JP2022102795A - Information processor, exposure device, and manufacturing method of article - Google Patents

Abstract

To provide an information processor capable of efficiently determining an exposure amount at an exposure device when forming a pattern on a substrate.SOLUTION: An information processor utilizes a learning model obtained by a mechanical learning to determine information related to an exposure amount in an exposure device when forming a pattern on a substrate.SELECTED DRAWING: Figure 7

Description

The present invention relates to an information processing device.

Conventionally, in order to form a pattern having a desired line width on a substrate, it has been required to expose the substrate with an optimum exposure amount in an exposure apparatus.
Patent Document 1 discloses an exposure condition setting device that sets an exposure amount based on at least one of a resist type and film thickness, a pattern type, a minimum dimension, and a density.

Japanese Unexamined Patent Publication No. 4-72711

In recent years, in order to manufacture a 3D-NAND memory or the like, an exposure apparatus may be exposed to form a pattern on a substrate coated with a high-viscosity resist so as to form a thick film.
It is known that such a high-viscosity thick-film resist has a film thickness distribution, but the exposure condition setting apparatus disclosed in Patent Document 1 does not consider the film thickness distribution of the resist. It is difficult to set the exposure amount according to such exposure.

Furthermore, regarding the line width of the pattern formed on the substrate coated with the thick film resist, in addition to the above parameters, there are many parameters related to the coating and developing conditions including the usage of the rinsing solution and the concentration control of the developing solution. Affect.
On the other hand, in the exposure condition setting device disclosed in Patent Document 1, it is difficult and often difficult to set the exposure amount by accumulating and collating the exposure data corresponding to such a wide variety of parameters. It takes time.

Therefore, an object of the present invention is to provide an information processing apparatus capable of efficiently determining the exposure amount in the exposure apparatus when forming a pattern on a substrate.

The information processing apparatus according to the present invention is characterized in that information regarding an exposure amount in an exposure apparatus when forming a pattern on a substrate is determined by using a learning model acquired by machine learning.

According to the present invention, it is possible to provide an information processing apparatus capable of efficiently determining an exposure amount in an exposure apparatus when forming a pattern on a substrate.

The schematic sectional view which shows the structure of the exposure apparatus which comprises the information processing apparatus which concerns on this embodiment. The figure which showed the example of the film thickness distribution of the thick film resist on the substrate exposed by an exposure apparatus. The figure which showed the example of the dependence of the line width of a pattern on the substrate exposed by an exposure apparatus with respect to the film thickness of a thick film resist. Schematic block diagram of a coating and developing device. Schematic block diagram of the board processing system. The flowchart which showed the method of acquiring the learning model in the information processing apparatus which concerns on this embodiment. The flowchart which showed the method of determining the exposure amount using the learning model in the information processing apparatus which concerns on this 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.
In the following, the direction parallel to the optical axis AX of the projection optical system 101 is the Z axis, and the direction (scanning direction) in which the reticle R and the substrate W are scanned during exposure in a plane perpendicular to the Z axis is the Y axis. The direction perpendicular to the Y axis (non-scanning direction) is defined as the X axis.

Semiconductor devices such as LSIs and devices such as flat panel displays (FPDs) are manufactured through a photolithography process.
In the photolithography process, an image of a pattern drawn on an original plate called a mask or reticle is applied to a wafer or a wafer to which a photosensitive agent called a resist is applied via a projection optical system composed of a lens, a mirror, or the like. It includes an exposure process of projecting onto a substrate such as a glass plate.

In the exposure step, it is required to resolve the pattern drawn on the original plate with a desired line width on the photosensitive agent applied to the substrate.
Then, in order to resolve the pattern with a desired line width, it is required to perform exposure with an optimum exposure amount, that is, when exposure is performed with a non-optimal exposure amount, the pattern can be formed with a desired line width. It will be difficult.

Further, if the exposure is not performed with the optimum exposure amount, not only the line width but also the cross-sectional shape of the photosensitive agent is affected in the formed pattern.
Specifically, although it depends on the type of photosensitive agent such as positive resist and negative resist, hemming may occur or a constricted shape may occur at the boundary surface of the substrate instead of the rectangular shape.

