JP2005337927A - Film thickness measuring method and film thickness measuring instrument - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To measure a film thickness in complicated structure such as film thickness on a gate of a transistor, after CMP (Chemical Mechanical Polishing). <P>SOLUTION: When measuring the film thickness on the gate of the transistor, an optical model is prepared to constitute a measuring object by the composition of respective different film structures of a gate area, an active area and an element separation area, and the surface film thickness on the gate is calculated by fitting. When calculating the film thickness, the film thickness is measured highly precisely by fitting optical constants of a polycrystal Si, a metal siliside and the like constituting the gate, at the same time. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to the measurement of the thickness of an optically transparent film and the film thickness management technology, and more particularly to the measurement of the film thickness of the surface of a semiconductor wafer in a film forming process and a planarization process of a semiconductor device manufacturing line. It is related to effective technology when applied.

  A semiconductor device undergoes processes such as film formation, exposure, and etching, and a semiconductor element and a wiring pattern are formed on a semiconductor wafer and manufactured. In recent years, in order to achieve high accuracy and high density, semiconductor elements and wiring patterns have been advanced in the direction of miniaturization and multilayering.

  Generally, when an insulating film is formed after forming a semiconductor element or wiring, the surface of the insulating film forms irregularities corresponding to the semiconductor element or wiring in the lower layer. As the pattern becomes finer, the depth of focus in exposure decreases, so that the surface of the insulating film of the semiconductor wafer needs to be flattened for accurate exposure.

  For the planarization, for example, CMP (Chemical Mechanical Polishing) that is polished by a chemical or mechanical action is used.

  Since the polishing rate of CMP differs depending on the density of semiconductor elements and wiring patterns, the pad state of CMP, the pressure state of the wafer holder, etc., the accuracy of the film thickness after CMP polishing is reduced, the variation in flatness in the semiconductor chip, Non-uniform film thickness occurs in the processed wafer surface.

Therefore, in the CMP process, the film thickness of the semiconductor wafer after the CMP process is measured and the polishing rate is calculated in order to accurately manage and control the film thickness and optimize the CMP process. The film thickness is measured on a large dummy pattern such as a pattern of wiring 61 formed on the silicon (Si) substrate 44 as shown in FIG. Thus, the film thickness d ₁ of the SiO ₂ film 46 was measured.

  However, since the dummy pattern is different in size and pattern density from the actual semiconductor element and wiring pattern, the polishing rate is different and the measured film thickness is also different from the actual pattern. Therefore, in order to accurately control the film thickness, it is difficult to obtain the polishing rate from the film thickness on the dummy pattern, and it is necessary to measure the film thickness on the actual pattern.

Until now, several methods have been proposed as techniques for measuring the film thickness on an actual pattern. For example, as shown in FIG. 17B, the film thickness measurement of the interlayer insulating film targeting the film thickness d ₁ of the SiO ₂ film 46 on the actual pattern by the wiring 61, or the element as shown in FIG. When forming the isolation region, there is a film thickness measurement of an STI (Shallow Trench Isolation) structure for the film thickness d ₁ of the Si ₃ N ₄ film 62.

  Japanese Patent Laid-Open No. 2001-74419 (Patent Document 1) discloses a predetermined structure for a structure having at least one cycle composed of at least two adjacent elements having different optical properties with respect to incident light. A technique for measuring the film thickness of a structure by irradiating incident light in the wavelength range and optimizing an optical model with respect to detected reflected light spectral data is disclosed.

  Japanese Patent Laid-Open No. 2001-21317 (Patent Document 2) discloses reflected light spectral data obtained by irradiating a substrate having a pattern structure composed of various pattern types with probe light having a plurality of wavelength components. A technique for specifying a measurement position by specifying a pattern type and measuring a film thickness by comparing with a waveform calculated in advance or measured in advance is disclosed.

Furthermore, with regard to this type of film thickness measurement technology, for example, in CMP processing, it is possible to measure the film thickness on an actual pattern using a fitting method for a processed semiconductor wafer, device design information, Some film thickness management measurement points are automatically determined based on a detected image of the wafer surface and a spectral waveform detected at the measurement points, thereby enabling highly accurate film thickness management (see Patent Document 3).
JP 2001-74419 A JP 2001-21317 A JP 2003-207315 A

  However, the film thickness measurement technique in the semiconductor device as described above has the following problems.

  The polishing rate in CMP processing is small in a region where the pattern density is high, and is large in a region where the pattern density is low. Therefore, the film thickness after CMP processing varies greatly depending on the density of elements and wirings in the semiconductor chip, and does not become completely flat.

  Further, there is no pattern around the semiconductor wafer, or the film thickness varies over the entire surface of the semiconductor wafer due to the pressure distribution of the wafer holder during CMP processing, and uniformity over the entire surface of the semiconductor wafer cannot be obtained. Therefore, in order to measure the film thickness after CMP processing, it is necessary to measure the film thickness on an actual pattern.

  However, conventional interlayer insulating films and STI structures for measuring film thickness on actual patterns are relatively simple in structure.

  In Patent Document 1, a structure having at least one cycle composed of at least two adjacent elements having different optical properties is targeted. However, what is actually applied is an interlayer insulating film or plug of a wiring layer. It is intended for relatively simple structures such as processes. In addition, there is no mention of an irregular structure in which the size and pitch of elements change as a structure.

