阅读说明：本技术 用于数字光刻工具的基于模型的动态位置校正 (Model-based dynamic position correction for digital lithography tools ) 是由 塔纳·科斯坤 穆罕默德·波伊拉兹 钟钦 帕查·蒙哥尔旺格罗恩 于 2020-01-16 设计创作，主要内容包括：本公开内容一般涉及光刻系统以及用于校正光刻系统中的位置误差的方法。在第一次启动光刻系统时,所述系统进入稳定期。在所述稳定期期间,在所述系统印刷或曝光基板时,收集位置读数和数据,诸如温度、压力和湿度数据。基于所收集的数据和位置读数来产生模型。在后续稳定期中,使用所述模型来估计误差,并且在所述后续稳定期期间动态地校正所估计的误差。(The present disclosure relates generally to a lithography system and a method for correcting position errors in a lithography system. At the first start-up of the lithography system, the system enters a stabilization phase. During the stabilization period, position readings and data, such as temperature, pressure, and humidity data, are collected as the system prints or exposes substrates. A model is generated based on the collected data and the location readings. In a subsequent stationary phase, an error is estimated using the model, and the estimated error is dynamically corrected during the subsequent stationary phase.)

1. A method, comprising:

starting a photoetching system and entering a stable period;

collecting data and position readings while the lithography system is printing during the stabilization period;

generating a model from the data and the location readings; and

the model is used to dynamically correct estimation errors during a subsequent stabilization period.

2. The method of claim 1, wherein the collected data is temperature data, and wherein the temperature data is collected using a plurality of temperature sensors disposed throughout the lithography system.

3. The method of claim 2, wherein the model is formed using one or more parameters selected from the group consisting of: a position at which the position reading should be located without thermal effects during the stabilization period, a position at which the position reading actually is located due to thermal effects, an approximation of a disturbance of the position reading, an initial temperature of the lithography system, and a measured temperature change from the initial temperature after a predetermined time.

4. The method of claim 2, wherein the temperature data is collected during a heating period and a cooling period of the stabilization period.

5. The method of claim 1, wherein the collected data is pressure data, or wherein the collected data is humidity data.

6. The method of claim 1, wherein the model is a set of cascaded transient models.

7. A method, comprising:

starting a photoetching system and entering a stable period;

collecting temperature data and position readings while the lithography system is printing during the stabilization period, wherein the temperature data is collected during a heating period and a cooling period;

generating a model from the temperature data and the position readings;

correcting the model;

estimating an error using the corrected model during a subsequent stabilization period; and

dynamically correcting the estimated error during the subsequent stabilization period.

8. The method of claim 7, wherein pressure data is further collected and the model is generated based on the pressure data, or wherein humidity data is further collected and the model is generated based on the humidity data.

9. The method of claim 7, wherein the model is a set of cascaded transient models.

10. The method of claim 7, wherein the temperature data is collected using a plurality of temperature sensors disposed throughout the lithography system.

11. The method of claim 7, wherein the model is formed using one or more parameters selected from the group consisting of: the position readings should be during the stabilization period without thermal effects, the position readings are actually an approximation of disturbances of the position readings due to thermal effects, an initial temperature of the lithography system, and a measured temperature change from the initial temperature after a predetermined time.

12. A method, comprising:

starting a photoetching system and entering a stable period;

collecting temperature data and position readings while the lithography system is printing during the stabilization period;

generating a model from the temperature data and the position readings;

forming an optimization problem to determine a thermal capacitance and a transfer coefficient of the lithography system;

estimating an error using the model and the optimization problem in a subsequent stabilization phase; and

dynamically correcting the estimated error during the subsequent stabilization period.

13. The method of claim 12, wherein pressure data is further collected and the model is generated based on the pressure data, or wherein humidity data is further collected and the model is generated based on the humidity data.

14. The method of claim 12, wherein the temperature data is collected during a heating period and a cooling period during the stabilization period, and wherein the model is a set of cascaded transient models.