As a method for determining the exposure amount, for example, there are the following methods.
That is, first, an exposure is performed in which a predetermined pattern is printed on the substrate coated with the photosensitive agent.
Then, such exposure is performed on a plurality of substrates while changing the exposure amount by a predetermined step amount.

At this time, the exposure performed by changing the focus step by step at the same time as the change in the exposure amount is called FEM (Focus Exposure Matrix).
Then, after the exposure is performed, PEB (Post Exposure Bake) or development is performed on the substrate depending on the exposure wavelength and the type of resist, and the pattern is resolved.

Next, the line width of each of the plurality of substrates having different exposure amounts and having the patterns resolved is generally measured by using SEM (Scanning Electron Microscope).
Then, for example, by fitting the exposure amount dependence of the line width obtained from the measurement result with a linear straight line, a quadratic curve, or the like, the exposure amount at which the pattern is formed with a desired line width can be determined.

On the other hand, in recent years, 3D-NAND memory has been manufactured in order to improve the storage capacity of the memory, and a thick film resist is used in the manufacture.
At that time, since a high-viscosity resist is applied to the substrate so as to form a thick film, it is difficult to control the film thickness distribution of the resist on the substrate with high accuracy.

Since the exposure amount described above also depends on the film thickness of the resist, if the film thickness distribution of the resist is not controlled with high accuracy, the distribution of the line width of the pattern in the substrate should also be controlled with high accuracy. Becomes difficult.
Since the film thickness distribution occurs when the thick film resist is applied to the substrate in this way, it is difficult to control the line width formed on the substrate by the conventional exposure amount determination method, and many other problems are solved. Occurs.

Further, regarding the line width of the pattern formed on the substrate to which the thick film resist is applied, in addition to the above parameters, there are many matters relating to the coating and developing conditions including the method of using the rinsing solution and the concentration control of the developing solution. Parameters also affect.
On the other hand, in the conventional exposure amount determination method, it is difficult and takes a lot of time to determine the exposure amount by referring to the exposure data accumulated corresponding to such a wide variety of parameters.

Therefore, the present embodiment provides an information processing apparatus capable of determining the exposure amount so that the line width of the pattern formed on the substrate can be controlled with high accuracy even when a thick film resist is used in the exposure apparatus. The purpose is to do.

FIG. 1 is a schematic cross-sectional view showing the configuration of an exposure apparatus 100 including an information processing apparatus according to the present embodiment.

The exposure apparatus 100 is a projection type exposure apparatus that projects (transfers) an image of a pattern formed (drawn) on the reticle R on the substrate W by, for example, a scan-and-repeat method.
The exposure device 100 includes a projection optical system 101, a measuring device (measuring means), a reticle stage 103, a control unit, a substrate stage 105, and an illumination optical system 106.

The illumination optical system 106 illuminates the reticle R by adjusting the light emitted from a light source that generates pulsed light such as an excimer laser (not shown) and then guiding the light to the reticle R.

The reticle stage 103 moves parallel to at least the X-axis and the Y-axis while holding the reticle R, which is an original plate made of quartz glass, for which a pattern to be transferred (for example, a circuit pattern) is formed on the substrate W. can do.
When the reticle stage 103 is exposed, the reticle stage 103 scans and moves at a constant speed in the direction parallel to the Y axis in the XY plane perpendicular to the optical axis AX of the projection optical system 101, and maintains the target position. The correction movement is appropriately performed in the direction parallel to the X axis so as to perform scanning.
Further, the reticle stage 103 is provided with a bar mirror 120. Then, the first interferometer 121 emits light toward the bar mirror 120 and measures the distance between the first interferometer 121 and the bar mirror 120, so that the position information in the direction parallel to each axis of the reticle stage 103 can be constantly measured.

The projection optical system 101 guides the light that has passed through the reticle R onto the substrate W so that the image of the pattern formed on the reticle R is projected onto the substrate W at a predetermined magnification (for example, 1/2 times). It glows.
As a result, the image of the pattern formed on the reticle R is projected onto the substrate W made of, for example, single crystal silicon, on which a resist (photosensitive agent) is coated on the surface.
Here, the image plane of the projection optical system 101 has a relationship perpendicular to the Z axis (that is, the direction parallel to the optical axis AX of the projection optical system 101).

