WO2017148322A1

WO2017148322A1 - Device and method for measuring overlay error

Info

Publication number: WO2017148322A1
Application number: PCT/CN2017/074508
Authority: WO
Inventors: 彭博方; 陆海亮; 王帆
Original assignee: 上海微电子装备（集团）股份有限公司
Priority date: 2016-02-29
Filing date: 2017-02-23
Publication date: 2017-09-08
Also published as: TWI635373B; CN107340689B; TW201802621A; CN107340689A

Abstract

Disclosed is a device and method for measuring an overlay error. A measuring light adjustment assembly for adjusting measuring light to be centrosymmetric relative to the optical axis of a microobjective (46) is added to the device, so that the measuring light forms positive and negative level diffraction light by passing through this device and through an overlay measurement mark, and finally a diffraction spectrum of the positive and negative level diffraction light is displayed on a detector (411), wherein the spectrum of the positive level diffraction light and the negative level diffraction light on the diffraction spectrum are staggered to each other. A control system calculates an overlay error according to the diffraction spectrum, such that a broadband light source can be used during resource selection. Therefore, by means of this device and method, the wavelength range of measuring light is wider and any light spot-shaped light source can be used, such that the acquired measuring signals are more abundant, the precision of measurement is improved and the utilization rate of light energy is higher. A small-sized overlay measurement mark also can receive measuring light, and the device and method are also suitable for small-sized measured objects, thereby being better adapted to fine semiconductor products.

Description

Device and method for measuring casing error

Technical field

The present invention relates to the field of semiconductor lithography, and more particularly to an apparatus and method for measuring a splicing error.

Background technique

According to the lithography measurement technology roadmap given by ITRS (International Technology Roadmap for Semiconductor), as the key dimensions of lithography patterns enter the 22nm and below process nodes, especially the double exposure (Double Patterning) technology Application, measurement accuracy requirements for lithography process parameters overlay has entered the sub-nano field. Due to the limitation of imaging resolution limit, the traditional imaging- and image-based imaging-imaging overlay (IBO) has gradually failed to meet the requirements of the new process node for the engraving measurement. Diffraction-Based overlay (DBO) based on diffraction light detection is gradually becoming the main means of engraving measurement.

Since the diffraction angle of the diffracted light changes with the incident angle of the incident light, the incident light of different angles is diffracted by the mark having the grating structure to form diffracted light of each diffraction order, and the intensity distribution of the diffracted light of each diffraction order is formed. For the reflected light angle resolution spectrum. For example, in Chinese patent CN1916603A (application number: 200510091733.1, published on February 21, 2007), in an annular illumination mode, the angular resolution spectrum of the reflected light formed by each diffraction order diffracted light is on the CCD detector. Distribution.

Based on the above principles, a DBO technique is disclosed in U.S. Patent No. 7,779,727, issued to U.S. Pat. In the angular resolution spectrum of the diffracted light, the asymmetry between the same diffraction orders gives the engraving error of the engraved mark.

The device structure diagram of the technical solution is disclosed in the patent. As shown in FIG. 1, the light emitted by the light source 2 sequentially passes through the lens group L2 and the filter device 30 to form a narrow-bandwidth incident light, and the objective lens L1 concentrates the incident light onto the substrate. 6 sets of engraved marks. The detector 32 is located on the back focal plane of the objective lens L1, and the diffracted light of the engraved mark is collected by the objective lens L1 and reflected by the reflecting surface 34 to be received by the detector 32. The detector 32 measures the angular resolution of the reflected light formed by the diffraction and reflection of the light at each angle by the engraved marking. In order to obtain a wide range of angular resolution spectra, a large numerical aperture (NA) objective lens L1 is used in this scheme. Since the diffraction angles of diffracted light of different wavelengths are different, in order to prevent overlap between different wavelength angle spectra, the scheme adopts a filter. Device 30 filters the light emitted by source 2 to form a narrow bandwidth of measurement light. In principle, the solution can only measure the angular resolution of the reflected light at one wavelength at a time. For multi-wavelength measurements, a scheme for splitting the pupil plane 40 of the objective lens L1 provided in the patent can be used to simultaneously measure angular resolution spectra at a plurality of discrete wavelengths. Despite this, the patent can only measure a limited number of discrete wavelengths.

It can be seen that the wavelength range of the measurement light used for the measurement of the overlay error in the prior art is limited, and in the face of complicated semiconductor manufacturing processes, there may be certain process adaptability problems. For example, if the wavelength of the light of the measuring light is exactly four times the film thickness of the thin film covered by the semiconductor, the interference effect is likely to occur, and the reflectance of the reflected light from the engraved mark is greatly reduced, thereby causing a decrease in measurement accuracy. Second, the large NA objective lens scheme used in the prior art has a smaller depth of focus range. In general, the effective numerical aperture used for measuring light is greater than 0.9, and the effective focal depth range is less than 1 μm, calculated at a typical measuring light wavelength of 600 nm. Therefore, the position of the focus surface must be controlled with high precision during the measurement process, which will affect the measurement speed and accuracy; in addition, in this case, if the focal plane control is weak, the spot of the measurement light is easily diffused to the measured Outside the engraved mark, a large amount of stray light is formed, which seriously interferes with the measurement process.

