JPH0755702A - Crystal-defect measuring device and semiconductor manufacturing apparatus using the device - Google Patents

Abstract

PURPOSE:To distinguish the crystal defect causing the defect of a semiconductor device from the foreign matter on the surface of a crystal and to measure the depth position and the size of the defect in the crystal in a high sensitivity. CONSTITUTION:Coherent laser light 1 is split into two parts with a prism 50. One light is cast on a crystal sample 3, and the other light undergoes frequency modulation with a vibrating mirror 2. Thus the reference light is obtained. The reflected light from the surface of the sample 3 is removed with a polarization plate 51. The scattered light from a defect and the frequency modulated reference light are made to interfere. Only the frequency component, which is equal to the modulating frequency, is amplified with a lockin amplifier 18. The intensity and the phase of the component are recorded together with the data of the position of the sample 3 for every defect. The height of the sample 3 is changed with a sample moving stage 16, and the similar measurement is performed. The profile of the intensity of the scattered light from each defect in the depth direction is obtained, and the refractive index around the defect is derived.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus for measuring crystal defects existing on or in a crystal on which a semiconductor device is formed, and more particularly, to distinguish crystal defects from foreign substances on the crystal surface with high sensitivity. The present invention relates to a measuring device and a semiconductor manufacturing device using the same.

[0002]

2. Description of the Related Art In recent years, with the high integration of LSIs, a decrease in reliability and yield has become a serious problem. And
One of the main causes of causing these is a minute crystal defect of Si crystal. These defects are caused by precipitation of foreign substances in the crystal, stacking faults of crystal lattices, point defects, and the like. As a method for observing these minute defects, there are the following conventional techniques.

The first is to irradiate infrared rays that pass through the Si crystal from the cross-sectional direction of the Si crystal (direction parallel to the surface of the sample), and detect scattered light from minute defects in the Si crystal from the surface direction of the sample. Is the way. This method is called infrared scattering tomography, and is described in detail, for example, in Journal of Crystal Growth, Vol. 88 (1988), p. In this technique, the resolution in the depth direction can be obtained by scanning the beam in the depth direction. The resolution is determined by the irradiation beam diameter, and the limit is about the wavelength of the irradiation light (about 1 μm).

Next, there is a confocal differential heterodyne microscope method as a method suitable for detecting a minute defect because it is hardly affected by stray light from the surface, and Applied Opticus magazine, Vol. 29 (1990). ) 4244
Details on the page. This method is characterized in that two light beams having different frequencies (each frequency is ν1 and ν2) are crossed at the position where the defect exists, and the crossing region fluctuates with a beat frequency of ν1-ν2. Interference fringes are formed. As a result, the scattered light intensity from the defect is ν1
Since the frequency fluctuates at −ν2, it is possible to remove the influence of scattered light other than the crossing region for observation by detecting only this frequency fluctuation component. Also, by making the detection optical system a confocal optical system via a pinhole, it is possible to limit the detection area of scattered light to a part of the intersection area, and the resolution of this method is the same as that of a confocal microscope. It is a degree.

[0005]

Devices in LSI are often formed in a region within 0.5 μm from the surface. If a defect occurs in this region, a device defect is likely to occur, but even if it occurs in a region deeper than this, it is often unrelated to the device defect. Therefore, it is practically important to measure defects in the surface area.
At that time, it is necessary to distinguish the foreign matter attached to the sample surface from the defect in the crystal. This is because it is difficult to prevent foreign matter from adhering to the sample before and during measurement. However, since the conventional technique detects defects or foreign matters based on scattered light, it is impossible to distinguish them in the surface layer within 0.5 μm.
Further, since various structures are formed in the semiconductor device in the process of manufacturing, there are various undulations on the surface, and it has been a conventional problem that these undulations are measured as defects. There is also a problem that defects formed in the initial stage of the semiconductor device manufacturing process cannot be detected because they are minute.

