KR0136213B1 - Photomask Inspection Method and Apparatus - Google Patents

Abstract

In the photomask inspection method, the photomask has an orthogonal polarization direction and the polarization state of the elliptical polarization caused by the superposition of two linear polarizations passing through two different optical paths, and the target portion of the photomask is a linear polarization beam. The photomask is inspected on the basis of the difference between the polarization states of the elliptical polarizations that are generated when two linearly polarized beams overlap each other after being installed in one beam of light.

Description

Photomask Inspection Method and Apparatus

FIG. 1 (a) is a view for explaining the principle of the present invention, and is a state diagram when two linearly polarized beams of an optical system having an inspection object progress.

FIG. 1 (b) is a view for explaining the principle of the present invention, which is a state diagram when two linearly polarized beams of an optical system without an inspection object proceed.

FIG. 1C shows the polarization state of light generated by superposition of two light beams in a plane perpendicular to the traveling direction of light when there is the inspection object of FIG.

FIG. 1D shows the polarization state of light generated by superposition of two light beams in a plane perpendicular to the traveling direction of light when there is no inspection object in FIG.

2 shows the structure of a system according to an embodiment of the invention.

3 is a diagram showing the structure of a system according to another embodiment of the present invention.

4 is a graph showing the transmission characteristics of the etaron used in the first embodiment of FIG.

5 to 7 show a system structure of another embodiment of the present invention.

8 shows the principle of an elliptical polarization measuring method.

FIG. 9 shows the trajectory of the tip of an amplitude vector of elliptical polarization generated by combining two linearly polarized beams.

10 is a diagram showing the relationship between the positional difference of two linearly polarized beams that generate elliptical polarization and the polarization state of the elliptical polarization.

11 shows the elliptical polarization principle of the present invention.

FIG. 12 is a view showing the locus of the tip of an amplitude vector of an elliptical polarization viewed from the light intensity detector side of FIG.

FIG. 13 shows an embodiment of a measurement system based on the principle of FIG.

14 shows measurement results obtained when measuring elliptical polarization using the measurement system of FIG.

15 to 17 are views showing a photomask inspection apparatus according to an embodiment of the present invention.

FIG. 18 shows a method for calculating δ and ax in the embodiment of FIG. 17.

19 is a view showing a photomask inspection apparatus according to another embodiment of the present invention.

20 to 22 are views showing a passage adjusting device according to an embodiment of the present invention.

23 to 25 is a view showing a photomask inspection apparatus according to another embodiment of the present invention.

FIG. 26A is a graph showing amplitude and light intensity distribution obtained using a conventional photomask (NPM). FIG.

FIG. 26 (b) shows amplitude and light intensity distribution obtained using a phase inversion photomask (PPM). FIG.

* Explanation of symbols for the main parts of the drawings

201, 301: light source 202, 302: optical system

203: filter 204, 207: linear polarizer

205, 212, 403, 413, 414, 415: half mirror

206, 211, 404, 418: total reflection mirror

207, 405, 406: lambda / 2 wavelength plate 208 photomask

209: photomask portion 210: random pattern portion

213, 416, 417: polarization state measuring device 303: etaron

401: extracted monochromatic light 402: beam splitter

411, 412: Inspection photomask

The present invention relates to a photomask inspection method and apparatus.

Recently, various fine patterning techniques have been proposed to increase the degree of integration of semiconductor integrated circuits.

In general, lithography was used to pattern.

To form a fine pattern, the resolution of a reduced projection exposure apparatus called stepper used in lithography technology must be improved.

Techniques for increasing the numerical aperture of the projection lens and decreasing the wavelength of the light source have been used to increase the resolution of the reduced projection exposure apparatus.

However, the technology has already achieved resolution near theoretical limits.

For this reason, other techniques have recently been studied to improve the resolution.

As a method of overcoming the theoretical limit, there is a method of giving a phase difference of approximately 180 ° C. to light passing through the transmission part of the patterned photomask when the photomask is applied, and a method of changing a certain transparency and phase in the pattern portion.

Such photomasks are generally referred to as phase reversal masks.

This method was presented in the J4 of Proceedings of the 36th International Symposium on Electron, Ion and Photo Beams.

This technique improves the light intensity distribution on the trademark surface as a result of light interference in each part of the photomask.

Phase shift technology is described below in comparison with a technique using a normal photomask.

FIG. 26A shows the case of using a normal photomask (NPM) using a line-and-space pattern formed by selectively placing a mask member on a glass substrate. It shows the amplitude and light intensity distribution when forming an image by.

In the amplitude distribution characteristic shown in FIG. 26 (a), the dotted line represents the amplitude distribution of the light passing through the spacer, and the solid line represents the amplitude distribution of the light according to the interference.

The light intensity distribution is obtained by squaring the amplitude distribution represented by the solid line.

FIG. 26 (b) shows the amplitude and the light intensity distribution when the phase inversion mask PPM is used.

As in FIG. 26 (a), the dotted line shows the amplitude distribution of the light passing through each space, and the solid line shows the amplitude distribution of the light according to the interference.

The intensity distribution is obtained by squaring the amplitude distribution represented by the solid line.

In the case of the normal photo mask NPM shown in Fig. 26 (a), the light intensity of each light shielding portion does not become zero because it overlaps in the in-phase phase of the refraction image from neighboring transmission portions.

In contrast, in the phase inversion photomask PPM of FIG. 26B, the light intensity of each light-closing part becomes zero because the light refraction from neighboring transmission parts is superimposed in reverse phase.

As a result, the contrast of the image is improved by using the phase inversion photomask PPM of FIG. 26 (b).

It should be noted that the contrast improvement effect decreases when the phase of light from neighboring transparent portions deviates from π and the amplitude of light passing through the phase inversion member decreases due to absorption in the phase inversion member.

Therefore, in the phase inversion mask, it is very important to adjust the phase and amplitude transmittance of the light transmitted through the respective portions.

If the refractive index and extinction coefficient of each material used in the photomask are accurately obtained, the phase, amplitude transmittance and energy transmittance can be adjusted by adjusting the thickness of each material.

In many cases, however, it is difficult to accurately measure the refractive index and extinction coefficient of each material.

In addition, impurities are mixed in each material during the manufacturing process, causing errors in refractive index and extinction coefficient.

In addition, the thickness of each material is necessarily accompanied by manufacturing tolerances.

For this reason, the phase, amplitude transmittance and energy transmittance must be accurately measured for each part of the fabricated photomask, and the results should be fed back into the manufacturing process.

This operation is repeated to realize the desired photomask.

However, there is no effective means of measuring phase, amplitude transmission and energy transmission.

It is an object of the present invention to provide a photomask inspection method and apparatus which can easily and accurately observe or measure phase, amplitude transmittance and energy transmittance.

In order to achieve the above object, the present invention divides light into light rays passing through two different optical paths, and the two beams are linearly polarized beams polarized perpendicularly to each other in a plane perpendicular to the traveling direction of light. Convert.

In addition, the polarization state of the elliptical polarization generated by the superposition of two linearly polarized beams is observed.

Accordingly, the inspection target portion of the photomask is provided in one of the optical paths, and the polarization state of the elliptical polarization generated in accordance with the installation of the inspection target portion is observed.

The photomask is inspected by observing the polarization state of light generated by the overlapping change before and after the inspection target portion of the photomask is installed in the optical path.

This operation is obtained by observing the polarization state of the elliptical light generated by superimposing the linearly polarized beams perpendicular to each other on the amplitude transmittance, the energy transmittance, and the phase change amount for the light passing through each photomask part.

This principle is explained with reference to FIGS. 1 (a)-(d) as follows.

1 (a) and (b) show the optical system of the present invention.

FIG. 1 (a) shows the case where the inspection objects are located in one of the optical paths of the linear polarization beams perpendicular to each other, and FIG. 1 (b) shows the case where the inspection objects are not located in the optical path.

Referring to Figs. 1 (a) and (b), the advancing direction of light is set to a plane parallel to the ground.

In addition, the x-axis is the right direction of the ground and the y-axis is a direction from the lower surface of the ground toward the upper surface side.

Referring to Figures 1 (a) and (b), reference numeral 101 denotes a linearly polarized light oscillating ahead in the y-axis direction, 102 denotes a half-mirror that divides the light 101 into two light rays, 103 denotes a total reflection mirror for refraction of one of the light beams, 104 denotes a λ / 2 wavelength plate, 105 denotes a linear polarization beam into which an inspection target portion of a photomask is to be inserted, and 106 denotes a linear polarization beam passing through another optical path. .

The linearly polarized beam 106 oscillates in the y-axis direction and the linearly polarized beam 105 oscillates in the x-axis direction.

The two linearly polarized beams 105 and 106 overlap each other through the total reflection mirror 107 and the half mirror 108.

1 (c) and (d) show the polarization states in the plane perpendicular to the traveling direction of light, and the polarization states are obtained using the optical system shown in FIGS. 1 (a) and (b).

The linearly polarized beams 105 and 106, respectively,

Ex = ax cos (ωt-δ ₁ ). … (One)

Ex = ay cos (ωt-δ ₂ ). … (2)

In this case, the light generated by the superposition of two light beams is expressed as follows.

Provided that δ = δ _1- δ ₂ . … (4)

Ex is a linearly polarized beam oscillating in the x-axis direction at a given point closer to the traveling direction than the half mirror 108 before the photomask 109 is inserted, Ey is y at the same point as Ex is defined. The linearly polarized beam oscillating in the axial direction, ax denotes an amplitude of Ex, ay denotes an amplitude of Ey, δ ₁ denotes an initial phase of Ex, and δ ₂ denotes an initial phase of Ey.

When the photomask 109 to be inspected is inserted into the optical path of the linearly polarized beam 105 as shown in FIG. 1A, the linearly polarized beam Ex 'after the photomask 109 to be inspected is inserted is It is displayed as follows.

Ex '= ax' cos (ωt-δ ₁ ). … (5)

At this time, Ex 'is a linearly polarized beam oscillating in the x-axis direction at the same point Ex is defined after the photomask 109 is inserted, δ ₁ is the change of δ ₁ according to the insertion of the photomask 109 Is a value obtained by the change of ax due to the insertion of the photomask 109.

The light generated by the superposition of the two rays is represented by the following equation.