15. The method of claim 12, wherein the model is formed using one or more parameters selected from the group consisting of: the position readings should be during the stabilization period without thermal effects, the position readings are actually an approximation of disturbances of the position readings due to thermal effects, an initial temperature of the lithography system, and a measured temperature change from the initial temperature after a predetermined time.

Technical Field

Embodiments of the present disclosure generally relate to lithography systems (photolithography systems) and to methods for correcting position errors (positional errors) in lithography systems.

Background

Photolithography (Photolithography) is widely used to manufacture semiconductor devices and display apparatuses, such as Liquid Crystal Displays (LCDs). Large area substrates are commonly used to manufacture LCDs. LCDs or flat panels are commonly used for active matrix displays (active matrix displays) such as computers, touch panel devices, Personal Digital Assistants (PDAs), cell phones, television monitors, and the like. In general, a flat panel may include a layer of liquid crystal material forming a pixel sandwiched between two plates. When power from a power source is applied to the liquid crystal material, the amount of light passing through the liquid crystal material can be controlled at the pixel location to enable the generation of an image.

Microlithography techniques are commonly used to create electrical features that are incorporated as part of the layer of liquid crystal material forming the pixels. According to this technique, a light-sensitive photoresist (photoresist) is typically applied to at least one surface of the substrate. The pattern generator then uses the light to expose selected areas of the photosensitive photoresist as part of the pattern to cause chemical changes in the photoresist (photoresist) in the selected areas to prepare these selected areas for subsequent material removal and/or material addition processes to produce an electrical function.

However, the tools used for such microlithography may take 8 hours or more to completely stabilize the printing and patterning behavior, during which the pattern of the photoresist may become non-uniform due to various influences (e.g., thermal variations). This tool includes many heat sources and components with different conductivities and thermal capacitances, each of which can lead to variations that produce non-uniform patterning, negatively impacting total pitch and overlay correction repeatability.

In order to continue to provide display devices and other devices to consumers at the prices demanded by consumers, new apparatuses, methods, and systems are needed to accurately (default) and economically (cost-effectiveness) create patterns on substrates, such as large area substrates.

Disclosure of Invention

The present disclosure relates generally to a lithography system and a method for correcting position errors in a lithography system. When the photoetching system is started for the first time, the system enters a stable period. Position readings and data, such as temperature, pressure, and humidity data, are collected as the system prints or exposes the substrate during the stabilization period. A model is generated based on the collected data and the location readings. The model is used to estimate the error in the subsequent settling period, and the error estimated in the subsequent settling period is dynamically corrected.

In one embodiment, a method comprises: starting a photoetching system and entering a stable period; collecting data and position readings while the lithography system is printing during a stabilization period; generating a model based on the data and the location readings; and using the model to dynamically correct the estimated error during a subsequent stabilization period.

In another embodiment, a method comprises: the lithography system is started and enters a stabilization period during which temperature data and position readings are collected while the lithography system is printing. Temperature data was collected during the heating and cooling periods. The model further includes generating a model based on the temperature data and the location readings; calibrating the model; estimating an error using the calibrated model during a subsequent stabilization period; and dynamically correcting the estimated error in a subsequent settling period.

In yet another embodiment, a method comprises: starting a photoetching system and entering a stable period; collecting temperature data and position readings while the lithography system is printing during a stabilization period; generating a model based on the temperature data and the location readings; forming an optimization problem for determining thermal capacitance and transfer coefficient of the lithography system; estimating an error using the model and the optimization problem in a subsequent stabilization phase; and dynamically correcting the estimated error during a subsequent stabilization period.

Drawings

So that the manner in which the above recited features of the present disclosure can be understood in detail, a more particular description of the disclosure, briefly summarized above, may be had by reference to embodiments, some of which are illustrated in the appended drawings. It is to be noted, however, that the appended drawings illustrate only exemplary embodiments and are therefore not to be considered limiting of its scope, for the invention may admit to other equally effective embodiments.

