JP2003207315A - Film thickness measuring method and device, manufacturing method and device for thin film device using the measuring method and device - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable to measure the film thickness on micro-patterns as minute as below 1 μm and manage the film thickness accurately by deciding the measuring point for management of the film thickness quickly without requiring skillfulness. <P>SOLUTION: For example in a CMP processing, the film thickness on a pattern as minute as below 1 μm of a processed wafer is made measurable using the footing method, and the measuring point for management of film thickness to enable high accuracy film thickness management is decided automatically on the basis of the design information of the intended device, the sensed image of the wafer surface, and the spectral waveform sensed at the measuring point. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to measurement of thickness and thickness distribution of a transparent film and film thickness control, and for example, in a method and a manufacturing line for manufacturing a semiconductor device on a silicon wafer, a film forming step or a film forming step. A method to measure the outermost surface film thickness on a wafer in the surface flattening process.
The present invention relates to an apparatus, a flattening processing apparatus, and a processing control method.

As an example of the transparent film, other than the above, DVD, T
It also includes a resist film, an insulating film, and the like in the manufacturing process of thin film devices such as FT and LSI reticles.

[0003]

2. Description of the Related Art For example, consider CMP processing in a semiconductor device manufacturing line. A semiconductor device is manufactured by forming a semiconductor element and a wiring pattern on a silicon wafer by film formation, exposure, etching and the like. In recent years, in order to realize high precision and high density, the trend toward miniaturization and multi-layering has been advanced. This increases irregularities on the wafer surface. Since such unevenness on the wafer makes it difficult to perform exposure, which is indispensable for forming wirings, the wafer surface is flattened.

As this flattening process, a method (CMP: Chemical Mechanical Polishing) of polishing the surface by chemical and physical actions to realize flattening is used. CMP is a processing method known in the art.

An important issue in CMP processing is film thickness control. In the past, this was managed by processing time. Generally, the polishing rate is calculated from the polishing amount obtained by measuring the film thickness before and after the CMP processing and the polishing time when the actual processing is performed, and this is fed back to the next processing time. .

Further, in order to confirm whether or not the processed film thickness is within a desired film thickness range, the film thickness is controlled by measuring a predetermined measuring point. When the film thickness is measured, it is measured on a pattern (dummy pattern) having a size that can be sufficiently measured by a conventional film thickness measuring device formed in the peripheral portion of the chip or the like. JP-A-6-25211
Japanese Unexamined Patent Application Publication No. 3 and Japanese Unexamined Patent Publication No. 9-7985 disclose an in-situ measurement system capable of measuring the film thickness on an actual device pattern (fine circuit pattern of an actual product). In Japanese Patent Laid-Open No. 9-109023, after processing,
There is disclosed an in-line measurement system that realizes an improvement in throughput by measuring the film thickness while maintaining it in water without cleaning. In Japanese Unexamined Patent Publication No. 6-252113, in measuring the film thickness on an actual device pattern, the spectral distribution of the interference light due to the film is frequency-analyzed, and the relationship between the frequency component of the spectral waveform and the film thickness is focused to determine the absolute film thickness. The method of calculating the value is disclosed. JP-A-9-193995
In Japanese Patent Laid-Open Publication No. HEI-2, an in-situ measurement system for detecting the end point of processing by detecting the position (wavelength) of the extreme value of the spectral waveform,
Japanese Patent Application Laid-Open No. 000-241126 discloses a method of calculating a film thickness by fitting a detected spectral waveform and a theoretical waveform based on a model. On the other hand, JP-A-9-7
In Japanese Patent Publication No. 985, a change in the interference light intensity due to a laser (single wavelength) film due to processing time is detected, and the film thickness is calculated from the frequency component of the waveform. In addition, JP
According to Japanese Patent Laid-Open No. 000-9437, if the area ratio of the pattern in the measurement visual field (the ratio of the area of the wiring pattern to the local in-plane area) is a certain value or more, film thickness measurement at an arbitrary position is performed. A disclosure of possible methods has been made.

[0007]

Considering, for example, a wiring process in a semiconductor device manufacturing process, the surface is often not completely flat even if CMP is performed.
This is because the ratio (area ratio of the pattern) of the wiring pattern of the lower layer in the local plane is not uniform. It is generally known that there is a correlation between the area ratio of the lower layer pattern and the film thickness after processing.

If there is a large variation in the film thickness after processing, it may cause a defect in the subsequent exposure process and etching process. Therefore, it is necessary to control the film thickness after processing.

Japanese Unexamined Patent Publication No. 2000-9437 discloses a method of measuring the film thickness on a device pattern of submicron order in order to evaluate the film thickness variation. The detected spectral data is frequency analyzed to determine the film thickness. However, this method makes it difficult to measure the film thickness when the film thickness to be measured is small with respect to the detection wavelength band. On the other hand, a film thickness measuring method by fitting, which is effective for a film having a relatively small film thickness, is disclosed in Japanese Patent Laid-Open No. 2000-241126.
It is disclosed in the publication. However, according to the technique disclosed herein, when measuring the thickness of a thin film formed on a fine pattern, the measurement accuracy is greatly affected by the density of the pattern, and the measurement is accurately performed on various patterns. It was not suitable to do.

Therefore, it is an object of the present invention to provide a film thickness measuring method and apparatus by fitting capable of measuring the film thickness even for a fine pattern, and a thin film device manufacturing method and manufacturing apparatus using the same. To do.