Provided that δ '= δ' _1- δ ₂ . … (7)

Therefore, according to the basic principle of the present invention, the elliptical polarization generated by the superposition of two linearly polarized beams before the photomask 109 is inserted is observed or measured, so that first, ax and δ are obtained.

Then, after the photomask 109 is inserted, the elliptical polarization generated by the superposition of two linearly polarized beams is observed or measured to obtain ax 'and δ.

When θ is the amount of phase change generated by the photomask 109, t is amplitude transmittance, and T is energy transmittance, these values can be obtained as follows.

θ = δ′-δ... … (8)

t = ax '/ ax... … (9)

T = (ax '/ ax) ² . … 10

2 shows an embodiment of the present invention.

Referring to FIG. 2, reference numeral 201 denotes a light source such as a low pressure mercury lamp for lithography, 202 an optical system having a condensing and aiming function, and 203 extracts monochromatic light such as a g-line or i-line of light emitted from the light source 201. The filter 204 is a linear polarizer, 205 is a half mirror, 206 and 211 are total reflection mirrors, and 207 is a λ / 2 wave plate.

Light emitted from the light source 201 is converted into monochromatic light by the optical system 202, the filter 203 and the linear polarizer 204, and is linearly polarized after being aimed at.

This linearly polarized light is amplitude-separated through the half mirror 205 into two light rays.

One light ray travels in a straight line, and its aiming direction is rotated 90 degrees by the λ / 2 wave plate 207.

The other ray travels in the direction moved 90 ° from the direction of the ray and is totally reflected by the total reflection mirror 206 and reflected 90 ° to travel toward the total reflection mirror 211.

The linearly polarized light that has reached the total reflection mirror 211 is reflected by 90 ° again and is superimposed on the linearly polarized light which has been straightened by the half mirror 212.

Reference numeral 208 denotes a photomask, 209 denotes a photomask portion composed only of a substrate portion exhibiting high transparency to exposure light, and 210 denotes an arbitrary pattern portion of the photomask 208.

In general, when inspecting the photomask, the transmittance and phase change amount of the arbitrary pattern portion 210 with respect to the photomask portion 209 composed only of the substrate portion exhibiting high transparency to the exposure light is required.

In this embodiment, the polarization state measuring device 213 is used to observe or measure the polarization state of the generated elliptical polarization.

At this time, the elliptical polarization occurs when the linearly polarized light passing through the photomask unit 209 including only the substrate part having a high light transmittance with respect to the exposure light is superimposed on the linearly polarized light traveling along the optical path outside the photomask.

By this observation or measurement, the ax value and (delta) value in Formula (1) and (4) are obtained.

Next, the polarization state of the elliptically polarized light generated when the linearly polarized light passing through the arbitrary pattern portion 210 overlaps the linearly polarized light traveling along the optical path outside the photomask is transmitted to the polarization state measuring device 213. Is observed or measured to obtain the ax 'and δ values in equations (5) and (7).

The phase change amount θ, amplitude transmittance t and energy transmittance T as target values are obtained from equations (8)-(10).

In the embodiment shown in FIG. 2, the amount of phase change, amplitude transmittance, and energy transmittance of each photomask portion with respect to the portion composed only of the substrate portion having high transparency with respect to the exposure light are obtained.

However, if the αx and δ values in the equations (1) and (4) can be obtained before measuring the polarization state of the elliptical polarization generated, the absolute value of the phase change amount, amplitude transmittance and energy transmittance of each photomask portion is Obtained.

At this time, the elliptical polarization occurs when two linearly polarized beams overlap with each other in a light path of a linearly polarized beam that proceeds in a straight line without arrangement of a photomask.

In addition, in the embodiment shown in FIG. 2, the photomask is measured while being arranged in the optical path of the polarizing beam running in a straight line of the two linearly polarized beams. However, even if the photomask is arranged in the optical path of another linearly polarized beam reflected 90 ° by the half mirror 212, the measurement can be performed without any problem.

3 shows another embodiment of the present invention.

In this embodiment, the optical path difference between the optical paths of the linearly polarized beam polarized in the orthogonal direction in the embodiment shown in FIG. 2 is less than the coherence length at which the light emitted from the light source can be changed into monochromatic light. It is set large.

In this state, a large photomask can be inspected as well.

In this case, the bandwidth of the linearly polarized beam polarized in the perpendicular direction should be reduced.

To this end, in this embodiment, the line spectrum of the voltage mercury lamp as a light source is used.

Referring to FIG. 3, reference numeral 301 denotes a light source made of a low pressure mercury lamp, 302 an optical system having a focusing and aiming function, and 303 an etaron.

Etalon is already known and is published in Max Borm and Emil Wolf's Principles of Optics, 6th edition, Pergamon Press

A plurality of different broad spectrums are emitted from the low pressure mercury lamp light source 301.

The etaron 303 passes only the necessary line spectrum.

As shown in FIG. 4, the etaron 303 is a band pass filter having a plurality of transmission bandwidths.

Referring to FIG. 4, the center width of the transmission band and the full width of one half of the maximum value of the transmission band are determined as the reflectance of the reflective film formed on the two planes and the distance between the opposing planes of the etaron. .

However, the position of the transmission band is determined by the angle defined by the plane parallel to the incident light.

Therefore, the installation angle of the etaron 303 is adjusted to transmit only the line spectrum corresponding to the i-line or g-line used in the lithography technique.

The configuration after the configuration for extracting monochromatic light is the same as the embodiment shown in FIG.

Assume that a laser oscillating at a wavelength substantially equal to the wavelength of the exposure light is used as the inspection light source, or that the exposure light itself is used as the inspection light source as in excimer laser lithography.

In this case, in the embodiments shown in FIGS. 2 and 3, the optical systems 202, 203, filters 203, and 303, which perform combined focusing and aiming operations for converting to monochromatic light, may be omitted. .

5 illustrates another embodiment of the present invention, and a plurality of photomasks can be inspected at one time.

This embodiment exemplifies an apparatus for simultaneously inspecting two photomasks.

Referring to FIG. 5, reference numeral 401 denotes monochromatic light extracted and aimed by the same configuration as the embodiment shown in FIG. 2 or FIG. 3, and 407 is a linear polarizer for converting incident light into linearly polarized light. A beam splitter for advancing light components corresponding to one-third intensity of monochromatic light in a straight line and reflecting 90 [deg.] Light components corresponding to two-thirds intensity of monochromatic light, and 403 is set to 90 by beam splitter 402. It means a lambda / 2 wave plate for inputting the reflected linearly polarized beam.

The light 401 is converted by the beam splitter 402, the half mirror 403 and the total reflection mirror 404 into three linearly polarized beams 408, 409 and 410 traveling in the same direction as the original traveling direction.

The linearly polarized beams 408 and 409 are light beams polarized in the same direction, and the linearly polarized beam 410 is light beams polarized in a direction perpendicular to the linearly polarized beam 408 or 409.

Referring to FIG. 5, 411 and 412 arranged at the emission sides of the λ / 2 wave plates 405 and 406 are photomasks to be inspected, and 413 are half mirrors arranged at the emission side of the total reflection mirror 404. , 414 is a half mirror arranged on the emission side of the inspection photomask 409, 415 is a half mirror arranged on the emission side of the inspection photomask 411, 418 is a straight through the half mirror 413 Total reflection mirrors for inputting linearly polarized light and reflecting them at 90 DEG, 416 and 417, refer to polarization state measuring devices, respectively.

Above all, the polarization state of the elliptically polarized light generated by the superposition of the linearly polarized beam 408 and the linearly polarized beam 410 and the elliptically polarized light generated by the superposition of the linearly polarized beam 409 and the linearly polarized beam 410. The polarization state is observed or measured by the polarization state measuring devices 416 and 417 while the photomasks 411 and 412 are not arranged.

Then, the polarization state of the elliptical light generated by the superposition of the linearly polarized beams 408 and 410 and the elliptical polarization state generated by the superposition of the linearly polarized beams 409 and 410 is While the photomasks 411 and 412 are arranged, they are observed or measured by the polarization state measuring devices 416 and 417, and the target phase change amount (θ), amplitude transmittance (t) and energy transmittance (T). It is calculated | required by this Formula (8)-(10).

6 shows an embodiment illustrating a basic configuration diagram of an apparatus for comparing / inspecting two photomasks.

Referring to FIG. 6, reference numeral 501 denotes a collimated monochromatic light extracted by the same configuration as the embodiment shown in FIG. 3, and 506 converts the light 501 into linearly polarized light. 502 is a half meter for inputting linearly polarized light from the linear polarizer 506, and 503 and 504 are total reflection mirrors for inputting and totally reflecting the linearly polarized beams divided by the half mirror 502, respectively. , 505 is a λ / 2 wave plate for inputting the linearly polarized light totally reflected by the total reflection mirror 503, 507 and 508 are the inspection photomask, 509 and 510 are transparent to the linearly polarized light, all the same material Support tables for photomasks 507 and 508 having the same thickness, 516 and 517 are total reflection mirrors arranged on the emission sides of photomasks 507 and 508, respectively, and 511 are total reflection mirrors 516 and The linearly polarized beam totally reflected by 517 Force to the half mirror, 512 and 513 are linearly polarized light beam, and 515 to each other, oscillating in a direction perpendicular to the polarization state measuring means, respectively.

First of all, two linearly polarized beams 512 and 513 pass through portions of the photomasks 507 and 508 having the same optical characteristics and pattern shape.

The polarization state of the elliptical polarization generated by the superposition of the two linearly polarized beams 512 and 513 is measured by the polarization state measuring device 515.

Thereafter, the support tables 509 and 510 are moved synchronously.

If the photomasks 507 and 508 are identical to each other, the two linearly polarized beams 512 and 513 change in the same phase and in the same amplitude.

Therefore, the polarization state of the elliptical polarization generated by the superposition does not change.

If such photomasks 507 and 508 have different portions, the polarization state of the elliptical polarization changes when the linearly polarized beams 512 and 513 pass through the portions.

Assume that one of the two photomasks 507, 508 is the original photomask and the other is a duplicated photomask.

In this case, if the copying is done completely, the state of the elliptical polarization generated by the overlap of the light beams from any part of the two masks shows the same polarization state as when initially installed.

However, if the defect portion generated in the replication process is located in the optical path, the polarization state of the elliptical polarization generated by the superposition is changed.