  Furthermore, in Patent Document 2, although various patterns are targeted, STI and damascene and relatively simple processes are targeted as application processes.

  When manufacturing a semiconductor device, an element isolation field is formed in an Si substrate by an STI process, an element transistor or capacitor is formed in an active region of the Si substrate, an interlayer insulating film is formed thereon, and then A multilayer wiring layer is formed together with an interlayer insulating film.

  These two inventions are directed to a relatively simple structure such as an STI process before forming a semiconductor element, an interlayer insulating film in a wiring layer after forming the element, and a damascene and plug process for the interlayer insulating film.

  On the other hand, an insulating film formed on a semiconductor element immediately after element formation has a complicated structure such as a transistor serving as an element or a field for element isolation in a lower layer. In the two inventions described above and the invention of Patent Document 3, the semiconductor wafer having such a complicated structure is measured with the film thickness after CMP processing as an actual pattern, for example, on the gate of a transistor. Is not stated.

  In the case where an insulating film immediately after formation of a semiconductor element such as a transistor is measured on the gate, a material whose crystal state changes depending on process conditions such as polycrystalline Si is used for the gate. The optical constants of these materials vary greatly depending on the film formation conditions or annealing treatment, but such cases are not mentioned.

  An object of the present invention is to increase the film thickness on a real pattern such as a gate after CMP processing on a sample having a complicated structure in which an insulating film is formed on a transistor or a field for element isolation as a semiconductor element. An object of the present invention is to provide a technique for measuring with high accuracy, calculating the polishing rate of the target wafer with high accuracy based on the measurement result, and optimizing the CMP processing conditions.

  The above and other objects and novel features of the present invention will be apparent from the description of this specification and the accompanying drawings.

  Of the inventions disclosed in the present application, the outline of typical ones will be briefly described as follows.

  The present invention irradiates a sample on which a film having an optically transparent surface is formed with white light, detects reflected light from the surface of the sample due to the irradiation of the white light, and spectrums zero-order light of the detected reflected light. A film thickness inspection method for determining a film thickness based on a waveform, wherein an optical model of a sample is composed of a combination of three or more structures having different optical structures and materials, and the reflected light obtained from the optical model By fitting the spectral waveform to the spectral waveform of the 0th-order light of the detected reflected light, the surface film thickness and the area ratio of each structure of the sample are obtained.

  Moreover, the outline | summary of the other invention of this application is shown briefly.

  A film thickness measuring apparatus according to the present invention includes an irradiation unit that irradiates a sample on which a film having an optically transparent surface is formed, and zero-order light of reflected light emitted from the sample by the irradiation unit. Detecting means for detecting, optical model creating means for composing an optical model of the sample by combining three or more structures having different optical structures and materials, and detecting a reflected light spectrum waveform obtained from the optical model creating means Fitting processing means for fitting to the spectral waveform of the zero-order light of the reflected light, and film thickness calculation means for calculating the film thickness and area ratio of each structure of the sample by the fitting processing means.

  Among the inventions disclosed in the present application, effects obtained by typical ones will be briefly described as follows.

  (1) After a semiconductor element is formed, a sample having a complicated structure in which an insulating film is formed is synthesized as an optical model of the sample with three structures of a gate region, an active region, and an element isolation region. By fitting the spectral waveform of the model to the detected spectral waveform, the film thickness and area ratio of each region of the sample can be measured with high accuracy.

  (2) When calculating the film thickness, the optical constants of the material whose crystal state changes depending on the process conditions, such as polycrystalline silicon and metal silicide constituting the gate, are simultaneously fitted to obtain the optical constants with high accuracy. The film thickness can be measured.

  (3) By the above (1) and (2), it becomes possible to measure the film thickness on the gate of the transistor which cannot be measured nondestructively, the film thickness can be managed with high accuracy, the yield is improved, and Throughput can be improved.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. Note that components having the same function are denoted by the same reference symbols throughout the drawings for describing the embodiment, and the repetitive description thereof will be omitted.

(Embodiment 1)
FIG. 1 is a schematic diagram illustrating the apparatus configuration of a film thickness measuring apparatus according to Embodiment 1 of the present invention, FIG. 2 is a schematic diagram illustrating a processing method of the film thickness measuring apparatus in FIG. 1, and FIG. FIG. 4 is a schematic diagram showing a structure of a sample to be measured in the film thickness measuring apparatus in FIG. 1, and FIG. 5 is a sample in FIG. FIG. 6 is a flowchart showing an example of processing of an optical model creation process in the film thickness measuring apparatus of FIG. 1, and FIG. 7 is a detected spectral waveform in the film thickness measuring apparatus of FIG. FIG. 8 is a schematic diagram for explaining a fitting example of a spectral waveform by an optical model, FIG. 8 is a schematic diagram for explaining a display example of a film thickness and area ratio calculation result output to a monitor in the film thickness measuring apparatus of FIG. The sample of the sample to be measured in FIG. Schematic diagram showing the structure of a case where irregular, FIG. 10, the area of the sample to be measured in FIG. 4 is a schematic diagram showing the structure when the singular.

