measuring the thickness of an oxide film even when the film thickness is very thin, and is a standard sample group that can be used to determine the validity of the film thickness measuring device and stably manage the condition of the film thickness measuring device.

In this case, the multiple silicon substrates having different surface roughnesses can be three or more silicon substrates each having a different surface roughness.

By providing three or more levels of surface roughness, the validity of the film thickness values output by the measuring device can be evaluated more stably, in more detail, and with higher accuracy based on the relationship between the film thicknesses.

The present invention also provides a method for manufacturing a standard sample group for a film thickness measuring device, which includes preparing a plurality of silicon substrates having the same surface quality, performing SC1 cleaning on each of the plurality of silicon substrates under different SC1 cleaning conditions to produce a plurality of silicon substrates each having a different surface roughness, removing the SC1 oxide film formed on the surface of the plurality of silicon substrates having different surface roughness by the SC1 cleaning with hydrofluoric acid cleaning to expose the bare surface, and cleaning the plurality of silicon substrates having different surface roughness with the exposed bare surface with ozone water under the same conditions to form an ozone oxide film, thereby producing a plurality of silicon substrates each having an ozone oxide film with a different thickness, thereby manufacturing a standard sample group.

This manufacturing method for standard sample groups for film thickness measuring devices makes it possible to stably and accurately measure oxide films even when the film thickness is very thin, and it is possible to manufacture standard sample groups that enable stable and accurate management of film thickness measuring devices.

In this case, the surface roughness can be measured using either the Haze value measured using a laser scattering particle counter or the Sa value or Sq value measured using an atomic force microscope.

This type of roughness index allows for more suitable evaluation of surface roughness, including roughness components that affect the oxide film thickness, and can be evaluated quickly and easily, especially when a laser scattering particle counter is used.

In this case, the SC1 cleaning conditions can be one or more of the SC1 chemical concentration, cleaning temperature, or cleaning time.

By varying these conditions, the etching behavior of the SC1 cleaning can be controlled, making it easy to create a group of standard samples with slightly different film thicknesses.

In this case, the film thickness measuring device can be an ellipsometer.

The present invention is particularly suitable when using an ellipsometer.

Furthermore, a method for managing a film thickness measuring device is provided, which uses a film thickness measuring device to measure the film thickness of an oxide film of a standard sample and manages the film thickness measuring device based on the measurement results, and uses the standard sample group or a standard sample group manufactured by the manufacturing method for a standard sample group for the film thickness measuring device as the standard sample, and manages the film thickness measuring device based on the film thickness measurement results of the ozone oxide film of the standard sample group.

With this type of film thickness control method, even when controlling the film thickness range of an extremely thin oxide film, the validity of the film thickness measurement device can be determined with stable precision, and the device condition can be controlled stably. In particular, slight differences in film thickness can be evaluated with stable precision without being affected by the time the standard samples are left standing, and the validity of the film thickness value output from the film thickness measurement device can be determined based on this evaluation result.

At this time, when managing the film thickness measurement device based on the film thickness measurement results, the magnitude relationship of the film thickness measurement results of each of the standard samples can be compared.

Ozone oxide films, which are natural oxide films formed with ozone water, oxidize uniformly, so the relationship between film thicknesses within a sample group is maintained, making this type of management method more desirable.

In this case, the film thickness measuring device can be an ellipsometer.

The present invention allows for more efficient film thickness control when using an ellipsometer.

As described above, the standard sample group for the film thickness measuring device of the present invention is capable of stably and accurately measuring the film thickness of an oxide film even when the film thickness is very thin, and it is possible to judge the validity of the film thickness measuring device and manage the state of the film thickness measuring device. In addition, the manufacturing method of the standard sample group for the film thickness measuring device of the present invention makes it possible to manufacture the standard sample group for the film thickness measuring device as described above. Furthermore, with the management method of the present invention, it is possible to stably and accurately judge the validity of the film thickness measuring device even when managing the film thickness range of an extremely thin oxide film, and it is possible to stably manage the state of the device.