That is, the defect part of the duplicated mask is detected.

FIG. 7 shows an embodiment in which the polarization state measuring apparatus of the embodiment shown in FIG. 2 is composed of a rotation analyzer and a light intensity detector, and the observation is performed using a plurality of analyzers and detectors.

In this embodiment, n detectors are used.

Referring to FIG. 7, reference numeral 601 denotes polarized light generated by superposition, 631 is a first light intensity detector, 632 is a second light intensity detector, 633 is a third light intensity detector, and 634 is an nth light intensity. The detector 621 denotes a first analyzer, 622 denotes a second analyzer, 623 denotes a third analyzer, and 624 denotes an n-th analyzer.

These analyzers are fixed so that their directions move with each other by an angle of 360 ° / n.

Reference numeral 611 denotes a beam splitter for directing only light components corresponding to 1 / n intensity of incident light in a straight line and reflecting the remaining light components at 90 °, and 612 is a 1 / (n-1) intensity of incident angle. It is a beam splitter for reflecting only the corresponding light component at 90˚ and the remaining light component is proceeding in a straight line, and 613 reflects only light component corresponding to 1 / (n-2) intensity of incident light at 90˚. The remaining light components are beam splitters for traveling in a straight line, and 614 denotes a total reflection mirror.

The polarization state of the polarized light 601 generated by the superposition by this arrangement is immediately observed at an accuracy of 360 ° / n.

In the embodiment shown in Figs. 5-7, a plurality of masks are inspected.

However, it is clear that the three embodiments can be applied to inspecting a plurality of inspection target portions in one mask without any special explanation.

As described above, according to the embodiment of the present invention shown in FIGS. 2 to 7, the transmittance and the amount of phase change of each photomask portion can be obtained easily and accurately.

Further, according to these embodiments, the present invention can be applied not only to a photomask but also to any object exhibiting transparency with respect to light of a wavelength to be handled.

In addition, if the thickness of the object is known, the refractive index, absorption coefficient, or extinction coefficient of the object can be obtained from the obtained amplitude transmittance and the phase change direction.

Referring to the elliptical length measurement device used in the embodiment as follows.

8 shows a basic configuration of an elliptical polarization measuring device.

Referring to FIG. 8, reference numeral 801 denotes a compensator, 802 an analyzer, 803 a photodetector, 805 an elliptical polarization to be measured, and 806 a cyclic polarization generated by the conversion of the elliptical polarization 805 by the compensator 801. Indicates.

When obtaining the leading trajectory of the amplitude vector of the elliptical polarization generated by the superposition of two linearly polarized beams represented by equations (1) and (2), an ellipse as shown in FIG. 9 is obtained.

This ellipse is represented by Formula (3) described above.

FIG. 10 shows the change of the slope of the ellipse with the change of the value of δ expressed by the equation (3).

Referring to FIG. 10, the direction of rotation of the amplitude of the vector is indicated by an arrow.

9 and 10, the long axis (distance) direction of the ellipse is defined as the X axis direction, the short axis direction is defined as the Y axis direction, and 1/2 of the long axis length is defined. We define a as b and half of the shorter length as b.

In this case an ellipse is given by

Assuming that ψ is an angle defined by the X and Y axes, the following relationship is established between the variables.

a ² + b ² = ax ² + ay ² . … (12)

a ² -b ² = (ax ² -ay ² ) cos 2ψ + 2axay sin 2ψ cos. … (13)

ab = axay sin δ. … (14)

In addition, the stroke parameters S ₀ , S ₁ , S ₂ , and S ₃ representing the polarization state and the degree of polarization of the elliptical polarization are expressed using the variables a, b, and ψ as follows.

S ₀ = ax ² + ay ² = a ² + b ² . … (15)

S ₁ = ax ² -ay ² = (a ² -b ² ) cos ² ψ. … (16)

S ₂ = 2axay cos δ = (a ² -b ² ) sin ² ψ. … (17)

S ₃ = 2axay sin δ = 2ab. … (18)

In this case, the parameters a / b and ψ and direction of rotation and wave to determine the polarization state of the elliptical polarization

The parameters ax / ay and ψ and direction of rotation or parameters S ₁ , S ₂ , S ₃ must be determined.

Referring to FIG. 8, a method of determining the polarization state of elliptical polarization is as follows.

Elliptical polarization is divided into two perpendicular linearly polarized beams represented by equations (1) and (2), respectively.

In this case, the compensator 801 must forcibly adjust the phase difference between these two linearly polarized beams to π / 2 or 3π / 2.

At this time, the compensator 801 adjusts the value of ax / ay to 1.

The compensator's construction is detailed in Max Born and Emil Wolf's Principles of Optics, 6th edition, Pergamon Press.

Assume that cyclic polarization 806 is obtained by appropriate polarization.

In this case, even if the analyzer 802 rotates, the polarization state of the cyclic polarization 806 does not change.

As a result, the output of the photodetector 803 remains the same.

Therefore, the phase difference of the two linearly polarized beams, expressed by equation (4), can be obtained from the phase difference given by the compensator 801.

Also, the original value of ax / ay can be known from the magnitude of the adjustment value given to adjust the value of ax / ay to 1.

However, this method requires the following unnecessary operation.

First, the adjustment amount of the phase difference and ax / ay given by the compensator 801 is set to a given value.

The analyzer 802 rotates 1/2 turn to check whether the output of the photodetector 803 depends on the angle of the analyzer 802.

These operations must be repeated until the dependency is zero.

In addition, the compensator 801 must be fixed while the compensator 801 is rotating.

For this reason, the adjustment amount given by the compensator 801 must be set to a discrete value.

Strictly speaking, this problem indicates that it is impossible to convert elliptical polarization to circular polarization unless the amount of adjustment necessary to convert the elliptical polarization to circular polarization is consistent with the set value as the adjustment value given by the compensator 801. .

That is, the accuracy in determining various parameters representing the polarization state of the elliptical polarization depends not only on the measurement accuracy but also on the degree of conversion of the elliptical polarization to the circular polarization.

In order to solve the above problem, an embodiment in which various parameters can be determined quickly and accurately without converting the elliptical polarization to be measured into other polarized light is shown in FIGS. 11 to 14.

In the embodiments shown in FIGS. 11 to 14, the parameters for the elliptical polarization are determined only by four measurements.

Four measurements are made with the maximum and minimum light intensity obtained when the analyzer is rotated 1/2 turn, the angle of the analyzer relative to the reference direction corresponding to the maximum light intensity, and the λ / 4 waveplate mounted on the front of the analyzer. The angle of the analyzer relative to the reference direction corresponding to the maximum light intensity of the case.

The determination method will be described in detail with reference to FIG.

11 shows a principle of measuring elliptical polarization according to an embodiment of the present invention.

Referring to FIG. 11, reference numeral 851 denotes a λ / 4 wave plate, 852 an analyzer, 853 an optical intensity detector, 855 an elliptical polarization, and 856 an polarized light obtained through an analyzer 852.

The method of determining the polarization state of the elliptical polarization using this measuring system is as follows.

First, when the elliptical polarization 805 to be measured is incident on the analyzer 852, the analyzer 852 rotates about the optical axis of the elliptical polarization counterclockwise when viewed from the light intensity detector 853 side.

In this operation, the maximum light intensity of the elliptic length 856 emitted from the analyzer 852, the rotation angle ψ of the analyzer 852 when the light intensity is maximum, and the minimum light intensity of the polarized light 856 are obtained. Lose.

Next, the λ / 4 wave plate 851 is disposed on the incidence side of the analyzer 852, and when the elliptical polarization 855 is sequentially incident on the λ / 4 wave plate 851 and the analyzer 852, the analyzer ( 852 rotates counterclockwise with respect to the optical axis of the elliptical polarization 855.

In this operation, the rotation angle ψ of the analyzer 852 set when the light intensity of the wave polarization 855 emitted from the analyzer 852 is maximum is obtained.

12 shows the trajectory of the tip of the amplitude vector of the elliptical polarization when viewed from the light intensity detector 853 side.

Referring to FIG. 12, the x-axis, y-axis, x-axis, y-axis and the ax, ay, a, b and ψ values are all the same as in FIG.

Reference θ indicates the direction of the analyzer 852 measured from the X axis.

Assume that (± m, ± n) is the intersection between the line represented by Eq. (11) and the line represented by Equation (19) below and the X- and Y-coordinates, given by the amplitude vector in the θ direction. Assume that energy is proportional to (m ² + n ² ) × 2.

Y = X tan? … (19)

More precisely, m ² and n ² are expressed as

m ² = a ² 2b ² / (b ² + a ² tan ² θ). … 20

n ² = a ² b ² tan ² θ / (b ² + a ² tan ² θ). … (21)

In this case, the optical component obtained by multiplying the amplitude vector by cos α in the (θ + α (−π / 2 <α <π / 2)) direction passes through the analyzer 802.

Therefore, for convenience, assuming that the proportionality constant is K, the total energy I (θ) passing through the analyzer 802 is given as follows.

Integrating when θ = 0 and π / 2, the total energy is

I (0) = Kπa ² b / (a + b). … (23)

I (π / 2) = Kπa ² b / (a + b). … (24)

I (0) means the maximum value indicated by the light intensity detector, and I (π / 2) means the minimum value indicated by the light intensity detector.

The equations (23) and (24) are solved for a and b as follows.

a ² = I (0) (I (0) + I (π / 2) / KπI (π / 2) ... (25)

b ² = I (π / 2) + (I (0) + I (π / 2) / KπI (0) …… (26)

therefore,

a / b = I (0) / I (π / 2)... … (27)

Solving equations (15) and (16) with respect to ax and ay and substituting equations (25) and (26) as follows:

Δ of equation (4) is according to equation (14)

sin ² δ = a ² b ² / ax ² ay ² . … (31)

Substituting equations (25), (26), (28) and (29) into equation (31) is as follows.

In the case of, the measurable amounts are I (0), I (π / 2) and ψ, and four solutions that can be obtained for δ are obtained using equation (32).

Since the value of ψ is limited as shown in FIG. 10, the solution that can obtain this can be limited to two.

If ψ is positive, 0 <δ <π / 2 or 3π / 2 <δ <2π.

If ψ is negative, then π / 2 <δ <π or π <δ <3π / 2.