The substrate stage 105 holds the substrate W via a chuck (not shown), and has θ _X , θ _Y , and θ _Z , which are directions parallel to the X-axis, Y-axis, and Z-axis, and rotation directions around each axis. It can move (rotate) in a direction.
Further, the substrate stage 105 is provided with a bar mirror 123. Then, the second interferometer 124 emits light toward the bar mirror 123 and measures the distance between the second interferometer 124 and the bar mirror 123, so that the position information in the direction parallel to each axis of the substrate stage 105 can be constantly measured.

The measuring device (measuring means) is positioned and tilted in a direction parallel to the Z axis on the surface of the substrate W held by the substrate stage 105 (hereinafter, these may be collectively referred to as "plane position"). To measure.
The measuring device includes a light projecting system that projects a light beam for measurement toward the substrate W and a light receiving system that receives the light beam reflected from the substrate W.

Specifically, the floodlight system of the measuring device is composed of a measurement light source 110, a collimator lens 111, a slit member 112, a floodlight side optical system 113, and a floodlight side mirror 114.
The measurement light source 110 is a light source such as a lamp or a light emitting diode, and emits a light flux for measurement toward the substrate W.
The collimator lens 111 converts the luminous flux emitted from the measurement light source 110 into a parallel luminous flux having a substantially uniform intensity distribution in the cross section.

The slit member 112 is, for example, a pair of prisms bonded so that their slopes face each other, and the bonded surfaces are light-shielded formed of chromium or the like having a plurality of openings (for example, nine pinholes). A membrane is provided. Then, the parallel light flux from the collimator lens 111 is divided into, for example, nine light fluxes.
The floodlight-side optical system 113 is both telecentric optical systems, and guides nine light fluxes that have passed through the slit member 112 to nine measurement points on the substrate W via the floodlight-side mirror 114, respectively. do.

Here, the plane having the pinhole of the slit member 112 and the plane including the surface of the substrate W satisfy the Scheimpflug condition with respect to the light projecting side optical system 113.
Further, in the exposure device 100, the nine luminous fluxes from the projection system of the measuring device are placed on the substrate W so that the incident angle Φ (angle formed by the optical axis AX of the projection optical system 101) is 70 ° or more. Incident.
Then, the nine light fluxes are θ (θ) with respect to the Y axis in the XY plane so that the light receiving system of the measuring device shown below can observe the nine measurement points on the substrate W independently of each other. For example, it is incident on the substrate W from the direction rotated by 22.5 °).

Next, the light receiving system of the measuring device is composed of a light receiving side mirror 115, a light receiving side optical system 116, a correction optical system 118, and a photoelectric conversion means 119.
The light receiving side optical system 116 is both telecentric optical systems, and guides nine light fluxes from the substrate W reflected by the light receiving side mirror 115 to the photoelectric conversion means 119.
The light receiving side optical system 116 is provided with a stopper diaphragm (not shown), and the stopper diaphragm is a high-order diffracted light generated by a pattern already formed on the substrate W at each of the nine measurement points. (Noise light) can be cut.

The correction optical system 118 includes nine correction lenses, and the nine light fluxes whose optical axes are parallel to each other by passing through the light receiving side optical system 116 are detected on the detection surface of the photoelectric conversion means 119. The light is focused so as to be spot light having the same magnitude as each other.
The photoelectric conversion means 119 is composed of, for example, a group consisting of nine one-dimensional CCD line sensors, and after detecting the intensity of each light flux incident on the detection surface, outputs the detection result to the arithmetic circuit 126.
As the photoelectric conversion means 119, one in which a plurality of two-dimensional position measuring elements are arranged may be adopted.

Further, regarding the light receiving side optical system 116, the correction optical system 118, and the photoelectric conversion means 119 in the light receiving system of the measuring device, each measurement point on the substrate W and the detection surface of the photoelectric conversion means 119 are conjugated with each other. The fall correction is performed in advance.
Therefore, it is possible to suppress the change in the position of the pinhole image on the detection surface of the photoelectric conversion means 119 caused by the local inclination at each measurement point on the substrate W.
Thereby, the change in height along the optical axis AX at each measurement point on the substrate W can be detected from the change in the pinhole image on the detection surface of the photoelectric conversion means 119.