  In the first embodiment, for example, in the semiconductor device manufacturing line shown in FIG. 3, a transistor is formed, an insulating film is formed by the film forming apparatus 131, and then polished by the CMP apparatus 13. The thickness measurement apparatus 133 measures the actual device pattern. After the measurement, contact holes are formed by the exposure device 134 and the etching device 135.

  As shown in FIG. 1, the film thickness measuring device 133 includes a detection optical system 1 and a film thickness measurement processing unit 30. The detection optical system 1 includes an XY stage 11 on which a semiconductor wafer (sample) 2 is placed, an objective lens 12, a half mirror 13, an imaging lens 17, a field stop 18, relay lenses 19 and 21, a spatial filter 20, and a spectrometer 22. , An illumination light source 16, an aperture stop 15, and a condenser lens 14.

  The illumination light source (irradiation means) 16 is a white illumination light source such as a xenon lamp or a halogen lamp, and irradiates the semiconductor wafer 2 with the white illumination light through the aperture stop 15, the condenser lens 14, the half mirror 13, and the objective lens 12. To do.

  As shown in the enlarged view of the semiconductor wafer in the lower part of FIG. 2, the semiconductor wafer 2 is obtained by forming an insulating film after forming a transistor and polishing the unevenness on the surface of the insulating film by CMP. The reflected light from the semiconductor wafer 2 is guided to the spectroscope 22 through the objective lens 12, the half mirror 13, the imaging lens 17, the field stop 18, the relay lens 19, the spatial filter 20, and the relay lens 21.

  The field stop 18 guides only the reflected light from the detection field on the semiconductor wafer 2 to the spectrometer 22. The spatial filter (detection means) 20 removes the diffracted light due to the repetitive pattern and the scattered light due to the unevenness of the sample surface from the reflected light from the semiconductor wafer 2, and guides only the 0th-order light to the spectroscope 22.

  The spectral waveform 4 spectrally separated by the spectroscope 22 is input to the film thickness measurement processing unit 30 as an electrical signal, and the film thickness and area ratio are calculated. In the film thickness measurement processing unit 30, a spectral waveform is input in the spectral waveform input processing 31, and in the optical model creation processing (optical model creation means) 32, the optical model 5 shown in the center of FIG. Fitting to the spectral waveform 4 detected by the fitting processing means) 33, the film thickness / area ratio calculating process (film thickness calculating means) 34 calculates the film thickness and area ratio, and outputs them to the monitor 35.

  Next, an outline of processing in the film thickness measurement processing unit 30 will be described.

  In the spectral waveform input process 31, the reflectance spectral waveform of the semiconductor wafer from which the spectral intensity characteristic of the illumination light source 16 is removed by dividing the spectral waveform of silicon (Si) as the reference spectral waveform with respect to the input spectral waveform. obtain.

  In the optical model creation process 32, the optical model of the sample is configured by combining three structures of a gate region, an active region, and an element isolation region. In the fitting process 33, fitting is performed on the detected spectral waveform using the film thickness and area ratio of the spectral waveform obtained by the optical model as parameters, and a condition for minimizing the square error is obtained. In the film thickness / area ratio calculation process 34, the film thickness and area ratio of each region are calculated from the obtained conditions.

  The optical model creation process 32 will be described in more detail with reference to FIG. 2, FIG. 4, and FIG.

As an example of constructing an optical model of a sample film structure in which an insulating film is formed after forming a transistor as a semiconductor element, an SiO ₂ film 45 as an insulating film for element isolation is formed on a Si substrate 44 as shown in FIG. Consider a structure in which an element isolation region 43 is formed, a transistor 57 including an active region 42 and a gate region 41 is formed, and the transistor 57 and the element isolation region 43 are covered with an SiO ₂ film 46 which is an insulating film.

In this example, the gate region 41 includes the Si substrate 44, the SiO ₂ film 47, the poly (poly) Si film 48, the SiO ₂ film 46, and the active region 42 includes the Si substrate 44, the SiO ₂ film 46, and the element isolation region 43. , Si substrate 44, SiO ₂ film 45, and SiO ₂ film 46.

  The measurement field 91 of the semiconductor wafer 2 can be set by adjusting the field stop 18. As an optical model for such a sample, the gate region 41, the active region 42, and the element isolation region 43 of the sample have a film structure and a first structure 51, a second structure 52, and a third structure 53 having different film thicknesses. Assume that it consists of a structure.

At this time, each structure is represented by a conventional multilayer film structure as shown in FIG. In general, as shown in FIG. 5, the surface reflectance R _j 162 of the j layer 161 is expressed by the following (Equation 1).

Therefore, if the thickness, optical constant (complex refractive index) of each layer in each structure is known in the first structure, the second structure, and the third structure, the surface reflection by the first structure, the second structure, and the third structure alone. The rates R ₁ , R ₂ and R ₃ can be calculated respectively.

  In general, when the pattern of each structure is large, each structure is an independent structure, and reflected light from each structure does not interfere. Therefore, the reflectance R of the entire film shown in FIG. 4 is the first structure, the second structure, and the second structure. It is obtained by multiplying the surface reflectance of the three structures by the area ratio of each structure to obtain the sum.

