JP2022092537A - Information processor, exposure apparatus, and method for manufacturing goods - Google Patents

Abstract

To provide an information processor capable of efficiently predicting temporal changes of imaging characteristics of a projection optical system disposed in an exposure apparatus.SOLUTION: An information processor of the present invention predicts temporal changes of imaging characteristics of a projection optical system disposed in an exposure apparatus by using a learning model acquired by machine learning.SELECTED DRAWING: Figure 7

Description

The present invention relates to an information processing apparatus, an exposure apparatus, and a method for manufacturing an article.

Conventionally, it has been required to correct fluctuations in imaging characteristics due to absorption of exposure energy in a projection optical system provided in an exposure apparatus.
Regarding changes in the imaging characteristics of the projection optical system, the change in the exposure energy density distribution in the projection optical system according to the exposure conditions such as the type of circuit pattern formed on the original plate and the effective light source distribution of the illumination light. It is also known that it will be different with it.

Patent Document 1 discloses a method of performing correction by calculating a correction coefficient of a prediction model for predicting a time change of an imaging characteristic of a projection optical system according to such a difference in exposure conditions.

Japanese Unexamined Patent Publication No. 2009-32875

In recent years, in exposure equipment, the exposure energy in the projection optical system is due to the frequent replacement of the original plate in response to the increase in the types of original plates on which the circuit pattern of the semiconductor device is formed due to the wide variety of products such as semiconductor devices. It is known that the density distribution of optics also changes in various ways.
Then, in order to deal with such a situation, it takes a lot of time to calculate the correction coefficient of the prediction model for each of a huge number of exposure conditions by using the method disclosed in Patent Document 1. turn into.

Therefore, an object of the present invention is to provide an information processing apparatus capable of efficiently predicting time changes in the imaging characteristics of the projection optical system provided in the exposure apparatus.

The information processing apparatus according to the present invention is characterized in that the time change of the imaging characteristics of the projection optical system provided in the exposure apparatus is predicted by using the learning model acquired by machine learning.

INDUSTRIAL APPLICABILITY According to the present invention, it is possible to provide an information processing apparatus capable of efficiently predicting a time change of an imaging characteristic of a projection optical system provided in an exposure apparatus.

The schematic diagram of the exposure apparatus provided with the information processing apparatus which concerns on 1st Embodiment. The figure which showed the example of the time change of the imaging characteristic of the projection optical system with exposure. The figure which showed the example of the circuit pattern drawn on the reticle. The figure which showed the example of the effective light source distribution formed by the illumination optical system. The figure which showed the characteristic of the density distribution of the exposure energy in the projection optical system by X dipole illumination. The figure which showed the characteristic of the density distribution of the exposure energy in the projection optical system by Y dipole illumination. The flowchart of the process performed in the information processing apparatus which concerns on this embodiment. The schematic diagram which shows the state of calculating the correlation degree in the information processing apparatus which concerns on this embodiment. The block diagram of the exposure processing system provided with the information processing apparatus which concerns on 2nd Embodiment.

Hereinafter, the information processing apparatus according to this embodiment will be described in detail with reference to the attached drawings. The drawings shown below may be drawn at a scale different from the actual one so that the present embodiment can be easily understood.

Further, in the following description, it is assumed that the imaging characteristic includes at least one of magnification, focus, distortion, astigmatism, spherical aberration, coma aberration and wavefront aberration. Here, the wavefront aberration is expressed as each term developed by the Zernike polynomial of the wavefront shape.
In the following, the above-mentioned imaging characteristics may be collectively referred to as aberration.

In the following, the direction perpendicular to the substrate mounting surface of the substrate stage is the optical axis direction (Z direction), and the direction in which the substrate mounted on the substrate mounting surface is scanned is the scanning direction (Y direction), and the substrate. The direction perpendicular to the scanning direction in the plane parallel to the mounting surface is referred to as the non-scanning direction (X direction).

[First Embodiment]
Conventionally, in the manufacturing process of semiconductor elements such as LSIs and superLSIs, an exposure apparatus that performs baking formation by projecting and exposing a circuit pattern drawn on a reticle (mask) onto a substrate (wafer) coated with a photosensitive agent has been used. It is used.

In such an exposure apparatus, the optical characteristics such as the refractive index change in temperature due to the heat generated by the absorption of a part of the energy of the exposure light in the projection optical system due to the repetition of the projection exposure. It has been known.
If the projection optical system is continuously irradiated with the exposure light for a long time, the imaging characteristics of the projection optical system will fluctuate, and it will be ignored for the line width resolution of the circuit pattern and the alignment accuracy for accurately superimposing the patterns over a plurality of steps. There is a risk that an unreasonable amount of deviation will occur.

Therefore, a method for correcting fluctuations in imaging characteristics according to the energy of the exposure light applied to the projection optical system has been proposed.

For example, such fluctuations in imaging characteristics (aberration) are predicted using a prediction model with exposure amount, exposure time, non-exposure time, etc. as variables, and based on the prediction results, the field of the substrate stage or projection optical system An exposure device that performs correction by moving a lens is known.
Then, by predicting the aberrations included in the imaging characteristics of the projection optical system, for example, the fluctuations of magnification, focus, distortion, astigmatism, and wavefront aberration by using a prediction model, the fluctuations of each imaging characteristic are corrected. be able to.

Further, conventionally, when the fluctuation amount of the image formation characteristics of the projection optical system deviates from the prediction, a method of changing the correction coefficient of the prediction model by using the result of directly measuring the fluctuation amount of the image formation characteristics is known.
Further, when a plurality of reticles having different circuit patterns are used, the deviation of the predicted value of the fluctuation amount of the image formation characteristics of the projection optical system due to the difference in the circuit pattern is acquired for each reticle. A method of making a prediction using the correction coefficient of is known.

In addition, by calculating the correction coefficient of the prediction model for predicting the fluctuation of the imaging characteristics of the projection optical system according to the exposure conditions such as the type of circuit pattern formed on the reticle and the effective light source distribution of the illumination light. A method of making corrections is known.

In the lithography process included in the manufacturing process of semiconductor devices and the like, it is required to control the amount of fluctuation of the imaging characteristics due to the exposure energy in the projection optical system within an allowable range.
In recent years, with the increasing variety of products such as semiconductor devices, the types of reticles on which circuit patterns of semiconductor devices are formed have increased, and in the lithography process, exposure is performed while frequently exchanging reticles for each product lot. It is done.

