JP2001007174A - Method and device for measuring pattern density, film- thickness measuring method, and process control method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To easily provide pattern information by measuring at least one selected among brightness, color, and edge position of the pattern on a substrate on which a pattern structure is provided. SOLUTION: A semiconductor element 1 is irradiated (6) with a lighting source 5, and the image of a pattern on the semiconductor element 1 is acquired as 2-dimensional information using an imaging element 3. At an image processing device 4, the edge position of a specific pattern which is to be measured is detected from the image of the pattern. The detected edge position is measured to provide the form of the specific pattern. If there is a distribution in color or brightness of the pattern image, the image processing device 4 partitions the pattern image into a plurality of regions based on the difference in color or brightness. Thus, the pattern density of a device pattern on a semiconductor wafer is provided at high speed.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method and an apparatus for measuring a pattern density on a substrate, in particular, a pattern density of a device pattern on a semiconductor wafer, and an insulating layer or a metal electrode of a semiconductor element in a semiconductor device manufacturing process. The present invention relates to a method for measuring a film thickness of a film or a coating resist layer, and a method for measuring a process end point in a film forming process or a removing process of the layer and the film, and the like.

[0002]

2. Description of the Related Art Densification of semiconductor devices continues to advance without showing a limit. As the densification increases, multilayer wiring of semiconductor devices and accompanying formation of interlayer insulating films and formation of electrodes such as plugs and damascenes are performed. Technology is becoming more and more important, and the device structure is becoming more and more complicated. In a manufacturing process of a semiconductor device having a complicated structure, it is increasingly important to know the thickness of a constituent film and a resist layer of a semiconductor element in an intermediate process.

In the measurement of the film thickness of a semiconductor element having such a complicated device structure, generally, a minute blank portion having no pattern (a place having no two-dimensional distribution of the film thickness) is designated, and the designated place is designated. The thickness of the other portion is calibrated and calculated from the measured value based on the calibration data obtained in advance. In order to specify and probe the measurement position, not only an image sensor for capturing the measurement position and a precision alignment mechanism, but also image processing software and the like are required. It becomes large and expensive in terms of software. Further, the time required for image processing, position search, and position determination greatly increases the measurement time of the measurement device.

In order to solve the above problems, a method of directly measuring a pattern portion, that is, a fine structure portion, and separately measuring the thickness of the portion has been developed.
-250071, Japanese Patent Application No. 10-289175, Japanese Patent Application No. 1
No. 0-150963, Japanese Patent Application No. 11-047485, and the like, and irradiates a semiconductor element surface with a light spot having a light diameter covering a relatively large measurement range, and has a nonuniform film thickness in the irradiated surface. The film thickness is measured even if a portion, that is, a pattern exists.

[0005] Since the measurement position is not specified or probed,
A precise alignment function is not required, and measurement at various positions on the semiconductor element can be performed at high speed. For this reason, the present technology uses a planarization process (CMP: chemical-mech) for a semiconductor element in a semiconductor device manufacturing process.
The method can also be used for detecting a process end point in a process such as analytic polishing, thin film formation, or etching.

[0006]

In the method of measuring the required film thickness of the irradiated portion by irradiating light to the entire semiconductor element having a pattern structure, that is, the fine structure portion, as described above, It is important to obtain the pattern information of the element. The most important of the pattern information is the two-dimensional information of the pattern structure, and a method for easily obtaining the two-dimensional information has not been found conventionally.

The present invention provides a method and apparatus for easily obtaining the pattern information, a film thickness measuring method and a film thickness measuring apparatus using the same, and a process for measuring a process end point in a film forming process and a removing process. It relates to a control method.

[0008]

Accordingly, the present invention proposes a method for easily obtaining necessary pattern information for the purpose of knowing the thickness of a semiconductor element wafer.
Therefore, first, a pattern density measuring method characterized by comprising a step of measuring at least one selected from luminance, color, and edge position of a pattern on a substrate having a pattern structure on the surface. 1) ".