The control unit includes a reticle position control system 122, a board position control system 125, an arithmetic circuit 126, and a main control unit 127.
The main control unit 127 is composed of, for example, a computer including a processor and a memory, and is connected to each component of the exposure apparatus 100 (and a control system for controlling them) via a line, so that the main control unit 127 is connected to each component (and a control system for controlling them) according to a program or the like. It controls the operation of each component in an integrated manner.
The reticle position control system 122 controls the operation of the reticle stage 103 and the first interferometer 121 based on the drive command from the main control unit 127.

The board position control system 125 controls the operation of the board stage 105 and the second interferometer 124 based on the drive command from the main control unit 127.
The arithmetic circuit 126 calculates the intensity value of each luminous flux reflected from the nine measurement points on the substrate W based on the detection result obtained from the photoelectric conversion means 119.

Further, the main control unit 127 can adjust the position of the substrate W in the XY plane and the position in the direction parallel to the Z axis so as to form the slit image of the reticle R in a predetermined region on the substrate W. ..
The position in the XY plane referred to here means a position in the direction parallel to the X axis and the Y axis and a rotation angle θ _Z around the Z axis, and the position in the direction parallel to the Z axis. , The height in the direction parallel to the Z axis and the rotation angles θ _X and θ _Y around the X axis and the Y axis.

Then, the main control unit 127 transmits a drive command to the reticle position control system 122 and the board position control system 125 to scan the reticle stage 103 and the board stage 105 in synchronization with each other and scan the pattern on the reticle R on the board. The scanning exposure formed on W is performed.
At this time, when the reticle stage 103 is scanned in the Y direction, the main control unit 127 scans the substrate stage 105 in the Y direction at a speed corrected by the reduction magnification of the projection optical system 101.
The scanning speed of the reticle stage 103 is highly productive based on the width of the masking blade (not shown) provided in the illumination optical system 106 in the scanning direction, the sensitivity of the resist applied to the surface of the substrate W, and the like. It is predetermined to be advantageous.

Next, the control for aligning the pattern formed on the reticle R on the substrate W, which is performed by the control unit in the exposure apparatus 100, will be described.

The main control unit 127 first acquires the position data of the reticle stage 103 and the substrate stage 105 from the first interferometer 121 and the second interferometer 124 for each position in the XY plane, and also acquires the substrate from an alignment microscope (not shown). Acquire the position data of W.
Next, the main control unit 127 creates control data based on the acquired position data. Then, the main control unit 127 sends a drive command to the reticle position control system 122 and the board position control system 125 based on the created control data, thereby appropriately moving the reticle stage 103 and the board stage 105 to desired positions. ..

On the other hand, the main control unit 127 controls the position (height) in the direction parallel to the Z axis, that is, the focus operation, the intensity of each light flux reflected from the nine measurement points on the substrate W from the arithmetic circuit 126. Get the value.
Next, the main control unit 127 creates control data based on each acquired intensity value. Then, the main control unit 127 appropriately moves (changes) the board stage 105 to a desired position (or posture) by transmitting a drive command to the board position control system 125 based on the created control data.

Specifically, the main control unit 127 determines the position of the substrate stage 105 in the direction parallel to the Z axis based on the measurement results from the measuring device (and the arithmetic circuit 126) arranged in the vicinity of the exposure slit (not shown). calculate.
Then, the main control unit 127 controls the substrate stage 105 so that the exposure position on the substrate W becomes the optimum image plane position based on the calculated position in the direction parallel to the Z axis.

The control unit may be provided integrally with the other components of the exposure apparatus 100, that is, in the same housing as the housing in which the other components of the exposure apparatus 100 are provided. Further, the control unit may be provided separately from the other components of the exposure device 100, that is, in a housing different from the housing in which the other components of the exposure device 100 are provided.

Next, the processing by the information processing apparatus according to this embodiment will be described. The information processing apparatus according to the present embodiment is assumed to be combined with the above-mentioned main control unit 127, but the information processing apparatus is not limited to this, and may be configured by a control unit or the like different from the main control unit 127.

