JP2010025575A - Film thickness measurement method and apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a film thickness measurement apparatus and a method for measuring a film thickness distribution in a wide area such as the entire surface of a wafer at a high speed. <P>SOLUTION: The method for measuring a film thickness of a thin film formed on a surface of a solid includes: an obtaining procedure for obtaining a plurality of reflection images of the surface of the thin film on different optical conditions; and a calculating procedure for calculating the two-dimensional film thickness distribution in the area corresponding to the reflection image of the fixed surface based on a fitting calculation between an intensity value obtained from each pixel for composing the reflection image on each optical condition and a multilayer film interference expression corresponding to the optical condition. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a film thickness measuring method and a film thickness measuring apparatus for measuring the film thickness of a multilayer thin film formed on a wafer surface or the like.

  To measure the thickness of the multilayer thin film formed on the wafer surface, the ellipsometry (polarization measurement method) is used to measure the polarization rotation angle. The technology that demands is applied.

  In the conventional technique using ellipsometry, the rotation angle of polarized light is measured for the reflected light from one measurement point on the wafer, and the film thickness is measured. And if the film thickness calculated | required from the result measured sequentially about several measuring points is in tolerance level, it was supposed that the film thickness of the whole wafer surface was in tolerance level.

In addition, a method has been proposed in which the spectral characteristics (for example, spectral transmittance) of a thin film formed on the surface of an optical element such as a lens is measured, and the film thickness is estimated based on the measurement result (Patent Document 1). reference).
JP 2001-174226 A

  In a conventional film thickness measurement technique using ellipsometry or a film thickness measurement technique using spectral characteristics, the film thickness as an average value of a minute region at one measurement point is obtained by one measurement.

  By the way, in recent semiconductor manufacturing technologies, miniaturization of circuits formed on a wafer has progressed, and improvement in performance such as high-speed operation is also required. Along with this, there is a tendency that the permissible range for fluctuations in the film thickness within the wafer becomes small.

  In particular, in an image sensor or the like, since it is required to equalize the sensitivity of all the pixels arranged in two dimensions, local variation of the film thickness in one chip may be a problem. Such local fluctuations in the film thickness may be caused by troubles in the apparatus for forming the thin film or foreign matter adhering to the wafer surface, or may be caused by a step on the wafer surface.

  On the other hand, if the conventional film thickness measurement technique described above is applied as it is to measure the film thickness distribution on the entire wafer surface, it is necessary to repeat the measurement for a large number of measurement points, which takes a lot of time for the measurement. End up.

  It is an object of the present invention to provide a film thickness measuring method and apparatus capable of measuring a film thickness distribution over a wide area such as the entire wafer surface at high speed.

  The above-described object can be achieved by the film thickness measuring method and apparatus disclosed below.

  This film thickness measurement method is characterized by a method for measuring the film thickness of a thin film formed on a solid surface, an acquisition procedure for acquiring a plurality of reflection images with different optical conditions for the surface of the thin film, and a reflection image The two-dimensional film of the region corresponding to the reflected image of the solid surface is obtained by fitting calculation of the luminance value obtained corresponding to each optical condition for each pixel constituting the pixel and the multilayer film interference equation corresponding to the optical condition. And a calculation procedure for calculating the thickness distribution.

  In the film thickness measuring method configured in this way, for example, reflection images corresponding to a wide area corresponding to the entire wafer surface are acquired collectively for different optical conditions, and these reflection images are subjected to fitting calculation processing. Thus, a two-dimensional film thickness distribution is calculated. For example, in the multilayer thin film interference type reflecting the reflectance of the solid surface and the refractive index of the constituent material of each layer for each optical condition so as to match the luminance value of each pixel constituting each reflected image, the film of each layer A two-dimensional film thickness distribution can be obtained by performing fitting calculation for changing the thickness and obtaining the film thickness of each layer for a minute region corresponding to each pixel.

  According to the film thickness measuring method and apparatus according to the present invention, it is possible to measure the film thickness distribution over a wide region such as the entire wafer surface at high speed, so that it can be applied to the total inspection of wafers.

Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.
(Embodiment 1)
FIG. 1 shows an embodiment of a film thickness measuring apparatus according to the present invention.

  The illumination device 11 shown in FIG. 1 includes, for example, a white light source having substantially uniform spectral characteristics, wavelength selection means for selectively extracting a single wavelength, means for making the extracted light uniform, illumination Intensity control means for controlling the intensity of light.