Therefore, the operator of the exposure apparatus creates control parameters for setting the exposure conditions in the exposure apparatus according to the type of the reticle.
Such control parameters include, for example, a reticle ID for identifying a reticle, an illumination mode ID for identifying an illumination light setting, an exposure region to which exposure light is applied in a projection optical system, an exposure amount, and the like. ..

In any of the methods using the prediction model for predicting the fluctuation of the imaging characteristics of the projection optical system shown above, it is necessary to calculate the correction coefficient for each exposure condition.
Therefore, it takes a lot of time to obtain the correction coefficient for each of the reticle that has been diversified in recent years by using such a conventional method.

Therefore, it is an object of the present embodiment to provide an information processing apparatus capable of efficiently predicting fluctuations in the imaging characteristics of the projection optical system for each exposure condition by adopting the following configuration. ..

FIG. 1 shows a schematic view of an exposure apparatus 1 including an information processing apparatus 20 according to the first embodiment.

The exposure apparatus 1 is a lithography apparatus that exposes a substrate so as to transfer a pattern formed on an original plate (reticle, mask) to a substrate (wafer) by a step-and-scan method.
The exposure device 1 includes an information processing device 20, a laser light source 101, a laser light source control device 102, a main control device 103, an illumination optical system 104, and an illumination optical system control device 108 according to the present embodiment.
Further, the exposure device 1 includes a projection optical system 114, a wafer stage 116, a wafer stage control device 120, a reticle stage 123, a reticle stage control device 124, and a projection optical system control device 129.

A gas such as KrF is sealed in the laser light source 101, and laser light having a far ultraviolet region, for example, a wavelength of 248 nm is emitted.
The laser light source control device 102 controls the gas exchange operation of the laser light source 101, the control for wavelength stabilization, the control of the discharge applied voltage, and the like.
The main control device 103 is connected to the exposure device 1 by an interface cable, and generates an instruction to control the entire exposure device 1.

The illumination optical system 104 guides the beam emitted from the laser light source 101 to the reticle 113 mounted on the reticle stage 123.
Specifically, the beam emitted from the laser light source 101 is shaped into a predetermined shape (circular shape, annular shape, quadrupole shape, double pole shape, etc.) via the beam shaping optical system 126, and is an integrator. A secondary light source is formed by incident on the lens 105.
The luminous flux from the secondary light source is directed by the condenser lens 107 having a function of changing the illuminance distribution of the reticle 113, so that the variable slit 110 is illuminated by Koehler illumination.
The variable slit 110 has a mechanism that can change the aperture width, and by controlling the aperture width, the intensity distribution (illuminance distribution) of the slit-shaped light (exposure light) in the non-scanning direction is made uniform. There is.

An aperture diaphragm 106 having a substantially circular aperture is provided between the integrator lens 105 and the condenser lens 107.
Then, by controlling the diameter of the opening of the aperture stop 106 by the illumination optical system control device 108, the numerical aperture (NA) of the illumination optical system 104 can be set to a desired value.
Further, since the ratio of the numerical aperture of the illumination optical system 104 to the numerical aperture of the projection optical system 114 is the coherence factor (σ value), the illumination optical system control device 108 controls the aperture throttle 106 of the illumination optical system 104. Then, the σ value can be set.

A half mirror 111 is arranged on the optical path of the illumination optical system 104, specifically between the condenser lens 107 and the variable slit 110, and a part of the exposure light illuminating the reticle 113 is reflected by the half mirror 111. It is taken out by.
A photo sensor 109 is arranged on the optical path of the reflected light from the half mirror 111, and the photo sensor 109 generates an output corresponding to the intensity (exposure energy) of the exposure light.
The output of the photosensor 109 is converted into exposure energy per pulse by an integration circuit (not shown) that integrates each pulse emission of the laser light source 101, and is input to the main control device 103 via the illumination optical system control device 108. Will be done.

A circuit pattern of a semiconductor element for printing is formed on the reticle 113, and the reticle 113 is illuminated by the illumination optical system 104.
A variable blade 112 composed of a light-shielding member that can move in a two-dimensional direction, that is, in an XY plane perpendicular to the optical axis direction, is provided between the variable slit 110 and the reticle stage 123, and is provided on the pattern surface of the reticle 113. The irradiation area is arbitrarily set.

The projection optical system 114 guides the exposure light that has passed through the reticle 113 to the wafer 115 mounted on the wafer stage 116, so that the exposure light is applied to one shot region on the wafer 115 coated with the photoresist on the reticle 113. The pattern of is arranged so as to be imaged and projected.
When a part of the pattern on the reticle 113 is reduced and exposed on the wafer 115 at a reduction magnification β (for example, 1/4), the reticle stage 123 and the wafer stage 116 are the same as the reduction magnification β with respect to the slit-shaped light. Scan in opposite directions to each other in the scanning direction at the speed ratio of.
Then, by repeating the multi-pulse exposure by the pulse emission from the laser light source 101, the pattern formed on the entire surface of the reticle 113 can be transferred to the shot region on the wafer 115.

The projection optical system 114 is provided with a field lens 127. The field lens 127 is held by the lens barrel 130, and the lens barrel 130, that is, the field lens 127 can be moved in the optical axis direction by a drive mechanism 128 composed of pneumatics, a piezoelectric element, and the like.
The projection optical system control device 129 controls the position of the field lens 127 on the optical axis to suppress the decrease of various aberrations of the projection optical system 114, improve the projection magnification, and reduce the distortion error. ..

The wafer stage 116 can move in the X and Y directions orthogonal to each other in the optical axis direction (Z direction) of the projection optical system 114 and in the plane perpendicular to the Z direction while holding the wafer 115.
The laser interferometer 118 can measure the position of the wafer stage 116 in the XY plane by measuring the distance between the moving mirror 117 fixed to the wafer stage 116.
The wafer stage control device 120 moves the wafer stage 116 to a predetermined position in the XY plane by controlling the drive mechanism 119 such as a motor based on the position of the wafer stage 116 measured by the laser interferometer 118.