Further, Japanese Unexamined Patent Publication No. 2000-9437 discloses a method for evaluating a film thickness distribution in a chip, which enables accurate film thickness evaluation without requiring skill.

However, in order to evaluate the film thickness distribution in the chip, for example, when trying to measure a total of 10 × 10 100 points, the operator had to set the positioning of the 100 measurement points. .

Therefore, an object of the present invention is to provide a method and apparatus for automatically determining measurement points, and a method and apparatus for manufacturing thin film devices using the method.

[0014]

In order to solve the problem of film thickness measurement by fitting, a new model is assumed in the present invention. That is, when the dimension of the pattern formed on the sample to be measured is the same as or smaller than the wavelength of the light used for measurement, the boundary area considering the pattern edge portion (pattern step portion) is set and the fitting is performed. The influence of this boundary area is taken into consideration.

Further, in order to solve the problem in selecting the film thickness measurement point, the measurement point under a desired condition is automatically determined based on the design information, the image around the measurement point, and the spectral waveform detected at the measurement point. .

Further, the process condition of each processing apparatus in the semiconductor device manufacturing line is controlled by using the information of the film thickness which is accurately measured by the above fitting.

[0017]

BEST MODE FOR CARRYING OUT THE INVENTION As an example of an embodiment of the present invention,
Targeting CMP in the manufacturing process of semiconductor devices,
An example applied to the film thickness control after processing of the film formed on the wafer surface is shown.

First, a film thickness measuring method by fitting will be described.

The fitting method has been conventionally used for film thickness measurement. Consider a film having a uniform structure in the measurement visual field, which is composed of a plurality of layers, as shown in FIG. If the film thickness and material (refractive index and absorption coefficient) of each layer are known, the surface reflectance R _j in the j layer is represented by the theoretical formula shown in Formula 1. That is, the surface reflectance R _N of the film can be obtained by sequentially applying Formula 1 from the lower layer.

[0020]

[Equation 1]

Therefore, the film thickness can be calculated by calculating the theoretical reflectance with the film thickness to be measured as an unknown number, and determining the unknown number so that the error with the actually detected reflectance becomes the smallest. At this time, not only the film thickness but also the material of the film may be set as the unknown.

However, this method assumes that the layer structure is uniform in the measurement visual field, and therefore cannot be applied to the case where the layer structure is not uniform in the measurement visual field as shown in FIG.
In order to be able to measure the film thickness even when the layer structure is not uniform in the measurement visual field, it is necessary to assume a model in which a plurality of layer structures are mixed as a theoretical formula.

Therefore, let us consider a case where two layered structures as shown in FIG. 35 coexist. FIG. 35 schematically shows the cross section of the target film.

The theoretical formula differs depending on the detection optical system and the fineness of the pattern to be measured. First, consider an optical system for condensing light as shown in FIG. The light from the light source 103 is collimated by the lens system 105 and is applied to the sample 101.
The reflected light from the sample 101 is condensed via the lens system 105. A spatial filter 106 is provided at this position to reduce the reflected light to 0.
Allows only the next light component to pass. The spatial filter 106 has an effect of removing the influence of scattered light on the sample surface and diffracted light due to the pattern. The light that has passed through the spatial filter 106 becomes parallel light again by the lens system 107, is condensed by the lens system 109 and is dispersed by the spectroscope 110, and spectral data can be obtained. At this time, by inserting the field stop 108 between the lens system 107 and the lens system 109, it is possible to set a detection region of any size.

In the case of this optical system, only the specularly reflected light from the measurement visual field is condensed at one point, so the model formula is given by

[0026]

[Equation 2]

Next, consider an image forming system as shown in FIG. 37 as an optical system. The light emitted from the light source 203 is applied to the sample 201 via the objective lens 220. The reflected light from the sample 201 is imaged again via the lens systems 211 to 213.
By inserting the field stop 208 at this image forming position, a detection region of any size can be set.

This light is dispersed by the spectroscope 210 to obtain spectral data. At this time, by providing the spatial filter 206 on the Fourier transform surfaces of the lens systems 211 to 213 and passing only the 0th-order light, it is possible to remove the influence of scattered light on the sample surface or diffracted light due to the pattern as in the above.

The model formula for this optical system differs depending on the size of the pattern to be measured and the width of repetition.

When the size of the pattern to be measured and the width of repetition are sufficiently larger than the detection wavelength, for example, as shown in FIG.
The repetitive pattern 115 having a size of several μm as shown in FIG.
With a measurement field size of φ10 μm
When measured at ˜800 nm, the model formula is as shown in Formula 3.

[0031]

[Equation 3]

When measuring a pattern having a size of about several μm or more as shown in FIG. 38, each layer structure can be considered as an independent structure (FIG. 39). That is, in this case, it can be considered that the reflected lights from different positions on the plane do not interfere with each other. Therefore, the mathematical expression 3 of the model formula shows the surface reflectance in the case of a single layer structure, Weighting is performed by the ratio occupied by the layer structure (area ratio a ₁ , a _{2 of the} pattern),
Just take the sum of them.

On the other hand, when the size of the pattern to be measured or the width of repetition is equal to or smaller than the detection wavelength, for example, a repetitive pattern 118 having a size of 1 μm or less as shown in FIG. When the measurement is performed in the detection wavelength band of 400 to 800 nm in the field of view size, each layer structure cannot be considered as an independent structure, unlike the case shown in FIG.