  The light irradiated from the illuminating device 11 becomes a parallel light beam by the concave mirror 12 and illuminates the surface of the wafer 13 under a uniform angle condition. The reflected light from the wafer 13 is incident on the imaging device 15 via the concave mirror 14. As a result, the image pick-up device 15 can take a reflected image from the wafer 13 over the entire surface of the wafer 13 under the same angle condition. Based on the image signal obtained by the imaging device 15, the image processing unit 16 forms image data corresponding to the entire wafer surface, and the formed image data is held in the image storage unit 17.

  The wavelength selection unit and the intensity control unit, the imaging device 15 and the image processing unit 16 of the illumination device 11 described above, according to instructions from the measurement control unit 18, respectively, a wavelength selection operation, an illumination intensity control operation, an imaging operation, and An image data forming operation is executed. Thereby, for example, the wavelength of the light emitted from the illuminating device 11 is changed variously, and the reflected image when the wafer 13 is illuminated with the light of each wavelength is photographed by the imaging device 15 and the corresponding image data is acquired. Can do.

  In this way, image data corresponding to the reflected image acquired by the imaging device 15 and the image processing unit 16 under illumination light of various wavelengths is sequentially stored in the image storage unit 17 corresponding to the wavelength of the illumination light. And is used for the processing of the fitting calculation unit 20.

  Hereinafter, the fitting calculation process performed by the fitting calculation unit 20 will be described.

  As described above, in an optical system in which both the illumination side and the imaging side are telecentric, the multi-layered film interference method applying the same angle condition is applied to the entire region of the reflected image of the wafer to be imaged. be able to.

  Here, the complex refractive index for each wavelength of the base material (for example, Si) of the wafer 13 and the complex refraction for each wavelength of the constituent material of each layer constituting the multilayer thin film formed on the base material of the wafer 13. The rate, the reference film thickness of each layer, and the estimated fluctuation range are known.

  Therefore, for every combination in which the thickness of each layer of the multilayer thin film is changed from the reference film thickness within the estimated fluctuation range for each wavelength at which the reflected image is taken, the reflectance when the film thickness of each layer is the combination is calculated in advance. Can be kept.

For example, assuming that each layer of a multilayer thin film composed of two layers varies within a range of plus or minus 100 nm, assuming that the thickness of each layer is changed in increments of 10 nm, 21 × 21 (= 441) What is necessary is just to calculate the reflectance of each wavelength about a street combination. Therefore, the thickness of each layer of the multilayer film consisting of N layers, when examined by dividing the variation range of each film thickness in the M phase, the combination of street M ^n, of at least N + 1 different wavelengths (lambda _1, ... , Λ _{N + 1} ,..., And the reflectances (R (λ ₁ ),..., R (λ _{N + 1} ),...) Are calculated, and the fitting calculation described below is performed. The film thickness of each layer can be determined.

  In the fitting table 21 shown in FIG. 1, each wavelength applied as illumination light of the film thickness measuring apparatus is calculated as described above corresponding to the multilayer thin film formed on the wafer 13 to be measured. A reference reflectance set is stored corresponding to a combination of film thicknesses of the respective layers. As shown in FIG. 1, when the entire surface of the wafer 13 to be measured is illuminated by the telecentric illumination under the same angle condition, in the fitting calculation for each pixel constituting the reflected image corresponding to the entire surface of the wafer 13, Since a common fitting table can be used, a reference reflectance set can be created with a very small amount of calculation.

  FIG. 2 is a flowchart showing the fitting calculation process.

  Hereinafter, with reference to FIG. 1 and FIG. 2, a method of obtaining the film thickness distribution on the entire surface of the wafer 13 by individually specifying the film thickness in the minute region of the wafer 13 corresponding to each pixel constituting the reflection image. explain.

  The reflectance calculation unit 22 shown in FIG. 1 reads out the luminance values of the pixel of interest from the reflection image corresponding to each wavelength held in the image storage unit 17 (step S1 in FIG. 2), and uses these luminance values. Based on the information, such as the intensity of illumination light, the sensitivity of the imaging device, the reflectance of the concave mirror, and the like held in a storage unit (not shown), the reflectance for each wavelength is calculated (step S2). The measured reflectance set including the reflectance for each wavelength calculated in step S2 is passed to the distance calculation unit 22 shown in FIG.