The present invention has been made to solve the above problems, and a crystal capable of detecting a crystal defect in the vicinity of the crystal surface in distinction from foreign particles present on the surface of the crystal and measuring a finer defect An object of the present invention is to provide a defect measuring device and a semiconductor manufacturing device using the same.

[0007]

In order to achieve this object, in the present invention, the refractive index around the defect existing on the crystal surface or in the crystal is measured to measure the defect. From the refractive index around the defect, the defect in the crystal and the fine particles (dust) existing on the crystal surface are distinguished. The refractive index is measured by narrowing the irradiation light into the crystal and scanning the crystal to measure the intensity distribution of the scattered light from the defect. Also, by measuring the phase of the scattered light,
The depth position of the defect in the crystal is measured.

As a concrete apparatus, linearly polarized coherent light is divided into two by a prism, one is narrowed into a crystal for irradiation, the other is frequency-modulated by an oscillating mirror, and a quarter wave plate is used. It is transmitted twice and the polarization direction is rotated 90 °. And of the scattered light from the defect,
Only the polarization component orthogonal to the polarization direction of the irradiation light is detected through the polarizing plate and interferes with the frequency-modulated light. Then, only the beat frequency component of the interference light is amplified as a signal. Further, the frequency-modulated light is divided into two, one of which is interfered with the scattered light from the defect to be a detection signal, and the other is detected as a reference light and subtracted from the detection signal.

When the crystal is other than silicon oxide, the refractive index of the defect existing in the crystal is assumed to be the same as that of silicon oxide, and the intensity of scattered light from the defect is measured to derive the size of the defect. To do.

Further, the above-mentioned crystal defect measuring device is incorporated into a semiconductor manufacturing apparatus for physically or chemically adding various elements onto a semiconductor substrate, and is generated on the surface of a semiconductor wafer or within a wafer crystal in the process of manufacturing a semiconductor element. Crystal defects are measured and evaluated.

[0011]

First, the principle of the refractive index measuring means around the defect in the crystal defect measuring apparatus according to the present invention will be described with reference to FIGS.

As shown in FIG. 2 (a), when the irradiation light is focused into the solid having the refractive index n in the vertical direction from the surface and enters, the width of the irradiation intensity distribution at the beam waist (width Δx 'in the horizontal direction). And the vertical width Δz ′). Δ
Although x ′ hardly depends on the refractive index n, Δz ′ increases as the refractive index increases. That is, the three-dimensional irradiation intensity distribution shape in the beam waist shown by the black portion in FIG. 2A becomes longer in the depth direction as the refractive index increases. FIG. 3 (a) shows the NA of light with a wavelength of 1.06 μm.
2 is a graph in which the sizes of Δz ′ and Δx ′ when irradiated under the condition of = 0.6 are calculated with respect to the refractive index of the sample. Here, Δz ′ is the half-value width of the irradiation intensity distribution in the vertical direction, and Δx ′ is the beam size in the horizontal direction (width that attenuates to 1 / e). From this graph, it can be seen that Δx ′ has almost no dependency on the refractive index, and Δz ′ changes in proportion to the refractive index. FIG. 3B shows a beam waist in a solid when light having a wavelength of 1.06 μm is narrowed down at various solid angles with respect to the normal direction of the solid surface, that is, when narrowed down at various numerical apertures (NA). 6 is a graph showing how the ratio of the horizontal and vertical widths of the irradiation intensity distribution in Fig. 3 changes as a function of the refractive index of a solid. The above analysis results can be used as a basic principle of measuring the refractive index around a defect. That is, by scanning the crystal sample or the beam in the horizontal direction and the vertical direction while measuring the scattered light intensity from the defect in the crystal, Δz ′ and Δx ′ are detected for each defect.
Is measured and the refractive index around the defect is determined from the ratio. The relational expression between this ratio and the refractive index is expressed as follows.