FIG. 1A is a perspective view of a lithography system according to one embodiment.

FIG. 1B is a perspective view of a lithography system according to another embodiment.

Fig. 2 is a perspective schematic view of an image projection device according to embodiments disclosed herein.

FIG. 3 illustrates a method of modeling and calibrating system behavior to estimate positional disturbances occurring during a stabilization period, according to embodiments disclosed herein.

Fig. 4A-4F illustrate exemplary graphs of data measurements according to embodiments disclosed herein.

Fig. 5 illustrates an alignment configuration of a first bridge component and a second bridge component, each having a plurality of eyes disposed thereon, according to embodiments of lithography herein.

Fig. 6A-6C illustrate exemplary graphs of data measurements and position readings at 200mm/s platform speed (stage speed) according to embodiments disclosed herein.

7A-7C illustrate exemplary graphs of data measurements and position readings at a platform speed of 100mm/s according to embodiments disclosed herein.

To facilitate understanding, identical reference numerals have been used, where possible, to designate identical elements that are common to the figures. It is contemplated that elements and features of one embodiment may be beneficially incorporated in other embodiments without further recitation.

DETAILED DESCRIPTION OF EMBODIMENT (S) OF INVENTION

The present disclosure relates generally to a lithography system and a method for correcting position errors in a lithography system. When the lithography system is first started, the system enters a stable period. During this stabilization period, position readings and data (e.g., temperature, pressure, and humidity data) are collected as the system prints or exposes the substrate. A model is generated based on the collected data and the location readings. This model is then used to estimate the error in the subsequent stationary phase and to dynamically correct the estimated error (estimated errors) during the subsequent stationary phase.

Fig. 1A is a perspective view of a lithography system 100 according to embodiments disclosed herein. The lithography system 100 includes a bottom frame 110, a plate 120, a stage 130, and a processing apparatus 160. The bottom frame 110 may be placed on the floor of a manufacturing facility and support the plate 120. The passive air isolator 112 is located between the bottom frame 110 and the plate 120. In one embodiment, the plate 120 is a monolithic piece of granite, and the platform 130 is disposed on the plate 120. The substrate 140 is supported by the stage 130. A plurality of holes (not shown) are formed in the platform 130 for allowing a plurality of lift pins (not shown) to extend therethrough. In some embodiments, the lift pins may be raised to an extended position, for example, from one or more transfer robots (not shown), to receive the substrate 140. The one or more transfer robots are used to load and unload the substrate 140 from the stage 130.

Substrate 140 comprises any suitable material, such as an Alkaline Earth Boro-Aluminosilicate (alkali Earth Boro-Aluminosilicate), used as part of a flat panel display. In other embodiments, the substrate 140 is made of other materials. In some embodiments, the substrate 140 has a photoresist formed thereon. The photoresist is sensitive to light radiation. Positive photoresist (positive photoresist) includes portions of photoresist exposed to light radiation that will be respectively soluble to a photoresist developer provided to the photoresist after the pattern is written into the photoresist. Negative photoresist (negative photoresist) includes portions of the photoresist exposed to light radiation that will be insoluble to a photoresist developer provided to the photoresist, respectively, after the pattern is written into the photoresist. The chemical composition of the photoresist determines whether the photoresist will be a positive photoresist or a negative photoresist. Examples of photoresists include, but are not limited to, at least one of diazonaphthoquinone (diazonaphthoquinone), phenolic resin (phenol formaldehyde), poly (methyl methacrylate), poly (methylglutaronide), and SU-8. In this manner, a pattern may be created on the surface of the substrate 140 to form an electronic circuit.

The system 100 includes a pair of supports 122 and a pair of rails 124. The pair of supports 122 are disposed on the plate 120, and in one embodiment, the plate 120 and the pair of supports 122 are a one-piece material. The pair of rails 124 is supported by the pair of supports 122, and the stage 130 moves along the rails 124 in the x-direction. In one embodiment, the pair of tracks 124 is a pair of parallel magnetic channels. As shown, each track 124 of the pair of tracks 124 is linear. In another embodiment, air bearings are used for high precision non-contact motion, and linear motors are configured to provide forces in the x-direction and y-direction to move the platform 130 back and forth. In other embodiments, one or more of the tracks 124 are non-linear. The encoder 126 is coupled to the stage 130 to provide position information to a controller (not shown).