  The distance calculating unit 22 calculates the distance between the measured reflectance set obtained as described above and each reference reflectance set corresponding to the combination of the film thicknesses of the respective layers stored in the fitting table 21 (step S3).

  Based on the distance to each reference reflectance set obtained in this way, the candidate extraction unit 24 shown in FIG. 1 sets the combination of film thicknesses that gives a reference reflectance set close to the measured reflectance set as the nearest point candidate. And the extracted nearest point candidate is held in the candidate holding unit 25 (step S4). For example, the candidate extracting unit 24 can extract a combination of film thicknesses corresponding to the reference reflectance set whose distance obtained in step S3 is equal to or smaller than a predetermined threshold value as the nearest point candidate. The closest point candidate extracted in this way is subjected to a verification process described later.

  Further, based on the distance calculated in the process of step S4 described above, the combination of film thicknesses that gives the reference reflectance set closest to the measured reflectance set by the candidate extraction unit 24 is the fitting result using the fitting table 21. (Step S5). The film thickness data indicating the combination of the film thicknesses is also provided to a verification process to be described later via the film thickness data holding unit 26.

  For each pixel constituting the reflected image, the processes in steps S1 to S5 described above are repeated, and when the processes for all the pixels are completed, the process proceeds to step S7 as an affirmative determination in step S6. A thickness distribution verification process is performed.

  Here, the process for verifying the film thickness distribution will be described by taking as an example the case where the film thickness of a single-layer thin film is to be specified from the reflectance for two wavelengths.

When the film thickness of the thin film to be measured is changed from the film thickness t ₁ corresponding to the lower limit of the estimated fluctuation range to the film thickness t ₂₀ corresponding to the upper limit, the reflectance R (λ ₁ ) with respect to the wavelengths λ ₁ and λ ₂ , R (λ ₂ ) may change as shown in FIG. In such a case, the distance between the reference reflectance set corresponding to the film thicknesses t ₂ and t ₃ and the reference reflectance set corresponding to the film thickness t ₂₀ that is significantly different from these film thicknesses is small. There is.

At this time, attention to are measured reflectance set obtained for the pixel (indicated by rectangles hatching in FIG. 3) corresponds to the reference reflectance set and the thickness t ₂₀ corresponding to the thickness t ₂ When obtained as a value approximately in the middle of the reference reflectance set, the specified film thickness varies greatly depending on which of the closest points is determined.

  In such a case, the film thickness data obtained for the pixels in the vicinity of the target pixel is used to verify the film thickness data of the target pixel, and the film thickness corresponding to the target pixel is considered in consideration of the surrounding film thickness data. Is effective.

  FIG. 4 is a flowchart showing the film thickness verification operation.

  The verification processing unit 27 reads the film thickness data corresponding to the target pixel and its neighboring pixels from the film thickness data holding unit 26, and compares the film thickness data of the target pixel with the film thickness data of the neighboring pixels (step). S11, S12).

  Based on the comparison result, the verification processing unit 27 determines whether or not the film thickness of the target pixel is an abnormal value far from the surrounding film thickness (step S13).

For example, in the example shown in FIG. 3, the film thickness t ₂ is obtained as a fitting result corresponding to the reflectance set corresponding to the target pixel (shown by a shaded rectangle in FIG. 3). (3, shows a rectangular outline) reflectivity set corresponding to a pixel as a fitting result corresponding to, greater different thicknesses t ₁₉ or the thickness t ₂₀ is obtained from the thickness t _2. In this way, when the difference between the film thickness corresponding to the target pixel and the average value of the film thickness corresponding to the neighboring pixels is equal to or greater than a predetermined threshold, the verification processing unit 27 performs the film thickness corresponding to the target pixel. Is an abnormal value (affirmative determination in step S13), and correction processing in steps S14 to S16 is performed.

  In this case, the verification processing unit 27 first reads out the nearest point candidate held in correspondence with the target pixel from the candidate holding unit 25 (step S14), and then, from these nearest point candidates, the film thicknesses of neighboring pixels. The one closest to is identified (step S15). Further, by updating the film thickness data corresponding to the target pixel using the specified film thickness, the film thickness data corresponding to the target pixel is corrected (step S16).

As in the example shown in FIG. 3, if the reflectance data corresponding to the pixel of interest, has become a middle vicinity of the reference reflectance data corresponding to the film thickness t ₃ and the thickness t ₂₀ with the film thickness t ₂ is The film thickness data corresponding to the reference reflectance data is extracted as the nearest point candidate. In this case, these thicknesses data read in step S14, the verification processing unit 27, are specified film thickness t ₂₀ close to the fitting result of the pixels in the neighborhood, the original fitting in thickness data identified By replacing the result, the abnormal value of the film thickness data is corrected.