[0013]

[Equation 1]

However, NA is the numerical aperture of the irradiation lens. Equation 1 is "Basics of Optoelectronics" by Yariff.
(Maruzen, 1974) 3.2-11 formula and 3.2-18 formula, and the definition formula of NA.

Next, the means for measuring the position of the defect in the crystal will be described with reference to FIGS. 1 and 2 (b).

First, as shown in FIG. 1, the scattered light from the defect in the sample 3 is caused to interfere with the frequency-modulated light by the prism 50, and the interference light is detected by the photodetector 7 using the lens and the pinhole 8. To detect. FIG. 2B shows this state with mathematical expressions and a schematic diagram. In FIG. 2B, the light is shown in a complex amplitude display, and the intensity is shown as the square of the absolute value. A defect existing at a depth Z from the surface of the sample is irradiated with light of frequency ω ₀ . Letting f (Z) be the intensity distribution of this irradiation light in the depth direction, the half-value width of this f (Z) is Δz ′.
Equivalent to. The scattered light intensity | Us | ^ 2 from the defect is the irradiation light intensity | U ₀ | ^ 2 multiplied by the scattering cross section σ (z) of the defect, and the phase of the complex amplitude is the phase of the surface reflected light. In comparison, the round trip to the defect position is offset by 4πnz / λ. However, Z is the depth position of the defect. This scattered light U
Reference light Ur obtained by modulating the same light as the irradiation light on s by frequency ω
Interfere with. The detected light intensity | Ud | ^ 2 is as follows.

[0017]

[Equation 2]

However, δ ″ −δ ′ is the phase difference between the scattered light and the reference light, and I ₀ is the irradiation light intensity, which is equal to | U ₀ | ^ 2.
Formula 2 shows the time change of the detection signal. Therefore, if only the signal change of the frequency equal to the modulation frequency of the reference light is amplified by the lock-in amplifier 18, high sensitivity detection becomes possible. The phase is measured for each defect to obtain 4πnZ / λ. However, the phase reference is taken so that the phase of the scattered light from the fine particles on the surface becomes zero. Since the refractive index around the defect is determined by the above measuring method, the position Z in the depth direction can be obtained from the phase. However, the position determination from this phase has an uncertainty of an integral multiple of λ / n. As a method of eliminating this uncertainty, there are considered a method of eliminating the uncertainty by changing the height of the sample and utilizing the distribution shape of f (z), and a method of performing similar phase measurement at different irradiation wavelengths. To be
The latter method will be described below. Generally, when the two irradiation wavelengths are λ and λ ′, and the phases of the same defect measured at each wavelength are Ω and Ω ′, λ / λ ′ =
If m ′ and m are obtained under the condition of (Ω ′ + 2πm ′) / (Ω + 2πm) and the condition that m ′ and m are integers, the defect depth is (Ω + 2πm) λ / 4πn or (Ω ′ + 2π).
m ') given by λ' / 4πn.

The size of the defect can be obtained from the refractive index of the defect itself, the refractive index around the defect and the scattered light intensity of the defect. When the refractive index of the defect itself is unknown, it is convenient to temporarily do the following. Regarding the defects in the silicon wafer, the grain size is calculated assuming the refractive index of silicon oxide. The particle size of foreign matter on the surface of the sample is calculated on the assumption that it is the same as the refractive index of polystyrene standard particles. Regarding the defects in the silicon oxide thin film formed on the silicon wafer, the particle size is calculated on the assumption that it is the same as the refractive index of the atmosphere.

Next, the enhancement of the sensitivity of defect detection will be described.

Reflection light from the crystal surface is one of the factors that lower the detection sensitivity in measuring defects on the crystal surface or inside the crystal. As a method of eliminating the reflected light, a method of detecting the polarized light in the direction perpendicular to the polarized direction of the irradiation light is used by utilizing the fact that the polarization direction of the reflected light from the material surface hardly changes due to the reflection. In FIG. 1, the polarizing plate 5
1 is installed so that scattered light having a polarization direction component different from the polarization direction of the irradiation light is detected and reflected light is removed.