  Here, the area ratio represents the ratio of the area to the measurement visual field 91 of each structure. However, as shown in FIG. 4, when there are a plurality of gate regions 41, active regions 42, and element isolation regions 43 with respect to the measurement visual field 91, the area ratio is obtained by calculating the sum of the areas in the plurality of regions. Expressed as a ratio to the area of the field of view 91.

On the other hand, when the pattern is small, the reflected light of each structure interferes with each other at the boundary portion of each structure. Therefore, the optical model of the film structure shown in FIG. 4 is represented by the sum of the single structure portions R ₁ , R ₂ , R ₃ that do not interfere with each other and the boundary portion R ₁₂₃ that interferes with each other, as shown in (Equation 2). The

  From the above, in the optical model creation process 32, as shown in FIG. 6, single structure models of the structure 1, the structure 2, and the structure 3 are created (steps S151 to S153). Subsequently, boundary models of the structure 1, structure 2, and structure 3 are created (step S154), the single structure and the boundary model are synthesized (step S155), and the reflectance R of the entire sample is obtained.

  In the fitting process 33, the spectral waveform 231 detected as shown in FIG. 7 is fitted using the nonlinear least square method with the film thickness and area ratio of the spectral waveform 232 obtained by the optical model as parameters, and the square Find the film thickness and area ratio that minimize the error.

  Next, a display example of the film thickness / area ratio calculation result output to the monitor 35 in the film thickness / area ratio calculation process 34 is shown in FIG.

  FIG. 8A shows an example of displaying the film thickness and area ratio of each detection location. The semiconductor chip 172 measured on the semiconductor wafer 171 at the upper left of the monitor 35 screen, and the semiconductor chip 180 on the semiconductor chip 180 at the lower left of the screen. The upper measurement position 181 is shown. The film thickness / area ratio is displayed in the film thickness / area ratio display area 173 on the right side of the screen.

The semiconductor chip 172 to be measured on the semiconductor wafer 171 and the measurement position 181 on the semiconductor chip 180 are specified in advance by the operator before the measurement. After the measurement is started, spectral data is acquired at the designated position, fitting is performed, and the film thickness and area ratio are calculated. After calculating the film thickness, the operator clicks on the semiconductor chip 172 to be measured on the semiconductor wafer 171, so that the film thicknesses d ₁ , d ₂ , d of the gate region, the active region, and the element isolation region in the designated semiconductor chip. ₃ and the area ratios m ₁ , m ₂ , m ₃ are displayed.

  FIG. 8B shows an example of displaying the distribution of the film thickness and area ratio on the semiconductor wafer. Similar to FIG. 8A, the semiconductor chip 172 to be measured on the semiconductor wafer 171 on the left side of the monitor 35 screen, The measurement position 181 on the semiconductor chip 180 is shown, and the film thickness / area distribution display area 182 on the right side of the screen displays the distribution of the film thickness and area ratio on the semiconductor wafer.

Similarly to FIG. 8A, the semiconductor chip 172 to be measured on the semiconductor wafer 171 and the measurement position 181 on the semiconductor chip 180 are specified in advance by the operator before the measurement, and after the film thickness is calculated, the operator outputs the items. By specifying one of the gate region, active region, element isolation region film thicknesses d ₁ , d ₂ , d ₃ , or area ratios m ₁ , m ₂ , m ₃ in 183, the film thickness / area distribution display The film thickness or area ratio distribution on the semiconductor wafer corresponding to the region 182 is displayed.

  In the present embodiment, as shown in FIG. 4, the sample is intended to repeat the same pitch of the gate region 41, the active region 42, and the element isolation region 43 with respect to the measurement visual field 91, but as shown in FIG. A case where the pitch of each region is irregular can be dealt with.

  Further, the case where only one gate region 41, active region 42, and element isolation region 43 exists in the measurement visual field 91 as shown in FIG. As can be seen from FIG. 10, the area ratio of the active region 42 and the element isolation region 43 is larger than the measurement visual field 91, but the gate region 41 is present in the measurement visual field 91 alone or in two structures. The area ratio of the region is rarely increased. Therefore, when measuring the film thickness on the gate, it is often possible to measure only by modeling with three structures as shown in this embodiment.

  Thus, according to the first embodiment, after forming transistors and capacitors, a sample having a complicated structure in which an insulating film is formed is used as an optical model of the sample as a gate region, an active region, and an element isolation. By synthesizing with three structures of regions and fitting the spectral waveform of the optical model to the detected spectral waveform, the film thickness and area ratio of each region of the sample can be measured with high accuracy.

(Embodiment 2)
FIG. 11 is a schematic diagram showing the gate structure of the sample in Embodiment 2 of the present invention, FIG. 12 is a schematic diagram showing the temperature dependence of the optical constant of the gate material in the sample of FIG. 11, and FIG. FIG. 14 is a schematic diagram for explaining the temperature-optical constant correspondence table of the gate material of the sample in the second embodiment, and FIG. 14 shows a processing example in the case where the optical constant of the gate material in the second embodiment of the present invention is obtained by fitting. It is a flowchart.

  In Embodiment 2, when the film thickness and area ratio of each region are calculated with respect to the sample on which the insulating film is formed after the formation of the transistor which is the semiconductor element shown in FIGS. A method for simultaneously obtaining the optical constant of the gate material in the gate region by fitting will be described.