In the exposure apparatus 1, the reticle 113 and the wafer 115 are positioned so as to have a predetermined positional relationship with each other.
Then, each control for scan exposure is performed by the laser light source control device 102, the wafer stage control device 120, and the reticle stage control device 124 based on the synchronization signal from the main control device 103.
As a result, the circuit pattern formed on the entire surface of the reticle 113 is transferred to the chip region of the wafer 115.

Then, after the wafer 115 is moved by a predetermined amount in the XY plane by the wafer stage 116 based on the step-and-scan method, the pattern formed on the reticle 113 is similarly projected and exposed on the other chip regions of the wafer 115. The operations to be performed are sequentially performed.

The information processing device 20 according to the present embodiment is a computer as an information processing means, and learning that it is possible to efficiently predict fluctuations in the imaging characteristics of the projection optical system 114 according to the exposure conditions in the exposure device 1. You can run the program.
In the exposure device 1, the information processing device 20 is configured to be able to transmit and receive data related to the exposure device 1 via a main control device 103 connected by an interface cable.
The information processing apparatus 20 can control the amount of fluctuation in the imaging characteristics of the projection optical system 114 within an allowable range for each exposure condition of the exposure apparatus 1.

Next, a time-dependent prediction model of the imaging characteristics of the projection optical system 114 used in the information processing apparatus 20 according to the present embodiment will be described.

FIG. 2 shows an example of the time change of the imaging characteristic at a predetermined image height of the projection optical system 114 caused by the exposure by the exposure apparatus 1, that is, the amount of aberration (aberration amount) F.
The aberrations referred to here include, for example, magnification, focus, distortion, astigmatism, spherical aberration, coma aberration, wavefront aberration and the like.
Further, at time t ₀ , it is assumed that the aberration amount F is the initial aberration amount F ₀ , and the difference between the aberration amount F and the initial aberration amount F ₀ , that is, FF ₀ is defined as the aberration fluctuation amount ΔF. The aberration fluctuation amount ΔF takes a different value for each image height.

As shown in FIG. 2, when the exposure operation is started by the exposure light emitted from the laser light source 101 passing through the projection optical system 114 at time _t0 , the projection optical system 114 has thermal energy from the exposure light. The amount of aberration F fluctuates as time t elapses by absorbing the above.
Then, at the time t ₁ when a predetermined time has elapsed from the time t ₀ , the aberration amount F reaches the aberration amount (hereinafter, referred to as the maximum fluctuation aberration amount) F _m in the maximum fluctuation.

After that, even if the exposure light passes through the projection optical system 114, the thermal energy absorbed by the projection optical system 114 and the thermal energy emitted by the projection optical system 114 reach an equilibrium state with each other, so that the aberration amount F is maximum. There is almost no change from the amount of fluctuation aberration F _m .
Then, when the exposure operation is stopped at _time t2, the thermal energy absorbed by the projection optical system 114 is released, so that the aberration amount F fluctuates so as to approach the initial aberration amount _F0 again as time t elapses. At time t3 _, the aberration amount F reaches the initial aberration amount F ₀ .

Here, when the aberration amount at a predetermined time is F _k , the aberration amount F _{k + 1} after the exposure operation for the time Δt is calculated by the following equation using the maximum fluctuation aberration amount F _m and the time constant _Th . It is approximated as in 1).
F _{k + 1} = F _m- (F _m -F _k ) x exp ( _-Δt / Th) ... (1)
Similarly, when the aberration amount at a predetermined time is F _k , the aberration amount F _{k + 1} after the exposure operation is not performed for the time Δt is calculated by the following equation using the initial aberration amount F ₀ and the time constant T _c . It is approximated as in 2).
F _{k + 1} = F _0- (F ₀ -F _k ) x exp (-Δt / T _c ) ... (2)

The time constants _Th and _T in each of the equations (1) and (2) are equivalent to the time constants on the heat transfer characteristics of the projection optical system 114, and are values peculiar to the projection optical system 114, and further have aberrations. It is a different value depending on the type.
That is, in order to accurately predict the fluctuation amount of the imaging characteristics of the projection optical system 114 according to the exposure energy, that is, the aberration fluctuation amount ΔF, the measurement is performed for each projection optical system 114 and each aberration. You need to get the appropriate value.

Further, for the maximum fluctuation amount _Fm , the following equation (3) is used using the aberration fluctuation amount K per unit light amount (unit exposure energy) and the parameter Q for determining the exposure energy incident on the projection optical system 114. ) Can be calculated.
F _m = K × Q ・ ・ ・ (3)
The parameter Q is determined from conditions such as exposure time, exposure amount, scanning speed, and exposure area information.

Then, as shown in FIG. 2, by modeling the time change of the aberration amount F of the projection optical system 114 with the curve 201 by the functions represented by the above equations (1) to (3) (hereinafter, (Called a "prediction model"), it is possible to predict the time change of the aberration amount F according to the exposure energy.

From the above, in order to predict the time change of the imaging characteristics of the projection optical system 114 with high accuracy, it is necessary to appropriately determine the aberration fluctuation amount K per unit exposure energy for each exposure condition.
This is due to the following reasons. That is, the light (diffracted light) diffracted according to the effective light source distribution formed by the illumination optical system 104 and the circuit pattern of the reticle 113 changes according to the exposure conditions at the time of exposure. Further, the irradiation region of the pattern surface of the reticle 113 controlled by the variable blade 112 changes according to the exposure conditions at the time of exposure.
This is because the energy density distribution of the exposure light incident on the projection optical system 114 changes, so that the magnitude of the aberration fluctuation amount ΔF of the projection optical system 114 and its image height dependence change.
The image height dependence referred to here is the amount of fluctuation in the imaging characteristics (aberration) between the non-scanning direction (X direction) and the scanning direction (Y direction) on the image plane of the projection optical system 114, that is, the amount of aberration fluctuation. It means that ΔF is different from each other.

As described above, in the information processing apparatus 20 according to the present embodiment, the prediction models shown in the equations (1) and (2) are used for the time change of the imaging characteristics of the projection optical system 114 according to the exposure energy. It is used to calculate correction coefficients such as F _m , F ₀ , _Th and T _c .
Then, in the information processing apparatus 20 according to the present embodiment, machine learning is used to appropriately and efficiently predict the correction coefficient of the prediction model according to the combination of the exposure conditions of the product lots that are diversified in the lithography process. It is characterized by making estimates.