Therefore, it is necessary to invent a device and a method for measuring a registration error, which can not only adapt to a wider range of measurement light, but also better adapt to a semiconductor that tends to be refined.

Summary of the invention

In order to solve the above problems, the present invention provides an apparatus and method for measuring a registration error, in which a measurement light adjustment component capable of adjusting measurement light to be symmetric with respect to a center axis of a microscope objective is added, so that measurement light passes through the sleeve. After the device, the positive and negative order diffracted light is formed by the engraved measurement mark, and finally the diffraction spectrum of the positive and negative order diffracted light is displayed on the detector, and the control system calculates the engraving error according to the diffraction spectrum, so the device and method can The wide-band light source is used to measure the registration error, so that the measurement wavelength range is wider, and the diffraction spectrum is the spectrum of the positive and negative order diffracted light, so the obtained measurement signal is more abundant, the measurement accuracy is improved, and the utilization of light energy is utilized. The rate is higher, and it can adapt to small-sized objects to be measured, making it more suitable for modern semiconductor products that tend to be refined.

In order to achieve the above object, the present invention provides an apparatus for measuring a registration error, including

An illumination system for generating measurement light;

a measuring light adjustment component;

a measuring beam splitter for splitting the measuring light, the lighting system, the measuring light adjusting component, and The measuring beamsplitters are sequentially arranged on the first straight line;

a microscope objective for collecting measurement light and projecting onto the surface of the object to be measured with the engraved measurement mark of the grating structure;

a detector for detecting a diffraction spectrum formed by diffraction after the measurement light is incident on the measurement mark, the detector is located on the facet of the microscope objective, and the detector, the measurement beam splitter, the microscope objective, and The objects to be measured are sequentially arranged on a second straight line, and the first straight line intersects the second straight line;

a control system coupled to the detector signal for calculating a registration error based on a diffraction spectrum formed by the engraved measurement mark displayed on the detector;

The measuring light adjusting component is configured to adjust the measuring light to be symmetrical about a center axis of the microscope objective such that a spectrum of the positive-order diffracted light and a spectrum of the negative-order diffracted light in the formed diffraction spectrum are shifted from each other.

Advantageously, the line passing through the center of the measuring light adjustment assembly and perpendicular to the measuring light adjustment assembly and the optical axis of the microscope objective are symmetrical about the normal to the measuring beam splitter.

Preferably, the measuring light adjusting component is an annular diaphragm or slit, and the annular diaphragm is composed of two quarter-rings symmetric about a center of the circle.

Advantageously, an incident beam splitter is also included between the illumination system and the measurement light adjustment assembly.

Advantageously, said measuring light adjustment component is a fiber optic cluster having a linear exit surface.

Preferably, the exit faces of the fiber bundles are linearly arranged fiber end faces.

Advantageously, the exit face of the fiber optic cluster further includes a collimating assembly for illuminating the light exiting the exit face of the fiber optic cluster, the collimating assembly for collimating light exiting the fiber optic cluster.

Preferably, the collimating assembly is a concave lens array or a self-focusing system.

Preferably, the incident surface of the fiber cluster is a two-dimensional rectangular surface or a three-dimensional structure.

Preferably, the three-dimensional structure is hemispherical or ellipsoidal.

Preferably, a monitoring light assembly is further disposed on the optical path after the measuring light penetrates the measuring beam splitter, and the control system further forms a measuring light signal and a signal formed by reflecting and diffracting the measuring light from the object to be measured. The monitoring optical signal formed by the reflected or diffracted measurement light on the monitoring optical component is normalized.

Preferably, the monitoring optical component comprises:

a lens group located on the optical path after the measuring light penetrates the measuring beam splitter;

a monitoring optical element located on the optical path after the measuring light penetrates the lens group for reflection or diffraction The measurement light is emitted and at least a portion of the reflected light or diffracted light thus generated is passed through the lens group.

Preferably, the monitoring optical element is a monitoring grating, the period of the monitoring grating is the same as the period of the grating of the engraved measurement mark, and the monitoring grating is placed obliquely so that only the diffracted light is diffracted from the monitoring grating - The first order diffracted light energy passes through the lens group, passes through the lens group, reaches the measuring beam splitter, and is reflected by the measuring beam splitter to the detector.

Preferably, the monitoring optical element is two mirrors placed perpendicularly to each other, and the measuring light penetrating through the lens group is reflected by the two mirrors to the measuring beam splitter, and is used by the measuring beam splitter Reflected to the detector.

Preferably, the object to be tested is carried by a film stage.

Preferably, a biasing means is also included, located on the path of the light from which the measuring light is incident.

Preferably, the polarizing device comprises:

a polarizer disposed between the measurement light adjustment component and the measurement beam splitter;

An analyzer is located between the measuring beam splitter and the detector.

Preferably, a compensator is further disposed on the optical path of the measuring light from the polarizer for measuring the reflectance change and the phase change of the measuring light having the polarization state.

Preferably, the measurement light generated by the illumination system is at least one of ultraviolet light, visible light, and infrared light.