Another factor that lowers the detection sensitivity is fluctuation in light intensity. Generally, the magnitude of fluctuation of light intensity is proportional to the square root of light intensity. In the present invention, when a defect is detected, the intensity fluctuation derived from the irradiation light intensity also becomes noise. In Formula 2, this corresponds to a change of | Us | ^ 2 + | Ur | ^ 2. In this case, the reference light intensity | Ur | ^ 2 is overwhelmingly larger than the scattered light intensity | Us | ^ 2, and the fluctuation of the reference light intensity dominates the overall strength fluctuation. A method for reducing this fluctuation will be outlined below. Using the polarizing prism 52 as shown in FIG.
The polarized scattered light is incident only on one of the detectors 7 ', and the reference light is divided into two components of equal intensity and orthogonal polarization directions, and is incident on both detectors 7, 7'. By taking the difference between the signals detected by the two detectors, the fluctuation of the reference light intensity is eliminated and the influence of the irradiation intensity fluctuation is reduced.

[0023]

【Example】

(Embodiment 1) FIG. 1 is a diagram showing a device configuration of a first embodiment of a crystal defect measuring device according to the present invention.

First, as a light source, N having a wavelength of 1.06 μm is used.
After linearly polarized light using the d: YAG laser 11, the prism 50 divides the beam into two parts, one of which vibrates the mirror 2 fixed to the piezoelectric element 41 to 1 MHz for frequency modulation, and is placed in front of the mirror 2. The polarization direction of the reflected light is rotated by 90 ° through the ¼ wavelength plate 53 to be used as the reference light. Also,
The other light is applied to a defect on the surface or inside of the sample 3, which is a crystal, by the prism 50 and the objective lens 5'having an NA of 0.6. This irradiation region is set to a depth of several μm from the sample surface. Here, the reflected light from the surface of the sample 3 is removed by the polarizing plate 51 as described in the section of the operation. The scattered light from the defect is passed through the polarizing plate 51 by the prism 50 so as to overlap with the reference light polarized perpendicularly to the irradiation light, narrowed down by the lens 5, and interfered through the pinhole 8. Then, the photodetector 7 detects the interference light.
The detection signal is converted into a voltage signal by the current-voltage conversion circuit 21,
The lock-in amplifier 18 takes in the reference signal from the driver 42 of the piezoelectric element 41, amplifies only the signal strength of the 1 MHz signal component of the frequency, and takes the amplitude and phase into the control computer 20. The data is the phase information signal from the sample scanning controller and the control computer 2 simultaneously.
It is taken in 0 and memorized. The above measurement is performed by scanning in a horizontal plane of a fixed height of the sample, and then the same measurement is sequentially performed while changing the height of the sample by 0.3 μm.

Next, for the same defect, a half-width is determined by obtaining a light profile in the depth direction from the values of scattered light intensity measured at different depth positions. Then, the ratio of the distribution in the lateral direction to the full width at half maximum is obtained, and the refractive index is determined for each defect from Expression 1. Further, for each defect, the depth position is determined from the change and the phase of scattered light intensity by the measurement performed by changing the height of the sample. The measured value of the half width of the scattered light intensity distribution is calibrated so that the refractive index obtained by measuring the polystyrene standard particles adhered on the sample surface is 1.0. FIG. 4 shows the measurement results of this example using a silicon wafer on which a thermal oxide film having a thickness of 0.5 μm was formed as a sample.
They are summarized in (a), (b) and (c).

As described above, in this embodiment, it is possible to measure foreign matter on the surface of the sample and defects inside the sample separately. It is also possible to obtain the depth distribution of defects inside the sample and the particle size distribution. Needless to say, the present invention is effective even if it is used only for measuring fine particles on the surface of a sample. In order to keep the sample clean, at least a portion of the sample moving stage 16 in contact with the sample 3 is
It is desirable to use a material containing few heavy metal impurities, for example, high-purity quartz.