The processing apparatus 160 comprises a support 162 and a processing unit 164. The support 162 is disposed on the plate 120 and includes an opening 166. The opening 166 allows the platen 130 to pass under the processing unit 164. The processing unit 164 is supported by the support 162. In one embodiment, the processing unit 164 is a pattern generator configured to expose photoresist in a photolithography process. In some embodiments, the pattern generator is configured to perform a maskless lithography process. The processing unit 164 includes a plurality of image projection devices (shown in fig. 2). In one embodiment, processing unit 164 includes up to 84 image projection devices. Each image projection device is provided in the box 165. Processing device 160 may be useful for performing maskless direct patterning.

During operation, this stage 130 is moved in the X direction from the loading position as shown in fig. 1A to the processing position. The processing position means one or more positions of the stage 130 when the stage 130 passes under the processing unit 164. During operation, the platen 130 is lifted by a plurality of air bearings (not shown) and moved from a loading position along a pair of rails 124 to a processing position. A plurality of vertical guide air bearings (not shown) are coupled to the platform 130 and positioned adjacent to the inner wall 128 of each support 122 to stabilize the movement (moment) of the platform 130. This stage 130 is also moved in the Y-direction by moving along a track 150 to process and/or index (indexing) the substrate 140. The stage 130 is capable of independent operation and may scan the substrate 140 in one direction and step in another direction.

The metrology system measures the X and Y lateral position coordinates of each stage 130 in real time so that each of the plurality of image projection devices can accurately position a pattern written into the photoresist covering the substrate. The metrology system also measures the angular position of each platform 130 about the vertical or z-axis in real time. The angular position measurement may be used during scanning by the servomechanism to keep the angular position constant, or the angular position measurement may be used to correct the position of the pattern written on the substrate 140 by the image-projecting device 270, as shown in FIG. 2. These techniques may also be used in combination.

FIG. 1B is a perspective view of a lithography system 190 according to embodiments disclosed herein. System 190 is similar to system 100; however, the system 190 includes two platforms 130. Each of the two stages 130 is capable of independent operation and may scan the substrate 140 in one direction and step in the other direction. In some embodiments, when one of the two stages 130 scans the substrate 140, the other of the two stages 130 unloads the exposed substrate and loads the next substrate to be exposed.

Although fig. 1A-1B depict two embodiments of a lithography system, other systems and configurations are also contemplated herein. By way of example, lithographic systems that include a stage of any suitable order of magnitude are also contemplated.

FIG. 2 is a perspective schematic view of an image-projecting device 270 according to one embodiment, where such an image-projecting device 270 is useful for a lithography system, such as system 100 or system 190. The image-projecting device 270 includes one or more spatial light modulators (spatial light modulators)280, an alignment and inspection system (alignment and inspection system)284 having a focus sensor 283 and a camera 285, and projection optics (projection optics) 286. The components of the image projection apparatus vary depending on the spatial light modulator used. Spatial light modulators include, but are not limited to, micro-leds, digital micro-mirror devices (DMDs), Liquid Crystal Displays (LCDs), and vertical-cavity surface-emitting lasers (VCSELs).

In operation, the spatial light modulator 280 is used to modulate one or more characteristics (e.g., amplitude, phase, or polarization) of light projected by the image projection apparatus 270 to a substrate (e.g., substrate 140). Alignment and inspection system 284 is used to align and inspect the components of image-projecting device 270. In one embodiment, the focus sensor 283 includes a plurality of laser lights that are directed through the lens of the camera 285 and through the back side of the camera 285 and imaged onto the sensor to detect whether the image-projecting device 270 is in focus. The camera 285 is used to image a substrate (e.g., substrate 140) to ensure that the alignment of the image-projecting device 270 and the lithography system 100 or 190 is correct or within a predetermined tolerance. Projection optics 286 (e.g., one or more lenses) are used to project light onto a substrate (e.g., substrate 140).