The present invention also provides a method for measuring a registration error, the measurement light is emitted by the illumination system, and a measurement light adjustment component is disposed between the illumination system and the measurement beam splitter, and the measurement light adjustment component shapes the measurement light into the After the optical axis of the microscope objective is symmetrical, it is reflected by the measuring beam splitter and is incident on the object to be measured after passing through the microscope objective. The measured light is diffracted by the object to be measured to form positive and negative order diffracted light, positive and negative order diffraction. The light sequentially passes through the microscope objective and the measuring beam splitter reaches the detector to form a diffraction spectrum in which the spectrum of the positive-order diffracted light and the spectrum of the negative-order diffracted light are mutually staggered, and the control system calculates the engraving of the object to be measured according to the diffraction spectrum on the detector. error.

Preferably, the monitoring light component is disposed on the optical path after the measuring light is emitted from the measuring light adjusting component and penetrates the measuring beam splitter, and the measuring light is sequentially incident on the monitoring light after passing through the measuring light adjusting component and the measuring beam splitter. The component is reflected or diffracted by the monitoring light component, and at least a portion of the reflected light or diffracted light generated thereby passes through the measuring beam splitter and is reflected by the measuring beam splitter to the detector, and is reflected and diffracted from the object to be measured. Measuring light signals formed by light and reflecting or diffracting light patterns from the monitoring light assembly The integrated monitoring optical signal is normalized to eliminate the interference of the light intensity fluctuation on the measurement of the registration error.

Preferably, the method comprises the following steps:

Step 1: Calculate the sensitivity of each pixel on the diffraction spectrum displayed by the detector, set a threshold of sensitivity, and filter out pixels of the diffraction spectrum whose sensitivity is less than the threshold;

Step 2: generating a relationship diagram between the difference between the intensity of the positive-order diffracted light and the intensity of the negative-order diffracted light that characterizes the measured light and the error value of the step unit;

Step 3: Iteratively recursively according to the relationship between the difference between the intensity of the positive-order diffracted light and the intensity of the negative-order diffracted light obtained by the second step and the error value of the step unit, each of which produces an intermediate image, when iteratively calculates When a symmetric intermediate image is obtained, the iteration is completed, and the total step size formed by the iteration is the engraving error of the engraved measurement mark.

Preferably, two engraved measurement marks arranged in a row are formed on the object to be measured, which are respectively a first engraved measurement mark and a second set of engraved measurement marks, and the first set of engraved measurement marks sets the first engraved measurement mark. The engraved error of the patterned photoresist and the upper patterned photoresist is 0, and the second engraved measurement mark sets the patterned photoresist of the upper layer and the patterned photoresist of the second engraved measurement mark. The offset is Δ.

Preferably, the method for calculating the sensitivity of each pixel point on the diffraction spectrum displayed by the detector in step one is

Where Left_Intensity is the intensity of the light illuminating the first set of measurement marks, and Right_Intensity is the intensity of the light illuminating the second set of measurement marks.

Preferably, when the relationship between the difference between the positive-order diffracted light intensity and the negative-order diffracted light intensity of the measuring light and the step unit set-out error value is generated in the second step, the formula is:

Where OV_step_map is the image of the amount of change in light intensity received by the detector corresponding to the detector when the stepping unit OV_step is run.

Preferably, the intermediate image value corresponding to the intermediate image generated in each iteration in step 3 is pad_ov=Left_Intensity-m*OV_step_map, where m is the number of iteration cycles, and when the pad_ov value is 0, the value corresponding to m is n. Then, the engraving error OV_value=n×OV_step of the measurement mark is set.

Compared with the prior art, the beneficial effects of the present invention are: the present invention provides a measurement of the registration error Device, including

An illumination system for generating measurement light;

a measuring light adjustment component;

a measuring beam splitter for splitting the measuring light, the lighting system, the measuring light adjusting component and the measuring beam splitter being sequentially arranged on a first straight line;

The present invention provides an apparatus and method for measuring a registration error, in which a measurement light adjustment component capable of adjusting measurement light to a center symmetry about an optical axis of the microscope objective is added, so that the measurement light passes through the set of devices The engraved measurement mark forms positive and negative order diffracted light, and finally displays a diffraction spectrum of positive and negative order diffracted light on the detector, and the spectra of the positive order diffracted light and the negative order diffracted light are staggered from each other on the diffraction spectrum, Do not interfere with each other, the control system calculates the engraving error according to the diffraction spectrum, so that when the light source is selected, a wide-band light source such as infrared light, ultraviolet light, visible light or a combination of these kinds of light can be used, and With a measuring light adjustment component, a surface light source, a line light source or a point light source can be used. Therefore, the device and method can measure a wider wavelength range of light and can use any spot. The shape light source is more abundant in the measurement signal, and the measurement accuracy is improved, and since the utilization of the light energy is high, the measurement light can be received for the small-sized engraved measurement mark, so the device and method are also adapted to small The size of the object to be measured makes it more suitable for modern semiconductor products that tend to be refined.

DRAWINGS

1 is a schematic structural view of a device for measuring a registration error in the prior art;

2 is a schematic structural diagram of an apparatus for measuring a winding error according to an embodiment of the present invention;

3 is a schematic cross-sectional view of a measured object according to an embodiment of the present invention;

4 is a schematic diagram showing changes in the measured diffracted light intensity as a function of a set of engraved values according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of a first set of measurement marks according to an embodiment of the present invention; FIG.