Specifically, input data is created by modeling and quantitatively classifying the characteristics of the density distribution of the exposure energy in the projection optical system 114.
Next, the teacher data is created by actually measuring the time change of the imaging characteristic (aberration) of the projection optical system 114 at the time of exposure and calculating the correction coefficient of the prediction model.

In addition, learning data composed of input data and teacher data is repeatedly created in a product lot to perform machine learning, that is, a learning model is created.
Then, by acquiring output data showing the correction coefficient of the prediction model according to the actual exposure conditions from the created learning model, the time change of the imaging characteristics of the projection optical system 114 at the time of exposure is appropriately and efficiently changed. Can be predicted.

Next, a method of quantitatively classifying the characteristics of the density distribution of the exposure energy in the projection optical system 114 from the drawing information of the circuit pattern formed on the reticle 113 will be described.

3A and 3B show an example of a circuit pattern drawn on the reticle 113.

Specifically, by actually measuring the circuit pattern drawn on the reticle 113 or referring to the design data, the pitch (interval) and the vertical and horizontal of the line and space (hereinafter referred to as L & S). The line width ratio (HV ratio) is obtained.
Here, the L & S pitch is a factor that determines the diffraction angle of the diffracted light according to the circuit pattern, and the HV ratio is a factor that determines the directivity of the diffracted light.

In the example shown in FIG. 3A, a vertically long vertical L & S pattern (hereinafter, may be referred to as a vertical pattern) 301 and a horizontally long horizontal L & S pattern (hereinafter, may be referred to as a horizontal pattern) 301. .) 303 and is shown.
Here, the vertical direction is the scanning direction (Y direction), and the horizontal direction is the non-scanning direction (X direction).

In the vertical pattern 301, L & S are arranged along the horizontal direction so that the distance between the adjacent wirings is the pitch 302.
Then, the pitch 302 and the areas V1, V2, V3, V4 and V5 of each vertical wiring are calculated from the drawing information of the vertical pattern 301.

Further, in the horizontal pattern 303, L & S are arranged along the vertical direction so that the distance between adjacent wirings is a pitch 304.
Then, the pitch 304 and the areas H1, H2, H3, H4, and H5 of each horizontal wiring are calculated from the drawing information of the horizontal pattern 303.

In FIG. 3B, a rectangle 305 showing an exposure region to which the exposure light is actually irradiated in the reticle 113 is shown.
The area of the vertical pattern 301 that is actually irradiated with the exposure light is inside the rectangle defined by the sides 306 and 307 of the rectangle 305, and the area of each vertical wiring included in the exposed area. The total, that is, the total area V is obtained from the areas V1 to V5.
Similarly, the area of the horizontal pattern 303 that is actually irradiated with the exposure light is inside the rectangle defined by the sides 306 and 308 of the rectangle 305, and the area of each horizontal wiring included in the exposed area. That is, the total area H is obtained from the areas H1 to H5.

Then, in order to quantify the directivity of the diffracted light due to the circuit pattern drawn on the reticle 113, the ratio between the horizontal wiring and the vertical wiring, that is, the HV ratio is quantified using the following equation (4). do.
HV ratio = H / (H + V) ... (4)

Then, in the information processing apparatus 20 according to the present embodiment, the diffracted light is directed by the circuit pattern formed in the reticle 113 based on the above-mentioned L & S pitch and HV ratio using the table shown in Table 1 below. Classify gender.

For example, when the pitches of the vertical pattern 301 and the horizontal pattern 303 are each 600 nm in the circuit pattern, they are classified as “500 nm or more and less than 700 nm”.
Further, when the HV ratio of the vertical pattern 301 and the horizontal pattern 303 is 60% in the circuit pattern, it is classified as "50% or more and less than 75%".

Next, a method of quantitatively classifying the characteristics of the density distribution of the exposure energy in the projection optical system 114 from the effective light source distribution formed by the illumination optical system 104 will be described.

Specifically, the XY intensity ratio of the illumination light is obtained by actually measuring the light intensity in the effective light source distribution formed by the illumination optical system 104.
That is, as shown in FIG. 1, it is shaped into a predetermined shape (circular shape, annular shape, quadrupole shape, double pole shape, etc.) via the beam shaping optical system 126 in the illumination optical system 104. The illumination light is measured by using the effective light source measuring device 135 arranged on the wafer stage 116. As a result, the effective light source distribution of the illumination light can be obtained.

FIG. 4A is a diagram showing how the effective light source measuring device 135 acquires the effective light source distribution.
As shown in FIG. 4A, a very small pinhole 403 having a diameter of several tens of μm is arranged on the image plane 402 of the effective light source distribution 401 in the projection optical system 114.

Then, after passing through the extremely small pinhole 403, the illumination light that has passed through the projection optical system 114 has a light intensity distribution equivalent to that of the effective light source distribution 401, that is, an effective light source distribution 404 and is applied to the light receiving unit 405 of the effective light source measuring device 135. Be irradiated.
The light receiving unit 405 is, for example, a two-dimensional CCD sensor in which CCD pixels are arranged in a grid pattern of 512 × 512, and output by photoelectric conversion is performed according to the brightness of the illuminated illumination light, so that the illumination light is effective. The light source distribution 404 can be measured.

FIG. 4B shows a top view of the two-dimensional CCD sensor 405, showing a state in which the effective light source distribution 404 of the illumination light having a ring-shaped shape is irradiated.
Here, the light intensity measured by the CCD pixels (small point light sources) arranged at the positions X and Y is represented as A (X, Y), and when A (X, Y) = 0, the CCD pixel. Is not illuminated with illumination light.

Further, the light receiving region of the two-dimensional CCD sensor 405 is divided by four diagonal lines in the 45-degree direction, the 135-degree direction, the 225-degree direction, and the 315-degree direction centered on the optical axis.
Then, each region sandwiched between the adjacent diagonal lines, that is, the X-direction right portion region 407, the Y-direction upper region 408, the X-direction left portion region 409, and the Y-direction lower region 410 (hereinafter, XR region 407, YU region 408, XL, respectively). Region 409 and YD region 410) are defined.