1 is a flowchart showing a method for producing a standard sample group for a film thickness measuring device. 1 is a graph showing the haze increase and etching amount when a bare silicon substrate is cleaned at different SC1 cleaning temperatures. 1 is a graph showing the results of film thickness evaluation one day and three months after sample production, in a case where SC1 cleaning, HF cleaning, and ozone water cleaning were performed (A), and in a case where only SC1 cleaning was performed (B). 1 is a graph showing the difference in film thickness for each cleaning temperature when SC1 cleaning, HF cleaning, and ozone water cleaning were performed (A), and when only SC1 cleaning was performed (B), with the SC1 cleaning temperature of 35° C. being used as a reference. 1 is a flowchart showing an investigation of the relationship between wafer surface roughness and oxide film thickness. 6 is a graph showing the relationship between the surface roughness (haze) of the silicon substrate, which has been roughened by CMP and SC1 cleaning in FIG. 5, and the native oxide film (A) and the 5 nm thermal oxide film (B). 6 is a graph showing the relationship between the surface roughness (haze) and the native oxide film (A) and the 5 nm oxide film (B) of a silicon substrate subjected to the roughening treatment of FIG. 5 by single-wafer cleaning. FIG. 6 is a graph showing the thickness of the natural _oxide film (ozone _oxide film) and the _{amount of} haze deterioration (ΔHaze value) before and after SC1 cleaning after the roughening treatment in FIG. 5 was performed by batch cleaning with a liquid composition of NH 4 OH:H 2 O 2 :H 2 O = 1:1:10, a cleaning time of 3 minutes, and a cleaning temperature of 35 to 80° C.

The present invention is described in detail below, but is not limited to these.

In this specification, the term "native oxide film" includes oxide films formed when a silicon substrate is left in the air, as well as oxide films formed during polishing and oxide films formed by cleaning a silicon substrate (sometimes called chemical oxide films). When such natural oxide films are to be further differentiated based on differences in film quality (structure), for example, those formed by ozone water cleaning are called "ozone oxide films," and those formed by SC1 cleaning are called "SC1 oxide films." As described below, it is known that "ozone oxide films" and "SC1 oxide films" have different film qualities (structures).

As described above, there was a need to provide a method for managing a film thickness measurement device that can accurately measure oxide films even when the oxide film is very thin, as well as a group of standard samples for performing such management and a method for manufacturing the same.

After extensive research into the above problem, the inventors discovered that a standard sample group for a film thickness measuring device includes a plurality of silicon substrates having different surface roughness, each of which has an ozone oxide film of a different thickness on its surface. This standard sample group for a film thickness measuring device can stably and accurately measure the thickness of an oxide film even when the film thickness is very thin, and can be used to determine the validity of a film thickness measuring device and to manage the condition of the film thickness measuring device, thus completing the present invention.

The inventors have also discovered that a method for manufacturing standard samples for a film thickness measuring device, which involves preparing multiple silicon substrates having the same surface quality, performing SC1 cleaning on each of the multiple silicon substrates under different SC1 cleaning conditions to produce multiple silicon substrates with different surface roughnesses, removing the SC1 oxide film formed on the surfaces of the multiple silicon substrates with different surface roughnesses by the SC1 cleaning with hydrofluoric acid cleaning to expose bare surfaces, and cleaning the multiple silicon substrates with different surface roughnesses with the exposed bare surfaces under the same conditions with ozone water to form an ozone oxide film, thereby producing multiple silicon substrates with ozone oxide films with different thicknesses, makes it possible to stably and accurately measure oxide films even when the film thickness is very thin, and produces standard samples that allow for stable and accurate management of the film thickness measuring device, and has completed the present invention.

The following explanation will be given with reference to the drawings.

First, we will discuss the relationship between the various wafer surface roughnesses formed in the silicon wafer manufacturing process and the thickness of the oxide film. Figure 5 is the investigation flow chart. A roughening process was performed on the prepared silicon substrate by changing the CMP processing conditions and SC1 cleaning conditions to form roughness. Next, the oxide film was completely removed by hydrofluoric acid cleaning in a batch cleaning machine, and then an oxide film was formed by cleaning with ozone water. In both the CMP level and the SC1 cleaning level, the oxide film formed during the roughening process was completely peeled off by hydrofluoric acid cleaning, and the ozone oxide film formed by the subsequent ozone water cleaning was formed, so it can be interpreted that the oxide film was formed by the same method. Next, haze measurement was performed using a particle counter. After that, some of the wafers were thermally oxidized with a target thickness of 5 nm, and the thickness of the natural oxide film (ozone oxide film) and the thermal oxide film with a target thickness of 5 nm (hereinafter referred to as "5 nm thermal oxide film") were evaluated by spectroscopic ellipsometry.

Figure 6 shows the film thickness of the native oxide film (A) and the 5 nm thermal oxide film (B) for each level of haze. At the CMP level (■), the film thickness of the native oxide film and the 5 nm oxide film were the same even when the haze value exceeded 10 ppm, but at the SC1 cleaning level (●), there was a tendency for both the native oxide film and the 5 nm thermal oxide film to become thicker as the haze increased. Since the oxide films were formed using the same method, it was assumed that the film thickness would be the same, as with the CMP level, but this was not the case with the SC1 cleaning level.