6 is a schematic diagram of a second set of measurement marks according to an embodiment of the present invention;

Figure 7 is a graph showing the relationship between positive and negative order diffracted light and diffraction efficiency according to an embodiment of the present invention;

FIG. 8 is a schematic diagram showing changes after filtering out pixel points with poor sensitivity according to an embodiment of the present invention; FIG.

FIG. 9 is a diagram showing relationship between the engraving error generated by the first embodiment of the present invention and the difference between the positive and negative sub-light intensity;

FIG. 10 is a schematic structural diagram of a fiber cluster according to an embodiment of the present invention; FIG.

Figure 11 is a schematic view showing the arrangement of the end faces of the optical fibers of Figure 10;

Figure 12 is a schematic view showing the structure of the incident surface of Figure 10;

FIG. 13 is a diagram showing changes in light intensity when a stepping unit of a stepping unit is performed according to an embodiment of the present invention; FIG.

14 is a diagram showing a measurement of an optical signal and a monitoring optical signal on a detector according to an embodiment of the present invention;

15 is a schematic structural diagram of a fiber cluster according to Embodiment 2 of the present invention;

16 is a schematic structural view of a third aperture according to an embodiment of the present invention;

17 is a schematic structural diagram of an apparatus for measuring a winding error according to Embodiment 4 of the present invention;

18 is a display diagram of a monitoring optical signal and a measuring optical signal according to Embodiment 4 of the present invention;

19 is a display diagram of a monitoring optical signal and a measuring optical signal according to Embodiment 5 of the present invention;

FIG. 20 is a diagram showing the display of the monitoring optical signal and the measuring optical signal according to Embodiment 6 of the present invention.

In Fig. 1: 2-light source, 30-filter device, 32-detector, 34-reflecting surface, 40-optical surface, 6-substrate, L1-objective lens, L2-lens group;

The invention shows: 41-light source, 43-fiber cluster, 432-incident surface, 437-fiber, 44-exit surface, 45-measuring beam splitter, 46-microscope objective, 47-measured object, 471-first Sleeve measurement mark, 472-second set Inscribed measurement mark, 48-seat stage, 49-lens group, 410-monitoring grating, 411-detector, 414-monitoring spectrum, 4151-first diffraction spectrum, 4152-second diffraction spectrum, 4153-third diffraction spectrum 4154-4th diffraction spectrum, 4171-first aperture, 419-mirror.

detailed description

The above described objects, features and advantages of the present invention will become more apparent from the aspects of the appended claims.

Embodiment 1

Referring to FIG. 2, the present invention provides an apparatus for measuring a registration error, including

An illumination system for generating measurement light, the illumination system comprising at least one light source 41, the light source 41 being a wide-band light source, which may be a surface light source, a line light source or a point light source, and a light source having other spot shapes, generated by the illumination system The measuring light is at least one of ultraviolet light, visible light, and infrared light.

A measuring light adjusting component, which passes through the center of the measuring light adjusting component and perpendicular to the line of the measuring light adjusting component and the optical axis of the microscope objective 46 is symmetric with respect to the normal of the measuring beam splitter 45, using such an incident angle so that the measuring beam is split The light reflected on the mirror 45 can be incident perpendicularly to the microscope objective 46 to ensure maximum collection of incident light.

In this embodiment, the measurement light adjustment component is a fiber cluster 43. Referring to FIG. 11, the fiber cluster 43 has a linear exit surface, and specifically, the end faces of the optical fiber 437 are linearly arranged.

A measuring beam splitter 45 is used for splitting the measuring light, and is sequentially arranged in a first straight line with the illumination system and the measuring light adjusting component.

A microscope objective 46 is provided for collecting the measurement light and projecting it onto the surface of the object 47 to be measured with the engraved measurement mark of the grating structure.

The engraved measurement mark on the object 47 is used for reflecting and diffracting the measurement light, and the engraved measurement mark is generally on the mask plate, and is located on the non-pattern area on the mask plate, and the engraved measurement mark includes two rows arranged in a row. The first set of inscribed measurement marks 471 and the second set of inscribed measurement marks 472 are set to be the first set of inscribed measurement marks 471 on the previous version of the mask when the mask is made. The sizing error is 0, and the second set of measurement marks 472 is set to be Δ with a preset offset of the second set of measurement marks 472 on the previous version of the mask;

Also included is a detector 411 for detecting a diffraction spectrum formed by the engraved measurement mark, the detector The 411 is located on the surface of the microscope objective 46, and is sequentially arranged in a second line with the measuring beam splitter 45, the microscope objective 46, and the object 47 to be measured, and the first straight line intersects the second straight line, thereby forming the second straight line of FIG. structure;

Also included is a control system (not shown) coupled to the detector 411 for calculating a registration error based on a diffraction spectrum formed by the engraved measurement mark displayed on the detector 411;

The measuring light adjusting component is also a light source shaping system for shaping the measuring light into a center symmetry about the optical axis of the microscope objective 46, so that the spectrum of the positive-order diffracted light and the spectrum of the negative-order diffracted light in the formed diffraction spectrum are staggered from each other. It is to make the two do not interfere with each other, so that the control system can avoid many errors in the calculation.