  The gate is often made of a material whose crystal state changes depending on the process conditions such as polycrystalline Si, metal silicide, etc., but these optical constants vary greatly depending on the film formation conditions and annealing conditions, and can be used to calculate the film thickness. Affect.

  Therefore, at the time of fitting, the optical constant of the gate material at the film thickness measurement location is obtained simultaneously. Here, the method will be described with reference to FIGS. 11, 12, 13, and 14.

Here, as a gate material, a case where the gate electrode is constituted by a two-layer structure of a polySi film 48 and a WSi ₂ film 64 as shown in FIG. 11 will be described.

The crystal state of the polySi film 48 and the WSi ₂ film 64 changes depending on the film forming conditions or annealing conditions, and the optical constants change accordingly. Here, the WSi ₂ film 64 will be described as an object.

FIG. 12 shows changes in the refractive index (n) of the optical constant and the extinction coefficient (k) when the temperature is changed as the film formation condition of the WSi ₂ film 64. As shown in the figure, the refractive index 101 and the extinction coefficient 103 change to the refractive index 102 and the extinction coefficient 104, respectively, as the temperature rises.

These changes are also caused by heating by annealing after film formation. When calculating the thickness d ₁ of SiO ₂ on the gate by fitting, the optical constants of the underlying polySi film 48 and WSi ₂ film 64 are used. Therefore, if the optical constants of the polySi film 48 and the WSi ₂ film 64 at the time of calculating the film thickness are different from the optical constants of the actual film, the correct film thickness cannot be obtained.

Therefore, when the respective film thicknesses d ₁ , d ₂ , d ₃ and the area ratio of the gate region 41, the active region 42, and the element isolation region 43 are obtained, the optical constant of the WSi ₂ film 64 is also obtained by fitting.

  The method will be described below.

  The film thickness measurement apparatus according to the second embodiment is the same as the film thickness measurement apparatus shown in FIG. 1 except for the optical model creation process 32 and the fitting process 33 of the film thickness measurement processing unit 30.

The optical model created by the optical model creation processing 32 is the same as the optical model of the first embodiment except that the optical constant of the WSi ₂ film 64 of the gate material becomes a variable as a parameter.

In the fitting process 33, fitting is performed using the optical constant of the WSi ₂ film 64 as a parameter as well as the film thickness and area ratio. In general, when fitting using optical constants as parameters, the refractive index and extinction coefficient are expressed by a dispersion formula that is a function of wavelength.

  As the dispersion formula, there is a method in which the refractive index and extinction coefficient are expressed by an nth-order polynomial of wavelength as in the Cauchy dispersion formula used for a transparent sample, or a vibrator model used for a sample having absorption.

  When using the dispersion formula for the refractive index and extinction coefficient, by fitting the coefficients in the dispersion formula in the same manner as the parameters of the film thickness and area ratio, the coefficient of the dispersion formula that minimizes the square error, and the film thickness, The area ratio can be obtained.

  However, using the dispersion formula increases the number of parameters for fitting. Therefore, as shown in FIG. 12, since the refractive index and extinction coefficient change depending on the film forming temperature, temperature is used as a parameter for the refractive index and extinction coefficient.

  Since it is difficult to express the refractive index and the extinction coefficient as a function of temperature, a temperature-optical constant correspondence table 191 as shown in FIG. 13 is created, and fitting is performed using the temperature-optical constant correspondence table 191.

  The temperature-optical constant correspondence table 191 uses a spectroscopic ellipsometer or the like to determine the wavelength dependence of the refractive index and extinction coefficient for each film forming temperature using a dummy pattern of the sample before film thickness measurement.

  Next, the details of the fitting process 33 will be described with reference to the process flow shown in FIG.

  The process is a two-stage process. First, an approximate value of optical constants and then a detailed value are obtained.

  First, the increment of the parameter (temperature) for scanning the temperature-optical constant correspondence table 191 with the large optical constant parameter increment setting is set (step S201). Here, the increment is set to 150 ° C.

In the temperature-optical constant correspondence table 191, the optical constant of the WSi ₂ film 64 is set for each temperature from 500 ° C. to 150 ° C., the film thickness and the area ratio are fitted (step S202), and the processing up to the final data is performed. Confirmation (step S203).

  Subsequently, the optical constant of the temperature that minimizes the square error is determined as an approximate value (step S204), and the parameter (temperature) increment for determining the detailed value of the optical constant is set (step S205). Here, it is 50 degreeC.

  Thereafter, optical constants before and after the temperature determined in the process of step S204 are set, and the film thickness and area ratio are fitted (step S206). After confirming the end of the data (step S207), the optical constant of the temperature that minimizes the square error, the film thickness, and the area ratio are respectively determined (steps S208 and S209).

  Thereby, in the second embodiment, the optical constant is obtained by simultaneously fitting the optical constants of the material whose crystal state changes depending on the process conditions such as polycrystalline Si and metal silicide constituting the gate when calculating the film thickness. And the film thickness can be measured with high accuracy.

  In the second embodiment, 50 ° C. is set as the increment in the process of step S205, but the increment is made smaller than that, and the optical constant at each temperature is interpolated from the data of the temperature-optical constant correspondence table 191. You may ask for it.

  Here, the optical constant of the gate material was measured for three structural parts as a sample. However, if there is a uniform and large pattern on the specimen, such as a dummy pattern that is the same as the gate, only the optical constant is measured at that part. A method of measuring alone is also conceivable. In this case, the measurement time can be shortened.