Further, the integrated values of the light intensities measured by the CCD pixels included in the _XR , _YU , _XL and _YD regions defined in this way are defined as LXR, LYU, LXL and LYD, respectively.
Then, the XY intensity ratio in the effective light source distribution of the illumination light acquired by the two-dimensional CCD sensor 405 is obtained from the equation (5) shown below.
XY intensity ratio = (L _XL + L _XR ) / (L _YU + L _YD + L _XL + L _XR ) ... (5)
In the effective light source distribution 404 of the illumination light having a ring-shaped shape, the _{LXR, LYU, LXL, and LYD have substantially the} same value, so that the XY intensity ratio is close to 50%.

FIG. 4C shows a state in which the two-dimensional CCD sensor 405 is irradiated with illumination light having a double pole shape in the X direction, that is, an effective light source distribution 412 (hereinafter referred to as X dipole illumination). ..
Further, FIG. 4D shows a state in which the two-dimensional CCD sensor 405 is irradiated with illumination light having a double pole shape in the Y direction, that is, an effective light source distribution 413 (hereinafter referred to as Y dipole illumination). ..

In such a case, the XY intensity ratio actually measured in the effective light source distribution 412 by the X dipole illumination is a value close to 100% from the equation (5).
On the other hand, the XY intensity ratio actually measured in the effective light source distribution 413 by Y dipole illumination is a value close to 0% from the equation (5).

Then, in the information processing apparatus 20 according to the present embodiment, the characteristics of the effective light source distribution of the illumination light are quantitatively classified based on the above-mentioned XY intensity ratio using the table shown in Table 2 below.

For example, when the XY intensity ratio is measured as 55% in the effective light source distribution 404 of the illumination light having a ring band shape, it is classified as "50% or more and less than 60%".
Further, when the XY intensity ratio is measured as 98% in the effective light source distribution 412 of the X dipole illumination, it is classified as "90% or more and 100% or less".
Further, when the XY intensity ratio is measured as 1% in the effective light source distribution 413 of the Y dipole illumination, it is classified as "0% or more and less than 10%".

Next, the relationship between the exposure conditions in the information processing apparatus 20 according to the present embodiment, the characteristics of the density distribution of the exposure energy in the projection optical system 114, and the imaging characteristics of the projection optical system 114, specifically, the fluctuation of the magnification. Will be explained.

FIGS. 5A and 5B schematically show the effective light source distribution 501 of the illumination light and the circuit pattern 502 formed on the reticle 113 as examples of the exposure conditions, respectively.

As shown in FIG. 5A, the effective light source distribution 501 is an effective light source distribution of X dipole illumination, and has a feature that the light intensity is strong in the X direction, that is, the XY intensity ratio is close to 100%. There is.
Further, as shown in FIG. 5B, in the circuit pattern 502, the vertical L & S pattern is larger than the horizontal L & S pattern, that is, the HV ratio is less than 50%, and the diffracted light is directed to be scattered in the X direction. Has sex.

FIG. 5C shows how the illumination light is irradiated on the pupil surface 503 of the projection optical system 114.

Specifically, the illumination light having the effective light source distribution 501 irradiates the 0th-order light 504 that includes the optical axis AX of the projection optical system 114 and is symmetric in the X direction with a YZ cross section parallel to the Y direction.
Then, in the vicinity of the 0th-order light 504, diffracted lights 505 and 506 generated by diffracting the illumination light by the vertical L & S pattern included in the circuit pattern 502 are irradiated.
Here, typically, the amount of light of the 0th order light 504 is larger than that of the diffracted lights 505 and 506.

When the exposure is continuously performed based on the density distribution of the exposure energy in the projection optical system 114 formed from the effective light source distribution 501 and the circuit pattern 502, the 0th-order light 504, the diffracted light 505 and 506 are often irradiated. The heat is absorbed and the temperature rises.
As a result, the lens included in the projection optical system 114 is thermally deformed, that is, thermally expanded along the X direction.

FIG. 5D shows a chip region where projection exposure is performed on the wafer 115.

As described above, the lens included in the projection optical system 114 is thermally deformed along the X direction, so that, as shown in FIG. 5D, the design chip region 507 is oriented in the X direction. Projection exposure is performed in the region 508 where the magnification error occurs.
That is, when the density of the exposure energy is higher in the X direction than in the Y direction, the magnification variation is also larger in the X direction than in the Y direction, and a direction difference (anisotropic) is generated in the magnification variation.

FIGS. 6A and 6B schematically show the effective light source distribution 701 of the illumination light and the circuit pattern 702 formed on the reticle 113 as another example of the exposure conditions, respectively.

As shown in FIG. 6A, the effective light source distribution 701 is an effective light source distribution of Y dipole illumination, and has a feature that the light intensity is strong in the Y direction, that is, the XY intensity ratio is close to 0%. There is.
Further, as shown in FIG. 6B, in the circuit pattern 702, the horizontal L & S pattern is larger than the vertical L & S pattern, that is, the HV ratio is larger than 50%, and the diffracted light is scattered in the Y direction. have.

FIG. 6C shows how the illumination light is irradiated on the pupil surface 503 of the projection optical system 114.

Specifically, the illumination light having the effective light source distribution 701 irradiates the 0th-order light 704 that includes the optical axis AX of the projection optical system 114 and is symmetric in the Y direction with the XZ cross section parallel to the X direction interposed therebetween.
Then, in the vicinity of the 0th-order light 704, diffracted lights 705 and 706 generated by diffracting the illumination light by the horizontal L & S pattern included in the circuit pattern 702 are irradiated.
Here, typically, the amount of light of the 0th order light 704 is larger than that of the diffracted lights 705 and 706.

When the exposure is continuously performed based on the density distribution of the exposure energy in the projection optical system 114 formed from the effective light source distribution 701 and the circuit pattern 702, the 0th-order light 704, the diffracted light 705 and 706 are often irradiated. The heat is absorbed and the temperature rises.
As a result, the lens included in the projection optical system 114 is thermally deformed, that is, thermally expanded along the Y direction.

FIG. 6D shows a chip region projected and exposed on the wafer 115.

As described above, the lens included in the projection optical system 114 is thermally deformed along the Y direction, so that, as shown in FIG. 6D, the design chip region 707 is oriented in the Y direction. Projection exposure is performed in the region 708 where the magnification error occurs.
That is, since the density of the exposure energy is higher in the Y direction than in the X direction, the magnification fluctuation is also larger in the Y direction than in the X direction, and a direction difference is generated in the magnification fluctuation.