Furthermore, Figure 7 shows the thickness of the native oxide film (A) and the 5 nm thermal oxide film (B) when the roughening process was performed by repeating a cycle of hydrofluoric acid cleaning and ozone water cleaning with single-wafer cleaning. The hydrofluoric acid cleaning and ozone water cleaning after the roughening process were also performed by the single-wafer method rather than the batch method, so the oxide film formation method is different from the batch method ozone water cleaning. Therefore, it is not possible to compare the film thickness of the above-mentioned SC1 level and CMP level with the single-wafer cleaning level, but it is possible to discuss the effect of haze within the single-wafer cleaning level. As a result, the single-wafer cleaning level had the same thickness of the native oxide film and the 5 nm thermal oxide film even when the haze changed, just like the CMP level.

In summary, the above results show that the roughness created by CMP and single-wafer cleaning does not affect the thickness of the oxide film, while the roughness created by SC1 cleaning increases the thickness of the oxide film.

Here, the results of an additional investigation on the SC1 cleaning level will be described. The roughening treatment was performed by batch cleaning with a liquid composition of NH ₄ OH:H ₂ O ₂ :H ₂ O=1:1:10, a cleaning time of 3 min, and a cleaning temperature of 35 to 80° C., and the oxide film was completely removed by hydrofluoric acid cleaning, and then cleaning with ozone water was performed. FIG. 8 is a graph showing the thickness of the natural oxide film (ozone oxide film) and the amount of haze deterioration (ΔHaze value) before and after SC1 cleaning. Focusing on the SC1 temperature from 35° C. to 60° C., the higher the temperature, the higher the ΔHaze value, and the thicker the thickness of the natural oxide film tends to be. This is because the film thickness is thicker due to the influence of the wafer surface roughness described above. Next, focusing on 60° C. and 80° C., the ΔHaze value increased at 80° C., which is a high temperature at which the etching reaction is likely to proceed, but the thickness of the natural oxide film was the same. It can be seen that when a roughness of a certain value or more is formed, the effect of thickening the native oxide film due to this surface roughness saturates. In other words, it can be interpreted that the surface roughness at a temperature of 80° C. is higher than that at 60° C., but the specific roughness components that thicken the film are equivalent.

From the above findings, it can be interpreted that wafer surface roughness is a factor that affects the oxide film thickness.

In the present invention, roughness components that increase the thickness of this oxide film are intentionally formed, and multiple samples with slightly different oxide film thicknesses are used as a standard sample group, with the aim of stably and accurately evaluating slight differences in film thickness, which was previously difficult, and managing the film thickness measurement device based on the evaluation results. The details are described below. Note that the standard sample group of the present invention is particularly advantageous when evaluating and managing extremely thin oxide film thicknesses, but it can also be applied to cases where the film thickness is large, and is not limited to thin oxide films of about 1 nm.

(Standard sample group)
The standard sample group according to the present invention includes a plurality of silicon substrates having different surface roughnesses, each of which has an ozone oxide film of a different thickness on its surface. The standard sample group according to the present invention is capable of stably and accurately measuring the thickness of an oxide film even when the film thickness is very thin, and is a standard sample group capable of judging the validity of a film thickness measuring device and managing the state of the film thickness measuring device. By relatively comparing the film thicknesses between the levels of such a standard sample group, it is not necessary to use the absolute value of the film thickness obtained from the film thickness measuring device as an index. In addition, the oxide film formed by ozone water can be oxidized uniformly even after being left for a predetermined time, and can be used for a long period of time.

In such a group of standard samples according to the present invention, it is preferable that the multiple silicon substrates having different surface roughnesses are three or more silicon substrates each having a different surface roughness. By providing three or more levels of surface roughness, the validity of the film thickness value output from the measuring device can be evaluated more stably, in detail, and with high accuracy based on the magnitude relationship of the film thicknesses.

The index representing the surface roughness is not particularly limited, but the Haze value measured by a laser scattering particle counter, the Sa value measured by an atomic force microscope, the Sq value, etc. can be used as an index. Such surface roughness indexes can more appropriately evaluate roughness including roughness components that affect the oxide film thickness. In particular, the use of a laser scattering particle counter allows for quick and easy evaluation.

Details of the advantageous effects of the standard sample group according to the present invention will be explained later in the explanation of the management method of the film thickness measurement device with reference to Figures 3 and 4, etc.

(Method of manufacturing standard samples)
Next, referring to FIG. 1, a method for manufacturing a standard sample group used for managing a film thickness measuring device will be described. First, a plurality of silicon substrates are prepared. In this case, there is no restriction on the conductivity type, diameter, or sample shape. The silicon substrate to be prepared preferably has a small surface roughness, for example, a surface roughness of 0.5 nm or less in terms of Sa value is more preferable. The reason for this is that if the surface roughness is 0.5 nm or less, the oxide film thickness calculated by spectroscopic ellipsometry is stable at a value of about 1 nm. In addition, as a general polishing process for silicon wafers, DSP (double-sided polishing) is followed by CMP (single-sided polishing) on the front side, and cleaning is performed after each polishing process. Generally, the Sa value of the front side of a silicon wafer after CMP is 0.1 nm or less, and the Sa value of the back side (DSP surface) is about 0.2 to 0.4 nm, so that a wafer after at least DSP processing and CMP processing is preferable for the production of the standard sample group of the present invention. Furthermore, a wafer after CMP processing with a smaller Sa value is more preferable.