In the case shown in FIG. 40, it is necessary to consider an intermediate model formula between the two models shown in FIGS. 35 and 39. That is, as shown in FIG. 41, it is necessary to consider a boundary structure in which two structures coexist in addition to a single layer structure. In this boundary structure, it is assumed that reflected lights from spatially different positions interfere with each other. The model formula of the entire reflected light can be obtained by the model formula as shown in Formula 4 by weighting each structure (including the boundary structure) with the area ratio in the measurement visual field and taking the sum. .

[0035]

[Equation 4]

If the above equation is applied as a model equation for fitting in some cases, a desired film thickness can be obtained.

As a parameter for fitting, it is possible to set the material of the film and the area ratio in the measurement visual field of each layer structure in addition to the film thickness.

The area ratio of each layer structure is preferably set in advance. However, if the set area ratio differs from the actual area ratio due to the displacement of the measurement visual field, etc., an error may occur in the calculation result. Therefore, if the area ratio is also obtained by fitting as a parameter, it is not necessary to set it in advance, and it is possible to prevent the occurrence of errors due to misalignment or the like.

When the material is used as a parameter, an approximate expression such as Cauchy's expression in which information on the material (refractive index, absorption coefficient, etc.) is shown in Equation 5 may be used.

[0040]

[Equation 5]

As a method for determining the parameter by fitting, the least square method is generally used. When the model formula is complicated as in the formulas 2 to 4, the Levenberg-Markt method, which is a non-linear method, is often used.

Further, even when the number of mixed layer structures is three or more, the desired film thickness can be obtained by applying this method. When the number 2, the number 3 and the number 4 are expanded and generalized when the number of layered structures is three or more, the number 6, number 7 and number 8 are respectively obtained.

[0043]

[Equation 6]

[0044]

[Equation 7]

[0045]

[Equation 8]

The number of parameters to be fitted is
Since it affects the calculation time and calculation error, it is desirable that the number be small. However, in the case of three or more structures, the number of parameters becomes extremely large, especially in the equation (8). Therefore, by ignoring factors that have a small influence, such as the area ratio occupying the measurement field of view being very small, and when a certain film thickness is obtained by a function with another film thickness as a variable, the film thickness is used as a fitting parameter. By introducing the above function instead of, the total number of parameters can be reduced.

FIG. 44 compares the measurement target shown in FIG. 35 with light in the same wavelength band and the case where the boundary structure is taken into consideration and the case where it is not taken into consideration. It can be seen from FIG. 44 that the error between the detected value and the theoretical value is reduced by considering the boundary structure.

Depending on the structure of the object to be measured, the boundary structure is taken into consideration.
If the accuracy is less than or equal to nm and the conditions are adjusted, a measurement accuracy of about 1 to 4 nm can be obtained. If the boundary structure is not taken into consideration, the measurement accuracy is lowered due to the above reason, and the error becomes from several tens of nanometers to several tens of nanometers. In the worst case, the detection value cannot be fitted and the measurement becomes impossible.

When the measurement target becomes smaller than the detection wavelength band, the ratio of the boundary structure in the measurement visual field increases. Therefore, when the film thickness on a fine pattern of 1 μm or less is measured by the imaging optical system, it is indispensable to consider the boundary structure in the fitting calculation.

Next, a method of setting the film thickness measurement point will be described.

In Japanese Patent Laid-Open No. 2000-9437, it is possible to measure the film thickness on an arbitrary pattern as long as the area ratio of the pattern in the measurement visual field is a certain value or more. It is described that the film thickness distribution can be obtained. By utilizing this, even an unskilled operator can accurately evaluate the film thickness distribution in the chip or in the wafer (for a detailed description of the film thickness measuring method and the like, see Japanese Patent Laid-Open No. 2000-9437). . However, JP 20
In the method disclosed in Japanese Unexamined Patent Publication No. 00-9437, for example, when measuring a total of 100 points (10 points × 10 points) in the chip, the operator needs to determine each measurement point.

On the other hand, in the present invention, the measurement point is automatically determined. The method for automatically determining the measurement points is shown below. Also,
Using this method, a method for determining measurement points for film thickness control and a film thickness control method using the method will be described.

In order to automatically determine the measurement point, it is first determined whether the region to be measured can be measured, that is, whether the area ratio of the pattern on the desired pattern and in the measurement visual field is sufficient. There is a need. As one of the determination methods, a determination method based on the spectral waveform detected in the region to be measured will be shown.

First, the case where the film thickness measuring method using frequency analysis described in Japanese Patent Laid-Open No. 2000-9437 is used will be described.

The detected spectral waveforms have different characteristics in the measurable region and the non-measurable region.
For example, in the case of the sample after the CMP process in the wiring process of the LSI, an example of the spectral waveform detected in the measurable region is shown in FIG.
2 shows a typical example of the spectral waveform detected in a region where measurement is impossible, respectively. The characteristics of the waveform shown in FIG.
It has a waveform in which a low frequency component due to the interference in the film on the pattern and a high frequency component due to the interference in the lower layer structure are superimposed. On the other hand, in the case of the waveform shown in FIG. 2, the high frequency component in FIG. 1 is dominant, and the low frequency component can hardly be confirmed. Such a difference is caused by the difference in whether the interference component due to the reflected light from the pattern is dominant or the interference component due to the reflected light from the lower layer is dominant (the measurement visual field is occupied by the pattern to be measured). If it is, it goes without saying that only the low frequency component due to the interference on the pattern).