FIG. 2 shows an example of the film thickness distribution of the thick film resist on the substrate W exposed by the exposure apparatus 100.

As shown in FIG. 2, the film thickness of the thick film resist is generally non-uniform on the substrate W.
This is because the thick film resist applied to the substrate W generally has a high viscosity, and it is difficult to spread the thick film resist uniformly over the entire substrate W.
Specifically, when the thick film resist is discharged for coating on the substrate W, the film thickness of the thick film resist tends to increase in a region landed on the substrate W, for example, in the center of the substrate W.

On the other hand, in general, in order to uniformly apply the resist on the substrate W, the resist is prepared by preliminarily applying the same rinse solution as the solvent of the resist on the substrate W to make the surface of the substrate W lipophilic. Measures are taken to facilitate uniform coating on the substrate W.
However, when a thick film resist having a high viscosity is applied on the substrate W, the rinse liquid having a low viscosity is applied on the substrate W earlier than the thick film resist.
That is, when the thick film resist and the rinsing liquid are applied on the substrate W, only the rinsing liquid is shaken off and spreads at the same time as the rotation, so that the thick film resist is not uniformly applied on the substrate W. It becomes.

Further, the film thickness of the thick-film resist tends to increase not only at the central portion of the substrate W but also at the edge portion of the substrate W.
This is because when a low-viscosity resist is applied onto the substrate W, the portion that flies to the outside of the substrate W does not fly to the outside of the substrate W and stays at the edge of the high-viscosity thick-film resist. This is because it becomes easier to do.
As described above, the film thickness of the thick-film resist generally tends to increase at the central portion and the edge portion of the substrate W.

FIG. 3 shows an example of the dependence of the line width of the pattern on the substrate W exposed by the exposure apparatus 100 on the film thickness of the thick film resist.
The dotted line shown in FIG. 3 is a straight line obtained by fitting the line width of the pattern as a linear function of the film thickness of the thick film resist.

As shown by the dotted line in FIG. 3, the line width of the pattern varies depending on the film thickness of the thick film resist.
That is, if thick resists having different film thicknesses are exposed to the same exposure amount, the line widths of the patterns will be different from each other.

FIG. 4 shows a schematic configuration diagram of a coating and developing apparatus 400 used for coating a resist on a substrate W and developing an exposed substrate W.

As shown in FIG. 4, the coating and developing apparatus 400 is provided with the coating unit 401 and the developing unit 402 separately from each other.
The coating unit 401 has a coating cup unit 403 and a coating bake unit 404.

Further, a plurality of coating cup units 403 are provided so as to be used properly according to the type of resist coated on the substrate W.
Further, a plurality of coating bake units 404 are also provided so that they can be used properly according to the temperature at which the substrate W is baked.

Similarly, the developing unit 402 is provided with a plurality of developing bake units 405 and developing cup units 406 that perform PEB after exposure and before development.
As described above, in the coating and developing apparatus 400, it is common that a plurality of units are provided in order to achieve a high throughput.

FIG. 5 shows a schematic configuration diagram of the substrate processing system 1000.

As shown in FIG. 5, in the substrate processing system 1000, the exposure apparatus 100 and the coating and developing apparatus 400 are connected to each other via the interface block 500.
In the substrate processing system 1000, the substrate W coated with the resist by the coating and developing apparatus 400 is conveyed to the exposure apparatus 100 by passing through the interface block 500.
The substrate W exposed by the exposure apparatus 100 is conveyed to the coating and developing apparatus 400 by passing through the interface block 500 again, and then developed by the coating and developing apparatus 400.

Further, as shown in FIG. 5, in the substrate processing system 1000, information on each other is transmitted and received between the exposure apparatus 100 and the coating and developing apparatus 400 via the network 501.
The network 501 may be a wired connection or a wireless connection, and may be further connected to another device.

FIG. 6 is a flowchart showing a method of acquiring a learning model for outputting output data representing an exposure amount by machine learning in the main control unit 127, which is an information processing apparatus according to the present embodiment.
Specifically, here, a method of acquiring a learning model in which the information on the coating and developing conditions in the coating and developing apparatus 400 and the information on the pattern forming conditions in the exposure apparatus 100 are input data and the exposure amount is output data will be described.