Next, SC1 cleaning is performed on the prepared silicon substrates under different cleaning conditions. The surface roughness formed by SC1 cleaning makes the oxide film thickness slightly thicker, so SC1 cleaning is performed to prepare samples with slightly different film thicknesses. As cleaning conditions, it is desirable to change one or more of the SC1 chemical solution concentration, cleaning temperature, and cleaning time. These are conditions that are easy to change in practical operation. For example, the chemical solution concentration may be adjusted in the range of NH ₄ OH:H ₂ O ₂ :H ₂ O=1:1:5 to 1:1:100. Also, for example, the cleaning temperature may be adjusted in the range of 30 to 60° C. As shown in FIG. 8, for example, when the SC1 temperature is set to 40° C., 50° C., and 60° C., a standard sample group with slightly different ozone oxide film thicknesses can be obtained. Also, for example, the cleaning time may be adjusted in the range of 0.5 to 10 minutes. The etching behavior of the chemical solution changes depending on these condition changes. As a result, the surface roughness of the silicon substrate changes, and the thickness value of the native oxide film obtained by the film thickness measuring device differs slightly depending on the roughness. In this case, it is more preferable to prepare three or more samples with different surface roughnesses for three or more levels (conditions). If there are three or more samples, the magnitude relationship between the samples can be evaluated.

In addition, according to the inventors' investigation, the conditions under which the thickening effect of the ozone oxide film due to the wafer surface roughness becomes saturated and the amount of thickening vary depending on the liquid composition and cleaning time. Therefore, it is more preferable to carry out a basic investigation as described above (Figures 5 to 8) in advance and determine the SC1 cleaning conditions for the standard sample group based on the results of this basic investigation.

In addition, the index representing the surface roughness is not particularly limited, but the Haze value measured by a laser scattering particle counter, the Sa value measured by an atomic force microscope, the Sq value, etc. can be used as an index. Such surface roughness indexes can more appropriately evaluate roughness including roughness components that affect the oxide film thickness. In particular, the use of a laser scattering particle counter allows for quick and easy evaluation.

As an example, FIG. 2 shows the surface roughness and etching amount when SC1 cleaning was performed on a bare silicon wafer with a chemical concentration of NH ₄ OH:H ₂ O ₂ :H ₂ O=1:1:10, a cleaning time of 3 min, and a cleaning temperature of 35 to 55°C. The surface roughness was measured using the haze value obtained by a laser scattering particle counter SP3 as an index, and the increase after cleaning (ΔHaze) was calculated. The larger this value, the rougher the surface is and the greater the surface roughness. The etching amount was calculated by using an SOI wafer (Si layer/SiO ₂ layer/Si substrate) and the difference in Si layer thickness before and after cleaning. As a result, both haze and etching amount tended to increase as the SC1 cleaning temperature increased. This can be interpreted as the etching amount increasing due to the etching reaction proceeding more at higher temperatures, and as a result, the roughness index haze also increased. In this way, it can be seen that the etching behavior changes by changing the SC1 cleaning conditions, and the roughness formed as a result changes.

Next, the silicon substrate after SC1 cleaning is cleaned with hydrofluoric acid to completely remove the oxide film and expose the bare surface. After that, multiple silicon substrates with different surface roughness with exposed bare surfaces are cleaned with ozone water under the same conditions to form an ozone oxide film, and a group of standard samples is produced by making multiple silicon substrates with ozone oxide films of different thicknesses. The ozone oxide film formed with ozone water is thicker than the SC1 oxide film formed by SC1 cleaning, and oxidation progresses uniformly even after leaving it for a specified time. Therefore, after SC1 cleaning to form roughness, the oxide film is completely removed by cleaning with hydrofluoric acid, and then the ozone oxide film is formed by cleaning with ozone water. Note that there are no restrictions on the hydrofluoric acid cleaning conditions as long as the oxide film can be completely removed. For example, the hydrofluoric acid concentration can be 0.3 to 5.0 wt%, the temperature can be 10 to 30°C, and the cleaning time can be 30 to 360 seconds. The concentration of ozone water used is in the range of 3 to 25 ppm, the temperature is 10 to 30°C, and the cleaning time is 30 to 360 seconds. If a batch-type cleaning machine is used, the process can be carried out in one batch to reduce the amount of work. This is the method for producing the standard samples.