A second aspect is that the pattern structure includes two or more layers of each pattern, and further includes a step of measuring not only the patterns but also the degree of overlap of each pattern. The described pattern density measurement method (Claim 2) is presented. Thirdly, "the measurement of each pattern is performed sequentially after the formation of each pattern layer, and the measurement result of the lower layer pattern is used in the measurement of the upper layer pattern. A density measurement method (Claim 3) is presented.

A fourth feature is that the measurement of each of the patterns is collectively performed after forming a multilayer pattern by using a detection mechanism capable of imaging two or more layers from the upper layer to the lower layer at a time. A pattern density measuring method according to claim 2 (claim 4) is presented. Fifthly, there is provided a "pattern density measuring method according to claim 4 wherein the pattern is separated into each layer by further using the difference in imaging conditions".

A sixth aspect of the present invention provides a "pattern density measuring method according to the fourth aspect of the present invention, wherein the pattern is further separated into each layer using the color information of the acquired image." . Seventh, a “substrate having an imaging lens, an imaging device, and an image processing device, wherein the imaging device is formed on the light receiving surface of the imaging device by the imaging lens, and has a pattern structure on the surface. Obtaining an image of the pattern above, the image processing apparatus, the luminance of the obtained image of the pattern, color, pattern pattern characterized by measuring the pattern density of the pattern from one or more selected from the edge position Density measuring device (Claim 7) "is presented.

Eighthly, "when the pattern density measuring method according to any one of claims 1 to 6 is provided, the spectral density of a pattern surface of a substrate having a pattern structure is further determined by using the pattern density. A film comprising a step of performing a simulation calculation, a step of measuring a spectral characteristic of the substrate to be measured, and a step of calculating a film thickness by performing a fitting calculation between the simulated spectral characteristic and the measured spectral characteristic. Thickness measurement method (Claim 8) "is presented.

A ninth aspect is that “the step of providing the pattern density measuring method according to any one of claims 1 to 6 further includes the step of determining a process condition of a semiconductor manufacturing process using the pattern density. Process control method (claim 9) characterized in that

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a measurement according to the present invention will be described below with reference to the drawings, but the present invention is not limited to the drawings. [Embodiment 1] This embodiment relates to acquisition of pattern information of a pattern structure of a semiconductor device. Here, the pattern structure means that the semiconductor element has a two-dimensional structure when viewed from observation light for film thickness measurement and the like. This two-dimensional structure has, for example, a metal electrode layer having a pattern (a shape formed by fine processing or the like).

The two-dimensional information of the pattern structure of the semiconductor element includes a plurality of elements. In the present embodiment, the area ratio of the pattern existing portion (electrode portion, etc.), that is, the pattern density is determined as the object to be measured. I do. For example, in the case where a simply modeled electrode pattern (having an antireflection layer laminated on an electrode metal) having a cross section as shown in FIG. This corresponds to the ratio of the area of a certain part A to the area of a part B having no electrode layer. For example, mathematically, the area of the i-th electrode layer is A _i , and the area of the i-th electrode is Assuming that the area of the non-existing portion is B _i , it is expressed by the following equation.

Pattern density = ΣA _i / (ΣA _i + ΣB _i ) Here, ΣA _i means taking the sum of the areas of all the electrode parts, and ΣB _i is the sum of the areas of the parts without all the electrodes. Means to take. The pattern density can be calculated from a design drawing of a semiconductor element (usually digitized by CAD), but the calculation procedure is generally quite complicated. Further, in a semiconductor element manufactured in an actual process, the width of an electrode pattern often differs from a design value (so-called pattern narrowing), and it is more preferable to measure the width from the actual semiconductor element itself.

Therefore, in the present embodiment, the pattern density is calculated using the actual image of the semiconductor device. For this purpose, an appropriate element observation system (optical system + imaging system) is provided, and the obtained image is subjected to image processing to calculate the pattern density. FIG. 5 is a diagram of the pattern density measuring device of the present embodiment. In FIG. 5, reference numeral 1 denotes a semiconductor element having a pattern structure to be measured, 2 denotes an imaging lens, 3 denotes an image sensor,
Is an image processing apparatus, 5 is an irradiation light source, and 6 is irradiation light.