To limit the solution to δ, which is limited to 2, to 1, the λ / 4 wave plate 851 is placed so that the phase of one of the two linearly polarized beams in the x and y directions advances by π / 2. It may be arranged on the incident side of the analyzer 852.

It is assumed that the phase of the linearly polarized beam in the y-axis direction is earlier by? / 2 than the linearly polarized beam in the x-axis direction.

In this case, π / 2 increases with respect to δ according to equations (1) and (3).

At this time, ψ is measured again.

If ψ given before insertion of the λ / 4 wave plate 851 is positive, 0 <δ <π / 2 or 3π / 2 <δ <2π.

If? is still positive after insertion of the λ / 4 wave plate 851, it can be determined that 3π / 2 <δ <2π is correct.

If ψ becomes negative, it can be determined that 0 <δ <π / 2 is correct.

Similarly, if π / 2 <δ <π or π <δ <3π / 2 is correct before the λ / 4 wave plate 851 is inserted, and ψ remains negative after the λ / 4 wave plate 851 is inserted, π / 2 <δ <π is determined.

If ψ is positive, then it is determined that π <δ <3π / 2 is correct.

Therefore, the original value of δ is uniquely determined from the sign of ψ before and after insertion of the λ / 4 waveplate.

As described above, since I (0), I (π / 2) and ψ are measurable quantities, all parameters related to elliptical polarization are represented by equations (15)-(18), (25)-(30) and (32). Can be determined by a determination step using

13 shows a measurement system used to carry out a method for measuring elliptical polarization according to the present invention.

Referring to FIG. 13, reference numeral 911 denotes a light intensity detector, 921 denotes a He-Ne laser for generating polarized light at an output of 1 mW, 922 denotes a polarizer arranged on the emission side of the He-Ne laser 921, 923 denotes a half mirror arranged at the emission side of the polarizer 922, and 924A 924B reflects the beam 932 of 90 between two linearly polarized beams 931 and 932 respectively divided by a half mirror 923. The total reflection mirror 925 denotes a λ / 2 wave plate arranged between the total reflection mirrors 924A and 924B, respectively.

The laser emitted from the He-Ne laser 92 is converted into linearly polarized light by the polarizer 922, and the traveling direction is rotated 90 ° with the linearly polarized beam 931 traveling in a straight direction by the half mirror 923. It is divided into a linear polarization beam 932.

The linearly polarized beam 932 is reflected back by 90 ° by the total reflection mirror 924A and proceeds in a direction parallel to the linearly polarized beam.

The oscillation direction of the linearly polarized beam 932 is rotated 90 degrees through the λ / 2 wave plate 925 to be perpendicular to the oscillation direction of the linearly polarized beam 931.

The traveling direction of the linearly polarized light beam 932 in which the linearly polarized light oscillation direction is rotated 90 degrees is further rotated 90 degrees by the total reflection mirror 924B.

As a result, the linearly polarized beam 932 is combined with the linearly polarized beam 931 by the half mirror 923 to generate newly polarized light 935.

The newly polarized light 935 is typically elliptical polarization.

However, in a special case, for example, when the phase difference between the linearly polarized beams 931 and 932 is 0 or π, the polarized light 935 becomes linearly polarized light.

When the phase difference between the linearly polarized beams 931 and 932 is π / 2 or 3π / 2, and the amplitude ratio between the beams is 1, the polarized light 935 becomes circularly polarized light.

FIG. 14 shows measurement results obtained using the measurement system shown in FIG.

Referring to FIG. 14, the x and y axes are parallel to the oscillation directions of the linearly polarized beams 931 and 932, respectively.

Assume that the angle defined by the analyzer 902 and the x-axis is θ as shown in FIG.

In this case, it is measured that 0≤θ≥π.

For π <θ <2π, the measured value shifts symmetrically with respect to the original value and further measurement is performed.

In this measurement, ψ = -5 °, I (0) = 78.55 µW, and I (π / 2) = 26.6 µW.

Four obtainable solutions for δ 77.2 °, 102.8 °, 257.2 °, and 282.8 ° are obtained from these values and equation (32).

Since ψ = -5 °, two obtainable solutions are limited to 102.8 ° and 257.2 ° from four possible solutions.

The λ / 4 wave plate 921 is inserted in front of the analyzer 902 so that the phase of the linearly polarized light is advanced by π / 2 in the y-axis direction.

In this case, ψ is negatively checked and δ = 102.8 °.

That is, δ is determined to be a unique value.

According to equations (27) and (30), since a / b = 2.953 and ax / ay = 2.489, all parameters related to elliptical polarization are determined.

As described above, according to the embodiments shown in Figs. 11 to 14, the rotation angle of the analyzer corresponding to the maximum light intensity, the minimum light intensity, and the rotation of the analyzer corresponding to the maximum light intensity when the λ / 4 waveplate is inserted The angle is obtained, and the polarization state of the elliptical polarization is determined based on these four values.

Therefore, all the parameters necessary for determining the polarization state of the elliptical polarization can be determined quickly.

In addition, since the elliptical polarization to be measured is measured without conversion of the elliptical polarization to another polarization, all parameters for the elliptical polarization can be determined quickly and accurately.

A method of measuring the polarization state of the elliptical polarization according to another embodiment of the present invention will be described with reference to FIGS. 15 and 16.

In the embodiment shown in FIG. 15, reference numeral 1201 is an observation light source, 1202 is a condenser lens, 1203 is a first real image of a light source 1201 formed by condenser lens 1202, and 1204 is brightness. Breaker, 1205 is first illumination lens, 1206 is second illumination lens, 1207 is field stop, 1208 is visible field defined by field stop, 1209 is half mirror, 1210 is first and second illumination lens 1205, In the second image of the light source 1201 formed by 1206, 1211 denotes an objective lens and 1212 denotes a photomask that is an observation object.

The second real image of the light source 1201 corresponds to the collection point surface of the objective lens 1211.

Reference numeral 1213 denotes an actual image of the visible field 1208 formed by the second illumination lens 1206 and the objective lens 1211, and 1214 denotes an actual image of the photomask 1212.

The component arranged between the light source 1201 and the actual image 1214 serves to enlarge / observe the portion of the photomask 1212 to be inspected.

Since the configuration is the same as that of a general microscope for metal observation, illumination light and light rays for forming the actual image 1214 of the photomask are omitted.

Reference numeral 1215 is a test light source, 1216 is a polarizer, and 1217 is a half mirror.

The light of the light source 1215 is divided into a first linearly polarized beam 1218 and a second linearly polarized beam 1219 by the half mirror 1217.

Reference numeral 1220 denotes a total reflection mirror, 1221 denotes a lens, and 1222 denotes a plate.

The plate 1222 has a transmittance that can sufficiently transmit the light of the observation light source 1201.

The premature point of the lens 1221 coincides with the center of the brightness stop 1204.

Therefore, the first linearly polarized beam 1218 overlaps and condenses on the first chamber of the observation light source at the center of the brightness blocker 1204.

In addition, the first linearly polarized light 1218 is again focused on the second real image 1210 portion of the observation light source.

Since the portion of the second real image 1210 of the observation light source coincides with the focal point plane of the objective lens 1211, the first linearly polarized beam 1218 passes through the parallel beam guide photomask 1212 through the objective lens 1211. .

On the other hand, the optical path of the second linearly polarized beam 1219 is refracted by the mirrors 1223a-1223c and the light beam 1219 passes through the λ / 2 wave plate 1224.

The mirrors 1223a-1223c serve to adjust the optical path length difference between the first linearly polarized beam 1218 and the second linearly polarized beam 1219 to be equal to or smaller than the coherence length.

Further, when the second linearly polarized beam 1219 passes through the λ / 2 wave plate 1224, the polarization direction of the second linearly polarized beam 1219 becomes perpendicular to the polarization direction of the first linearly polarized beam 1218. .

The first and second linearly polarized beams 1218 and 1219 overlap each other by the half mirror 1215 to form elliptical polarization.

The polarization state of the elliptical polarization is observed by the polarization state observation device 1226.

In the inspection of the photomask, the transmittance and phase change amount of any portion are required only for the portion of the substrate which exhibits high transparency to the exposure light source.

Referring to the embodiment shown in FIG. 15, the polarization state observing apparatus 1226 may measure the polarization state of the elliptical polarization, wherein the elliptical polarization passes through only a portion of the substrate exhibiting high transparency with respect to the exposure light source. This occurs when the one linearly polarized beam 1218 overlaps the second linearly polarized beam 1219.

The measurements are set to ax and δ in equations (1) and (4).

Subsequently, the polarization state observing apparatus 1216 measures the polarization state of the elliptical polarization, which occurs when the first linearly polarized beam 1218 passing through the arbitrary pattern overlaps the second linearly polarized beam 1219. do.

The measurement is set to ax 'and δ in equations (5) and (7).

The target phase change amount, amplitude transmittance, and energy transmittance are obtained using equations (8), (9), and (10).

In the embodiment shown in FIG. 15, the deviation change amount, amplitude transmittance, and energy transmittance of each photomask are obtained only for the portion of the substrate exhibiting high transparency to the exposure light source.

The polarization state of the elliptical polarization generated by the superposition of the two linearly polarized beams can be measured without placing the photomask on the optical path of the first linearly polarized beam, that is, without the photomask 1212, and the measured value is expressed by Equation (1) and If (x) is set to ax and δ, the absolute value of the phase change amount, amplitude transmittance and energy transmittance of each mask part can be obtained.

In the embodiment shown in FIG. 15, the light emitted from the inspection light source 1215 is parallel light.

However, light emitted from a laser oscillated at the wavelength to be inspected, such as inspection light, may be used.

In contrast, parallel light is obtained by extracting 365 mm of light emitted from a lamp, that is, a mercury lamp, by a filter or a monochromator and aiming it by an aiming optical system.

When a very small area of the photomask 1212 is inspected, a relay optical system is installed between the inspection light source 1215 and the polarizer 1216 or between the polarizer 1216 and the half mirror 1217.

In this configuration, the cross sectional areas of the first and second linearly polarized beams 1218 and 1219 will be reduced.

FIG. 16 shows another embodiment of the present invention corresponding to the embodiment shown in FIG.