(Management method of film thickness measuring device)
Next, a method for managing a film thickness measuring device using a standard sample group according to the present invention will be described. The method for managing a film thickness measuring device according to the present invention is a method for managing a film thickness measuring device that measures the film thickness of an oxide film of a standard sample using a film thickness measuring device and manages the film thickness measuring device based on the measurement result. The standard sample group according to the present invention or a standard sample group manufactured by the manufacturing method for a standard sample group according to the present invention is used as a standard sample, and the film thickness measuring device is managed based on the film thickness measurement result of an ozone oxide film of the standard sample group. In particular, by comparing the magnitude relationship of each film thickness measurement result of the standard sample group, it is possible to manage the film thickness measuring device more accurately and stably. Note that the film thickness measuring device to which the method for managing a film thickness measuring device according to the present invention can be applied is not particularly limited, but is more suitable when an ellipsometer is used.

Here, in order to explain why an ozone oxide film is formed by cleaning with ozone water, in addition to the standard sample group of the present invention that was subjected to the above-mentioned conditions, i.e., SC1 cleaning, HF cleaning, and ozone water cleaning under the SC1 cleaning conditions, a sample group (hereinafter referred to as the "control sample group") that was subjected to only SC1 cleaning under the SC1 cleaning conditions, without HF cleaning or ozone water cleaning, was also prepared.

The SC1 cleaning conditions for forming the surface roughness are the same for both, but the difference is whether or not HF cleaning and ozone water cleaning are used. Since HF cleaning and ozone water cleaning do not roughen the surface like SC1 cleaning, it can be interpreted that the two sample groups have the same surface roughness at the same SC1 cleaning level, but the cleaning methods for forming the natural oxide film are different. Specifically, the standard sample group of the present invention is an ozone oxide film, which is a natural oxide film formed by ozone water cleaning, while the other control sample group is an SC1 oxide film, which is a natural oxide film formed by SC1 cleaning, and the film quality (structure) is different. Furthermore, since the cleaning temperature (cleaning conditions) is different with only SC1 cleaning, it can be interpreted that the film quality is different between each level of the control sample group with only SC1 cleaning. On the other hand, the standard sample group of the present invention, which has been subjected to ozone water cleaning, is subjected to ozone water cleaning under the same conditions, so the film quality is interpreted as being equivalent.

To summarize the above, when comparing both sample groups within each sample group, the standard sample group that underwent ozone water cleaning has different surface roughness between levels due to different SC1 cleaning temperatures, but the oxide film quality is the same between levels because the ozone water cleaning was performed under the same conditions. The control sample group that only underwent SC1 cleaning has different surface roughness between levels due to different SC1 cleaning temperatures, and because the ozone water cleaning was not performed under the same conditions, it can be interpreted that the oxide film (SC1 oxide film) quality is also different due to the influence of the SC1 cleaning temperature. Furthermore, when comparing both sample groups at the same SC1 cleaning level, it can be interpreted that the surface roughness is the same but the oxide film quality is different because the SC1 cleaning conditions are the same.

The thickness of the native oxide film was evaluated for these sample groups (standard sample group according to the present invention, control sample group) using a spectroscopic ellipsometer, which is a film thickness measuring device. In this case, in order to investigate the change over time, measurements were taken one day and three months after the date of sample preparation. The results are shown in Figure 3. First, looking at the measurement results one day after the sample preparation, the standard sample group (A) that underwent SC1 cleaning, HF cleaning, and ozone water cleaning according to the present invention, and the control sample group (B) that only underwent SC1 cleaning, both showed a tendency for the film thickness to increase slightly as the SC1 cleaning temperature increased. More specifically, for the standard sample group that underwent ozone water cleaning, the SC1 cleaning temperature was 1.281 nm at 35°C and 1.313 nm at 55°C, and the difference in film thickness between the two was 0.032 nm. In contrast, for the control sample group that only underwent SC1 cleaning, the SC1 cleaning temperature was 1.052 nm at 35°C and 1.111 nm at 55°C, and the difference in film thickness between the two was 0.059 nm. The difference in film thickness for the standard sample group that was cleaned with ozone water can be interpreted as being influenced only by surface roughness, since the surface roughness differs but the film quality is the same, as in the comparison within the sample group described above. In contrast, the difference in film thickness for the control sample group that was only cleaned with SC1 can be interpreted as being influenced by both film quality and surface roughness, since both the surface roughness and film quality differ, as in the comparison within the sample group described above. Therefore, the control sample group that was only cleaned with SC1, which has two influencing factors, has a larger film thickness difference than the standard sample group that had an ozone oxide film formed.

The actual film thickness of the standard sample group, which had been cleaned with ozone water, was thicker than that of the control sample group, which had only been cleaned with SC1. This is also interpreted as a difference in the method of forming the natural oxide film, and it is thought that the oxidation reaction proceeds more easily when ozone water, which has a stronger oxidizing power, is used, resulting in a thicker film.