(Embodiment 2) FIG. 5 shows an apparatus configuration of a second embodiment according to the present invention.

In the first embodiment, the frequency modulation is performed by vibrating the mirror 2, but in the present embodiment, the ultrasonic modulators 15 and 15 'are used. Laser light source 11
Light is split into two by the half mirror 9, one of which is the element 1
It is modulated at 80 MHz by 5 ', reflected by the polarizing prism 52', and irradiated on the sample. The other light is emitted by element 15
The reference light is modulated at 9 MHz, interferes with the scattered light from the defect, and is detected by the photodetector 7. The reflected light from the sample surface is removed by the polarizing plate 51 and the polarizing prism 52 'in this embodiment. Of the detected optical signal intensity, only the frequency component of 1 MHz is selectively amplified by the frequency filter 40 or the lock-in amplifier 18 and taken into the control computer 20. Although the frequency filter 40 and the lock-in amplifier 18 are switched in FIG. 5, only one of them may be used. Other measurement methods and data processing methods are the same as in the first embodiment.

(Embodiment 3) FIG. 6 shows an apparatus configuration of Embodiment 3.

In this embodiment, a light source 11 '(Nd: YLF laser, wavelength 131) having a different wavelength from the structure of the first embodiment is used.
9 nm) is added. As described in the action, there is an uncertainty in obtaining the defect position from the phase, but in order to eliminate this uncertainty, the same intensity and phase as those in the first embodiment are used for each wavelength at a plurality of wavelengths. Take measurements. The method of obtaining the depth position of the defect from the phase is as described in the operation. The reflected light from the sample surface is removed using the polarizing plate 51. It is effective to add the light sources 11 'having different wavelengths to the second embodiment as in the third embodiment.

(Embodiment 4) FIG. 7 shows an apparatus configuration of Embodiment 4.

In this embodiment, the configuration of the second embodiment has two signal detectors. The scattered light that has passed through the polarizing plate 51 and is polarized is made incident on only one of the detectors 7 ′ by the polarizing prism 52, and the reference light is circularly polarized by the ¼ wavelength plate 53. It is divided into two components of equal intensity and orthogonal polarization directions. The respective polarization components of the reference light are incident on the detectors 7 and 7 '.
The subtraction circuit 27 takes the difference between the signals of the two detectors. This makes it possible to remove the influence of the intensity fluctuation of the reference light. This principle is as described in the operation. The light reflected from the sample surface is reflected by the polarizing plate 51 and the polarizing prism 5.
2'and remove only the scattered light from the defect.

(Fifth Embodiment) In this embodiment, the defect measuring apparatus of the first embodiment is used to perform process control in manufacturing a semiconductor device.

First, of a series of semiconductor device manufacturing steps,
After the high-temperature heat treatment process such as thermal oxidation and annealing is completed, a semiconductor device in the process of manufacture or a semiconductor substrate for process evaluation is used to detect defects existing on the surface or inside of the semiconductor device by using the measuring device of the first embodiment. measure. When the number or size of detected defects exceeds a predetermined management standard, the semiconductor device in the process of manufacturing is discarded as a defective product. As a result, it is possible to eliminate the waste of adding more steps to defective products. Further, there is also an advantage that the defect can be detected without waiting for the completion of the semiconductor device, and the cause can be investigated and the problem can be improved at an early stage. The above process control method is an example, and it is desirable to employ various control methods according to the characteristics of the process.