In the measurement, first, the semiconductor element 1 having the pattern structure is irradiated 6 with the illumination light source 5. Next, an image of the pattern on the semiconductor element is acquired as two-dimensional information using the image sensor 3, and then the edge (boundary) position of the specific pattern to be measured is detected from the image of the pattern by the image processing device 4. . The edge position is detected by, for example, scanning the acquired two-dimensional image in a one-dimensional direction across the specific pattern as shown in FIG. Is determined by regarding the edge position. This edge position is also obtained for the other end of this specific pattern, and the difference between them is used to obtain the width of this specific pattern. This scanning is also performed for the other part of the specific pattern or in the direction orthogonal to the specific pattern, and the area of the specific pattern is obtained by measuring the edge position to obtain the shape of the specific pattern.
The above measurement is performed for all patterns of the same type as this specific pattern, the area of each pattern is obtained, and the total area of all the specific patterns to be measured is calculated by taking the sum of these areas. The density of the specific pattern is calculated from the ratio between the total area of all the specific patterns to be measured and the area corresponding to the entire measurement visual field taken in. If there are other different types of patterns, the pattern density is obtained in the same manner.

The image obtained by the above method may be an image for white light or an image for monochromatic light, but if this pattern density is used for film thickness measurement, it is preferable to measure the film thickness. Irradiation light having the same spectrum as is preferred. For calculating the pattern density, color or luminance information of a pattern image acquired as two-dimensional information can be used. In this case, the image sensor can detect the color or luminance of the image. If there is a distribution in the color or luminance of the acquired pattern image, the image processing device 4
The image of the pattern is divided into a plurality of regions based on the difference in color or luminance. Specifically, for example, the acquired image is divided into a plurality of regions based on appropriately provided reference values of color or luminance. Each section includes a portion corresponding to the ground other than the pattern, but includes one or more types of patterns. The sum of the number of pixels belonging to the section corresponding to the specific type of pattern is calculated, and the total number of pixels (
By calculating the ratio to the total number of pixels in the entire measurement field of view, each pattern, that is, each pattern density of the specific pattern is calculated. If there are other different types of patterns, the pattern density can be obtained in a similar manner. The above-described method using edges and the method using colors or luminance may be used alone or in combination. When used in combination,
For example, it is possible to roughly determine a boundary based on color, and thereafter detect the boundary with high accuracy by using an edge to improve measurement accuracy.

The pattern density measurement in the case of a multilayer structure in which the electrode patterns of the semiconductor element are arranged in multiple layers is performed as follows. When the pattern is a multilayer, it is necessary to sufficiently consider a signal acquisition system and to perform detailed simulation calculations in order to correctly measure each pattern density on the semiconductor element.

When the pattern is a multilayer, the pattern density of each layer must be obtained as a value obtained by removing the overlapping portion of another pattern that is superimposed on the pattern when calculating the pattern density of each layer. It is calculated by removing the overlapping portion at the top of each image.
When a pattern is a multilayer, one or more of the following methods are used for separation evaluation of each pattern.

First, each time a pattern is formed on a semiconductor device, the pattern just formed is observed at the same location on the device by edge, color, or brightness. The pattern density information of each pattern is obtained by this observation. At this time, if the lower layer can also be observed, information on the lower layer is obtained together with the upper layer. By repeating this observation, it is possible to calculate the pattern density from which the overlapping portion of each pattern has been removed, using both the image information of the lower layer already acquired and the image information of the multilayer element newly acquired.

Generally, the pattern is formed such that the device electrode is embedded in a dielectric layer (such as SiO ₂ ).
At a light wavelength used in a normal observation optical system, the dielectric layer is transparent, so that the lower layer can be seen therethrough. When it is difficult to obtain pattern information by observation in the visible light range in a semiconductor device having a structure very finer than the wavelength of visible light, if a pattern of two or more layers from the upper layer to the lower layer is formed by using a UV microscope, etc. Collective observation becomes possible. In the calculation by observation for each layer described above, the lower layer and the upper layer do not necessarily need to be visible at the same time,
Collective observation is effective for position confirmation for measurement at the same place on a semiconductor element.