Next, looking at the measurement results three months after sample preparation, both sample groups tended to have a thicker film thickness after three months than after one day. This is because oxidizing species in the air diffused into the natural oxide film during the exposure period, and the oxidizing species that reached the interface between the oxide film and the silicon substrate promoted the oxidation reaction. The second axis of Figures 3(A) and (B) shows the difference in film thickness after three months and one day at the same SC1 cleaning level. When comparing the difference in film thickness between the sample groups, the control sample group with only SC1 cleaning had a larger difference in film thickness than the standard sample group of the present invention, which was cleaned with SC1, HF, and ozone water, and oxidation had progressed more. This is because the control sample group with only SC1 cleaning had a thinner natural oxide film thickness one day after sample preparation, which made it easier for oxidizing species in the air to reach the substrate interface and promote oxidation. Since the amount of film thickening after a specified exposure time depends on the thickness of the natural oxide film before exposure, it can be determined that the main cause of the thickening of the natural oxide film by exposure to air, especially at about 1 nm, is the oxidation reaction. The influence of the deposits described in Patent Document 4 cannot be completely denied, but if the deposits were the main cause, the thickness would not depend on the thickness of the native oxide film before it was left standing, and should be the same thickness. The film thickness of 10 nm disclosed in Patent Document 4 is probably much thicker than 1 nm, so even if it was left standing, the oxidizing species would not reach the substrate interface, and it is presumed that the deposits are the main cause.

Next, when the aforementioned difference in film thickness was compared between the sample groups, the standard sample group up to the ozone water cleaning of the present invention was about 0.05 nm, independent of the SC1 cleaning temperature, and oxidation had progressed uniformly, whereas the control sample group with SC1 cleaning only fluctuated widely within a range of about 0.15 to 0.20 nm, and oxidation had progressed non-uniformly. Furthermore, when comparing the levels within the control sample group with SC1 cleaning only, after three months, the film thickness was thinnest with an SC1 cleaning temperature of 40°C. After one day, the film thickness was thinnest with 35°C, and a reversal of the relationship in film thickness between the levels had also occurred.

Therefore, it can be seen that the standard sample group that has been subjected to the ozone cleaning of the present invention can reduce the amount of thickening due to oxidation even if left for a specified time, and furthermore, by allowing the oxidation to proceed uniformly, the relationship in film thickness between sample levels can be maintained even after the specified time has been left for a specified time, even if there is a slight difference in film thickness. From the above, in the present invention, it is necessary to make the final natural oxide film an ozone oxide film formed with ozone water, which is less affected by the time it is left for.

A more detailed comparison between the standards was carried out. Figure 4 shows the results of calculating the difference in film thickness one day after sample preparation and three months after sample preparation, using the SC1 cleaning temperature of 35°C, which was the thinnest film thickness one day after sample preparation, as the standard. As shown in Figure 4 (A), in the standard sample group up to the ozone water cleaning of the present invention, the difference in film thickness was larger at higher SC1 cleaning temperatures, regardless of the standing time, and the absolute value of the difference in film thickness at the same cleaning temperature was also equivalent. From this result, it can be seen that even a slight difference in film thickness of about 0.01 nm can be detected, and when evaluating the film thickness with a film thickness measuring device, if the film thickness of the standard sample group of the present invention is evaluated and the same difference in film thickness is obtained, it can be determined that even a film thickness difference of 0.01 nm between the samples to be evaluated is a significant difference. In this way, by managing the film thickness measuring device using the standard sample group of the present invention, it is possible to ensure reliability even for film thickness results of about 1 nm, which is particularly thin and difficult to manage in the past.

In contrast, as shown in Figure 4 (B), for the sample group that was only cleaned with SC1, the relationship between the film thicknesses was reversed, especially at SC1 cleaning temperatures of 40 and 45°C, so the differential film thickness was negative. Furthermore, although the differential film thickness values at SC1 cleaning temperatures of 50 and 55°C were positive, there was a large discrepancy in the differential film thickness values; for example, at 55°C, the film thickness was 0.059 nm after one day, while it was 0.036 nm after three months, for a difference of 0.023 nm. From these results alone, it is difficult to determine whether such a difference in film thickness is due to the film thickness measurement device or the cleaning conditions. Therefore, sample groups in which the relationship between the film thicknesses is reversed due to the exposure time are not suitable for device management.

However, from the measurement results of the standard sample group of the present invention, it can be determined that the difference of 0.023 nm is due to little to the equipment, and is due to the cleaning conditions and the time left standing. Furthermore, since it can be used even after being left standing for three months, by using the same sample for a long period of time, it is possible to stably manage the equipment without the trouble of making new standard samples. If the film thickness evaluation results of the standard sample group of the present invention are significantly different, for example, if the relationship between the film thicknesses is reversed, the equipment can be inspected to check for abnormalities, and after making appropriate repairs, the film thickness can be evaluated again to confirm whether the results are equivalent to the film thickness of the standard sample group of the present invention (film thickness difference between levels), and the equipment can be managed.