First, in step S601, the main control unit 127 acquires the coating and developing conditions of the resist on the substrate W by the coating and developing apparatus 400 from the coating and developing apparatus 400 via the network 501.
The coating and developing conditions acquired here include, for example, the type of resist, the size of the viscosity of the resist, the size of the etching resistance of the resist, and the type of the layer to which the resist is applied, such as line and space and contact holes. Is included.
The coating development conditions to be acquired include the type of coating method according to the thickness of the resist, the number of rotations of the coating cup unit 403, the rotation sequence of the coating cup unit 403 (rotational speed profile), and the rinse used at the time of coating. Includes liquid type.

The coating development conditions to be acquired include the discharge timing (discharge sequence, discharge profile) of the rinse liquid, the shape of the discharge portion of the resist, the configuration of the coating cup unit 403, and the configuration of the coating bake unit 404.
Further, the acquired coating development conditions include the temperature distribution in the coating bake unit 404 and the film thickness distribution (average value and dispersion of the film thickness) of the resist on the substrate W.
The acquired coating and developing conditions include changes between the substrates W such as the film thickness when the resist is applied on the plurality of substrates W, and the film thickness distribution (average value of the film thickness) when the lot of the resist is switched. And dispersion) fluctuations are included.
The acquired coating and developing conditions include the rate of resist coating defects on the substrate W, the machine difference between each unit, the temperature difference between each unit, the presence or absence of foreign matter on the substrate W, and the foreign matter on the substrate W. The location and uneven coating of the resist on the substrate W are included.

The applied and developed conditions to be acquired include a developing method such as a sequence of adding a developer to the substrate W, the number of rotations of the developing cup unit 406, and the rotation sequence (rotational speed profile) of the developing cup unit 406.
Further, the acquired coating and developing conditions include the type of rinsing liquid for removing the developing liquid (for example, pure water) and the discharge timing of the rinse liquid (discharge sequence, discharge profile).
The applied and developed conditions to be acquired include the shape of the developer discharge portion, the configuration of the developing cup unit 406, the configuration of the developing bake unit 405, and the temperature distribution in the developing bake unit 405.

The coating and developing conditions to be acquired include changes between the substrates W such as the film thickness when developing a plurality of substrates W, and the film thickness distribution (average value and dispersion of film thickness) when the lot of the developer is switched. ) Fluctuations are included.
Some of the information mentioned above may be stored in the main control unit 127 in advance.

Next, in step S602, the main control unit 127 acquires the pattern forming conditions of the substrate W in the exposure apparatus 100.
The pattern forming conditions of the substrate W acquired here include, for example, lighting conditions, focus, layer information of the reticle R, and the like.

Further, the acquired pattern formation condition of the substrate W also includes the exposure amount data used as the teacher data of the learning model.
Here, the exposure amount data is created by actually measuring the line width of the pattern on the substrate W developed as described above and acquiring the exposure amount that achieves the desired line width. be able to.
Further, the exposure amount referred to here includes the distribution of the exposure amount (the average value and the dispersion of the exposure amount) in the exposure region on the substrate W.

Next, in step S603, the main control unit 127 determines whether or not the processes of steps S601 and S602 have been executed for all the substrates W.
Then, when the processes of steps S601 and S602 are executed for all the substrates W (Yes in step S603), the process proceeds to step S604.

On the other hand, if there is a substrate W for which the processes of steps S601 and S602 have not been executed (No in step S603), the process returns to step S601.
That is, when there are coating development conditions and pattern forming conditions to be acquired for the plurality of substrates W, the processes of steps S601 to S603 are repeated for each of the substrates W.

Then, in step S604, the main control unit 127 acquires a learning model by learning the learning data composed of the input data and the teacher data acquired in steps S601 and S602.
Specifically, the main control unit 127 creates input data from the coating development conditions acquired in step S601 and the pattern forming conditions acquired in step S602.
Then, the main control unit 127 creates teacher data from the exposure amount data included in the pattern formation conditions acquired in step S602, and acquires a learning model by performing learning.