When using such an optical system capable of observing the upper layer and the lower layer at the same time, it is also a preferable method to utilize the difference in the optical conditions under which each layer is observed. For example,
In microscopic observation, a method of separating the pattern images of the respective layers according to the difference in the focus (focusing) condition on the observation surface is a preferable method. By sequentially performing focusing (focusing) on each layer and acquiring a pattern image in that layer, it is possible to calculate a pattern density in which an overlapping portion of each layer is removed.

Further, as a simpler method, there is a method using color information. The electrodes of each layer appear to have different colors due to the difference in the thickness of the dielectric layer on top, not only when different materials are used, but also for electrodes made of the same material. That is, since the thickness of the dielectric layer on the upper layer differs between the pattern in the lower layer and the pattern in the upper layer, the color of each pattern is different. Therefore, the pattern density of each layer can be easily calculated by acquiring the acquired image in color and separating the pattern area of each layer according to the color difference. This method is more effective when used in combination with the above-described focusing change method. These methods can be applied not only to the metal pattern but also to a pattern of a dielectric film such as a trench, and can be similarly calculated even when both are mixed.

The imaging device used for pattern density measurement is not particularly limited, but a solid-state imaging device, particularly, a CCD device is preferably used. As described above, the measured pattern density is used as important basic data for measuring the thickness of the dielectric film on the semiconductor element. The pattern density can be effectively used for determining a process condition in a semiconductor process other than the film thickness measurement described herein. For example, when evaluating the influence of an underlayer upon exposure of a resist in a lithography process, an electrode (an antireflection layer such as titanium nitride is usually provided on the electrode) is disposed at what ratio. Knowing what is done is useful information. [Embodiment 2] FIG. 6 is a diagram showing a film thickness measurement method of this embodiment.

In the present embodiment, the spectral reflectance (or spectral transmittance) of the semiconductor device is calculated by simulation using the pattern density obtained as the information on the pattern obtained in the first embodiment, and then the calculated value is compared with the semiconductor device to be measured. The film thickness is measured by fitting calculation with the actually measured spectral characteristics. When calculating the reflectivity of the pattern surface of the semiconductor element using the measured pattern density, Japanese Patent Application No. 11-047
485, under the condition that the coherence of irradiation light is controlled. For example, if the spatial coherence length is adjusted so that the irradiation light interferes over an irradiation surface including a sufficient number (usually several or more) of the basic units of the pattern,
The reflectance of the zero-order light with respect to normal incidence does not depend on the difference in pattern arrangement and the difference in fineness of the pattern, but only on the density of the pattern (for example, the density of the metal electrode layer). That is, if only the pattern density information is obtained, the spectral reflectance can be calculated.

The above-described simulation calculation of the spectral reflectance is performed by using the pattern density measured in the first embodiment and further forming a film having various thicknesses on the uppermost layer pattern. Do it. In the calculation, the optical constant and the film thickness of the constituent film obtained by another method are used. The simulation calculation value calculated for each film thickness value may be stored in the signal processing unit 8 in FIG.

In the next stage, a light source (not shown) of the optical measuring section 7 irradiates the pattern surface of the semiconductor element to be measured with white light (or a component obtained by spectrally separating the white light) vertically. The reason for this perpendicular incidence is that in the optical measurement of a semiconductor element in which a device pattern exists, when a light beam is obliquely incident, the reflected light and transmitted light are greatly affected by the pattern arrangement, and are generally used. This is because it becomes difficult to perform accurate measurement. For example, by making the light incident perpendicularly, it becomes easy to remove the diffracted component of the primary light or more from the reflected light or transmitted light. Of course, a method of irradiating from the back surface of the wafer as a semiconductor element (in this case, transmitted light is generally detected) may be used, but in this case, a multi-component wavelength light source in the infrared region is required. In this case as well, the light is incident vertically. The above irradiation light is irradiated with a spot diameter larger than the basic unit (minimum unit) of the pattern.
The light reflected from the pattern surface is separated by a spectroscope (not shown) in FIG. 6 and then detected by an optical detection device (not shown) in FIG. 6, and a spectral reflection signal is sent to the signal processing unit 8.