Referring to FIG. 16, reference numeral 1301 denotes an observation light source, 1302 a condenser lens, 1303 is a first image of a light source 1301 formed by a condenser lens 1302, 1304 is an erection breaker, and 1205 is a first light. Lens, 1306 is the second illumination lens, 1307 is the field stop, 1308 is the visible field defined by the field stop 1307, 1309 is the half mirror, 1310 is the first and second illumination lenses 1305, 1306 The second real image of the light source 1301 generated by 1311 denotes an objective lens and 1312 denotes a photomask to be observed.

The second real image 1310 of the observation light source coincides with the focal point plane of the objective lens 1311.

Reference numeral 1313 denotes an actual image of the visible field formed by the second illumination lens 1306 and the objective lens 1311.

Reference numeral 1314 denotes an actual image of the photomask 1312.

A component arranged between the actual image 1314 and the light source 1301 of the photomask 1312 serves as an apparatus for enlarging / observing the portion to be inspected of the photomask 1312.

Since the arrangement is the same as that of a general microscope for metal observation, the illumination rays and the rays for forming the real image 1314 of the photomask are omitted.

Reference numeral 1315 denotes an inspection light source, 1316 a polarizer, and 1317 a half mirror.

Light emitted from the light source 1315 is divided into a first linearly polarized beam 1318 and a second linearly polarized beam 1319 by the half mirror 1317.

Reference numeral 1320 denotes a lens and 1321 denotes a plate.

The plate 1321 has a sufficient transmittance so that a sufficient amount of light emitted from the light source 1301 is transmitted.

The focal point of the lens 1320 coincides with the center of the visible field 1308.

Therefore, the first linearly polarized beam 1318 is focused at the center of the field stop 1307 or at the center of the visible field 1308.

Since the actual image of the visible field 1308 is formed at the position of the photomask 1312 by the second illumination lens 1306 and the objective lens 1311, the first linearly polarized beam 1318 is focused on the photomask 1312. do.

Thus, a very small area of the photomask 1313 can also be inspected.

Reference numerals 1322, 1323, and 1316 denote first to third inspection lenses, and 1324 and 1325 denote mirrors.

The first inspection lens 1322 and the objective lens 1311 are arranged to be optically symmetric with respect to the plane including the photomask 1312, and the second inspection lens 1323, the second illumination lens 1306, and the mirror ( 1325 and plate 1321, third inspection lens 1326 and lens 1320 are also arranged to be optically symmetrical.

In this arrangement, the first linearly polarized beam 1318 passes through the third inspection lens 1326 and becomes parallel rays.

On the other hand, the optical path of the second linearly polarized beam 1319 is refracted by the mirrors 1327a and 1327b.

In addition, it rotates 90 degrees by the (lambda) / 2 wave plate 1328 in the polarization direction of the 2nd linear polarization beam 1319. As shown in FIG.

Reference numeral 1329 denotes a half mirror and 1330 denotes a polarization state observing apparatus.

The mirrors 1327a and 1327b are formed between the first and second linearly polarized beams 1318 and 1319 to be equal to or smaller than the interference length.

It controls the optical path length difference.

The first and second linearly polarized beams 1318 and 1319 overlap each other by the half mirror 1333 to form an elliptical polarization.

The polarization state of the elliptical polarization is observed by the polarization state observation device 1330.

The amount of phase change, amplitude transmittance and energy transmittance of each part of the photomask are determined in the same manner as in the embodiment in FIG.

In the embodiment shown in FIG. 16, the light emitted from the inspection light source 1315 is parallel light.

However, as such an inspection light source, light of a laser oscillating at the wavelength to be inspected may be used.

Alternatively, 365 nm light emitted from a lamp, or mercury lamp, may be extracted by a filter or a monochromator and aimed by a collimation optical system to obtain parallel light.

In the embodiments shown in FIGS. 15 and 16, chromatic aberration correction corresponding to the wavelength of the light of the observation light source is performed for each optical element through which only the light of the observation light source passes.

Chromatic aberration correction corresponding to the wavelength of linearly polarized light is performed for each optical element through which only linearly polarized light passes.

The chromatic aberration correction corresponding to the light of the observation light source and the wavelength of the linear polarization is performed for each optical element through which both the light of the observation light source and the linear polarization pass.

However, even if chromatic aberration correction corresponding to light of the observation light source and linearly polarized light is performed for all the optical elements, no particular problem occurs.

As described above, in the embodiments shown in FIGS. 15 and 16, the transmittance and the phase change amount of each photomask portion are easily and accurately obtained.

In addition, the present invention can be applied not only to a photomask but also to any object exhibiting transparency to light having a wavelength to be handled.

In addition, if the thickness of the object is known, the refractive index and absorption coefficient or extinction coefficient of the object can be obtained from the resulting amplitude transmittance and phase change amount.

Figure 17 illustrates another embodiment of the present invention with emphasis on the pattern involved.

Referring to FIG. 17, reference numeral 1401 denotes a light source, 1402 a polarizer for inputting light from the light source 1401 and converting it into linearly polarized light, 1403 is arranged on the emission side of the polarizer, and linearly polarizes the linearly polarized light of the polarizer 1402. A beam splitter for dividing the light propagating into the light beam and the 90 ° reflected beam, 1404 is a total reflection mirror for reflecting the 90 ° linearly polarized beam traveling in a straight line from the beam splitter 1403, and 1405 to 1408 are optical path adjustments Represents a total reflection mirror group constituting a device.

The total reflection mirror group reflects the linearly polarized beam 90 degree reflected by the beam splitter 1403, respectively, so that the linearly polarized beam is finally emitted in the same direction as the direction of the linearly polarized beam of the beam splitter 1403. do.

Reference numeral 1409 denotes an inspection object arranged on the emission side of the total reflection mirrors 1405-1408, a photomask, 1410 an objective lens arranged on the emission side of the photomask, 1411 a costume of the photomask 1409, and 1412 a total reflection mirror ( [Lambda] / 2 wavelength plate arranged at the emission side of 1404, 1413 is a half mirror, 1414 is a total reflection mirror, 1415 is an optical system having an equivalent position and arrangement with an objective lens 1410 when viewed from an image 1411, The virtual plane 1417, the position of which is equivalent to the photomask when viewed from the image 1411, represents an analyzer arranged in the optical path on the emission side of the half mirror 1413, respectively.

Phase 1411 is formed on the discharge side of the analyzer.

As the light source 1401 used in this embodiment, a light source that provides parallel light having a wavelength necessary for inspection may be used.

For example, if the inspection is done using the wavelength of the KrF excimer laser beam, KrF excimer laser may be used.

If the inspection is performed using a wavelength of 365 nm, a laser oscillating at 365 nm, or a mercury lamp, a monochromator, and a combination of sights may be used.

Parallel light emitted from the light source 1401 is converted into linearly polarized light by the polarizer 1402, and is divided into light rays traveling along two optical paths by the beam splitter 1403.

One linearly polarized beam passes through the optical path adjusting device constituted by the total reflection mirrors 1405-1408, and is then irradiated onto the photomask 1409 to form an image 1411 through the objective lens 1403.

On the other hand, the optical path of the linearly polarized beam that passes through the beam splitter 1403 is refracted by the total reflection mirror 1404 so that the polarization direction of the light beam is rotated 90 degrees through the λ / 2 wave plate 1412.

The linearly polarized beam then passes through the optical system 1415 and is then superimposed on the linearly polarized beam passing through the photomask by the total reflection mirror 1414 and the half mirror 1413.

For ease of explanation, it is assumed that one linearly polarized beam is irradiated to the photomask as the linearly polarized beam for inspection.

In addition, it is assumed that the other linearly polarized beam is a reference linearly polarized beam 1419.

If x and y are polarization directions of the linearly polarized beams 1418 and 1419, and delta x is a phase delay of the half mirror 1413, the linearly polarized beam 1418 is expressed as follows.

Ex = ax cos (ωt-δx)

When? y is the phase difference of the half mirror, the linearly polarized beam 1418 is expressed as follows.

Ey = ay cos (ωt-δy)

These equations correspond to equations (1) and (2).

The superimposed light is expressed as follows.

Where δ = δx-δy

These equations correspond to equations (3) and (4).

δx1 is called the phase delay of the linear polarization beam 1418 for inspection in the photomask 1409, δx2 is called the phase delay that occurs when light passes through the target portion (inspection portion) of the photomask, and δx3 is the photo delay The phase delay of the optical path extending to the upper surface after passing through the mask 1409 is assumed, and the beam splitter is assumed to be a starting point.

Let δy _{1 be} the phase delay of the reference linearly polarized beam 1418 in the virtual plane 1416, and δy _{3 be} the phase delay of the optical path extending from the virtual plane to the top plane.

In this case, the above formula corresponding to formula (4) is as follows.

δ = (δx ₁ + δx ₂ + δx ₃ )-(δy ₁ + δy ₂ ). … (33)

In this case, δx ₃ is a function of the position of the inspection linear polarization beam with respect to the photomask, is constant for the refracted light emitted at a given point in the photomask, and the reference passes through the corresponding point in the virtual plane 1416. Coincides with δy ₃ of the linearly polarized beam.

Therefore, equation (33) can be written as

δ = (δx ₁ + δx ₂ ) − δy ₁ . … (34)

In the embodiment shown in FIG. 17, a two-dimensional CCD camera is arranged in the image 1411. FIG.

While the analyzer 1417 is rotated, the output intensity of the CCD element that inputs the image of the relevant point of the photomask is monitored, and δ in equation (34) and deltax in equation (1) are Wolfgang Budde, Photo electric Analysis of Polarized Obtained by the method disclosed in Light, Applied Optics, No. 3, Vol. 1, May 1962.

More specifically, as shown in FIG. 18, the analyzer 1417 installed perpendicular to the optical axis rotates about 1 / n rotation x n times = 1 rotation (n is an integer) with respect to the optical axis, and the angle of the analyzer The output of the light intensity detector at the slope is read.

When the analyzer's direction set for the first rotation of the analyzer (i = 1, 2, 3 ……) is given (360 ° / n) × i = α ₁ (n = integer) with respect to the reference direction, the output of the light intensity detector Assuming I ₁ , K ₀ , K ₁ , and K ₃ are obtained.

In this case, the reference direction is assumed to be in a plane perpendicular to the optical axis.

As a result, δ and ax are expressed as follows.

Then, δ and ax are obtained by the following same procedure as described above in the absence of the photomask 1409.