As described above, the difference in the characteristic of the waveform includes the difference in the frequency component. Therefore, it is possible to extract the characteristic by frequency-analyzing the waveform and determine whether or not the measurement point can be measured.

FIG. 3 shows the result of frequency analysis of the waveform data shown in FIG. 1, and FIG. 4 shows the result of frequency analysis of the waveform data shown in FIG. As a method of determining whether or not measurement is possible from these results, for example, in FIG. 3, the ratio of the intensity 5 of the high frequency component and the intensity 6 of the low frequency component is calculated and compared with a preset threshold value. Then, there is a determination method in which it is possible to measure when it is larger than the threshold value, and it is impossible to measure when it is not.

Further, as shown in FIG. 4, the intensity of the high frequency component may be compared with a preset threshold value 7, and if it is larger than the threshold value, it may be determined that measurement is impossible, and if not, measurement is possible.

As the frequency analysis means, a frequency analysis method such as FFT (fast Fourier transform) or MEM (maximum entropy method) may be used.

Further, instead of performing the frequency analysis, the periodicity of the waveform may be extracted and determined by calculating the autocorrelation function.

As a method not paying attention to the periodicity of the waveform,
There are methods as shown in FIGS. For example, the local maximum value 8 of the waveform 1 is extracted, and the determination is made by comparing the magnitude of their variation with a preset threshold value. That is, since the low-frequency component is large in the spectral waveform when the film thickness can be measured, the variation becomes large when the measurement is possible (FIG. 5), and the variation becomes small when the measurement is impossible (FIG. 6). To use.

The above method is not limited to the measuring method disclosed in Japanese Unexamined Patent Publication No. 2000-9437, but can be naturally applied to the case of measuring a measurement object of about several tens of μm which can be measured by the conventional measuring device by the conventional measuring method.

A method of making a decision using a fitting method is also conceivable. For example, when a plurality of structures coexist in the measurement visual field and the layer structure including the film to be measured is known, a single theoretical waveform of the structure including the film to be measured is fitted to the detected waveform. It can be determined by doing.

When the error between the detected waveform and the fitted theoretical waveform is small (FIG. 7), it means that the area ratio of the layer structure including the film to be measured in the measurement visual field is large, and in the opposite case (FIG. 8), the area is large. It means that the rate is small. Therefore, it is possible to compare a preset threshold value with the above error and determine that the measurement is possible when the error is small and determine that the measurement is impossible when the error is large.

Next, a determination method in the case of using the film thickness measuring method by fitting according to the present invention will be described.

The error between the theoretical waveform and the detected waveform resulting from the fitting (the sum of squares of the error at each wavelength) is calculated and compared with a preset threshold. It is determined to be impossible, and if less than that, it is determined to be measurable.

FIG. 42 is a diagram for evaluating the calculation accuracy of the area ratio obtained by fitting. It can be seen that the area ratio can be obtained with a few percent error by fitting.
When determining the area ratio as a fitting parameter, compare the area ratio obtained by fitting with a preset threshold value.If the area ratio is less than or equal to the threshold value, it is determined that measurement is impossible. There is a method of determining that measurement is possible.

When the area ratio in the visual field and the measurement error are separately evaluated, as shown in FIG. 43, the measurement error tends to increase as the area ratio in the measurement visual field decreases. Empirically, when the area ratio in the visual field is less than about 20 to 30%, the measurement error increases.

Next, a method of searching for a measurement point based on the above determination will be shown. For example, as shown in FIG. 18, in the case of measuring 100 points of 10 points × 10 points in one chip, each measurement point set in a grid pattern is not necessarily a measurable point. Therefore, each measurement point is measured. It is necessary to set it to a possible point.
9 to 14 are diagrams showing an example of a procedure for searching for a measurement point. 9 and 10 show the search point interval and the number of points in the horizontal direction X.
And the vertical direction Y are set, the spectral waveforms are sequentially detected at that point, and the determination as to whether measurement is possible or not is shown by the method described above.

In FIG. 9, for example, 5 is set at an interval dX in the X direction.
The figure shows a case where a region of 3 points is searched at intervals of dY in the Y direction. In this case, an example is shown in which, with the initially set measurement point as the center, a search is performed sequentially from a point close to this point, and the point at which it is determined that measurement is possible is the search result and the film thickness is measured.

In the case of FIG. 10, an example is shown in which a region of 5 points at intervals dX and dY in the X and Y directions is searched spirally around the initially set measurement point.

In either case, all the search points may be once spectrally detected and the best judgment result may be selected, or the judgment results can be measured first by repeating the judgment operation in sequence. You can choose one that has become.

When the spectroscopic detection and determination processing can be performed at high speed, a method of performing detection and determination in real time as shown in FIG. 11 can be considered.

FIGS. 12 to 14 show a case where the processing which is performed in the square shape in FIGS. 9 to 10 is performed in a spiral line shape. Further, the same effect can be obtained by performing the same processing concentrically.

FIGS. 15 to 17 schematically show the display screen of the measurement region when the above-mentioned processing is actually carried out by the film thickness measuring device. FIG. 15 shows the search example shown in FIG. A thick circle 16 and a thin circle 17 shown in FIG.
Indicates the measurement point and the search area that are initially set.