  By repeating the process from step S11 to step S16 described above for all pixels, the film thickness data obtained by performing the fitting process individually for each pixel is verified based on the result corresponding to the neighboring pixels, and an abnormal value is obtained. Can be detected and corrected.

  After completion of such verification processing, in step S8 of FIG. 2, the film thickness data held in the film thickness data holding unit 26 is outputted, and the film thickness data corrected for the abnormal value is output, whereby the film corresponding to the entire surface of the wafer 13 is output. A thickness distribution can be provided to the user.

  As described above, in the film thickness measurement apparatus according to the present invention shown in FIG. 1, the reflection images corresponding to the entire surface of the wafer 13 to be measured are acquired in a lump for each illumination wavelength. The reflectance data necessary for the acquisition can be acquired in a short time. Therefore, the film thickness measuring apparatus according to the present invention can measure the film thickness distribution on the entire surface of the wafer in a very short time, and can be applied to 100% inspection of wafers.

  Further, in the film thickness measuring apparatus according to the present invention, the distribution of the film thickness including the film thickness variation can be measured for each layer of the multilayer thin film. It is also possible to detect minute fluctuations in film thickness. In particular, the film thickness data obtained for each pixel is verified based on the film thickness data corresponding to neighboring pixels, and the abnormal values are corrected to correct the influence of noise contained in the acquired reflected image. The film thickness of each layer can be determined with high accuracy. Further, by increasing the number of wavelengths for measuring the reflectance to be greater than the number indicated by the above-described conditions (that is, the number N + 1 of the multilayer thin films), the thickness of each layer can be obtained with higher accuracy by fitting calculation. it can.

  The optical condition to be changed when acquiring the reflected image is not limited to the wavelength described above, but may be the incident angle or polarization of the illumination light on the measurement target wafer, or a combination thereof. For example, prepare a film thickness measurement system equipped with multiple illumination systems that illuminate the wafer at different incident angles and an imaging system that images the wafer surface corresponding to each incident angle, and switch between these to obtain a reflected image By performing the above, it is possible to obtain reflected images having different optical conditions by multiplying the number of wavelengths available for measurement by the number of angle conditions. Then, using all of the obtained reflection images, fitting calculation considering each angle condition can be performed to obtain a film thickness distribution corresponding to the entire wafer surface. Such a configuration is effective when the wavelength that can be used as illumination light for the wafer to be measured is limited.

  In addition, the film thickness for obtaining the combination of film thicknesses that gives the reference reflectance set closest to the measured reflectance set was used, but interpolation was performed between the closest reflectance set and the reflectance set of the film thickness before and after that. By obtaining the interpolation point closest to the measurement reflectance set, the film thickness measurement resolution can be made smaller than the film thickness increment at the time of creating the reference reflectance set. In the example of FIG. 3, the points t18, t19, and t20 are approximated by a curve and interpolated to obtain the nearest interpolation point from the neighboring pixels (open rectangles) in FIG. 3, and the film thickness corresponding to the interpolation point is obtained. Is obtained from the film thickness values at t18, t19, and t20 in accordance with the position of the interpolation point. In a more complicated example than FIG. 3, for example, when the number of reflected images is 3, curve approximation or curved surface approximation may be performed in a three-dimensional space.

  Further, for example, after measuring the film thickness distribution of each layer up to the i-th layer as described above, the measurement result can be used for measuring the film thickness distribution of the newly formed i + 1-th layer.

  In this case, for each pixel of the reflected image of the wafer, a reference reflectance data set in consideration of the thickness of each layer up to the i-th layer for each stage in which the variation range of the film thickness of the (i + 1) th layer is divided in an appropriate step. Calculate. Then, instead of the fitting table 21 common to the entire wafer surface shown in FIG. 1, a fitting table made of a reference reflectance data set in consideration of the film thickness of each layer up to the i-th layer is used, whereby the i + 1-th layer film is obtained. Thickness can be determined.