As can be clearly seen from Eq. (34), the difference δ between the two values is a phase delay occurring when the linearly polarized beam passes through the photomask, and the ratio ax of the two values is the amplitude transmittance.

Energy transmittance is obtained by squaring this ratio.

Also, as can be clearly seen from equations (3) and (34), since the polarization state of the elliptical polarization is determined according to the optical characteristics of the pattern, an ellipse generated in a specific pattern, that is, the upper part of the pattern designed to give only phase inversion If the long axis direction of polarized light coincides with the direction of the analyzer 1417, an image is obtained as a specific pattern to be emphasized.

The optical path adjusting device composed of the total reflection mirrors 1405-1408 is driven by a high-precision displacement device using a piezoelectric element (not shown) for a specific pattern, i.e., a pattern designed to give only phase inversion, and δx in equation (34). Assume that ₁ , δ is adjusted to be set to nπ (n is an integer).

In this case, the polarization generated by the superposition on the upper portion of the specific pattern becomes linearly polarized light.

Therefore, when the linear polarization direction generated by the overlap coincides with the direction of the analyzer 1417, a more emphasized image of a specific pattern is obtained.

If the photomask 1409 is composed of a light shielding member, a substrate, and an object for retarding the phase of light with respect to light passing through the substrate portion, in equation (34), δx ₁ is δ for the substrate portion, where n is an integer. Is adjusted to be.

In this operation, the polarized light passing through the image in the substrate portion becomes linearly polarized light.

If the analyzer 1417 is installed in the direction perpendicular to the direction of linearly polarized light, an image of only an object for retarding the phase of the light with respect to the light passing through the substrate portion can be obtained.

19 shows another embodiment of the present invention.

Referring to FIG. 19, reference numeral 1501 denotes a light source, 1502 a polarizer arranged on the emission side of the light source, 1503 a beam splitter arranged on the emission side of the polarizer, and 1504 a polarizer by the beam splitter 1503. A total reflection mirror for reflecting a 90 degree linearly polarized light reflected from the light of (1502) by 90 degrees, 1505 is a photomask which is an inspection object arranged on the emission side of the total reflection mirror 1504, and 1506 is the emission side of the photomask The objective lens 1507 is an image of the photomask 1505, 1508 is a λ / 2 wave plate for inputting a linearly polarized beam traveling in a straight line through the beam splitter 1503, 1509 is a half mirror, and 1510 is total reflection. The optical system 1511 is an optical system whose position and arrangement are equivalent to the objective lens 1506 when viewed in the image 1507, the virtual plane whose position is equivalent to the photomask when viewed in the image 1507, and 1513 represent an analyzer, respectively.

As the light source 1501, a light source that provides parallel light having a wavelength required for inspection may be used.

For example, if the inspection is done using the wavelength of the KrF excimer laser beam, the excimer laser will be used.

If the inspection is performed using a 365 nm wavelength, a laser or mercury lamp oscillating at 365 nm, a monochromator and a sight may be used in combination.

Referring to FIG. 19, the parallel light emitted from the light source 1501 is converted into linearly polarized light by the polarizer 1502 and divided into linearly polarized beams traveling along two optical paths by the beam splitter 1503.

One linearly polarized beam is irradiated onto the photomask 1505 to form an image 1507 through the objective lens 1506.

The polarization direction of the linearly polarized beam traveling in a straight line through the beam splitter 1503 is rotated 90 degrees by the λ / 2 wave plate 1508.

The linearly polarized beam then passes through the optical system 1511 and overlaps the linearly polarized beam passing through the photomask 1505 by the total reflection mirror 1510 and the half mirror 1509.

In the embodiment shown in FIG. 19, a two-dimensional CCD camera is used as the correlation means, provided at the position of the image 1507. FIG.

The operation principle of this embodiment is the same as that of the embodiment shown in FIG. 17, but the polarization generated by the superposition at the upper portion corresponding to the specific pattern cannot generally be converted to linear polarization.

However, the embodiment shown in FIG. 19 is designed such that the optical path adjusting device in the embodiment shown in FIG. 17 is omitted.

For this reason, in this embodiment, since the phase change generated when the linear polarization beam passes through the photomask 1505 is ignored, there is no optical path difference between the linear polarization beams passing through the two optical paths and oscillating in the vertical direction.

Therefore, the restriction on the interference length emitted from the light source 1501 is very relaxed.

In the embodiment shown in FIG. 17 and FIG. 19, if the wavelength of the inspection light is closer to the ultraviolet region than the visible region, the plate is formed of a thin film metal film deposited such as aluminum or chromium, and at least the emitting surface is non-reflective. The members are used as beam splitters 1403 and 1503 and half mirrors 1413 and 1509.

Chromatic aberration correction is performed according to the waveguide of the linearly polarized light for each optical system through which the linearly polarized light passes.

As described above, according to the embodiments shown in FIGS. 17 and 19, the transmittance and the phase change amount of each photomask can be easily and accurately obtained.

In addition, since only a specific pattern can be formed as an image, or only an image having a specific highlighted pattern can be formed, the shape of the specific pattern can be inspected.

The present invention is generally applicable not only to photomasks but also to objects which exhibit transparency to light having a wavelength to be handled.

In addition, as long as the thickness of the object is a known value, the adjustment of each object and the absorption coefficient and extinction coefficient can be obtained as amplitude transmittance and phase change amount.

In an optical instrument such as the present invention, the light path has to be adjusted from time to time.

In this case, there is a need for a low cost device that does not impose a restriction on the light rays traveling along the optical path to be adjusted.

20 shows an embodiment suitable for solving such a problem.

This embodiment is basically designed such that the two plates are arranged symmetrically with respect to a plane perpendicular to the optical path to be adjusted.

When described in detail by the accompanying drawings of the present invention.

Referring to FIG. 2, reference numeral 1601 denotes an optical axis, 1602 a ray traveling along the optical axis, 1603 and 1604 are plates made of an optically anisotropic material exhibiting transparency or high transparency with respect to the wavelength of the ray 1602, respectively. Planes perpendicular to the optical axis 1601, 1606a-1606d, represent the multi-dielectric layers, respectively.

Plates 1603 and 1604 are arranged symmetrically about a plane 1605 perpendicular to the optical axis.

In this arrangement, the light rays 1602 deviate from the optical axis passing through one plate 1603, but the light rays 1602 travel along the original optical axis passing through the other plate 1604.

In other words, no deviation from the optical axis occurs.

In addition, since each of the plates 1603 and 1604 is made of an optically anisotropic material, the light beam 1662 may not be a linearly polarized beam.

While the plates 1603 and 1604 are symmetrical with respect to the plane 1605, the optical path is adjusted by adjusting the angle defined by the plates 1603, 1604 and the plane 1605 or the optical axis 1601. I can adjust it.

Surface reflected light, such as noise to be removed from the optics, is prevented by the multi-layers 1606a-1606d.

FIG. 21 shows a configuration designed to easily check the optical axis when the optical path is actually adjusted by the optical path adjusting device shown in FIG.

Like reference numerals in FIG. 21 denote like parts in FIG.

Reference numeral 1655 denotes a slide glass plate having a scattering flat surface, 1656 a mark like a graphic pattern formed on the surface of the slide glass plate 1655, 1657 a drawing rod, and 1658a and 1658b a drawing rod 1657. Fixing pieces 1659a and 1659b fixed to the fixation pins respectively.

The optical axis position checking operation, which is an important function in the embodiment shown in FIG. 21, is described as follows.

First, the drawing rod is pushed until the fixing piece 1658a contacts the fixing pin 1659a.

As a result, the slide glass plate 1655 is located in the optical path.

The position of the light rays 1602 on the slide glass plate 1655 is checked using the marks 1566 when there are no two plates 1603 and 1604.

Then, the plates 1603 and 1604 are provided to adjust the angles defined by the plates 1603 and 1604 and the long axis 1601 to obtain a predetermined optical path.

In this case, as described in the embodiment shown in FIG. 20, when the plates 1603 and 1604 are not symmetric with respect to the plane perpendicular to the optical axis 1601, the light rays on the slide glass plate 1655 The position of 1602 is displaced to the initial position.

If no displacement occurs, it indicates that the plates 1603 and 1604 are properly installed.

After checking the position of the light side, the drawing rod 1657 is pulled out until the fixing piece 1658b contacts the fixing pin 1659b.

The slide glass plate 1655 coupled with the drawing rod 1657 does not move out of the advertisement and interfere with the progress of the light rays 1602.

In the embodiment shown in FIG. 21, the slide glass plate 1655 is used as the optical axis position checking plate in the optical axis position detecting means.

However, a paper plate having a recessed protrusion on the surface, a metal plate or an object emitting scattering light can be used.

If the wavelength of the light ray 1602 is not the ultraviolet region, a material that emits fluorescence, that is, a paper plate containing a fluorescent dye or a metal plate coated with fluorescent material, ie, ethyl salicylate, may be used.

22 shows a modified embodiment of FIG.

The same reference numerals in FIG. 22 denote the same parts as in FIG.

An important function of this change is to electrically perform the optical axis position checking operation.

First, the two-dimensional position sensor 1665 is positioned in the optical path using a mechanism (not shown).

The position of the light beam 1602 on the two-dimensional position sensor 1665 is checked using a control circuit in the absence of the two plates 1603 and 1604.

The plates 1603 and 1604 are provided to adjust the angles defined by the plates 1603 and 1604 and the optical axis 1601 to obtain a predetermined pipeline.

In this case, as described in the embodiment shown in FIG. 20, if the plates 1603 and 1640 are not symmetrical with respect to the plane perpendicular to the optical axis 1601, the light rays 1602 on the two-dimensional position sensor 1655. The position of is shifted to the initial position.

The absence of this displacement indicates that the plates 1603 and 1604 are properly installed.

If the two-dimensional position sensor 1655 moves out of the optical path by the use of a mechanism (not shown) after checking the optical axis position, the two-dimensional position sensor 1665 does not interfere with the progress of the mine 1602.

Although the embodiment shown in FIG. 22 uses the two-dimensional position sensor 1665, any element capable of detecting the spot position of light electrically, such as a two-dimensional CCD element, may be used.

In the embodiment shown in Figs. 20 to 22, each plate has a multi-dielectric layer formed on the surface of the plate to prevent surface reflection.