It should be noted that, in the initial stage of the semiconductor device manufacturing process, the minute defect formed in the semiconductor substrate can be detected in a wide range for the first time by the defect measuring apparatus of the first embodiment. . Moreover, in Example 1, since high-purity quartz is used for the stage of the sample,
It has become possible to measure the semiconductor device itself in the process of manufacturing and add more steps after measurement to complete the finished product.
With these, the above process control can be realized. It should be noted that it is desirable to measure the semiconductor device in the process of manufacture using the first embodiment, before the cleaning process or the process in which contamination does not cause a serious problem such as plasma etching or ion implantation. This means that although Example 1 has a clean sample stage, the occurrence of defective semiconductor devices and process contamination due to sudden contamination due to mechanical troubles and human error is minimized. This is because.

(Embodiment 6) In this embodiment, the present technology is applied to a film thickness monitor in the thin film growth process.

First, an example of application to a device for monitoring a change in film thickness during the epitaxial growth process of Si on a silicon substrate will be described below. In this case, any of the devices in Examples 1, 2, 3 and 4 may be used. However, since the reflected light from the surface is used, the polarizing plate 51 is set to transmit the reflected light. The Z direction position of the surface is constantly monitored by the phase change of the reflected light. Conventional measurement methods include scanning the irradiation wavelength and measuring the film thickness from the wavelength at which the reflected light intensity is the strongest (Nanometrics' product name Nanospecs), and polarization analysis of the irradiation light and the reflected light. There is an ellipsometry method for deriving the thickness. However, the above technique aims at monitoring the film thickness growth of a substance different from that of the substrate, and it is theoretically impossible to monitor the film thickness growth of the same substance as that of the substrate. For example,
It is not possible to monitor the epitaxial growth of silicon on a silicon substrate. The advantage of using this technology as a monitor of the thin film growth process is that the film thickness change is recorded by the phase change of the reflected light from the film surface, so it is possible to measure even the film of the same substance, and the reflected light from the bottom surface of the film can be measured. That is, the film thickness can be accurately evaluated by the difference between the phase and the phase of the reflected light from the upper surface of the film. Further, in the present technology, since the phase of light is measured as the phase of an interference fringe that changes with time in a sinusoidal shape, the resolution is high. In this method, the time derivative of the change in brightness of the interference fringes due to the temporal change in the film thickness of the sample is detected, and the integrated amount thereof becomes the film thickness. On the other hand, when the interference fringes that do not change with time are used, the measurement resolution of the film thickness is zero in the light and dark peaks and valleys (positions where the derivative is zero), and the resolution becomes maximum in the middle. Therefore, the measurement resolution of the film thickness varies from zero to the maximum value. On the other hand, when the beat interference fringes are used, the light and dark changes temporally even if the film thickness is constant, and by detecting the phase thereof, the resolution in film thickness measurement can always be maximized. This is the reason for the high resolution.

For the above reasons, it is an important application example of the present measuring apparatus to incorporate the present measuring apparatus into an apparatus for growing various films on a semiconductor substrate or the like to form a film thickness monitor.

[0039]

As described above, in the crystal defect measuring device according to the present invention, the refractive index around the defect existing in the crystal is measured, and therefore the defect is measured separately from the foreign matter existing on the surface. It is possible to measure the phase difference between the scattered light from the defect and the reflected light from the crystal surface, and the depth position of the defect in the crystal can be measured with high resolution.

[Brief description of drawings]

FIG. 1 is a diagram showing a device configuration of an embodiment of a crystal defect measuring device according to the present invention.

2A is a diagram showing the relationship between the size (Δz ′, Δx ′) of the beam waist of the irradiation light in the sample and the refractive index n, and (b) the relationship between the phase of the detection signal and the depth of the defect. It is the figure shown.

FIG. 3 (a) Δz ′ and Δ when NA = 0.6
A graph showing the refractive index dependence of x ′, (b) (Δz ′)
6 is a graph showing the refractive index dependence of / (Δx ′) by changing the NA of the irradiation lens.