The learning model acquired here is a model in which internal random variables are optimized by using an algorithm such as an error back propagation method in a neural network composed of, for example, a multi-layer perceptron.
Then, as will be described later, by inputting the input data including the coating development conditions and the pattern forming conditions acquired for the new substrate W into the acquired learning model, the output data representing the exposure amount for the substrate W is input. Can be obtained.

If the coating development conditions and pattern forming conditions newly acquired or already acquired by the main control unit 127 include image information such as a film thickness distribution image, a foreign matter image, and a coating unevenness image, a convolutional neural network is used. You may acquire a learning model using.
Further, when the coating development condition and the pattern formation condition acquired by the main control unit 127 include time-series information, a recurrent neural network may be used when acquiring the learning model.
If the number of training data is small, a support vector machine may be used to acquire the training model.

Further, the main control unit 127 learns to acquire output data related to the coating and developing conditions by learning the coating and developing conditions acquired in step S601 as teacher data and the pattern forming conditions acquired in step S602 as input data. You may get a model.
That is, as will be described later, the main control unit 127 may acquire the output data regarding the coating development conditions of the substrate W by inputting the input data regarding the acquired pattern formation conditions into the acquired learning model.

In step S605, the main control unit 127 stores the information of the learning model acquired in step S604 in a storage device (not shown) provided in the exposure apparatus 100.
Here, the information of the learning model may include information representing the structure of the neural network such as the number of layers of the perceptron and the number of neurons, and information of the optimized random variable.

FIG. 7 is a flowchart showing a method of determining the exposure amount using the acquired learning model in the main control unit 127, which is the information processing apparatus according to the present embodiment.

First, in step S701, the main control unit 127 acquires the information of the learning model stored in step S605, that is, the information representing the structure of the above neural network and the information of the optimized random variable.

Next, in step S702, the main control unit 127 acquires the pattern forming conditions of the substrate W in the exposure apparatus 100.
The pattern forming conditions acquired here include the same conditions as the pattern forming conditions acquired in step S602, specifically, lighting conditions, focus, layer information of the reticle R, and the like.
In step S702, the pattern formation conditions for each of the substrates W may be sequentially acquired, or the pattern formation conditions for the plurality of substrates W may be collectively acquired at once.

Next, in step S703, the main control unit 127 acquires the coating and developing conditions of the resist on the substrate W by the coating and developing apparatus 400 from the coating and developing apparatus 400 via the network 501.
The coating and developing conditions acquired here include the same conditions as the coating and developing conditions acquired in step S601.

Then, in step S704, the main control unit 127 creates a learning model from the information acquired in step S701. Then, by inputting the input data composed of the pattern forming conditions and the coating development conditions acquired in steps S702 and S703 into the created learning model, the output data indicating the exposure amount is acquired.

Then, in step S705, the main control unit 127 outputs the acquired exposure amount, so that the exposure apparatus 100 exposes the substrate W based on the output exposure amount.
At this time, the main control unit 127 may calculate a statistical value indicating variation between the exposure amounts for each of the substrates W from the result output from the learning model.
As a result, it is possible to set a threshold value for the exposure amount based on the calculated statistical value and not to adopt it when the exposure amount output from the learning model exceeds the threshold value.

When using a learning model that outputs information about the coating and developing conditions, the pattern forming conditions acquired in step S702 are input to the learning model created from the information acquired in step S701 to obtain the coating and developing conditions. Output data about is output.
Then, the main control unit 127 outputs the acquired coating and developing conditions to the coating and developing apparatus 400, so that the coating and developing of the substrate W is performed in the coating and developing apparatus 400 based on the output coating and developing conditions.

The above series of processes is performed by the main control unit 127 provided in the exposure apparatus 100, but is not limited to this, and is not limited to this, for example, an information processing apparatus such as a control unit (not shown) provided in the coating / developing apparatus 400. It may be done using.

Further, the above series of processes may be performed by combining the main control unit 127 provided in the exposure apparatus 100 and the control unit provided in the coating and developing apparatus 400.
In this case, for example, the main control unit 127 provided in the exposure apparatus 100 acquires the learning model, and outputs the information of the acquired learning model to the control unit provided in the coating / developing apparatus 400.
Then, the control unit provided in the coating and developing apparatus 400 creates a learning model, and the exposure amount can be determined using the created learning model.