When the lithography system 100, 190 is first started up, the system 100, 190 enters a stabilization phase. The stabilization period refers to the time required to stabilize the printing and patterning behavior of the system (i.e., the time required for the system to fully warm up). During the settling period of the lithography system 100, 190, various effects and variations (e.g. thermal variations) occur, which may have a negative impact on the total pitch and overlay correction repeatability. In some cases, due to various effects and variations, the lithography system 100, 190 may take eight hours or more to stabilize the printing and patterning behavior. Furthermore, each of the lithography systems 100, 190 includes many heat sources and components with different conductivity coefficients and thermal capacitances, each of which can potentially cause variations, making it difficult to strictly monitor the systems 100, 190.

To use the system 100, 190 during the stabilization period to accurately and precisely expose the substrate, model-based software corrections may be used to correct any errors that occur during the stabilization period. The behavior of the system 100, 190 may be modeled and calibrated to estimate the potential changes that occur during the stabilization period, as shown in FIG. 3 below, to enhance total pitch and overlay correction repeatability (total pitch and overlay correction repeatability). This model may then be used during subsequent stabilization periods of the system 100, 190 to correct for overlay and total pitch errors. Using this model for dynamic position correction eliminates or reduces expensive hardware solutions. In addition, since position correction can be applied to digital masks (digital masks), this model can be used for dynamic position correction.

FIG. 3 illustrates a method 300 of modeling and calibrating system behavior to estimate positional disturbances occurring during a stabilization period, according to embodiments disclosed herein. The method 300 may be used with the lithography systems 100, 190 of fig. 1A and 1B, respectively.

The method 300 has an operation 302 in which the lithography system is started up and enters a stabilization period. During the stabilization period, the printing and patterning behavior of the system may be unstable due to various effects and variations (e.g., heat, pressure, and/or humidity variations). The stabilization period means the time required for the printing and patterning behavior of the system to stabilize (i.e., the time required for the system to fully warm up).

In operation 304, the lithography system collects data and position readings while printing or exposing the substrate during the stabilization period. Data is continuously collected as the system aligns and exposes the substrate to mimic the production line. In one embodiment, the collected data is temperature data. Temperature data may be collected using one or more temperature sensors disposed adjacent to components of known tools that fluctuate in temperature during the heating and cooling periods, such as encoders. For example, approximately 20 temperature sensors may be provided on the lithography tool to collect and monitor the temperature of the chuck (chuck), encoder (encoder), bridge/riser (riser), etc.

To collect position readings, alignment marks on the calibration plate or substrate may be captured periodically throughout the stabilization period (as shown in FIG. 5). The calibration plate may be used as a reference during the stabilization period. Additionally or alternatively, encoder count changes relative to interferometer readings may further be recorded periodically, with the interferometer serving as a reference. The relative change in position with respect to the reference used is then recorded.

Due to thermal effects and fluctuations that occur during the stabilization period, the pattern printing position on the substrate or calibration plate may be unintentionally disturbed. Thus, disturbances in the position readings on the substrate or calibration plate may be directly related to temperature fluctuations. Other effects may also cause interference in position readings, such as pressure, humidity, etc. In this case, a sensor configured to collect pressure data, humidity data, or the like may be used instead of or in addition to the temperature sensor. However, thermal effects will be exemplified throughout.