4A is a diagram showing the in-plane distribution of foreign matter on the surface, the grain size distribution and the depth position distribution, and FIG. 4B is a diagram showing the defect in-plane distribution, the grain size distribution and the depth position distribution in the oxide film. FIG. 4C is a diagram showing in-plane distribution of defects inside the crystal, grain size distribution, and depth position distribution.

FIG. 5 is a diagram showing a device configuration of a second embodiment.

FIG. 6 is a diagram showing a device configuration according to a third embodiment.

FIG. 7 is a diagram showing a device configuration according to a fourth embodiment.

[Explanation of symbols]

1 ... Laser light 2 ... Mirror 3 ... Sample 5 ... Lens 5 '... Objective lens 7 ... Photodetector 7' ... Photodetector 8 ... Pinhole 8 '... Pinhole 9 ... Half mirror 10 ... Mirror 11 ... Laser light source 11 ′ ... Laser light source 15 ... Frequency modulation element 15 ′ ... Frequency modulation element 16 ... Sample moving stage 18 ... Lock-in amplifier 20 ... Measurement control computer 21 ... Current-voltage converter 21 '... Current-voltage converter 25 ... Frequency modulation element Driver 26 ... Stage moving controller 31 ... Display device 40 ... Frequency filter 41 ... Piezoelectric element 42 ... Piezoelectric element driver 50 ... Prism 51 ... Polarizing plate 52 ... Polarizing prism 52 '... Polarizing prism 53 ... Quarter wave plate

─────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] December 24, 1993

[Procedure Amendment 1]

[Document name to be corrected] Drawing

[Name of item to be corrected] Figure 2

[Correction method] Change

[Correction content]

[Fig. 2]

Claims (9)

[Claims] 1. A crystal defect measuring device for detecting a defect existing on a crystal surface or in a crystal, wherein a refractive index around the defect is measured. 2. The defect existing in the crystal is measured by measuring the refractive index around the defect to be distinguished from the fine particles existing on the surface of the crystal. Crystal defect measuring device. 3. A means for squeezing the irradiation light into the crystal as a means for measuring the refractive index around the crystal defect, and a means for narrowing the crystal sample by moving the irradiation light relatively. 3. The crystal defect measuring device according to claim 1, further comprising: a means for scanning the crystal sample with irradiation light to measure an intensity distribution of scattered light from defects existing in the crystal. 4. A crystal defect measuring device, characterized in that the depth position of the defect is determined by measuring the phase of scattered light from the defect existing in the crystal. 5. A crystal defect measuring device for detecting a defect existing on a crystal surface or in a crystal, a light source for emitting linearly polarized coherent light, a means for irradiating light from the light source into the crystal, and Means for detecting only a polarization component orthogonal to the polarization direction of the irradiation light of the scattered light of,
Means for interfering the light from the light source, which is frequency-modulated by a constant frequency, with the scattered light from the defect, and detecting the interference light as a signal; and the beat frequency of the interference light in the detection signal. A crystal defect measuring device comprising means for amplifying only a component. 6. A means for dividing the frequency-modulated light to detect one of the divided lights as a signal, and providing means for obtaining a difference between the detection signal and the detection signal of the interference light. The crystal defect measuring device according to claim 5. 7. A crystal defect measuring device for detecting defects in a crystal, wherein the size of a defect existing in a crystal other than silicon oxide is assumed to have the same refractive index as that of silicon oxide. Then, the crystal defect measuring device is characterized in that the intensity of scattered light from the defect is measured and derived. 8. A semiconductor manufacturing apparatus for physically or chemically adding an element to a semiconductor substrate, wherein the crystal defect measuring apparatus according to any one of claims 1 to 7 is incorporated in the manufacturing apparatus. Semiconductor manufacturing equipment. 9. A semiconductor manufacturing apparatus characterized by measuring and evaluating a crystal defect on the surface of a semiconductor wafer or in a wafer crystal by the crystal defect measuring apparatus according to any one of claims 1 to 7.