However, a single dielectric layer may be formed in place of the multi dielectric layer.

As signed above, in each optical path adjusting device of the embodiment shown in FIGS. 20 to 22, two plates made of an optically anisotropic material exhibiting transparency or high transparency with respect to the wavelength of the incident light are connected to the optical axis of the incident light. It is installed symmetrically with respect to the vertical plane.

In this arrangement, no deviation of the optical axis occurs.

Since each plate is made of an optically anisotropic material, the light rays traveling along the optical axis may not be linearly polarized light.

In addition, each device has a simple configuration and can be obtained with low cost materials and low cost producers.

In addition, each device is allowed to check the position of the optical axis without disturbing the progress of the light beam.

Since the surface of each board is non-reflective, no noise is generated.

Figure 23 shows another embodiment of the present invention.

Reference numeral 1701 denotes a laser that emits unpolarized light, 1702 denotes a polarizer arranged on the emission side of the laser 1701, 1703 denotes a reduced optical system arranged on the emission side of the polarizer 1702, and 1704 denotes a reduced optical system 1703. Each of the beam splitters arranged on the emission side is shown.

The laser emitted from the laser 1701 is changed into linearly polarized light by the polarizer 1702.

The cross-sectional area of the linearly polarized light is reduced by the reduction optical system 1703 and divided by the beam splitter 1704 into beams traveling along two optical paths.

Reference numerals 1705 and 1706 denote split beams, 1707a and 1707b are total reflection mirrors arranged in one of the optical paths divided by the beam splitter 1704, and 1708 is arranged between the total reflection mirrors 1707a and 1707b. / 2 wavelength plate is shown, respectively.

The polarization plane of the beam 1705 is rotated by the λ / 2 wave plate 1708 in a direction perpendicular to the beam 1706.

Reference numeral 1709 denotes an inspection photo mask, 1710 denotes a stage supporting an inspection photo mask, and 1711 denotes a driving circuit of the stage 1710.

The photomask 1709 is moved in a plane perpendicular to the beam 1706 by the stage 1710 and the driving circuit 1711.

Reference numeral 1712 denotes a half mirror arranged at the emission side of the photomask 1709 so that the beams 1705 and 1706 overlap each other, and 1713 is a rotatable analyzer 1714 arranged at the emission side of the half mirror 1712. Light intensity detector 1715, arranged on the emitting side of analyzer 1713, represents a control computer for controlling the operation of drive circuit 1711, analyzer 1713 and light intensity detector 1714, respectively.

It is assumed that the polarization direction of the beam 1706 is the x direction, and the polarization direction of the beam 1705 passing through the wavelength plate 1708 of λ / 2 is the y direction.

In this case, when δx is the phase delay of the half mirror 1712, the beam 1706 is expressed as follows.

Ex = axcos (ωt-δx)

If the phase delay of the half mirror 1712 is delta y, the beam 1705 is expressed as follows.

Ey = aycos (ωt-Sy)

These equations correspond to equations (1) and (2). The light generated by the superposition is expressed as follows.

This equation corresponds to equation (3). At this time,

δ = δx−δy.

This equation corresponds to equation (4). In equation (3), the ax, ay, and δ values can be determined using the analyzer 1713 and the light intensity detector 1714.

When the δ and ax values obtained when the photomask 1709 is not provided on the stage 1710 by δ and ax, the amount of phase change θ when the beam 1706 passes through the photomask 1709 And energy transmittance (T) are given by

θ = δ-δ. … 40

T = (ax / ay). … (41)

Therefore, δ and ax are first found before the photomask 1709 is installed on the stage 1710, and then the photomask 1709 is installed on the stage 1710.

The control computer 1715 outputs a command signal to the driving circuit 1711 to control the stage 1710 so that the measurement start portion of the photomask 1709 overlaps the beam 1706, and then moves the stage 1710 to a corresponding position. To stop.

Δ and ax are then calculated according to equations (40) and (41).

The control computer 1715 then outputs a command signal to the drive circuit 1711 to control the stage 1710 so that the next measurement portion of the photomask 1709 overlaps the beam 1706 and then stages the stage 1710. Stop at the corresponding position.

Then, δ and ax are obtained, and θ and T are calculated according to equations (40) and (41).

By repeating the above process, the transmittance distribution and the phase change amount distribution of the photomask 1709 are obtained.

The defects caused by the factors in manufacturing the photomask 1709 can be seen by comparing the distribution actually obtained in the present invention with the transmittance distribution in design of the photomask 1709 and the amount of phase change in design. It is assumed that a predetermined area is set for the transmittance or the amount of phase change, and only a portion representing the measured value corresponding to that area is displayed. In this case, only patterns having an associated phase or transmittance are made into an image.

FIG. 24 shows the modified embodiment of FIG. Referring to FIG. 24, reference numeral 1801 denotes a laser emitting unpolarized light, 1802 a polarizer, 1803 a reduced optical system, and 1804 a beam splitter.

The laser beam emitted from the laser 1801 is converted into linearly polarized light by the polarizer 1802.

The cross-sectional area of the linearly polarized light is reduced by the reduced optical system 1803 and then divided by the beam splitter 1804 into beams traveling along two optical paths.

Reference numerals 1805 and 1806 denote split beams, 1807a and 1807b reflect total reflection mirrors, and 1808 denotes lambda / 2 wavelengths only.

The polarization plane of the beam 1805 is rotated in a direction perpendicular to the beam 1806 by the λ / 2 wave plate 1808.

Reference numeral 1809 denotes an inspection photo mask, 1810 a stage, and 1811 a stage 1810 driving circuit, respectively.

The photomask 1809 is moved in a plane perpendicular to the beam 1806 by the stage 1810 and the drive circuit 1811.

Reference numeral 1812 denotes a half mirror for superimposing beams 1805 and 1806, 1813 a rotatable polarizer, 1814 an optical intensity detector, 1815 a control computer, and 1816 an optical path control device.

The polarization direction of the beam 1806 is assumed to be the x direction, and the polarization direction of the beam 1805 passing through the λ / 2 wavelength plate 1808 is assumed to be the y direction.

In this case, the two beams 1805 and 1806 are represented in the same manner as in the embodiment shown in FIG.

Stage 1810 is adjusted to move to a position where the relative pattern of photomask 1809 overlaps beam 1806 and then stops at that position.

The phase delay δ y of the beam 1805 in the half mirror 1812 is adjusted by the optical path adjusting device such that δ is set to 0 or 2π in equation (4).

In this case, as can be clearly seen in equation (3), the polarization generated when the beams 1805 and 1806 overlap each other by the half mirror 1812 becomes linearly polarized light.

The analyzer 1813 is fixed in a direction perpendicular to the polarization direction of the linearly polarized light generated by the superposition.

The stage 1810 then moves continuously.

That is, the beam 1806 scans the photomask 1809 from the perspective of the photomask 1809.

When a suitably formed relationship pattern reaches the position of the beam 1806, light is not incident on the light intensity detector 1814.

For this reason, the output of the light intensity detector 1814 is zero.

When an improperly formed relationship pattern or a pattern other than the relationship pattern reaches the position of the beam 1806, light is incident on the light intensity detector 1814.

Therefore, if the portion corresponding to the output 0 from the light intensity detector 1814 is displayed, only a properly formed relation pattern is made into the image.

The defect can be detected by comparing the shape of the relationship pattern actually formed with the shape of the relationship pattern at design time.

25 illustrates another embodiment of the present invention.

Referring to FIG. 25, reference numeral 1901 denotes a laser emitting polarized light, 1902 an zoom optical system, 1903 a beam shape system, and 1904 a beam splitter. The laser beam, which is linearly polarized light emitted from the laser 1901, becomes a beam having a cross-sectional area enlarged by the zoom optical system 1902.

The beam is made into a beam having a spherical cross-sectional shape by the beam shape system 1903.

This beam is divided into beams traveling along two optical paths by the beam splitter 1904.

Reference numerals 1905 and 1906 denote split beams, 1907a and 1907b total reflection mirrors, 1908 are λ / 2 wavelength plates for rotating the polarized light plane of the beam 1905 in a direction perpendicular to the beam 1906, and 1909 an inspection photo. The mask 1910 represents a stage and 1911 represents a stage 1910 driving circuit, respectively.

The photomask 1909 is moved in a plane perpendicular to the beam 1906 by the drive circuit 1911.

Reference numeral 1912 denotes a half mirror for superimposing beams 1905 and 1906, 1913 a rotatable analyzer, 1914 a light intensity detector, and 1916 a control computer.

It is assumed that the polarization direction of the beam 1906 is the x direction, and the polarization direction of the beam 1905 passing through the λ / 2 wavelength plate 1908 is the y direction.

In this case, the two beams 1905 and 1906 are represented in the same way as in the embodiment shown in FIG.

Therefore, δ and ax are first found before the photomask 1909 is installed on the stage 1910.

In this case, since ax depends on the cross-sectional area of the beam 1906, ax is stored in the control computer 1915 in an amount per unit cross-sectional area.

Photomask 1909 is then installed on stage 1910.

The control computer 1915 outputs a command signal to the driving circuit 1911 to control the stage 1910 so that the measurement start portion of the photomask 1909 overlaps the beam 1906, and the stage 1910 at a corresponding position. To stop.

At this time, since the shape of the relationship pattern superimposed on the beam 1906 can be known by the control computer 1915, the beam shape system 1901 is controlled by the control computer 1915 to have the same cross-sectional shape as that of the relationship pattern. To make a beam.

Then δ and ax are obtained.

The quantity per unit area is calculated for ax.

The values of θ and T are calculated based on these values according to equations (40) and (41).

The control computer 1915 then outputs a command signal to the drive circuit 1911 to control the stage 1910 so that the next measurement portion of the photomask 1909 overlaps the beam 1906 and the stage at the corresponding portion. Stop 1910.

The beam shape system 1903 is controlled such that the relationship pattern superimposed on the beam 1906 matches the cross-sectional shape of the beam 1906.

Then δ and ax are obtained. The amount per unit cross-sectional area is calculated for ax.

θ and T values are calculated according to equations (40) and (41).

By repeating the above process, the transmittance distribution and the phase change amount distribution of each photomask 1909 are obtained.