With such a group of standard samples for film thickness measuring devices, the manufacturing method thereof, and the management method using the group of standard samples, it is possible to stably and accurately evaluate the oxide film thickness even when the film thickness is very thin, for example, about 1 nm. Furthermore, it is also possible to detect abnormalities in the management of the film thickness measuring device.

The present invention will be described in detail below with reference to examples, but the present invention is not limited thereto.

(Examples 1 to 5)
As silicon substrates for the standard sample group, a plurality of silicon wafers with a diameter of 300 mm were prepared, which were subjected to CMP polishing and single-wafer cleaning using a combination of hydrofluoric acid and ozone water. First, the haze value before cleaning was obtained using a KLA laser scattering particle counter SP3. Next, cleaning was performed in the order of SC1 cleaning, hydrofluoric acid cleaning, and ozone water cleaning using a batch cleaning machine under the cleaning conditions described below. The SC1 cleaning conditions were NH ₄ OH:H ₂ O ₂ :H ₂ O=1:1:10, cleaning time 3 minutes, and cleaning temperatures 35, 40, 45, 50, and 55° C. The hydrofluoric acid cleaning was performed at 0.5 wt%, cleaning temperature 25° C., and cleaning time 3 minutes, and the ozone water cleaning was performed at 20 ppm, cleaning temperature 25° C., and cleaning time 3 minutes. Next, the haze of the silicon wafers after cleaning with ozone water was evaluated using a laser scattering particle counter SP3. Thereafter, the increase in haze before and after cleaning was calculated. The results were 0.0067 ppm at an SC1 cleaning temperature of 35° C., 0.0093 ppm at 40° C., 0.0106 ppm at 45° C., 0.0142 ppm at 50° C., and 0.0201 ppm at 55° C. It was confirmed that the higher the temperature at which the etching reaction proceeds, the greater the increase in haze and the rougher the surface.

Next, the thickness of the natural oxide film (ozone oxide film) was evaluated using a J. A. Woollam M-2000V spectroscopic ellipsometer. The film thickness was evaluated one day and three months after the sample was made. The results are shown in Table 1. After one day, the film thickness tended to be slightly thicker as the SC1 cleaning temperature increased, and the difference in film thickness from the case of an SC1 cleaning temperature of 35°C was 0.001 nm at an SC1 cleaning temperature of 40°C, 0.007 nm at 45°C, 0.020 nm at 50°C, and 0.032 nm at 55°C, confirming that a film thickness difference of at least 0.01 nm could be detected. Next, when the results were evaluated after three months, the film thickness was thicker after three months than after one day due to natural oxidation during the leaving period, but the difference in film thickness within each SC1 cleaning temperature level was about 0.05 nm at all levels. Furthermore, after one day, there was a tendency for the film thickness to be slightly thicker as the SC1 cleaning temperature increased, and the difference in film thickness with an SC1 cleaning temperature of 35°C was 0.002 nm at an SC1 cleaning temperature of 40°C, 0.006 nm at 45°C, 0.021 nm at 50°C, and 0.035 nm at 55°C, roughly the same as after one day. From these results, it was possible to detect a film thickness difference of 0.01 nm even after three months, and it was confirmed that there were no problems with the equipment condition.

(Comparative Examples 1 to 5)
A plurality of silicon wafers having the same diameter of 300 mm as in the example were prepared. First, the haze value before cleaning was obtained using a KLA laser scattering particle counter SP3. Next, only SC1 cleaning was performed using a batch cleaning machine under the cleaning conditions described below. The SC1 cleaning conditions were NH ₄ OH:H ₂ O ₂ :H ₂ O=1:1:10, cleaning time 3 minutes, and cleaning temperatures 35, 40, 45, 50, and 55° C. Next, the haze of the silicon wafers after SC1 cleaning was evaluated using the laser scattering particle counter SP3. Thereafter, the amount of haze increase after cleaning was calculated. The results were 0.0071 ppm at an SC1 cleaning temperature of 35° C., 0.0091 ppm at 40° C., 0.0108 ppm at 45° C., 0.0141 ppm at 50° C., and 0.0204 ppm at 55° C. It was confirmed that the higher the temperature at which the etching reaction proceeds, the greater the amount of haze increase and the rougher the surface.