The signal processing section 8 compares the spectral reflection signal with the spectral reflectance for each film thickness calculated by simulation. Specifically, it is preferable to perform a fitting calculation, and the fitting calculation is most preferably a method using a cross-correlation coefficient. As a result of this calculation, the film thickness when the highest cross-correlation coefficient is obtained, which is equal to or larger than the reference value, is used as the measured value of the film thickness of the semiconductor element.

By performing such a measurement, it is possible to measure the thickness of a blank film in a conventional apparatus for measuring the film thickness using an optical interference phenomenon, but to solve the problem that the film thickness of a pattern film cannot be measured. did. As described above, the measurement target substrate is described as a semiconductor element in the first and second embodiments, but it goes without saying that the present invention can be applied to other general substrates.

[0032]

[Embodiment 1] FIG. 5 shows a pattern density measuring apparatus according to this embodiment. Here, a microscope was used as the imaging lens 2 and a CCD was used as the imaging device 3.

The semiconductor element 1 to be measured is a sensor device on a 6-inch wafer. This semiconductor element has a block structure composed of a plurality of blocks as shown in FIG. 2 in one chip (or semiconductor chip). Within each block, the density and fineness of the pattern are substantially uniform. In this example, an image with a magnification of about 300 times is acquired for the portion A (minimum line width of about 0.8 μm), which is the block to be measured, and the dimensions of the pattern are obtained by detecting the edge position, and the area ratio of the pattern (electrode) portion (That is, pattern density) was calculated to be about 24%. If the spot diameter of the irradiation light for measuring the film thickness in this chip is sufficiently smaller than the size of the portion A of the block and sufficiently larger than the minimum unit (basic size) of the pattern, the area ratio of the pattern portion is Density. The value of the pattern density calculated by dividing the same acquired image into a plurality of regions according to the difference in luminance also coincides with the value, and it was confirmed that the pattern density could be obtained easily and quickly by either method.

Using the value of 24% of the pattern density, the film thickness of the uppermost dielectric pair film arranged on the pattern of the portion A is variously changed, and the simulation calculation of the spectral reflectance is performed. The measurement of the film thickness of the dielectric film was performed by fitting the measured spectral reflectance signal and the spectral reflectance calculated by the simulation using a measuring device as shown in FIG. The spectral reflection characteristics and the simulation calculation were in good agreement, and highly accurate film thickness measurement (accuracy of 10 nm or more) became possible. [Embodiment 2] As in Embodiment 1, a pattern density measuring apparatus as shown in FIG. 5 was used, and a trench and a multilayer pattern of two electrode patterns were used as a pattern structure shown in an enlarged view in FIG. The pattern density of a device (semiconductor element) was measured. 4A is a schematic front view of the pattern, and FIG. 4B is a cross-sectional view taken along a cutting line MN in FIG.

Images were obtained while sequentially focusing on each layer. The color image obtained by the CCD image sensor is
Titanium silicide (Ti) formed on a silicon substrate E
The portion A of the Si) film, the trench portion B, the lower electrode portion C of titanium nitride (TiN), and the uppermost electrode portion D of titanium nitride (TiN) were observed in different colors. For this reason, the pattern and trench region of each layer can be divided by the separation routine of the same color region, whereby the pattern density of each layer is calculated, which is required at the time of actual film thickness measurement. A: B: C: D in FIG. , That is, the pattern densities of A, B, C, and D could be calculated at high speed. Here, A, B, C, and D are portions that are exposed when viewed from above the underlying substrate, the trench, the lower layer electrode, and the uppermost layer electrode, and of course, the overlapping portions are removed.

In the same manner as in Example 1, using these pattern densities, a dielectric pair film is arranged on the uppermost layer electrode pattern in the D portion, and the film thickness of the dielectric film is varied to change the spectral reflectance. The simulation calculation is performed, and the measurement of the thickness of the dielectric film is attempted by fitting using the cross-correlation coefficient between the spectral reflection signal measured by using the measuring device shown in FIG. 6 and the spectral reflectance calculated by the simulation. As a result, the actually measured spectral reflection signal and the simulation calculation were in good agreement, and the film thickness could be measured with a cross-correlation coefficient of 0.9 to 0.95 or more.