In this embodiment, referring to FIG. 10 and FIG. 11, the measurement light adjustment component is a fiber cluster 43 having a linear exit surface, and the fiber cluster 43 has a plurality of optical fibers 437. The optical fiber 437 generally has a small diameter of up to several hundred. Micron, because the direction of light emitted by the light passing through the optical fiber 437 is disordered, a collimating component (not shown) is disposed on the exit surface of the optical fiber tuft 43 to use a more common collimating component, such as a concave lens array or a self-focusing system. The collimating component shapes the light emerging from the optical fiber 437 into mutually parallel light such that the light incident on the object 47 is uniform during the measurement, and the error is reduced from the root.

Preferably, referring to FIG. 12, the incident surface of the optical fiber cluster 43 is a two-dimensional rectangular surface, that is, the end faces of the optical fibers 437 are arranged to form a rectangle.

Preferably, in order to improve the accuracy of the measurement, a device capable of providing reference light is provided, specifically a monitoring optical component. Referring to FIG. 2, the monitoring optical component is located on the optical path after the measuring light penetrates the measuring beam splitter 45, and the specific include

a lens group 49 located on the optical path after the measuring light penetrates the measuring beam splitter 45;

A monitoring optical element is located on the optical path after the measuring light penetrates the lens group 49 for reflecting or diffracting the measuring light and passing the reflected or diffracted light through the lens group 49. Specifically, the monitoring optical element in this embodiment is monitored. The grating 410 monitors the period of the grating 410 to be the same as the period of the grating of the engraved measurement mark, and the monitoring grating 410 is placed obliquely so that only -1 order diffracted light of the diffracted light diffracted from the monitoring grating 410 passes through the lens group 49. The optical paths of the remaining 0th and +1st order diffracted lights do not pass through the lens group 49, and when the -1st order diffracted light passes through the lens group 49, it reaches the measuring beam splitter 45 and is reflected by the measuring beam splitter 45 to the detector 411.

Referring to FIG. 14, the measurement light reflected and diffracted from the object 47 is formed as a measurement optical signal. In this embodiment, since the linear light source is used, the spectrum displayed on the detector 411 is the first diffraction spectrum 4151. Monitoring the optical component on the reflected and diffracted measuring light to form a monitoring optical signal at the detector 411 Displayed as monitoring spectrum 414, the monitoring optical signal is used as a reference contrast quantity, and the control system normalizes the measuring optical signal and the monitoring optical signal. After normalization, the calculated engraving error can eliminate the wide band. The influence of the disturbance of the intensity of the partial band in the light source on the measurement of the engraving error.

Referring to FIG. 2 and FIG. 3, the measuring device provided by the present invention is mainly used for measuring the engraving error of the object to be tested 47. The object 47 to be measured is a silicon wafer placed on the wafer table 48 of the workpiece table of the lithography machine. There are at least two layers of patterned photoresist on the chip, and the measured registration error is the coverage error between the two patterned photoresists at the same position. After the first layer of patterned photoresist is fabricated on the silicon wafer, a layer of photoresist needs to be applied again in the subsequent process, and the layer of photoresist is patterned, but due to the second patterned lithography When the glue is aligned, it may not be aligned with the pattern of the first patterned photoresist, so that the coverage error of the patterned photoresist described above is caused.

Referring to FIG. 4, when there is a registration error between the two patterned photoresists, the light diffracted from the photoresist changes, and when the engraving error value is equal to zero, the positive and negative high-order diffracted light intensities Equally, when the engraved error value is not equal to zero, the positive and negative high-order diffracted light intensities are not equal, and when the engraving error value is near zero, the diffracted light intensity is linearly related to the engraved error value.

Based on the above principle, the present invention provides a method for measuring a registration error based on the above-described measuring device, wherein the measuring light is emitted by the light source 41, and the measuring light is shaped by the measuring light adjusting component to be symmetrical about the optical axis of the microscope objective 46. The measurement beam splitter 45 reflects and passes through the microscope objective 46 and is incident on the engraved measurement mark of the object 47 to be measured. The measurement light is diffracted by the engraved measurement mark to form a diffracted light of positive and negative order, positive and negative order diffracted light. The diffraction spectrum of the spectrum of the positive-order diffracted light and the spectrum of the negative-order diffracted light is sequentially formed by the microscope objective 46 and the measuring beam splitter 45, and the control system calculates the measured spectrum according to the diffraction spectrum on the detector 411. The engraving error of the object 47.

Referring to FIG. 5 and FIG. 6, using the measurement method provided by the present invention, the object 47 is required to have at least two engraving measurement marks, which are a first engraved measurement mark 471 and a second engraved measurement mark 472, respectively. Two marks are respectively located on both sides of the silicon wafer, and are located in the non-pattern area with an effective pattern area interposed therebetween. When designing the mask, the first set of measurement marks 471 and the underlying photoresist are set. The splicing error is 0, but since there are necessarily alignment errors due to the two mask exposures, the overprint error ε must be generated, so the offset between the first set of measurement marks 471 and the underlying photoresist is 0+. ε=ε, which sets a preset offset Δ between the second set of measurement marks 472 and the underlying photoresist, and then the offset actually formed after the second mask exposure is ε+Δ. .