The pattern 18 having a horizontally long shape shown by a black line in the drawing is a pattern to be measured for film thickness. Since the initially set area is not on the pattern, measurement cannot be performed as a result of the determination.
After that, the measurement positions are sequentially shifted in accordance with the search procedure and the determination is performed as needed (see FIG. 16, the portion marked with “X” in the figure means that measurement is impossible as a result of the determination). At the time point of FIG. 17, the measurement area is on the pattern, so that it is determined that measurement is possible, and the film thickness is measured.

After that, the measurement area is moved to the next measurement point and the same processing is performed.

For example, FIG. 18 schematically shows an image of one chip (the black portion 22 shows a measurable area), but as shown in the figure, for example, 10 points × 10.
When a total of 100 measurement points are set, it is possible to automatically set all the points to be a measurable region by repeating the above processing (FIG. 19).

20 and 21 schematically show operation screens when the above-mentioned processing is actually carried out by the film thickness measuring apparatus. FIG. 20 shows a case in which an image for one chip detected in advance by some method is displayed. With reference to this screen, for example, the reference position 27 of the chip is recognized, and the point is used as a reference to set, for example, 10 × 10 measurement points.
After that, the above series of processing is performed to determine the film thickness measurement point. FIG. 21 is a screen for confirming the determined measurement point.

If the determination is not made automatically, the operator may make the determination manually. For example, as shown in FIG. 20, a desired measurement point is selected with a cursor 26 or the like, and an enlarged image of that portion or an implementation image is displayed (see FIG. 21). Therefore, a method in which the operator operates the stage or the like to determine the measurement point can be considered.

At this time, for confirmation, the operator may perform measurement determination processing at an arbitrary position and refer to the measurement point setting.

22 to 29 show an example in which the determination processing is performed based on the detected image around the measurement region. Figure 2
Reference numeral 2 schematically shows an image around the measurement region. In FIG. 22, the black horizontal line 32 indicates the pattern to be measured as a circle 33.
Indicates the measurement area. In the case of FIG. 22, the film thickness cannot be measured because the measurement region is not on the pattern. Therefore, for example, the edge detection processing (see FIG.
3), horizontal line extraction processing (FIG. 24), local density distribution extraction of the extracted horizontal lines (FIGS. 25 to 27), and the like to extract a sufficiently measurable pattern density region 36. By calculating the positional relationship between this area and the initially set position and moving the stage, the area where the film thickness can be measured can be automatically determined.

The measurement area can also be automatically determined by performing the same operation as described above based on the design information. That is, for example, the entire chip is divided into regions of an appropriate size, the area ratios of the patterns in those regions are calculated based on the design information, and the image processing is performed based on the calculated area ratios. A method of setting the measurement point is conceivable in the same manner as.

By performing the above-described processing, the film thickness distribution within the chip or within the wafer surface can be accurately evaluated regardless of the skill of the operator. If the coordinate information of the measurement area determined by the above process is saved, the above process need not be repeated for the same product.

Next, a method of determining the measurement point for film thickness management by evaluating the film thickness distribution in the above processing will be described. As described above, in order to accurately manage the film thickness with a small number of measurement points, it is most efficient to measure the maximum value and the minimum value of the film thickness. FIG. 30 shows a case where the film thickness of one chip is measured at 10 points × 10 points, ie, 100 points in total. 30 max
And the portions indicated by min indicate the maximum and minimum values of the film thickness. Therefore, these two points may be selected as the measurement points for film thickness management (see FIG. 31). When more accurate evaluation is required, as shown in FIG.
By evaluating the film thickness in each region of x and min at a finer interval and determining the maximum value and the minimum value position of the film thickness, more accurate management is possible.

Areas other than the maximum and minimum values of the film thickness may be used as the measurement points for controlling the film thickness.

FIG. 32 shows an example in which the film thickness control of the CMP is performed by the film thickness control measurement points determined by the above method. In FIG. 32, information (coordinates, measurement conditions) of film thickness management measurement points evaluated by one stand-alone type film thickness measurement device 42 is developed in the film thickness measurement device 44 mounted on the CMP device 43. The case is shown. One CM
A film thickness measuring device 44 mounted on the P device 43, or a plurality of Cs
All of the film thickness measuring devices 44, 44 ', 44 "mounted on the MP devices 43, 43', 43" may play a stand-alone type role. Further, the film thickness measuring device at the deployment destination does not necessarily have to be equipped with a CMP device.

Next, as shown in FIG. 45, in a semiconductor device manufacturing line, a processing apparatus for processing a semiconductor substrate,
For example, the process conditions for processing the substrate in each processing device are controlled using the results of measuring the film thickness of the film forming device, the CMP device, the exposure device, the etching device, etc. by the film thickness measuring technique according to the present invention. can do.

That is, a film forming process for forming a thin film on a substrate, a CMP process for processing a thin film formed on a substrate,
An exposure step of applying a resist on the processed thin film and exposing a pattern, and C using the exposed resist as a mask
In a manufacturing process of a semiconductor device including an etching process of etching a thin film processed by MP processing, the film thickness measurement according to the present invention is performed on a substrate that has undergone a film forming process or a substrate that has gone through a CMP process to perform optical measurement on the substrate. The thickness of a transparent thin film with an accuracy of 10 nm or less, and using the measurement result to control the process conditions of at least one of the film forming step, the CMP step, the exposing step, and the etching step. Will be possible.