[0037]

As described above, according to the present invention, there is provided a pattern density measuring method and a measuring apparatus capable of easily and quickly obtaining the pattern density of a substrate having a pattern structure, particularly the pattern density of a device pattern on a semiconductor wafer. I got it. As a result, not only detection of the film thickness on a semiconductor element having a pattern structure in a semiconductor manufacturing process, but also other process control can be performed with high accuracy and effectiveness.

[Brief description of the drawings]

FIG. 1 is a schematic cross-sectional view of a pattern, showing what the pattern density means.

FIG. 2 is a diagram illustrating one of devices to be measured in the first embodiment.
It is a schematic diagram of a chip.

FIG. 3 shows a fitting result between actual measurement data and a simulation calculation value in the first embodiment.

FIG. 4A is an image of a device having a multilayer pattern acquired by a CCD imaging device in Example 2, and FIG. 4B is a cross-sectional view thereof.

FIG. 5 is a conceptual diagram of a pattern density measuring device according to the first embodiment of the present invention.

FIG. 6 is a conceptual diagram of an optical film thickness measuring device of the present invention.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 substrate to be measured (semiconductor element) 2 imaging lens (microscope) 3 imaging element (CCD) 4 image processing device 5 irradiation light source 6 irradiation light 7 optical measurement unit 8 signal processing unit 9 display unit 10 silicon substrate 11 thermal oxide film 12 Electrode layer 13 Anti-reflection film 14 Dielectric film

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 2F065 AA00 AA12 AA22 AA30 AA56 AA58 BB02 BB17 CC19 CC31 DD06 FF01 FF04 GG24 HH12 HH13 JJ03 JJ09 JJ26 NN20 QQ25 QQ26 QQ27 QQ29 QQ31 CAQA4CA106A04A01 DH03 DH12 DH50 DJ38 DJ40 5B057 AA03 BA02 CA01 CA02 CA08 CA12 CA16 DA08 DB02 DB05 DB06 DB09 DC04 DC16

Claims (9)

[Claims] 1. A pattern density measuring method, comprising the step of measuring at least one selected from luminance, color, and edge position of a pattern on a substrate having a pattern structure on a surface. 2. The pattern density according to claim 1, wherein said pattern structure comprises two or more layers of each pattern, and further comprising the step of measuring not only said each pattern but also the degree of overlap of each pattern. Measurement method. 3. The pattern density measurement according to claim 2, wherein the measurement of each pattern is performed sequentially after the formation of each pattern layer, and the measurement result of the lower layer pattern is used in the measurement of the upper layer pattern. Method. 4. The measurement of each pattern is performed by using a detection mechanism that can image two or more layers from the upper layer to the lower layer at a time.
3. The pattern density measurement method according to claim 2, wherein the measurement is performed collectively after forming the multilayer pattern. 5. The pattern density measuring method according to claim 4, further comprising separating the pattern into each layer by using a difference in image forming conditions. 6. The pattern density measuring method according to claim 4, further comprising the step of separating the pattern into each layer using the color information of the acquired image. 7. An imaging device comprising an imaging lens, an imaging device, and an image processing device, wherein the imaging device forms an image on a light receiving surface of the imaging substrate by the imaging lens and has a pattern structure on a surface. Acquiring an image of the pattern, wherein the image processing apparatus measures the pattern density of the pattern from one or more selected from luminance, color, and edge position of the acquired image of the pattern. apparatus. 8. The method according to claim 1, further comprising the step of: performing a simulation calculation of spectral characteristics of a pattern surface of a substrate having a pattern structure using the pattern density. A method for measuring a film thickness, comprising: measuring a spectral characteristic of a substrate to be measured; and calculating a film thickness by fitting calculation between the simulated spectral characteristic and the measured spectral characteristic. 9. The method according to claim 1, further comprising the step of determining a process condition of a semiconductor manufacturing process using the pattern density. Process control method.