As described above, in the information processing apparatus 20 according to the present embodiment, the XY intensity ratio is calculated and classified in the effective light source distribution formed by the illumination optical system 104.
Further, the L & S pitch and the HV ratio are calculated and classified in the circuit pattern 702 formed in the reticle 113.

Then, input data is created by quantitatively classifying the characteristics of the density distribution of the exposure energy from the exposure conditions including the classified XY intensity ratio, L & S pitch, and HV ratio.
In other words, in the information processing apparatus 20 according to the present embodiment, input data is created by classifying the exposure conditions in the exposure apparatus 1 with respect to the anisotropy in the cross section parallel to the scanning direction and the non-scanning direction.

Next, the exposure apparatus 1 is exposed under the above exposure conditions to actually measure the time change of the imaging characteristics (aberration) of the projection optical system 114, and the correction coefficient of the prediction model is calculated to teach the teacher. Create data.

Then, the training data composed of the input data and the teacher data is repeatedly created, that is, the correlation between the feature amount of the density distribution of the exposure energy and the correction coefficient is repeatedly learned, and the learning model is created.
As a result, by acquiring output data showing the correction coefficient of the prediction model according to the actual exposure conditions from the created learning model, the time change of the imaging characteristics of the projection optical system 114 at the time of exposure is appropriate and efficient. Can be predicted.

Next, the processing actually performed in the information processing apparatus 20 according to the present embodiment will be described.

FIG. 7 shows a flowchart of processing performed in the information processing apparatus 20 according to the present embodiment.

First, in step S1, the feature amount of the effective light source distribution formed by the illumination optical system 104 is calculated.
For example, consider the case where the X dipole illumination shown in FIG. 5A is irradiated. At this time, the effective light source measuring device 135 measures the light intensity distribution equivalent to the effective light source distribution of the X dipole illumination, which indicates that the light intensity is strong in the X direction (non-scanning direction), for example, the XY intensity ratio is 98. Calculated as%.

Next, in step S2, the feature amount of the circuit pattern formed on the reticle 113 is calculated.
For example, consider the case where the circuit pattern shown in FIG. 5B is formed on the reticle 113.
At this time, a reticle line width measuring device (not shown) measures the circuit pattern formed on the reticle 113 to calculate, for example, that the L & S pitch is 600 nm. Further, it is calculated that the circuit pattern includes many vertical L & S patterns, for example, the HV ratio is 85%.
From this, it can be seen that the circuit pattern formed on the reticle 113 has directivity to generate diffracted light along the horizontal direction (X direction).

Then, in step S3, input data is created by classifying the feature amount of the density distribution of the exposure energy in the projection optical system 114 from the feature amount calculated in steps S1 and S2.
An example of the table used when creating the input data in step S3 is shown in Table 3 below.

From the table shown in Table 3, the XY intensity ratio = 98% calculated in step S1 is classified as "90% or more and 100% or less", and the L & S pitch = 600 nm calculated in step S2 is "500 nm or more". It is classified as "less than 700 nm".
Further, the HV ratio = 85% calculated in step S2 is classified as "75% or more and 100% or less".

Input data is created by classifying the feature quantities of the density distribution of the exposure energy in the projection optical system 114 in this way.
The steps S1 to S3 shown above are referred to as an input data creation step 801.

Next, in step S4, a correction coefficient suitable for the exposure condition, that is, a correction coefficient corresponding to the input data created in step S3 is estimated. Specifically, it is determined whether the training model already includes the correction coefficient.

If the training model does not include the correction coefficient (No in step S4), the process proceeds to step S5, and the correction coefficient of the preset design conditions, that is, the initial value of the correction coefficient is selected.
On the other hand, if the training model includes a correction coefficient (Yes in step S4), the process proceeds to step S6, and the correction coefficient having the highest correlation with the input data is selected (the correlation degree will be described later). ).
The steps S4 to S6 shown above are referred to as a correction coefficient estimation step 802.

FIG. 8A shows a prediction model 601 using the correction coefficients M _m , M ₀ , _Th and T _c of the design conditions selected in step S5.
Note that FIG. 8A shows the magnification M as the imaging characteristic of the projection optical system 114, that is, F _m and F ₀ in the equations (1) and (2) are replaced with M _m and M ₀ , respectively. There is.

As shown in FIG. 8A, the initial value of the magnification M, that is, the initial magnification is M ₀ , and the exposure light emitted from the laser light source 101 at time t ₀ is exposed by passing through the projection optical system 114. When the operation is started, the magnification M indicates a time change.
Then, after the magnification M reaches the maximum fluctuation magnification M _m at time t ₁ , the magnification M hardly changes from the maximum fluctuation magnification M _m .

Then, when the exposure operation is stopped at time t ₂ , the magnification M fluctuates so as to approach the initial magnification M ₀ again as the time t elapses, and reaches the initial magnification M ₀ at time t ₃ .
Note that in the prediction model 601 there is no difference in direction with respect to the time change of the magnification M of the projection optical system 114.

Next, in step S7, the time change of the magnification M of the projection optical system 114 is actually measured during the exposure processing of the product lot under the exposure conditions shown in FIGS. 5A and 5B.

FIG. 8B shows an example of the time change of the magnification M measured in step S7.
Specifically, the time change _MX1 to MX10 of the magnification M measured at the predetermined X coordinate of the projection optical system 114, and the time change _{MY1 to MY10 of} the magnification M measured at the predetermined _Y coordinate are present. It is shown.

Next, in step S8, the correction coefficient of the prediction model represented by the equations (1) and (2) for the data measured in step S7 is calculated.
Specifically, for the time change _MX1 to _MX7 (exposure characteristics) of the magnification M in the X direction from time t ₀ to t ₁ , fitting is performed by the prediction model represented by the equation (1). Calculate the correction coefficients MX and _{Th X} with. Then, the time change of the _{magnification} M in the _X direction from the time t ₂ to t ₃ (non-exposure characteristics) is corrected by fitting with the prediction model represented by the equation (2). Calculate the coefficients M ₀ and T _{c X.}

Similarly, the time change of the magnification M in the _Y direction from time t ₀ to t ₁ MY1 to _MY7 (exposure characteristics) is corrected by fitting with the prediction model represented by the equation (1). Calculate the coefficients _{MY and ThY .} Then, the time change _MY8 to _MY10 (non-exposure characteristics) of the magnification M in the Y direction from time t ₂ to t ₃ is corrected by fitting with the prediction model represented by the equation (2). Calculate the coefficients M ₀ and T _cY .
The predictive models thus obtained are shown as predictive models 602 and 603 in FIG. 8B, respectively.