As a result of the measurement of the film thickness, if the film thickness varies, the exposure process and the etching process may be affected. In the case of the exposure step, the exposure is performed by focusing on a portion where there is no shot, so if the variation in film thickness is large, it becomes defective. In this case, when the unevenness of the surface due to the variation of the film thickness within the shot is at least within the depth of focus of the exposure, the focus can be set to the optimum height by evaluating the variation of the film thickness within the shot. It is possible to reduce the occurrence of defects.

Also in the case of etching, when the variation in film thickness is large, defective penetration of holes may occur. Also in this case, if the variation in the film thickness is evaluated, the occurrence of defects can be reduced by optimizing the etching time or the like.

Even in the film forming step, for example, when the variation in the film thickness after polishing changes depending on the film thickness before polishing, the variation in the film thickness after polishing is evaluated to determine the thickness of the film formation. Can be optimized.

By reflecting the result of the measurement as described above not only in the CMP process but also in the conditions of various processes, the occurrence of defects can be reduced.

[0094]

According to the present invention, no skill is required for a short time.
It is possible to determine a measurement point for film thickness control that can be automatically and accurately evaluated. And, by using the above method,
It is possible to improve the yield and the throughput. For example, the method of manufacturing a semiconductor device on the above-mentioned silicon wafer and the CMP process in the manufacturing line enable highly accurate film thickness control, and the throughput of the process can be improved.

[Brief description of drawings]

FIG. 1 is a graph showing an example of a spectral waveform detected in a region where film thickness can be measured, showing a relationship between reflectance and wave number.

FIG. 2 is a graph showing an example of a spectral waveform detected in a region where film thickness measurement cannot be performed, showing a relationship between reflectance and wave number.

FIG. 3 is a graph showing a relationship between reflected light intensity and frequency in an example of a result of frequency analysis of a spectral waveform detected in a film thickness measurable region.

FIG. 4 is a graph showing an example of a result of frequency analysis of a spectral waveform detected in a region where film thickness measurement is impossible, showing a relationship between reflected light intensity and frequency.

FIG. 5 is a graph showing a relationship between reflectance and wave number in an example in which a maximum value of a spectral waveform detected in a region where film thickness can be measured is detected.

FIG. 6 is a graph showing a relationship between reflectance and wave number in an example in which a maximum value of a spectral waveform detected in a region where film thickness measurement cannot be performed is detected.

FIG. 7 is a graph showing an example of fitting a theoretical waveform to a spectral waveform detected in a film thickness measurable region.

FIG. 8 is a graph showing an example of fitting a theoretical waveform to a spectral waveform detected in a region where film thickness measurement cannot be performed.

FIG. 9 is a schematic view of a sample surface showing an example of a measurement search procedure according to the present invention (searching for a rectangular region in order from the closest one).

FIG. 10 is a schematic view of a sample surface showing an example of a measurement search procedure according to the present invention (a spiral search for a rectangular area).

FIG. 11 is a schematic view of a sample surface showing an example of a measurement search procedure according to the present invention (searching a rectangular region in a spiral line shape).

FIG. 12 is a schematic view of a sample surface showing an example of a measurement search procedure according to the present invention (searching in a spiral order).

FIG. 13 is a schematic view of a sample surface showing an example of a measurement search procedure according to the present invention (searching in a spiral order).

FIG. 14 is a schematic view of a sample surface showing an example of a measurement search procedure according to the present invention (searching in the order closer to a spiral line). FIG.

FIG. 15 is a front view of a display screen displaying a film thickness measurement region according to the present invention.

FIG. 16 is a front view of a display screen displaying a film thickness measurement region according to the present invention.

FIG. 17 is a front view of a display screen displaying a film thickness measurement region according to the present invention.

FIG. 18 is a front view of a display screen displaying a film thickness measurement region in one chip according to the present invention.

FIG. 19 is a front view of a display screen displaying a film thickness measurement region in one chip according to the present invention.

FIG. 20 is a front view of a display screen for performing an operation of setting a film thickness measurement point in one chip according to the present invention.

FIG. 21 is a front view of a display screen showing an example of an enlarged image displayed when setting a film thickness measurement point according to the present invention.

FIG. 22 is a diagram showing an example of an image around a measurement region according to the present invention.

FIG. 23 is a diagram showing an example of an image around the measurement region according to the present invention, which has been subjected to edge detection processing.

FIG. 24 is a diagram showing an example of an image around the measurement region according to the present invention, which has been subjected to horizontal line extraction processing.

FIG. 25 is a diagram showing an example of an image around a measurement region according to the present invention, showing an extracted horizontal line density distribution extraction processing procedure;

FIG. 26 is a diagram showing an example of an image around a measurement region according to the present invention and showing a density distribution extraction processing procedure of an extracted horizontal line.

FIG. 27 is a diagram showing an example of an image around a measurement region according to the present invention and showing a density distribution extraction processing procedure of extracted horizontal lines.

FIG. 28 is a diagram showing an example of measurement point determination processing using image processing according to the present invention.

FIG. 29 is a diagram showing an example of measurement point determination processing using image processing according to the present invention.

FIG. 30 is a film thickness distribution chart showing a film thickness distribution in one chip measured by the present invention.

FIG. 31 is a plan view of one chip showing the state of film thickness management measurement points in one chip selected according to the present invention.

FIG. 32 is a schematic plan view of a CMP processing line showing an example in which the present invention is applied to the CMP processing line.

FIG. 33 is a cross-sectional view of a multilayer film.