  Furthermore, in the second embodiment, when determining the optical constant of the gate material in the gate region, the optical constant is represented by the film formation temperature, but it may also be represented by a small number of parameters such as the composition ratio or annealing temperature. Is possible.

(Embodiment 3)
FIG. 15 is a schematic diagram showing the apparatus configuration of the film thickness measurement apparatus according to Embodiment 3 of the present invention, and FIG. 16 shows the reflected light at the film thickness measurement location detected by the image detection system of the film thickness measurement apparatus in FIG. It is the schematic explaining a detection image.

  In the third embodiment, as in the first and second embodiments, as shown in FIGS. 4, 9, and 10, the film thickness of the sample in which the insulating film is formed after the formation of the transistor as the semiconductor element, When calculating the area ratio, obtain the detection image of the sample from the reflected light of the sample, calculate the area ratio of each region from the obtained detection image, and use the calculated area ratio of each region to calculate the number of parameters in the fitting The method of reducing the calculation range in fitting or reducing fitting is shown.

  As shown in FIG. 15, the configuration of the film thickness measuring apparatus in the present third embodiment detects an image of the film thickness measurement location on the semiconductor wafer, compared to the structure of the film thickness measuring apparatus in the first embodiment. An image detection system for calculating the area ratio of each region is added.

  The film thickness measuring device includes a detection optical system 70 and a film thickness measurement processing unit 80.

  The detection optical system 70 is obtained by adding a half mirror 71, a mirror 72, an imaging lens 73, and a video camera 74 to the detection optical system of the first embodiment.

  As in the first embodiment, the illumination light source 16 irradiates the semiconductor wafer 2 with white illumination light via the objective lens 12, and reflects the reflected light from the semiconductor wafer 2 via the pinhole of the spatial filter 20. The zero-order light of the reflected light is guided to the spectrometer 22. The spectral waveform dispersed by the spectroscope 22 is input to the film thickness measurement processing unit 80 as an electric signal, and the film thickness and area ratio are calculated.

  On the other hand, in the image detection system added in the third embodiment, the reflected light from the semiconductor wafer 2 is divided by the half mirror 71, and the film thickness is reflected by the video camera 74 via the mirror 72 and the imaging lens 73. A photodetection image is obtained. The reflected light detection image detected by the video camera 74 is input to the film thickness measurement processing unit 80, and the area ratio of each region is calculated.

  In the film thickness measurement processing unit 80, the spectral waveform is input in the spectral waveform input process 31, the image signal of the detected portion is input in the image input process 81, and the area ratio of each region on the sample surface is calculated in the area ratio calculation process 82. In the optical model creation process 83, the optical model of the sample is composed of a composite of the three structures of the gate region, the active region, and the element isolation region. In the fitting process 84, the detected spectral waveform is obtained by the optical model. Fitting is performed by changing the parameters of the film thickness and area ratio of the obtained spectral waveform, and a condition for minimizing the square error is obtained.

  Next, main processes of the film thickness measurement processing unit 80 will be described.

  In the area ratio calculation process 82, as shown in FIG. 16, from the reflected light detection image 221 input in the image input process 81, the gate region 41, the active region 42, and the element isolation region 43 are divided by image processing. The area of each region is obtained from the number of pixels.

  As a method for distinguishing the gate region 41, the active region 42, and the element isolation region 43 from the reflected light detection image 221, for example, the reflectance of each region is different or the wavelength component of the reflected light of each region is different. The region is determined from the reflection intensity of the reflected light detection image.

  Further, the division of each region and the area calculation can be specified by the operator with a cursor on the monitor without using image processing. Of the calculated areas of the gate region 41, the active region 42, and the element isolation region 43, the area of the measurement visual field 91 shown in FIG. 16 is calculated, and the area ratio of each region to the measurement visual field 91 is calculated. .

  In the optical model creation process 83, an optical model is created, but the film structure, optical constants, and the like are the same as those in the first embodiment. However, since the area ratio of each region is known, the ratio of the single structure portion and the boundary portion in each region is used as the area ratio parameter. Therefore, the number of area ratio parameters can be reduced.

  In the fitting process 84, the spectral waveform of the reflected light obtained by the optical model is fitted to the spectral waveform of the reflected light detected from the sample, and the film thickness of the gate region, the active region, the element isolation region, and in each region The ratio of the single structure part and the boundary structure part is calculated.

  In the fitting, since the area ratios of the gate region, the active region, and the element isolation region are known, the number of parameters is reduced, and the calculation time can be reduced.

  In the third embodiment, when fitting the spectral waveform of the reflected light obtained from the optical model to the spectral waveform of the detected reflected light, a detection image of the sample is obtained from the reflected light of the sample, and each detected image is obtained from the obtained detection image. The area ratio of the region was calculated, and using the calculated area ratio of each region, the number of parameters in fitting was reduced, and the film thickness was calculated.

  Thereby, in the third embodiment, since the number of parameters in the fitting can be reduced, the film thickness can be measured with high accuracy while shortening the calculation time.