Fig. 4A-4F show exemplary graphs of data measurements and position readings. Fig. 4A-4F are merely examples of data measurements and are not intended to be limiting. Fig. 4A shows the temperature change in degrees celsius over a period of time at a platform speed of 200mm/s for the bridge component and riser component of the system. Fig. 4B shows the corresponding position marks (in microns) found along the y-axis during the heating period for the bridge and riser components at a platform speed of 200mm/s, illustrating the positional interference caused by thermal effects. Figure 4C shows the change in celsius over time of the first bridge member, the second bridge member and the riser member in the system at a platform speed of 100 mm/s. Fig. 4D shows the corresponding position marks (in microns) found along the y-axis during the heating period for the bridge and riser components at a platform speed of 100mm/s, further illustrating the y-axis position disturbance caused by thermal effects. FIG. 4E shows temperature readings in degrees Celsius of a master motor and a slave motor moving a stage in a lithography system over a period of time. FIG. 4F shows the position markers found over time along the x-axis and y-axis during the cool down period. Fig. 4A to 4F show that the system behavior during the stabilization period can be expressed mathematically.

Fig. 5 illustrates an alignment configuration 500 of a first bridge member 504 and a second bridge member 506, wherein each of the first bridge member 504 and the second bridge member 506 has a plurality of eyes 508 disposed thereon, according to one embodiment. The alignment configuration 500 may be used to collect data for the graphs shown in fig. 4A-4F above, and for the graphs shown in fig. 6A-6C and 7A-7C below. First bridge member 504 and second bridge member 506 are disposed above substrate or plate 502. The plate 502 includes a plurality of alignment marks 510. Although 32 alignment marks 510 are shown, any number of alignment marks may be used. Further, while two bridge members 504, 506 are shown, additional bridge members may be utilized in the lithography system, and each of the bridge members 504, 506 may have more than four eyes disposed thereon. The alignment arrangement 500 may include an exposure unit having a camera (not shown) for collecting position readings.

Fig. 6A-6C show exemplary graphs of data measurements and position readings at a platform speed of 200 mm/s. Fig. 7A-7C show exemplary graphs of data measurements and position readings at a platform speed of 100 mm/s. Fig. 6A-6C and 7A-7C are merely examples of data measurements and are not intended to be limiting. Any number of temperature sensors and any number of position markers placed on the plate may be used to collect or measure the temperature and position data shown in the graphs of fig. 6A-6C and 7A-7C.

Fig. 6A and 7A show position markers found along the x-axis (in microns) over a period of time (in hours) during a stabilization period, showing x-axis position disturbances caused by thermal effects. Fig. 6B and 7B show the temperature measured at two different locations on the chuck of the system over a period of time (in hours) during the stabilization period. Fig. 6C and 7C show the celsius temperatures of the first, second, and third encoders of the system over a period of time (in hours) during the stabilization period.

In operation 306, a model is generated based on the collected data and the location readings. Such a model may include more than one subset of data, such as with a model generated to account for temperature effects, pressure effects, and/or humidity effects, among others. In generating this model, the system is assumed to be linear or weakly non-linear. This model may use efficient thermal capacitance and transfer as model parameters. This model may further allow for operating the system in an iterative manner. The data depicted in one or more of fig. 4A-4F, 6A-6C, and 7A-7C may be used alone or in combination to help generate this model.

Furthermore, dynamic eye-to-eye (eye-to-eye) and/or bridge-to-bridge (bridge) models may be incorporated into the generated models. This model may capture the change in center of the eye relative to one another, or capture the drift in the spacing between bridges (e.g., first bridge component 504, second bridge component 506, and eye 508 of fig. 5). Dynamic eye-to-eye (eye-to-eye) and/or bridge-to-bridge (bridge-to-bridge) models may be empirical models, and model parameters may be calibrated based on experimental results.

Operations 302 and 304 may be repeated one or more times to collect a greater amount of data for use in generating such a model. This model may be a cascaded transient model (cascaded transient model) that relates position error to multiple sensor readings. The transient response of each component depends on the thermal capacitance or mass and the thermal transfer properties of the component. A cascaded empirical model can then be used to represent thermal effects in the system.