Next, the thickness of the natural oxide film (SC1 oxide film) was evaluated using a J. A. Woollam M-2000V spectroscopic ellipsometer. The film thickness was evaluated one day and three months after the sample was prepared. The results are shown in Table 1. One day after preparation, the higher the SC1 cleaning temperature, the slightly thicker the film thickness was. The difference in film thickness from an SC1 cleaning temperature of 35°C was 0.007 nm at an SC1 cleaning temperature of 40°C, 0.026 nm at 45°C, 0.035 nm at 50°C, and 0.059 nm at 55°C, confirming that a film thickness difference of at least 0.01 nm could be detected. The results after three months showed that the film was thicker after three months than after one day due to natural oxidation during the storage period, and the difference in film thickness within each SC1 level varied greatly within the range of 0.157 to 0.204 nm. Furthermore, the film thickness was thicker at an SC1 cleaning temperature of 35°C than at SC1 cleaning temperatures of 40 and 45°C, and there was no tendency for the film thickness to become slightly thicker with higher SC1 cleaning temperatures, as seen one day after preparation. The difference in film thickness with an SC1 cleaning temperature of 35°C was -0.022 nm at an SC1 cleaning temperature of 40°C, -0.021 nm at 45°C, 0.018 nm at 50°C, and 0.036 nm at 55°C, which was significantly different from one day after preparation, and in particular the difference in film thickness at SC1 cleaning temperatures of 40°C and 45°C was negative. Based on the results of only Comparative Example 1-5, it was not possible to determine whether the results obtained three months after preparation were due to sample or device causes, or the validity of the measured values.

In addition, because the evaluation of the working example and the comparative example were carried out at the same time, the results of the working example showed that there was no abnormality in the ellipsometry device, and the cause of the variation in the measured film thickness of the samples obtained in the comparative example due to the exposure period was found to be due to the sample. Specifically, it was determined that the natural oxide film (SC1 oxide film) formed by SC1 cleaning is thin and therefore prone to oxidation during the exposure period, and that the relationship between the film thicknesses was reversed because oxidation did not progress uniformly.

As described above, according to the embodiment of the present invention, it is possible to stably and accurately manage the device that measures the thickness of thin oxide films.

The present invention is not limited to the above-described embodiments. The above-described embodiments are merely examples, and anything that has substantially the same configuration as the technical idea described in the claims of the present invention and provides similar effects is included within the technical scope of the present invention.

Claims (9)

A group of standard samples for a film thickness measuring device,
A plurality of silicon substrates having different surface roughnesses;
A group of standard samples for a film thickness measuring device, wherein the plurality of silicon substrates each have an ozone oxide film of a different thickness on the surface thereof. The group of standard samples for a film thickness measurement device according to claim 1, characterized in that the plurality of silicon substrates having different surface roughnesses are three or more silicon substrates each having a different surface roughness. A method for producing a standard sample group for a film thickness measuring device, comprising the steps of:
A plurality of silicon substrates having the same surface quality are prepared, and the plurality of silicon substrates are subjected to SC1 cleaning under different SC1 cleaning conditions to produce a plurality of silicon substrates each having a different surface roughness;
removing the SC1 oxide film formed on the surfaces of the plurality of silicon substrates having different surface roughnesses by the SC1 cleaning with hydrofluoric acid cleaning to expose bare surfaces;
a plurality of silicon substrates each having a different surface roughness and each having an exposed bare surface, are cleaned with ozone water under the same conditions to form an ozone oxide film, thereby producing a plurality of silicon substrates each having an ozone oxide film with a different thickness, thereby producing a group of standard samples for a film thickness measuring device. The method for manufacturing a standard sample group for a film thickness measurement device according to claim 3, characterized in that the surface roughness is indexed to either a Haze value measured by a laser scattering particle counter or a Sa value or Sq value measured by an atomic force microscope. The method for manufacturing a standard sample group for a film thickness measurement device according to claim 3 or 4, characterized in that the SC1 cleaning conditions are one or more of the SC1 chemical concentration, cleaning temperature, and cleaning time. The method for manufacturing a standard sample group for a film thickness measurement device according to any one of claims 3 to 5, characterized in that the film thickness measurement device is an ellipsometer. A method for managing a film thickness measuring device, comprising the steps of: measuring a film thickness of an oxide film of a standard sample using a film thickness measuring device; and managing the film thickness measuring device based on the measurement results, comprising the steps of:
A method for managing a film thickness measuring device, comprising the steps of: using, as the standard samples, a standard sample group according to claim 1 or 2 or a standard sample group manufactured by a method for manufacturing a standard sample group for a film thickness measuring device according to any one of claims 3 to 6; and managing the film thickness measuring device based on a film thickness measurement result of the ozone oxide film of the standard sample group. The method for managing a film thickness measurement device according to claim 7, characterized in that, when managing the film thickness measurement device based on the film thickness measurement results, the magnitude relationship of the film thickness measurement results of each of the standard samples is compared. The method for managing a film thickness measurement device according to claim 7 or 8, characterized in that the film thickness measurement device is an ellipsometer.

Images