As another method of reducing the parameters at the time of fitting, there is a method of using the measurement result when the film thickness is measured in the STI process before the transistor is formed. In the STI process, for example, as shown in FIG. 17C, the film thicknesses d ₁ and d ₂ and the area ratio of the active region 42 and the element isolation region 43 are measured.

  Measurement is performed by designating a position on a semiconductor chip. Therefore, when measuring the film thickness of the insulating film after transistor formation, the number of parameters in the fitting can be calculated by using the measured active region at the same location on the semiconductor chip, the measured film thickness of the element isolation region, and the measured area ratio. Can be reduced.

  When determining the thickness and area ratio of the gate region, active region, and element isolation region by fitting, as a method of reducing parameters, when measuring the same location on a semiconductor wafer or multiple semiconductor chips, each semiconductor on the semiconductor wafer Since the same part of the chip has almost the same structure and dimensions, the area ratio of each region is almost the same, so the number of parameters in fitting is reduced, or the calculation range in fitting is reduced to reduce the film thickness and area ratio. There is a way to measure.

  As mentioned above, the invention made by the present inventor has been specifically described based on the embodiment. However, the present invention is not limited to the embodiment, and various modifications can be made without departing from the scope of the invention. Needless to say.

  The film thickness measurement technique of the present invention is suitable for a highly accurate film thickness measurement technique such as a resist film and an insulating film in a manufacturing process of a thin film device such as a DVD, a TFT, and an LSI reticle in addition to a semiconductor device such as silicon. .

It is the schematic which shows the apparatus structure of the film thickness measuring apparatus in Embodiment 1 of this invention. It is the schematic explaining the processing system of the film thickness measuring apparatus of FIG. It is the schematic which shows the line structure in the manufacturing line of the semiconductor device in Embodiment 1 of this invention. It is the schematic which shows the structure of the sample of the measuring object in the film thickness measuring apparatus of FIG. It is the schematic which shows the film | membrane structure of the single structure in the sample of FIG. It is a flowchart which shows an example of the process of the optical model creation process in the film thickness measuring apparatus of FIG. FIG. 2 is a schematic diagram illustrating fitting of a detected spectral waveform and a spectral waveform by an optical model in the film thickness measuring apparatus of FIG. 1. It is the schematic explaining the example of a display of the film thickness output to the monitor in the film thickness measuring apparatus of FIG. 1, and an area ratio calculation result. It is the schematic which shows a structure in case the pitch of the sample of the measuring object of FIG. 4 is irregular. It is the schematic which shows the structure in case each area | region of the sample of the measurement object of FIG. 4 is single. It is the schematic which shows the gate structure of the sample in Embodiment 2 of this invention. It is the schematic which shows the temperature dependence of the optical constant of the gate material in the sample of FIG. It is the schematic explaining the temperature-optical constant correspondence table | surface of the sample gate material in Embodiment 2 of this invention. It is a flowchart which shows the process example in the case of calculating | requiring the optical constant of the gate material in Embodiment 2 of this invention by fitting. It is the schematic which shows the apparatus structure of the film thickness measuring apparatus in Embodiment 3 of this invention. It is the schematic explaining the reflected light detection image of the film thickness measurement location detected by the image detection system of the film thickness measuring apparatus of FIG. It is the schematic explaining the film thickness measuring method which this inventor examined.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Detection optical system, 2 ... Semiconductor wafer (sample), 11 ... XY stage, 12 ... Objective lens, 13 ... Half mirror, 14 ... Condensing lens, 15 ... Aperture stop, 16 ... Illumination light source (irradiation means), 17 DESCRIPTION OF SYMBOLS ... Imaging lens, 18 ... Field stop, 19 ... Relay lens, 20 ... Spatial filter (detection means), 21 ... Relay lens (detection means), 22 ... Spectroscope, 30 ... Film thickness measurement processing part, 31 ... Spectral waveform Input process 32... Optical model creation process (optical model creation means) 33... Fitting process (fitting process means) 34... Film thickness / area ratio calculation process (film thickness calculation means) 35 35 Monitor 41. , 42 ... active region, 43 ... isolation region 44 ... silicon substrate, 45 to 47 ... SiO ₂ film, 48 ... poly-Si film, 51 ... first structure, 52 ... second structure, 53 ... third structure 57 Transistors, 64 ... WSi ₂ film, 70 ... detection optical system, 71 ... half mirror, 72 ... mirror, 73 ... imaging lens 74 ... video camera, 80 ... film thickness measuring unit, 81 ... image input process, 82 ... Area ratio calculation process, 83 ... Optical model creation process, 84 ... Fitting process, 85 ... Thickness calculation process, 86 ... Monitor, 91 ... Measurement field of view, 101,102 ... Refractive index, 103,104 ... Extinction coefficient, 131 ... Deposition apparatus 132 ... CMP apparatus, 133 ... Film thickness measuring apparatus, 134 ... Exposure apparatus, 135 ... Etching apparatus, 171 ... Semiconductor wafer, 172 ... Semiconductor chip, 173 ... Thickness / area ratio display area, 180 ... Semiconductor chip 181 ... Measurement position, 182 ... Film thickness / area distribution display area, 183 ... Output item, 191 ... Temperature-optical constant correspondence table, 231 ... Spectral waveform, 232 ... Optical waveform.