Using the above-mentioned engraved measurement mark, measuring the engraving error specifically includes the following steps:

Step 1: Please refer to FIG. 7 and FIG. 8 to calculate the sensitivity of each pixel on the diffraction spectrum displayed by the detector 411.

Where Left_Intensity is the intensity of light impinging on the first set of measurement marks 471, Right_Intensity is the intensity of light irradiated on the second set of measurement marks 472, and the positive and negative light intensities on the first set of measurement marks 471 are respectively For Left_Intensity_Positive=k·ε+b, Left_Intensity_Nagetive=-k·ε+b, the positive and negative light intensities on the second set of measurement marks 472 are Right_Intensity_Positive=k·(ε+Δ)+b, Right_Intensity_Nagetive= -k·(ε+Δ)+b, where k is the measurement process parameter of the device, b is the basic value of the light intensity, that is, the light intensity value when the engraving error value is 0, and ε is the engraving error value.

Please refer to FIG. 7 , where the sensitivity is poor, and the ordinate does not change greatly with the change of the abscissa, so it needs to be filtered out. The filtering method is to set the sensitivity threshold according to experience. The calculated pixel points whose sensitivity is less than the threshold value are filtered out, and filtered as shown in FIG.

Step 2: Referring to FIG. 9, a difference between the difference between the intensity of the positive-order diffracted light and the intensity of the negative-order diffracted light that characterizes the measured light and the map generated by the control system is generated between the map (image) generated by the control system. relationship.

The formula for generating the relationship diagram is:

OV_step_map is an image of the amount of change in light intensity received by the detector 411 corresponding to the running error of one step unit OV_step. The corresponding image is shown in FIG. 13 , where OV_step is a constant value.

Step 3: Using the iterative recursive calculation of the computer system to calculate the engraving error value, that is, iteratively recursively according to the relationship between the difference between the positive diffracted light intensity and the negative diffracted light intensity and the unit engraved error value obtained according to step 2, Specifically, the calculation value corresponding to the intermediate image map generated by each iteration is calculated as pad_ov=Left_Intensity-m*OV_step_map, where m is the number of iteration cycles, and when the calculated intermediate image map corresponding to pad_ov is close to no deviation, that is, pad_ov When the value is close to 0, the value corresponding to m at this time is n, and the engraving error OV_value=n×OV_step of the engraved measurement mark.

In addition, in order to improve the accuracy of the measurement, referring to FIG. 2 and FIG. 14, after the measurement light is emitted from the measurement light adjustment component, the monitoring light component is disposed on the optical path after the penetration of the measurement beam splitter 45, and the measurement light passes through the measurement light adjustment component in turn. After measuring the beam splitter 45, it is incident on the monitoring light component, and is reflected and reflected by the monitoring light component. The shot is taken to the measuring beam splitter 45 and reflected by the measuring beam splitter 45 to the detector 411, and the measuring light signal formed by reflecting and diffracting the measuring light from the object 47 to be measured is formed by reflecting and diffracting the measuring light from the monitoring light component. The monitoring optical signal is normalized to eliminate the interference of stray light on the measurement of the registration error.

Embodiment 2

Referring to FIG. 15, the difference between this embodiment and the first embodiment is that the incident surface 432 of the optical fiber cluster 43 has a three-dimensional structure, such as a hemispherical shape as shown in FIG. 15, or an ellipsoidal shape, and the spherical incident surface can be increased. The surface area of the large incident surface is capable of collecting more incident light.

Embodiment 3

Referring to FIG. 16, the difference between this embodiment and the first embodiment is that the measurement light adjustment component is an annular first aperture 4171, and the annular first aperture 4171 is composed of two quarter-rings symmetric about the center of the circle. That is, the two quarter rings are shaded, and the rest are light transmissive.

Embodiment 4

Referring to FIG. 17, the difference between this embodiment and the third embodiment is that the monitoring optical element is two mirrors 419 with an angle of 90° with each other, and the measuring light penetrating through the lens group 49 is divided by two mirrors 419. It is reflected to the measuring beam splitter 45 and is reflected by the measuring beam splitter 45 to the detector 411. Since the monitoring spectrum 414 is formed by the mirror 419, the formed monitoring spectrum 414 and the second diffraction spectrum 4152 formed by the measuring light signal are as shown in FIG. Shown.

Embodiment 5

Referring to FIG. 19, the difference between this embodiment and the third embodiment is that the shape of the second aperture (not shown) used is two half-chords symmetric about the center, and the monitoring optical element is the same as that of the fourth embodiment, thus obtaining The third spectrum of diffraction 4153 formed by the monitoring spectrum 414 and the measured light signal is as shown in FIG.

Embodiment 6

Referring to FIG. 20, the difference between the embodiment and the fifth embodiment is that the third aperture (not shown) used is formed by rotating the second aperture of the fifth embodiment clockwise by 90°, and the obtained monitoring spectrum 414 and the measurement light are obtained. The fourth diffraction spectrum 4154 formed is as shown in FIG.

Example 7

This embodiment differs from the first embodiment in that the measuring light adjusting component is a slit (not shown).

Example eight

The difference between this embodiment and the seventh embodiment is that there are two slits in the illumination system and the measurement light adjustment. The component is provided with an incident beam splitter (not shown) which splits the measuring light into two identical beams and respectively enters the two slits.