なおここで、最大変動倍率Ｍ_Ｘ及びＭ_ｙそれぞれについては、初期倍率Ｍ_０に対するずれの割合、すなわち（Ｍ_Ｘ－Ｍ_０）／Ｍ_０及び（Ｍ_Ｙ－Ｍ_０）／Ｍ_０で表している。 In addition, examples of the correction coefficients MX, _MY , _{Th X} , Th _Y , T _cX , and T _cY calculated in step S8 above are shown in Table 4 below.

Here, each of the maximum fluctuation magnifications _MX and My is expressed by the ratio of deviation to the initial _{magnification} M ₀ , that is, ( _MX − M ₀ ) / M ₀ and ( _MY − M ₀ ) / M ₀ . There is.

In this way, the teacher data corresponding to the input data shown in Table 3 above can be obtained from the table shown in Table 4.
The steps S7 and S8 shown above are referred to as a teacher data creation step 803.

Next, in step S9, the degree of correlation between the feature amount of the density distribution of the exposure energy in the projection optical system 114, which is the input data, and the correction coefficient, which is the teacher data and the output data, is calculated from the prediction error in the prediction model.

FIG. 8C is a schematic diagram showing how the above-mentioned correlation degree is calculated in step S9.
Specifically, in FIG. 8C, a prediction model 604 using a plurality of measurement points acquired by measuring the time change of the magnification M of the projection optical system 114 under a predetermined exposure condition and a predetermined correction coefficient. Is shown.

Here, in step S9, among the plurality of acquired measurement points, the value having the largest error from the prediction model 604 (difference from the measurement point most distant from the prediction model) is set as the prediction error D, and the prediction error D is set as the prediction error D. It is recorded as the degree of correlation of the prediction model 604.
That is, in the above example, the prediction error D is calculated from the prediction model 601 selected in step S5 and the prediction models 602 and 603 acquired in step S8 for the plurality of measurement points measured in step S7. And record it as the degree of correlation.

For example, the degree of correlation calculated in each of the prediction models 602 and 603, that is, the value of the prediction error D is shown in Table 4.
Here, regarding the degree of correlation, that is, the prediction error D, the ratio of the deviation from the magnification M _calc calculated from the prediction model of the magnification M _meas at the measurement point when calculating the prediction error D, that is, (M _meas -M _calc ) /. It is represented by M _calc .

This makes it possible to quantitatively learn from the input data according to the exposure conditions exemplified in Table 3 and the teacher data and the degree of correlation exemplified in Table 4.
The step S9 shown above is referred to as a learning step 804.

Then, when the exposure process is performed under the same exposure conditions from the next time onward, the correction coefficient can be estimated based on the correlation degree data acquired up to this time.
Further, in step S10, it is determined whether or not the product lot is repeatedly produced under the same exposure conditions.
Then, when production is performed (Yes in step S10), the process returns to step S4, and when production is not performed (No in step S10), the process ends.

Returning to step S4, the correction coefficient estimation step 802 is performed again. That is, the correction coefficient of the prediction model is estimated in step S4.
At this time, the learning model includes the correction coefficient learned in step S9 up to the previous time (Yes in step S4).
Therefore, after shifting to step S6, the correction coefficient having the smallest prediction error D, that is, the highest degree of correlation is selected from the correction coefficients accumulated in the learning model in step S6.

Then, by shifting to steps S7 to S10, learning is repeated until the production of the product lot under the same exposure conditions is completed.
In this way, by repeatedly machine learning the time change of the magnification M of the projection optical system 114 for each exposure condition and accumulating the correction coefficients during the exposure processing of the product lot, the magnifications M are mutually in the X direction and the Y direction. Even if different time changes are shown, the estimation accuracy of the correction coefficient can be improved.

Then, in the exposure apparatus 1, when correcting the magnification, distortion, astigmatism, spherical aberration, coma aberration, and wave surface aberration of the projection optical system 114 based on the estimated correction coefficient, the field lens 127 is moved according to the time. Let me.
Further, when the focus of the projection optical system 114 is corrected based on the estimated correction coefficient, the position of the wafer stage 116 in the optical axis direction is changed according to the time.

As described above, in the information processing apparatus 20 according to the present embodiment, learning using the exposure conditions in the exposure apparatus as input data and the correction coefficient of the prediction model for predicting the time change of the imaging characteristics of the projection optical system 114 as teacher data. I am creating a model.
Then, even if the number of exposure conditions increases, such as an increase in the types of original plates, by acquiring output data that gives the correction coefficient of the prediction model from the created learning model, the imaging characteristics of the projection optical system change over time. Can be predicted efficiently.

In the above, the time change of the magnification M has been described as the variation of the imaging characteristics of the projection optical system 114, but the present invention is not limited to this. For example, the information processing apparatus 20 according to the present embodiment can be applied to focus, distortion, astigmatism, spherical aberration, coma aberration, wavefront aberration and the like as imaging characteristics of the projection optical system 114.
Further, in the above description, the feature amount of the density distribution of the exposure energy by the X dipole illumination shown in FIGS. 5 (a) to 5 (d) has been described, but the present invention is not limited to this. For example, the information processing apparatus 20 according to the present embodiment can be applied to the feature amount of the density distribution of the exposure energy by the Y dipole illumination shown in FIGS. 6A to 6D.

[Second Embodiment]
FIG. 9 shows a block diagram of the exposure processing system 1000 provided with the information processing apparatus 220 according to the second embodiment.

As shown in FIG. 1, the information processing apparatus 20 according to the first embodiment is a so-called stand-alone type apparatus in which the exposure apparatus 1 is connected to the main control apparatus 103 by an interface cable.
On the other hand, as shown in FIG. 9, the information processing apparatus 220 according to the present embodiment is connected to a plurality of exposure apparatus including the exposure apparatus 1001 and the exposure apparatus 1002 via the network cable 1003 in the exposure processing system 1000. It is a so-called online type device.
Further, as shown in FIG. 9, the information processing apparatus 220 according to the present embodiment is also connected to the reticle pattern measuring device 1005 for measuring the circuit pattern of the reticle 113 via the network cable 1003.