FIG. 34 is a cross-sectional view of a measurement target in which a plurality of layer structures are mixed.

FIG. 35 is a cross-sectional view of a measurement target in which two layer structures are mixed.

FIG. 36 is a front view showing a schematic configuration of a detection optical system of a light collecting system of the film thickness measuring device according to the present invention.

FIG. 37 is a front view showing a schematic configuration of a detection optical system of the image forming system of the film thickness measuring apparatus according to the present invention.

FIG. 38 is a plan view of a pattern showing a relationship between a pattern having a size larger than the detection wavelength and a measurement visual field.

FIG. 39 is a sectional view of a pattern showing a method for setting a model formula of a pattern having a size larger than the detection wavelength according to the present invention.

FIG. 40 is a plan view of a pattern showing a relationship between a pattern having a size smaller than the detection wavelength and a measurement visual field.

FIG. 41 is a sectional view of a pattern showing a method of setting a model formula of a pattern having a size smaller than the detection wavelength according to the present invention.

FIG. 42 is a graph showing the relationship between the calculated area ratio and the pattern area ratio in the measurement visual field, which is an example of the calculation accuracy evaluation result of the area ratio calculated in the present invention.

FIG. 43 is a graph showing the relationship between the absolute value of the measurement error and the pattern area ratio in the measurement visual field according to the present invention.

FIG. 44 is a graph showing a relationship between reflectance and wavelength showing an example of an effect when a boundary structure according to the present invention is considered.

FIG. 45 is a block diagram showing an example of the configuration of a semiconductor device manufacturing line that reflects the film thickness measurement result according to the present invention in process conditions.

[Explanation of symbols]

1 …… Spectral waveform detected in the area where film thickness can be measured 2 …… Spectral waveform detected in the area where film thickness cannot be measured 3 …… Frequency analysis result of the spectral waveform detected in the area where film thickness can be measured 4 …… Film Frequency analysis result of the spectral waveform detected in the area where the thickness cannot be measured 5 …… High frequency component strength 6 …… Low frequency component strength 7 …… Strength threshold 8 …… Spectral waveform detected in the film thickness measurable area Maximum value of the spectral waveform detected in the region where the film thickness cannot be measured 10 The theoretical waveform fitted to the spectral waveform detected in the region where the film thickness can be measured 11 The region where the film thickness cannot be measured Theoretical waveform fitted to the spectral waveform detected in step 12 ... Measurement point initially set, 13 ... Search area, 14 ...
Search area 15 …… Search screen display window, 16 …… Measurement point initially set, 17 …… Search area 18 …… Measurement target pattern, 19 …… Possible patterns, 20…
… Display of non-measurable area 21 …… 1 chip image, 22 …… Measurement target pattern, 23 ……
Set measurement point 24 …… Measurement point setting window, 25 …… Set measurement point, 26
...... Cursor 27 …… Reference position mark, 28 …… Measurement point setting window, 29 …… Measurement area 30 …… Lower layer pattern, 31 …… Measurement target pattern, 32 ……
… Measurement target pattern 33 …… Measurement area, 34 …… Edge detected pattern, 35…
… Horizontal line extracted pattern 36 …… Extracted measurable area, 37 …… 1 chip film thickness distribution measurement result 38 …… Detailed film thickness distribution measurement result including minimum film thickness 39 …… Details including maximum film thickness Thickness distribution measurement results 40 …… Results of selecting film thickness control measurement points, 41 …… Measurement points for film thickness control 42 …… Stand-alone type film thickness measurement equipment, 43 ……
CMP device 44 …… CMP film thickness measuring device 101 …… Sample, 102 …… Stage, 103 …… Light source, 104 ……
Beam splitter, 105 …… lens system, 106 …… spatial filter, 107 …… lens system, 108 …… field stop, 109 …… lens system, 110 …… spectrometer, 111 …… lens system, 112 …… lens system , 113 …… Lens system, 114 …… Measuring field of view, 115 ……
Patterns of several μm size, 116 …… Layer structure 1,117 …… Layer structure 2,118 …… Patterns less than 1 μm, 119 …… Boundary structure,
120 ... Objective lens, 121 ... Deposition condition determination device, 122 ...
… CMP condition determination device, 123 …… Exposure condition determination device, 124
...... Etching condition determination device

   ─────────────────────────────────────────────────── ─── Continued front page    F term (reference) 2F065 AA30 BB02 BB22 CC19 DD06                       FF41 HH04 HH13 LL04 LL30                       QQ18 QQ31                 2G020 AA04 BA04 BA18 CA12 CB04                       CC23 CD16                 4M106 AA01 AA12 BA04 CA48 DH03

Claims (20)