Example nine

The difference between this embodiment and the eighth embodiment is that a shutter (not shown) is provided between the slit and the measuring beam splitter 45 for shielding the non-measuring light beam which interferes with the measurement.

Example ten

The difference between this embodiment and the first embodiment is that a filter device (not shown) is added to the device, and the filter device is located on the optical path on which the measurement light is incident on the object 47 to be measured, such as can be disposed on the illumination system. After the light source 41 emits the measuring light and passes through the filtering device, the measuring light having a narrow bandwidth can be filtered out, which is more advantageous for the measurement.

Embodiment 11

The difference between this embodiment and the first embodiment is that a biasing device (not shown) is also included, which is located on the optical path on which the measuring light is incident on the detector 411, so that the measuring light becomes light having a polarization state.

Specifically, the polarizing device includes:

a polarizer is disposed between the measuring light adjusting component and the measuring beam splitter 45;

An analyzer is located between the measuring beam splitter 45 and the detector 411.

A polarizer is added to the device so that the measurement light becomes polarized light having a TE mode or a TM mode, but the TE mode or the TM mode is selected according to the condition of the object 47 to be measured, since the TE mode and the TM mode are the same measured. The reflectivity of the object 47 is not the same. If the object to be measured is metal and has a linear grating structure, the TE mode is more easily absorbed by the linear grating, so the reflection efficiency is low, and finally, the measurement process parameter k is affected. In general, the larger k is, the larger the engraved error value ε is, which is reflected in the measured light intensity value, so that it is easily calculated by the control system, which improves the measurement accuracy, and therefore needs to be measured according to the measurement process and The object selects polarized light of different nature.

The present invention has been described in the above embodiments, but the invention is not limited to the embodiments described above, and it is obvious that those skilled in the art can make various modifications and changes to the invention without departing from the spirit and scope of the invention. Thus, it is intended that the present invention cover the modifications and variations of the invention as claimed.

Claims

A device for measuring a registration error, characterized in that it comprises

An illumination system for generating measurement light;

a measuring light adjustment component;

a measuring beam splitter for splitting the measuring light, the lighting system, the measuring light adjusting component and the measuring beam splitter being sequentially arranged on a first straight line;

a microscope objective for collecting measurement light and projecting onto the surface of the object to be measured with the engraved measurement mark of the grating structure;

a detector for detecting a diffraction spectrum formed by diffraction after the measurement light is incident on the measurement mark, the detector is located on the facet of the microscope objective, and the detector, the measurement beam splitter, the microscope objective, and The objects to be measured are sequentially arranged on a second straight line, and the first straight line intersects the second straight line;

a control system coupled to the detector signal for calculating a registration error based on a diffraction spectrum formed by the engraved measurement mark displayed on the detector;

The measuring light adjusting component is configured to adjust the measuring light to be symmetrical about a center axis of the microscope objective such that a spectrum of the positive-order diffracted light and a spectrum of the negative-order diffracted light in the formed diffraction spectrum are shifted from each other.
The apparatus for measuring a registration error according to claim 1, wherein a line passing through a center of said measuring light adjusting unit and perpendicular to said measuring light adjusting unit and an optical axis of said microscope objective are said Measure the normal symmetry of the beam splitter.
The apparatus for measuring a registration error according to claim 2, wherein said measuring light adjusting component is an annular diaphragm or slit, said annular diaphragm being composed of two quarter circles symmetric about a center of the circle. Ring composition.
The apparatus for measuring a registration error according to claim 1, further comprising an incident beam splitter between said illumination system and said measurement light adjustment component.
The apparatus for measuring a registration error according to claim 2, wherein said measuring light adjustment component is a fiber cluster having a linear exit surface.
The apparatus for measuring a registration error according to claim 5, wherein the exit face of the fiber bundle is a linearly arranged fiber end face.
The apparatus for measuring a registration error according to claim 5, wherein said fiber cluster is out The surface also includes a collimating assembly for measuring light exiting the exit face of the fiber optic cluster, the collimating assembly for collimating light exiting the fiber optic cluster.
The apparatus for measuring a registration error according to claim 7, wherein the collimating assembly is a concave lens array or a self-focusing system.
The apparatus for measuring a registration error according to claim 5, wherein the incident surface of the optical fiber cluster is a two-dimensional rectangular surface or a three-dimensional structure.
The apparatus for measuring a registration error according to claim 9, wherein the three-dimensional structure is hemispherical or ellipsoidal.
The apparatus for measuring a registration error according to claim 1, further comprising a monitoring light component located on an optical path after the measuring light penetrates the measuring beam splitter, the control system further facing from the The measurement light signal formed by the reflected and diffracted measurement light on the measurement object is normalized with the monitor light signal formed by reflecting or diffracting the measurement light from the monitoring light assembly.
The apparatus for measuring a registration error according to claim 11, wherein the monitoring light component comprises:

a lens group located on the optical path after the measuring light penetrates the measuring beam splitter;