That is, the information processing device 220 according to the present embodiment is a computer as an information processing means, and machine learning regarding the correction coefficient of the prediction model can be performed based on the training data in each of the plurality of exposure devices. can.
Specifically, input data is created from the data of the effective light source distribution measured in each exposure device and the data of the circuit pattern measured by the reticle pattern measuring device 1005.

Then, teacher data is created by calculating the correction coefficient of the prediction model from the measured values of the imaging characteristics of the projection optical system 114 acquired in each exposure device, and the correlation degree data between the input data and the teacher data is obtained. Calculate and accumulate.
By recording the correlation degree data for each exposure condition learned for the plurality of exposure devices in the recording device 30 and creating a learning model in this way, it is possible to further improve the estimation accuracy of the correction coefficient.

As described above, in the information processing apparatus 220 according to the present embodiment, learning using the exposure conditions in the exposure apparatus as input data and the correction coefficient of the prediction model for predicting the time change of the imaging characteristics of the projection optical system 114 as teacher data. I am creating a model.
Then, even if the number of exposure conditions increases, such as an increase in the types of original plates, it is possible to acquire output data that gives correction coefficients of the prediction model from the created learning model.
In addition, the information processing apparatus 220 according to the present embodiment more efficiently predicts the time change of the imaging characteristics of the projection optical system by creating a learning model using the learning data in a plurality of exposure devices. be able to.

[Manufacturing method of goods]
Next, a method of manufacturing an article using an exposure apparatus including the information processing apparatus according to any one of the first and second embodiments will be described.

The articles are semiconductor devices, display devices, color filters, optical components, MEMS and the like.
For example, a semiconductor device is manufactured by going through a pre-process for forming a circuit pattern on a wafer and a post-process including a processing step for completing the circuit chip produced in the pre-process as a product.

The pre-process includes an exposure step of exposing a wafer coated with a photosensitive agent using an exposure device provided with an information processing device according to any one of the first and second embodiments, and a developing step of developing the photosensitive agent. include.
Then, an etching step, an ion implantation step, and the like are performed using the developed photosensitive agent pattern as a mask, and a circuit pattern is formed on the wafer.

By repeating these steps such as exposure, development, and etching, a circuit pattern composed of a plurality of layers is formed on the wafer.
In the post-process, dicing is performed on the wafer on which the circuit pattern is formed, and chip mounting, bonding, and inspection processes are performed.

The display device is manufactured by going through a process of forming a transparent electrode. The step of forming the transparent electrode is a step of applying a photosensitive agent on a glass wafer on which a transparent conductive film is vapor-deposited, and an exposure apparatus provided with an information processing apparatus according to any one of the first and second embodiments. It includes a step of exposing a glass wafer coated with a photosensitizer. Further, the step of forming the transparent electrode includes a step of developing the exposed photosensitive agent.

According to the method for manufacturing an article according to the present embodiment, it is possible to produce an article having higher quality and higher productivity than before.

Although the preferred embodiments have been described above, the present invention is not limited to these embodiments, and various modifications and changes can be made within the scope of the gist thereof.
Further, although the information processing apparatus according to the present embodiment has been described above, the information processing method shown above, the program for implementing the method, and the recording medium on which the program is recorded can also be read by the computer. Included in the scope of embodiments.

1 Exposure device 20 Information processing device 114 Projection optical system

Claims (12)

An information processing device characterized by predicting time-dependent changes in the imaging characteristics of a projection optical system provided in an exposure device using a learning model acquired by machine learning. The input data of the training model is created by classifying at least one of the effective light source distribution in the exposure apparatus and the pattern drawn on the original plate with respect to the anisotropy in the cross section parallel to the scanning direction and the non-scanning direction. The information processing apparatus according to claim 1, wherein the information processing apparatus is to be used. According to claim 1 or 2, the teacher data of the learning model is created by calculating the correction coefficient by measuring the time change of the imaging characteristic of the projection optical system and fitting it with the prediction model. The information processing device described. The imaging characteristic according to any one of claims 1 to 3, wherein the imaging characteristic includes at least one of magnification, focus, distortion, astigmatism, spherical aberration, coma, and wavefront aberration. Information processing device. The information processing device is
Input data is created by classifying at least one of the effective light source distribution in the exposure apparatus and the pattern drawn on the original plate with respect to the anisotropy in the cross section parallel to the scanning direction and the non-scanning direction.
Teacher data is created by measuring the time change of the imaging characteristics of the projection optical system and calculating the correction coefficient by fitting with a prediction model.
A claim characterized in that the learning model is created by performing learning by calculating the degree of correlation between the measured time change of the imaging characteristic and the teacher data and the output data for estimating the correction coefficient. The information processing apparatus according to any one of Items 1 to 4. The information processing apparatus according to claim 5, wherein the information processing apparatus calculates a prediction error of the measurement points of the imaging characteristics farthest from the prediction model as the correlation degree. .. The information processing device according to claim 5 or 6, wherein the information processing device creates the learning model by performing the learning on a plurality of the exposure devices. One of claims 1 to 7, wherein the information processing apparatus predicts a time change in the imaging characteristics of the projection optical system provided in each of the plurality of exposure devices by using the learning model. The information processing device according to paragraph 1. An exposure apparatus that exposes the substrate so as to transfer the pattern drawn on the original plate to the substrate.
An exposure apparatus comprising the information processing apparatus according to any one of claims 1 to 8. A step of exposing the substrate by using the exposure apparatus according to claim 9.
The process of developing the exposed substrate and
The process of manufacturing an article from the developed substrate and
A method of manufacturing an article comprising. An information processing method comprising a step of predicting a time change of an imaging characteristic of a projection optical system provided in an exposure apparatus using a learning model acquired by machine learning. A computer-readable recording medium in which a program that causes a computer to perform information processing is recorded.
A computer that records a program that causes a computer to perform a process of predicting changes in the imaging characteristics of a projection optical system provided in an exposure apparatus using a learning model acquired by machine learning is read by a computer. Possible recording medium.

Images