[Claims] 1. A sample having a structure in which a pattern formed on a surface is covered with an optically transparent thin film is irradiated with light, and reflected light generated from the sample by irradiation of the light is detected via an optical system. A method for obtaining the film thickness of the optically transparent film by using the information of the spectral distribution waveform of the detected reflected light, which uses the information of the area ratio of the pattern in the detection visual field of the optical system. A method for measuring a film thickness of a thin film, which comprises obtaining a film thickness of an optically transparent film covering the pattern. 2. A sample in which a multilayer film is formed and the surface of which is covered with an optically transparent thin film is irradiated with light, and reflected light generated from the sample by irradiation of the light is detected via an optical system. Then
A method of obtaining the film thickness of the optically transparent film by using the information of the spectral distribution waveform of the detected reflected light, wherein the area model having a plurality of layer structures is set and the reflected light from the area model is set. A film of a thin film, characterized in that the thickness of the optically transparent film is obtained by fitting using the information of the calculated waveform and the information of the detected spectral distribution waveform of the reflected light. Thickness measurement method. 3. The film thickness measurement according to claim 2, wherein the area model includes an area model in which reflected light from the pattern and reflected light from a layer below the pattern are mixed. Method. 4. A plurality of layer structures are set in the area model,
The film thickness measuring method according to claim 2, wherein adjacent portions of the respective structures are set as different structures. 5. A sample having a structure in which a pattern formed on the surface is coated with an optically transparent thin film is irradiated with light, and reflected light generated from the sample by the irradiation of the light is detected through an optical system. A method for obtaining the film thickness of the optically transparent film using the information of the spectral distribution waveform of the detected reflected light, wherein the reflected light from the pattern and the reflected light from a layer below the pattern A region model in consideration of a mixed region is set, a waveform of reflected light from the set region model is calculated, and information of the calculated waveform and information of the spectral distribution waveform of the detected reflected light is used. A method for measuring a film thickness of a thin film, which comprises obtaining a film thickness of an optically transparent film covering the pattern. 6. The method for measuring the film thickness of a thin film according to claim 1, wherein the width of the pattern formed on the surface of the sample is 1 μm or less. 7. A film thickness measuring method for irradiating a wafer surface with light, detecting reflected light from the wafer surface by the irradiation of light, and measuring a film thickness based on the detected reflected light,
A film thickness measuring method, characterized in that a measuring point is determined by using the detected spectral data of reflected light. 8. The film thickness measuring method according to claim 7, wherein the measurement point is determined by using the information indicating the position and the position of the maximum value or the minimum value of the spectral data of the reflected light. 9. A frequency analysis is performed on the spectral data of the reflected light,
The film thickness measuring method according to claim 7, wherein a measuring point under a desired condition is determined based on a result of the frequency analysis. 10. The area ratio in the measurement field of view of the structure to be measured is calculated from the spectral data of the reflected light, and the measurement point is determined by using the information of the calculated area ratio. Film thickness measurement method. 11. An irradiation means for irradiating a sample having an optically transparent thin film formed on its surface with light, and a detection means for detecting reflected light emitted by the irradiation means from the sample through an optical system. And a film thickness calculating device for calculating a film thickness from the data detected by the detecting device, wherein the film thickness calculating device is an area of the pattern in a detection visual field of the optical system. A film thickness measuring device for a thin film, characterized in that the film thickness of the optically transparent film is obtained using information on the rate. 12. An irradiating means for irradiating a sample having an optically transparent thin film formed on its surface with light, and a detecting means for detecting reflected light emitted by the irradiating means from the sample through an optical system. And a film thickness measuring device for calculating a film thickness from the data detected by the detecting device, wherein the film thickness calculating device sets a region model having a plurality of layer structures. Calculating the waveform of the reflected light from the area model, and determining the film thickness of the optically transparent film by fitting using the information of the calculated waveform and the information of the spectral distribution waveform of the detected reflected light. A thin film thickness measuring device characterized by: 13. The detecting means has a wavelength band of 400-8.
13. The thin-film device manufacturing apparatus according to claim 11 or 12, which detects light of 00 nm. 14. The thin film device manufacturing apparatus according to claim 11, further comprising a measuring point determining unit that determines a measuring point using spectral data of reflected light detected by the detecting unit. . 15. The measurement point determining means determines the measurement point by using information on a position indicating the maximum value or the minimum value of the spectral data of the reflected light and the size thereof. Film thickness measuring device. 16. The measurement point determining means frequency-analyzes the spectral data of the reflected light, and determines a measurement point under a desired condition based on a result of the frequency analysis. Film thickness measuring device. 17. The measuring point determining means determines an area ratio of a structure to be measured in a measurement field of view from the spectral data of the reflected light, and determines the measuring point by using the obtained area ratio information. The film thickness measuring device according to claim 14. 18. A film forming step of forming a thin film on a substrate, a CMP step of processing the thin film formed on the substrate, and an exposing step of applying a resist on the processed thin film to expose a pattern. And a step of etching a CMP-processed thin film using the exposed resist as a mask, in the manufacturing process of a semiconductor device, the substrate that has undergone the film-forming process or the thin film on the substrate that has undergone the CMP process. Is measured with an accuracy of 10 nm or less, and the process result of the measurement is used to control the process condition of at least one of the film forming step, the CMP step, the exposing step, and the etching step. Manufacturing method of semiconductor device. 19. A film forming step of forming a thin film on a substrate, a CMP step of processing the thin film formed on the substrate, and an exposing step of applying a resist on the processed thin film to expose a pattern. And a step of etching a CMP-processed thin film using the exposed resist as a mask, in the manufacturing process of a semiconductor device, the substrate that has undergone the film-forming process or the thin film on the substrate that has undergone the CMP process. The thickness of the film is measured at each measurement point determined by using the spectroscopic data of the reflected light obtained by detecting the reflected light of the light applied to the substrate, and the film forming step using the measured result, A method of manufacturing a semiconductor device, which comprises controlling a process condition of at least one of a CMP step, an exposure step and an etching step. 20. An optically transparent thin film is formed on the substrate in the film forming step.
21. The method for manufacturing a semiconductor device according to 9 or 20.