On the other hand, it is possible to adopt an illumination system of a form other than the telecentric illumination system. In this case, instead of the fitting table 21 common to the entire surface of the wafer shown in FIG. 1, each fitting table may be prepared using a multilayer film interference equation under an angle condition at each pixel position. However, in this case, since the fitting table must be calculated for each pixel having a different illumination angle, the amount of calculation increases.
(Embodiment 2)
By the way, like a region where memory cells are arranged in a memory element or a region where pixels are arranged in an image sensor, the same fine structure is repeatedly arranged and physically separated by wiring around the periphery. A thin film formed in a certain region may have a certain tendency. For example, as shown in FIG. 6, there may be a tendency for a dish-shaped film thickness distribution in which the peripheral part of the region becomes thick.

  For such a region, it is possible to increase the measurement accuracy by modeling the film thickness distribution and fitting the entire region using the multilayer interference method.

  FIG. 5 shows another embodiment of the fitting calculation unit according to the present invention.

In the fitting calculation unit 20 illustrated in FIG. 5, the region extraction unit 31 generates image data of a region that can be modeled as described above based on the process information about the wafer 13 to be measured (see FIG. 1). Extracted from the image data of the reflected image stored in the storage unit 17 corresponding to each wavelength. Based on the image data received from the region extraction unit 31, the reflectance calculation unit 32 calculates a reflectance measurement value R _ij corresponding to each wavelength λ _j for the pixel i included in the extracted region. A measurement spectral reflectance set including measurement reflectance R _ij corresponding to each wavelength calculated for each pixel is held in the measurement value holding unit 33.

The characteristic value table 34 shown in FIG. 5 stores information about each layer of the multilayer thin film formed on the wafer 13 to be measured and the base material of the wafer 13. In addition, the reference reflectance calculation unit 35, based on the information stored in the characteristic table 34, for each layer of the multilayer thin film, at a position corresponding to each pixel in the region represented by the curved surface model applied to each layer. The reference reflectance corresponding to each wavelength is calculated from the calculated value of the film thickness. For example, the reference reflectance calculation unit 35 sets the film thickness z _ki of the thin film of the k-th layer at the coordinates (x _i , y _i ) of the pixel i in the region to 4 relating to the x and y coordinates as shown in Expression (1). It is possible to calculate the reference reflectivity r _ij for each wavelength λ _j using a model of a secondary function and using the thickness of each pixel and each layer indicated by the curved surface model. This calculation formula is represented by a function f _j (see formula (2)). In formula (2), K is the number of layers of the multilayer thin film.

z _ki = a _k (x _i -b _k ) ⁴ + c _k (y _i -d _k ) ⁴ + e _k (1)
r _ij = f _j (z _1i , z _2i , a _3i ,..., z _Ki ) (2)
The adaptive control unit 36 illustrated in FIG. 5 includes a reference spectral reflectance set including the reference reflectance r _ij obtained for each pixel by the reference reflectance calculating unit 35 and a measured spectral reflectance held in the measured value holding unit 33. The coefficients a _k , b _k , c _k , d _k , e _k (k = 1, 2,..., K) of Equation (1) that minimize the difference from the set are obtained. For example, the adaptive control unit 36 uses the least square method to minimize the sum of squares of the difference between the measured reflectance R _ij and the reference reflectance r _ij, and the coefficients a _k , b _k , c of each layer in Expression (1). _k , d _k , e _k (k = 1, 2,..., K) may be obtained.

Expression (3) is an expression for _obtaining the square sum S of the difference between the measured reflectance R _ij and the reference reflectance r _ij .

A simultaneous equation (see equation (4)) obtained by partial differentiation of equation (3) with coefficients a _k , b _k , c _k , d _k , e _k (k = 1, 2,..., K) is expressed as coefficient a _{If k} , b _k , c _k , d _k , and e _k (k = 1, 2,..., K) can be solved, the coefficients a _k , b _k , c _k , d _k , e _k (k = 1, 2,..., K) can be obtained analytically. If the above simultaneous equations (formula (4)) are complex and cannot be solved for the coefficients a _k , b _k , c _k , d _k , e _k (k = 1, 2,..., K), the coefficient a _k , B _k , c _k , d _k , e _k (k = 1, 2,..., K) are changed little by little to calculate the sum of squares S, and the coefficients a _k , b _k , _{c k, d k, e k} (k = 1,2, ..., K) may be searched for.

  Thus, after the fitting of the model curved surface for each layer is completed, the film thickness distribution specifying unit 37 calculates the film thickness of each pixel in the above-described region using the obtained model curved surface formula. It is output as the film thickness distribution of each layer in this region.