By comparing the transmittance distribution in the design of the photomask and the distribution of phase change in the design with the actual distribution obtained in the present invention, defects caused by the elements in the manufacturing of the photomask 1909 can be found.

It is assumed that a predetermined area is set for the transmittance or the amount of phase change, and only a portion representing the measured value corresponding to the area is displayed.

In this case, only patterns having a relevant phase or transmittance are made into the image.

In the embodiment shown in FIG. 25, the beam system produces a beam having a spherical cross section.

If the relationship pattern overlapping the beam 1906 is not spherical, the relationship pattern will be divided into spherical portions.

Alternatively, a beam-shaped system can be designed such that cells such as liquid crystal cells are arranged in a two-dimensional plane and the transmittance of each cell can be changed in accordance with an external signal.

With this beam shape system, the beam 1906 can be made such that the cross-sectional shape of the beam 1906 matches the shape of the relationship pattern.

As described above, according to the embodiment shown in FIGS. 23 to 25, the transmittance and the amount of phase change of each photomask portion are easily and accurately obtained.

Further, since only a specific pattern is made of an image, or only an image having a highlighted specific pattern is made, the refractive index and absorption coefficient or extinction coefficient of the object can be obtained from the obtained amplitude transmittance and phase change amount.

Claims (29)

Two linearly polarized beams 105, 106, (408, 409, 410), (512, 513), (931, 934), (218,) having orthogonal polarization directions and passing through two different optical paths. 219), (318, 319), (418, 419), (1705, 1706), (1805, 1806), (1905, 1906), and polarization state of the elliptical polarization caused by the superposition of one photomask Target parts 109, 209, 411, 412, 508, 1212, 1312, 1409, 1505, 1709, 1809, and 1909 are two straight lines A photomask inspection method comprising: inspecting a photomask based on a difference in polarization states of elliptical polarizations generated when two linearly polarized beams overlap each other after being installed in an optical path of one beam of polarized beams. The photomask inspection according to claim 1, wherein the optical path of the linearly polarized beam having the photomask is divided into a plurality of optical path portions, and the inspection target portions of the photomasks are located at each optical path portion so as to be simultaneously examined. Way. The photomask of claim 1, wherein the optical path of the linearly polarized beam provided with the photomask is divided into a plurality of optical path portions, and the inspection target portions of one photomask are positioned at each optical path portion so as to be simultaneously inspected. Inspection method. Steps provided on two linearly polarized beams 512 and 513 having polarization directions perpendicular to the photomasks 507 and 508 and passing through different optical paths, and overlapping of the two linearly polarized beams passing through each photomask. And a step of observing the polarization state of the polarization generated by the step, and comparing the two masks to inspect the mask. While rotating the analyzer 852 about the optical axis of the elliptical polarization 855, an elliptical polarization 855 to be measured is incident on the analyzer 852, corresponding to the maximum light intensity and the maximum light intensity of the polarization 856 emitted from the analyzer. A step of obtaining a rotation angle of the analyzer and a minimum light intensity of polarization, and a? / 4 wavelength plate 851 is provided on the incidence side of the analyzer, and the elliptical polarization is sequentially performed with the? / 4 wavelength plate 851 Rotating the analyzer with respect to the optical axis of the elliptical polarization while being incident on the analyzer 852 to obtain a rotation angle of the analyzer when the light intensity of the polarized light emitted from the analyzer is maximum, and the maximum light intensity and the maximum light intensity. Determining the polarization state of the ellipse polarization based on four values of the corresponding rotation angle of the analyzer, the minimum light intensity, and the analyzer's rotation angle corresponding to the maximum light intensity when the λ / 4 wavelength plate is inserted. Ellipsometry measurement method comprising the tab. The linearly polarized beam is divided into a first linearly polarized beam having a first polarization direction and a second linearly polarized beam having a second polarization direction different from the first polarization direction, and the first and second portions are applied to the target pattern of the target photomask. One of the linearly polarized beams is irradiated, and one linearly polarized beam that has passed through the target photomask is superimposed on another linearly polarized beam that does not pass through the target photomask to generate synthetic light. In the method for obtaining a photomask inspection, a step of moving a target photomask to stop at predetermined positions 1710, 1711, 1715, 1810, 1811, 1815, 1911, 1910, and 1915, and the amount of phase change from synthetic light And measuring the energy transmittance repeatedly. 7. The method of claim 6, wherein by adjusting the optical path of one of the first and second linearly polarized beams, the composite light is converted to linearly polarized light and the object photomask is continuously moved while the analyzer is held at a particular angle. Photomask inspection method. The method of claim 6, wherein the cross-sectional area of the linearly polarized beam is reduced or enlarged to form the same shape as the predetermined cross-sectional shape or the target pattern, and the photomask is moved to stop at a predetermined position, and then one of the two linearly polarized beams is removed. A photomask inspection method characterized by irradiating a target photomask pattern of a photomask. Means for generating a linearly polarized beam having a first polarization state as a reference ray, means for generating at least one linearly polarized beam having a second polarization state as the inspection ray, and means for superimposing one of the inspection rays on the reference ray And means for measuring the polarization state of the polarization generated by the superposition, wherein the detection beam and the reference beam have the same wavelength and the same wavelength distribution, and the polarization directions of the detection beam and the reference beam are perpendicular to each other, and a narrow band. And the wavelength distribution of is wide enough to have an interference length larger than the optical path difference between the optical paths through which the detection and reference light passes. Separation means for dividing light into light rays passing through the first and second optical paths, first polarization means for polarizing the light rays passing through the first optical path with a first linearly polarized beam, and a first polarized beam Second polarization means for polarizing the light beam in the direction to convert the light beam passing through the second optical path into a second linearly polarized beam, and photosynthetic means for superimposing the light beam passing through the first optical path on the light beam passing through the second optical path; Means for measuring a polarization state of light generated by superposition by said photosynthetic means, while a measurement object is installed in an optical path between one of said first and second polarizing means and said synthesizing means; A magnification / observation means for magnification / observation, the magnification / observation means having an observation light source and an optical system for irradiating observation light from an observation garden on an object, and an optical path and an object of the observation light source installed The photomask inspection apparatus characterized in that the optical path portion match. 11. An apparatus according to claim 10, wherein one of the first and second optical paths comprises phase adjusting means for changing the phase of the linearly polarized light passing through the optical path. The photomask inspection apparatus according to claim 10, wherein the optical path provided with the object is constituted by an optical system designed to aim linearly polarized light passing through the object's object position. 13. The apparatus of claim 12, wherein the optical system is designed such that linearly polarized light passing through the optical system is focused at a target position. 11. The photoconductor of claim 10, wherein the first and second optical paths comprise means for adjusting the optical distance therebetween such that the optical distance difference is not greater than the coherence length between the first and second linearly polarized beams. Mask inspection device. The optical element constituting the optical system at a portion where the optical path portion of the observation light coincides with the optical path portion on which the object is installed is subjected to chromatic aberration correction with respect to the wavelength of the observation light and linearly polarized light. Mask inspection device. A light source for emitting parallel linearly polarized light, a first half mirror for dividing the linearly polarized light emitted from the light source into an inspection light beam and a reference light beam having a polarization direction perpendicular to each other, an inspection light optical system through which the inspection light passes; A photomask inspection apparatus comprising: a reference optical system through which reference light passes, and a second half mirror and light detection means for synthesizing the inspection light and the reference light, wherein the inspection light optical system inspects the inspection light. And an optical system for forming an image of the inspection light passing through the photomask, wherein the reference optical system shares a portion of the optical axis with the image optical system, and has an upper portion of the optical system. An optical system equivalent to the image optical system, installed at the same position as that of the image optical system on the optical axis, The light detecting means includes an optical detector arranged above the image optical system and an analyzer arranged between the second half mirror and the optical detector, and the optical distance difference between the image inspection optical system and the reference optical system is A photomask inspection apparatus, characterized in that it is smaller than the interference length of linearly polarized light. The photomask inspection apparatus according to claim 16, wherein each of the photodetectors is a device obtained by two-dimensionally arranging conversion elements designed to convert the intensity of incident light into an electrical signal level. 17. The photoconductor according to claim 16, wherein the linearly polarized light is ultraviolet light, and the first and second half mirrors each have a thin metal film formed on a surface thereof, the plate having at least an antireflective emission surface. Mask inspection device. 17. An apparatus according to claim 16, further comprising means for adjusting the optical distance difference between the inspection light optical system and the reference light optical system. And an optically anisotropic material substantially transparent with respect to the wavelength of incident light, and comprising two plates arranged symmetrically with respect to a plane perpendicular to the optical axis of the incident light. 21. An optical path adjusting apparatus according to claim 20, wherein each plate has a thin film dielectric film formed on its surface. 21. The apparatus of claim 20, further comprising means for detecting the position of the optical axis. The optical path adjusting device according to claim 22, wherein the optical axis position detecting means has a scattering plane when the wavelength of incident light corresponds to a visible region. The optical path illumination device according to claim 22, wherein when the wavelength of incident light corresponds to an ultraviolet region, the optical axis position detecting means is a material having a plane of fluorescence having a mark. The optical path adjusting device according to claim 22, wherein the optical axis position detecting means is a two-dimensional photoelectric conversion element arranged in an optical path. 20. An apparatus according to claim 19, wherein said optical axis position detecting means is a mechanism capable of withdrawing from an optical path. Means for emitting a linearly polarized beam, means for reducing the cross-sectional area of the beam, means for dividing the reduced beam into two beams, and for rotating the polarization direction of one of the divided beams by 90 °. Means, and an inspection photomask installed on one of the beams in the two polarization directions, the stage being moved into a plane perpendicular to the beam without disturbing the propagation of the beam, the beam having a 90 ° polarization direction rotated and the polarization direction not being rotated. Means for synthesizing a beam, a rotatable analyzer and a light intensity detector arranged in an optical path of the synthesized light. 28. The photomask inspection apparatus according to claim 27, wherein the means for reducing the cross-sectional area of the linearly polarized beam is a means for reducing or expanding the cross-sectional area of the linearly polarized beam and making the cross-sectional area of the linearly polarized beam into a predetermined shape. 29. The device according to claim 28, wherein the means for making the cross section of the beam into a predetermined shape is constituted by cells arranged in two dimensions, and each cell is made of a material having a transmittance which can be controlled by an external signal. Photomask inspection apparatus.