A monitoring optical element is located on the optical path after the measuring light penetrates the lens group for reflecting or diffracting the measuring light and passing at least a portion of the resulting reflected or diffracted light through the lens group.
The apparatus for measuring a registration error according to claim 12, wherein said monitoring optical element is a monitoring grating, said monitoring grating having the same period as said grating of said engraved measurement mark, said monitoring grating Tilting such that only -1 order diffracted light energy diffracted from the monitoring grating passes through the lens group, passes through the lens group, reaches the measuring beam splitter, and is reflected by the measuring beam splitter to detector.
The apparatus for measuring a registration error according to claim 12, wherein said monitoring optical element is two mirrors placed perpendicularly to each other, and measuring light penetrating through said lens group is reflected by two mirrors. To the measuring beam splitter and reflected by the measuring beam splitter to the detector.
The apparatus for measuring a registration error according to claim 1, wherein said object to be measured is carried by a carrier.
The apparatus for measuring a registration error according to claim 1, further comprising a bias The device is located on the optical path where the measuring light is incident on the detector.
The apparatus for measuring a registration error according to claim 16, wherein said polarizing means comprises:

a polarizer disposed between the measurement light adjustment component and the measurement beam splitter;

An analyzer is located between the measuring beam splitter and the detector.
The apparatus for measuring a registration error according to claim 16, further comprising a compensator for measuring a reflectance of the measurement light having a polarization state on an optical path from which the measurement light is emitted from the polarizer. Changes and phase changes.
The apparatus for measuring a registration error according to claim 1, wherein the measurement light generated by the illumination system is at least one of ultraviolet light, visible light, and infrared light.
A method for measuring a registration error, wherein a measurement light is emitted by an illumination system, and a measurement light adjustment component is disposed between the illumination system and the measurement beam splitter, and the measurement light adjustment component shapes the measurement light into After the optical axis of the microscope objective is symmetrical, it is reflected by the measuring beam splitter and is incident on the object to be measured after passing through the microscope objective. The measured light is diffracted by the object to be measured to form positive and negative order diffracted light, positive and negative order diffraction. The light sequentially passes through the microscope objective and the measuring beam splitter reaches the detector to form a diffraction spectrum in which the spectrum of the positive-order diffracted light and the spectrum of the negative-order diffracted light are mutually staggered, and the control system calculates the engraving of the object to be measured according to the diffraction spectrum on the detector. error.
The method of measuring a slanting error according to claim 20, wherein the monitoring light component is disposed on the optical path after the measuring light is emitted from the measuring light adjusting component and penetrates the measuring beam splitter, and the measuring light passes through Measuring the light adjusting component, measuring the beam splitter, and then incident on the monitoring light component, and being reflected or diffracted by the monitoring light component, thereby generating at least a part of the reflected light or the diffracted light passes through the measuring beam splitter, and the measuring beam splitter Reflecting to the detector, normalizing the measurement light signal formed by the reflection and diffraction measurement light on the object to be measured and the monitoring light signal formed by the reflection or diffraction measurement light on the monitoring light component, for eliminating the intensity fluctuation Measure the interference of the overlay error.
The method of measuring a registration error according to claim 21, which comprises the following steps:

Step 1: Calculate the sensitivity of each pixel on the diffraction spectrum displayed by the detector, set a threshold of sensitivity, and filter out pixels of the diffraction spectrum whose sensitivity is less than the threshold;

Step 2: Generate the difference between the light intensity of the positive-order diffracted light and the light intensity of the negative-order diffracted light that characterize the measured light and step a diagram of the relationship between the unit set error values;

Step 3: Iteratively recursively according to the relationship between the difference between the intensity of the positive-order diffracted light and the intensity of the negative-order diffracted light obtained by the second step and the error value of the step unit, each of which produces an intermediate image, when iteratively calculates When a symmetric intermediate image is obtained, the iteration is completed, and the total step size formed by the iteration is the engraving error of the engraved measurement mark.
The method for measuring a registration error according to claim 22, wherein two engraved measurement marks arranged in a row are formed on the object to be measured, which are a first set of measurement marks and a second set of measurement marks, respectively. The first set of inscribed measurement marks sets the engraving error of the patterned photoresist and the upper patterned photoresist where the first engraved measurement mark is located to 0, and the second engraved measurement mark sets the upper layer of patterned photoresist The offset from the patterned photoresist where the second set of marks is located is Δ.
A method of measuring a registration error as recited in claim 23, wherein the method of calculating the sensitivity of each pixel on the diffraction spectrum displayed by the detector in step one is
Where Left_Intensity is the intensity of the light illuminating the first set of measurement marks, and Right_Intensity is the intensity of the light illuminating the second set of measurement marks.
The method for measuring a registration error according to claim 24, wherein the generating in step 2 is indicative of a difference between the intensity of the positive-order diffracted light and the intensity of the negative-order diffracted light of the measuring light and the error value of the step unit. When the relationship diagram is between, the formula is based on:
Where OV_step_map is the image of the amount of change in light intensity received by the detector corresponding to the detector when the stepping unit OV_step is run.
A method of measuring a registration error according to claim 25, wherein the intermediate image value corresponding to the intermediate image generated in each iteration in step three is

Pad_ov=Left_Intensity-m*OV_step_map, where m is the number of iterations of the loop. When the value of pad_ov is 0, the value of m corresponds to n, and the set error of the engraved measurement mark is OV_value=n×OV_step.