  In the fitting calculation using such a curved surface model, the film thickness distribution in the modeled area is obtained based on the measured values for all the pixels included in the area, so that the acquired reflection image is formed. To reduce the influence of noise included in the brightness value of each pixel on the film thickness measurement at the position corresponding to that pixel, suppress the occurrence of abnormal values in the measurement results, and obtain a highly accurate film thickness distribution Can do.

(Embodiment 3)
The film thickness distribution is measured over the entire surface of the wafer 13 to be measured by switching between the method of collectively fitting a region using the curved surface model described in the second embodiment and the method of performing fitting individually for each pixel. It is also possible.

  FIG. 7 shows another form of the fitting calculation unit according to the present invention.

  In the fitting calculation unit 20 shown in FIG. 7, based on the image data representing the reflection image corresponding to each wavelength stored in the image storage unit 17, the reflectance calculation unit 41 measures the reflectance corresponding to each wavelength. And the reflectance data is held in the reflectance data holding unit 43.

  Here, during the calculation process by the reflectance calculation unit 41 described above, a deviation from the setting value of the intensity of each wavelength in the illumination light of the illumination device 11 illustrated in FIG. 1 or in the image sensor provided in the imaging device 15. The accuracy of the reflectance data necessary for the film thickness distribution measurement can be improved by correcting the measured value of the reflectance in consideration of the sensitivity shift and the like.

  For example, prior to measurement, a reflection image from the surface of a reference sample having a shape equivalent to the wafer 13 to be measured and having a known spectral reflectance distribution is acquired for each wavelength, and these reflection images and this reference are obtained. Based on the spectral reflectance distribution of the sample, for each wavelength used for measurement, a correction value corresponding to each pixel is obtained and held in the calibration data holding unit 42, which can be used for the processing of the reflectance calculating unit 41. . As the reference sample described above, a silicon substrate itself or a silicon substrate formed with a uniform protective film using a substance having a known spectral reflectance such as silica can be used. When a silicon wafer on which a protective film is formed is used as the reference sample, the spectral reflectance distribution of the protective film can be measured in advance using, for example, a spectrometer.

  Similar to the region extraction unit 31 shown in FIG. 5, the region extraction unit 44 shown in FIG. 7 is based on the process information of the wafer 13 to be measured and the spectral reflectance data corresponding to the region that can be modeled. To extract. Based on the spectral reflectance data extracted by the region extraction unit 44, the region fitting 45 collectively performs the fitting process on the regions described in the second embodiment. The region fitting unit 45 includes, for example, the characteristic value table 34, the reference reflectance calculation unit 35, and the adaptive control unit 36 illustrated in FIG.

  On the other hand, for other spectral reflectance data, the individual fitting unit 46 performs a process of searching for the nearest point from the reference reflectance set held in the fitting table 21 for each pixel as described in the first embodiment. Is called. The individual fitting unit 46 includes, for example, at least the fitting table 21 and the distance calculation unit 23 illustrated in FIG.

  Further, the film thickness distribution inside the area obtained by the area fitting unit 45 is held in the film thickness data holding unit 47 corresponding to each pixel included in the area. The film thickness obtained for each pixel by the individual fitting unit 46 is held in the film thickness data holding unit 47 corresponding to each pixel. In this way, the measurement result obtained by the region fitting unit 45 and the measurement result obtained by the individual fitting unit 46 are synthesized in the film thickness data holding unit 47, whereby the film thickness corresponding to the entire surface of the wafer 13 to be measured. Distribution can be obtained.

It is a figure which shows embodiment of a film thickness measuring apparatus. It is a flowchart showing a fitting calculation process. It is a figure explaining the verification process of film thickness distribution. It is a flowchart showing film thickness verification operation | movement. It is a figure which shows another form of a fitting calculation part. It is a figure which shows the example of film thickness distribution. It is a figure which shows another form of a fitting calculation part.

Explanation of symbols

DESCRIPTION OF SYMBOLS 11 ... Illuminating device 12, 14 ... Concave mirror, 13 ... Wafer, 15 ... Imaging device, 16 ... Image processing part, 17 ... Image storage part, 18 ... Measurement control part, 20 ... Fitting calculation part, 21 ... Fitting table, 22 , 32, 41 ... reflectance calculation unit, 23 ... distance calculation unit, 24 ... candidate extraction unit, 25 ... candidate holding unit, 26, 47 ... film thickness data holding unit, 27 ... verification processing unit, 31, 44 ... region extraction 33: Measurement value holding unit, 34 ... Characteristic value table, 35 ... Reference reflectance calculating unit, 36 ... Adaptive control unit, 37 ... Film thickness distribution specifying unit, 42 ... Calibration data holding unit, 43 ... Reflectance data holding 45, region fitting unit, 46 ... individual fitting unit.