As described above, in the information processing apparatus according to the present embodiment, the conditions for applying the photosensitive agent to the substrate, the conditions for the photosensitive agent, the conditions for exposing the photosensitive agent, and the exposed photosensitive agent are used. The conditions for developing are used as input data.
Further, a learning model is created by using information on the exposure amount in the exposure apparatus 100 when forming a pattern on the substrate as teacher data.

As a result, even if the parameters under the above conditions included in the input data are very diverse, it is possible to efficiently acquire information on the exposure amount from the learning model created by machine learning.

[Manufacturing method of goods]
Next, a method of manufacturing an article using the exposure apparatus 100 including the information processing apparatus according to the present embodiment will be described.

The method for manufacturing an article including a semiconductor IC element, a liquid crystal display element, a MEMS, and the like includes a step of first using an exposure apparatus 100 to expose a substrate such as a wafer or a glass substrate coated with a photosensitive agent.
Further, the method for manufacturing an article includes a step of developing an exposed substrate (photosensitive agent) and a step of processing the developed substrate by another well-known processing step.
In addition, other well-known processing steps referred to here include etching, photosensitive agent peeling, dicing, bonding, packaging and the like.

According to the method for manufacturing an article according to the present embodiment, it is possible to manufacture an article of higher quality 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 method for determining the exposure amount shown above, the program for carrying out the method, and the computer-readable recording in which the program is recorded. The medium is also included in the scope of this embodiment.

100 Exposure device 127 Main control unit (information processing device)
W board

Claims (12)

An information processing device characterized by determining information on an exposure amount in an exposure device when forming a pattern on a substrate by using a learning model acquired by machine learning. The information processing apparatus according to claim 1, wherein the information regarding the exposure amount includes a distribution of the exposure amount in the exposure region on the substrate. The input data of the learning model is the conditions for applying the photosensitive agent to the substrate, the conditions for the photosensitive agent, the conditions for exposing the photosensitive agent, and the conditions for developing the exposed photosensitive agent. The information processing apparatus according to claim 1 or 2, wherein the information processing apparatus comprises at least one of them. The conditions for applying the photosensitive agent to the substrate include at least one of the rotation speed of the coating cup unit provided in the coating developing apparatus, the rotation sequence of the coating cup unit, and the discharge timing of the rinse liquid. The information processing apparatus according to claim 3, wherein the information processing apparatus is characterized by the above. The information processing apparatus according to claim 3 or 4, wherein the conditions of the photosensitive agent include at least one of the viscosity and the type of the photosensitive agent. The condition for exposing the photosensitive agent according to any one of claims 3 to 5, wherein the condition for exposing the photosensitive agent includes at least one of the illumination condition, the focus, and the layer information of the original plate in the exposure apparatus. Information processing equipment. The conditions for developing the exposed photosensitive agent are at least one of the number of rotations of the developing cup unit provided in the coating and developing apparatus, the rotation sequence of the developing cup unit, and the shape of the developer ejection portion. The information processing apparatus according to any one of claims 3 to 6, wherein the information processing apparatus comprises one. The information processing apparatus is among the conditions for applying the photosensitive agent to the substrate, the conditions for the photosensitive agent, the conditions for exposing the photosensitive agent, and the conditions for developing the exposed photosensitive agent. Input data is created by acquiring at least one from the exposure apparatus and the coating and developing apparatus.
Information on the line width of the pattern, which is output by measuring the pattern of each of the plurality of substrates on which the pattern is formed by performing exposure under the conditions of different exposure amounts by the exposure device, is acquired. , Create teacher data by determining the exposure amount corresponding to the predetermined line width,
The information processing apparatus according to any one of claims 1 to 7, wherein the learning model is created from the created input data and teacher data. 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 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. A method for determining an exposure amount, which comprises a step of determining information on an exposure amount in an exposure apparatus when forming a pattern on a substrate by using a learning model acquired by machine learning. A computer-readable recording medium on which a program that causes a computer to determine the exposure amount is recorded.
A computer can read a program recording a program characterized by having a computer perform a step of determining information about an exposure amount in an exposure apparatus when forming a pattern on a substrate using a learning model acquired by machine learning. Recording medium.

Images