Claims (12)

A sample having an optically transparent surface is irradiated with white light, reflected light from the sample surface due to the irradiation of the white light is detected, and a spectral waveform of the detected zero-order light of the reflected light is detected. A film thickness measuring method for determining the film thickness of the film based on
The optical model of the sample is composed of a combination of three or more structures optically different in structure and material, and the spectral waveform of the zero-order light of the reflected light obtained by detecting the spectral waveform of the reflected light obtained from the optical model A film thickness measuring method characterized in that the surface film thickness and the area ratio of each structure of the sample are obtained by fitting to the surface. In the film thickness measuring method according to claim 1,
An optical model of the sample is configured by combining three structures of a gate region, an active region, and an element isolation region, and a spectral waveform of reflected light obtained from the optical model is a spectral waveform of zero-order light of the detected reflected light. A film thickness measuring method, wherein the film thickness and the area ratio in the gate region, the active region, and the element isolation region are obtained by fitting to each other. In the film thickness measuring method according to claim 2,
By fitting the spectral waveform of the reflected light obtained from the optical model to the spectral waveform of the detected zero-order light of the reflected light, the film thickness, area ratio, and optical constant in the gate region, active region, and element isolation region The film thickness measuring method characterized by calculating | requiring. In the film thickness measuring method according to claim 3,
By fitting the spectral waveform of the reflected light obtained from the optical model to the spectral waveform of the detected zero-order light of the reflected light, the gate region, the active region, the film thickness in the element isolation region, the area ratio, and the gate region A method for measuring a film thickness, comprising: obtaining an optical constant of a gate material in In the film thickness measuring method according to claim 3,
A method for measuring a film thickness, wherein the optical constant in each region is obtained by fitting using a small number of parameters such as a film formation temperature or a composition ratio. In the film thickness measuring method according to claim 4,
A method for measuring a film thickness, characterized in that when an optical constant of a gate material in the gate region is obtained, fitting is performed using a small number of parameters such as a film formation temperature or a composition ratio. In the film thickness measuring method according to claim 2,
When obtaining the film thickness and area ratio in the gate region, the active region, and the element isolation region by fitting the spectral waveform of the reflected light obtained from the optical model to the spectral waveform of the detected zero-order light of the reflected light The detection image of the sample is obtained from the reflected light of the sample, the area ratio of each region is calculated from the obtained detection image, and the number of parameters in the fitting is reduced using the calculated area ratio of each region. Alternatively, a film thickness measurement method characterized by reducing a calculation range in fitting. In the film thickness measuring method according to claim 2,
By fitting the spectral waveform of the reflected light obtained from the optical model to the spectral waveform of the detected zero-order light of the reflected light, the film thickness and the area ratio in the gate region, the active region, and the element isolation region are obtained. At this time, by using the active area in the completed STI process, the measured area ratio of the element isolation region, and the measured film thickness, the number of parameters in fitting is reduced or the calculation range in fitting is reduced. Measurement method. In the film thickness measuring method according to claim 2,
By fitting the spectral waveform of the reflected light obtained from the optical model to the spectral waveform of the detected zero-order light of the reflected light, the film thickness and the area ratio in the gate region, the active region, and the element isolation region are obtained. In this case, the film thickness measuring method is characterized in that the number of parameters in the fitting is reduced or the calculation range in the fitting is reduced by using an area ratio at substantially the same position between the semiconductor chips on the semiconductor wafer. An irradiation means for irradiating the sample on which the surface is formed with an optically transparent film with white light;
Detecting means for detecting zero-order light of reflected light emitted from the sample irradiated by the irradiation means;
An optical model creating means configured to compose an optical model of the sample by combining three or more structures having different optical structures and materials;
Fitting processing means for fitting the reflected light spectral waveform obtained from the optical model creating means to the spectral waveform of the zero-order light of the reflected light detected;
A film thickness measuring apparatus comprising: a film thickness calculating means for calculating a film thickness of each structure of the sample and an area ratio from the fitting processing means. An irradiation means for irradiating the sample on which the surface is formed with an optically transparent film with white light;
Detecting means for detecting zero-order light of reflected light emitted from the sample irradiated by the irradiation means;
An optical model creating means configured to synthesize an optical model of the sample by combining three structures of a gate region, an active region, and an element isolation region;
Fitting processing means for fitting the reflected light spectral waveform obtained from the optical model creating means to the spectral waveform of the zero-order light of the reflected light detected;
A film thickness measuring apparatus comprising: a film thickness calculating means for calculating a film thickness and an area ratio in the gate region, the active region, and the element isolation region of the sample from the fitting processing means. An irradiation means for irradiating the sample on which the surface is formed with an optically transparent film with white light;
Detecting means for detecting zero-order light of reflected light emitted from the sample irradiated by the irradiation means;
An optical model creating means for synthesizing the optical model of the sample with three structures of a gate region, an active region, and an element isolation region, and configuring the optical constant of the gate material as a parameter;
Fitting processing means for fitting the spectral waveform of the zero-order light of the reflected light obtained by detecting the reflected light spectral waveform obtained from the optical model creating means as a parameter with a film thickness, an area ratio, and an optical constant;
A film thickness measuring device comprising: a film thickness calculating means for calculating the film thickness, area ratio, and optical constant of the gate material in the gate region, active region, and element isolation region of the sample from the fitting processing means .