Claims (12)

A method for measuring the thickness of a thin film formed on a solid surface,
An acquisition procedure for acquiring a plurality of reflection images having different optical conditions for the surface of the thin film;
For each pixel constituting the reflected image, a fitting calculation of a luminance value indicating brightness corresponding to an individual optical condition and a multilayer interference type corresponding to the optical condition is performed to calculate an area corresponding to the reflected image on the solid surface. A calculation procedure for calculating a two-dimensional film thickness distribution. In the film thickness measuring method according to claim 1,
In the film thickness measuring method, the acquisition procedure includes illuminating a thin film formed on the solid surface by telecentric illumination to acquire a two-dimensional reflected image from the thin film surface. In the film thickness measuring method according to claim 1,
In the film thickness measuring method, the obtaining procedure obtains two-dimensional reflected images larger than the number of layers constituting the thin film by changing optical conditions. In the film thickness measuring method according to claim 1,
The calculation procedure is as follows:
A film thickness measuring method comprising: a verification procedure for verifying a film thickness calculated corresponding to each pixel constituting the reflection image using a film thickness calculated for neighboring pixels. In the film thickness measuring method according to claim 1,
The calculation procedure is as follows:
A procedure for modeling a film thickness distribution for at least a partial region of the thin film using a curved surface function;
A step of identifying a coefficient of the curved surface function based on a fitting calculation between a distribution of luminance values indicating brightness corresponding to individual optical conditions in the region and the multilayer film interference equation. To measure the film thickness. In the film thickness measuring method according to claim 1,
Acquiring a reflected image for calibration for each optical condition for a surface of a reference sample having a known spectral reflectance distribution;
A procedure for calculating a correction value for each of the optical conditions based on the relationship between the reflection image for calibration corresponding to each of the optical conditions and the known spectral reflectance distribution, and for providing a film thickness calculation process according to the calculation procedure And a film thickness measuring method comprising: A film thickness measuring device for measuring a film thickness of a thin film formed on a solid surface,
An acquisition means for acquiring a plurality of reflection images having different optical conditions for the surface of the thin film;
Fitting of a luminance value indicating brightness corresponding to an individual optical condition for each pixel constituting the reflected image, and a multilayer film interference type reflecting the known characteristics of the solid surface and the thin film corresponding to the optical condition A film thickness measuring apparatus comprising: a calculating unit that calculates a two-dimensional film thickness distribution corresponding to the reflected image by calculation. In the film thickness measuring device according to claim 7,
The film thickness measuring apparatus according to claim 1, wherein the acquisition means includes illumination means for illuminating a thin film formed on the solid surface by telecentric illumination and for obtaining a two-dimensional reflected image by the thin film surface. In the film thickness measuring device according to claim 7,
The acquisition means switches the combination of optical conditions including the wavelength or direction of illumination light for illuminating the thin film, and provides different optical conditions more than the number of layers constituting the thin film when acquiring a reflected image. A film thickness measuring apparatus comprising: a condition switching means for performing In the film thickness measuring device according to claim 7,
The calculating means includes
A film thickness measuring apparatus comprising: verification means for verifying a film thickness calculated corresponding to each pixel constituting the reflection image using a film thickness calculated for neighboring pixels. In the film thickness measuring device according to claim 7,
The calculating means includes
Modeling means for modeling a film thickness distribution for at least a partial region of the thin film using a curved surface function;
A film comprising: coefficient calculation means for specifying a coefficient of the curved surface function based on a fitting calculation of a luminance value distribution indicating brightness corresponding to each optical condition and the multilayer film interference equation Thickness measuring device. In the film thickness measuring device according to claim 7,
Calibration image acquisition means for acquiring a reflected image for calibration with respect to the surface of a reference sample having a known spectral reflectance distribution for each optical condition;
Based on the relationship between the calibration reflection image corresponding to each optical condition and the known spectral reflectance distribution, a correction value for each optical condition is calculated and used for film thickness calculation processing by the calculation means A film thickness measuring device comprising: a value calculating means.

Images