The model may be formed using the following variables or parameters: the position reading should be (x, y) during the stationary phase without thermal effect, the position reading is actually (x ', y') due to thermal effect, the approximate value of the position reading disturbance is (Δ x, Δ y) (i.e., the difference between the position reading without thermal effect and the position reading with thermal effect), the initial temperature (T;)₀) And a temperature change relative to the initial temperature reading (Δ T). In one embodiment, at least the initial temperature (T) must be known₀) And a temperature change (Δ T) relative to the initial temperature reading to formAnd (4) modeling. To approximate where the position reading is actually due to thermal effects, equations 1-4 may be used.

Equation 1:

equation 2:

equation 3:

equation 4:

in the case of the equations 3 and 4,is a spatial mode, and ∈ is a model transition between temperature and global position change. The positional disturbances (Δ x, Δ y) are formulated as a function of all temperature sensor readings, including previous readings and current readings.

In operation 308, the model is calibrated. Calibrating the model may include continuously operating the lithography system to model a stabilization period, and leaving the lithography system idle (idle) after the stabilization period to model a cool-down period. The model is further calibrated by forming an optimization problem. An optimization problem is formed to obtain model parameters that minimize the total cost function (C) (shown in equation 7 below). The cost is defined as the sum of the differences between the measured and model predicted values at a plurality of locations (x, y) representing the transition and a plurality of temperature conditions. An optimization problem may be formed to minimize the cost function.

An optimization problem is developed to determine a plurality of thermal capacitances and transfer coefficients for the system during the stabilization period. The input to the optimizer is the temperature readings collected at a plurality of locations and the corresponding location errors. The output of the optimizer is a set of thermal capacitances and transfer coefficients. Optimizer minimalizing substrateThe difference between the measured position of (a) and the model estimated position. The model can be calibrated using equations 5 through 7. Equation 5 and equation 6 are used for model prediction error, where x'_measAnd y'_measThe measurement position changes due to thermal effects. Equation 7 is a cost function, where K is the number of data collected and L is the number of calibration parameters.

Equation 5: epsilon_x(x,y)＝x′_meas-x′

Equation 6: epsilon_y(x,y)＝y′_meas-y′

Equation 7:

in operation 310, an error in a subsequent settling period is estimated using the calibration model and the estimated error in the subsequent settling period is dynamically corrected. After the model is calibrated, this model can be used during a subsequent stabilization period to correct predicted position errors and disturbances due to thermal effects. As noted above, thermal effects are only one type of effect or change that may be considered and are not meant to be limiting examples. The estimated position error during the stabilization period may be corrected by dynamically adjusting the digital mask of the real-time (on-the-fly) lithography system, rather than correcting or changing the physical lithography system itself. The correction of the estimated position error may be a dynamic digital correction applied during exposure of the digital mask on each plate (plate) or each substrate.

The calibration model may further be used to monitor the stability of the lithography system. An alignment model that models alignment of the digital mask may be formed based on the calibration model. The alignment model may then be compared to the alignment of the digital mask during a subsequent stabilization period. Comparison of the alignment model with the alignment during the subsequent stabilization period may be used to determine a similarity measure. The similarity metric may be used to determine whether the subsequent stationary phase is the same as the initial stationary phase used to generate the model (i.e., whether the subsequent stationary phase experiences the same positional disturbance as the initial stationary phase). The similarity metric may determine the stability of the system by determining whether the same positional interference repeatedly occurs. If the system is stable, it will be easier to estimate the potential position error, since the same error will repeatedly occur at the same point in time during the stabilization.

In at least one embodiment, the model may be a machine learning model or a problem with model guidance, such as a neural network. For example, if a large amount of data is available, the lithography system may use multiple sensors and a large amount of data to actively correct for positional disturbances or errors that have a high frequency of occurrence during subsequent stabilization periods before the errors occur. The system may use data, sensors, and/or models to estimate or determine errors with high frequency of occurrence and compensate for potential errors before they occur. After compensating for disturbances or errors having a high frequency of occurrence, the system may further compare the current print position to the model to determine if the compensation actually corrected the potential error, and may make other adjustments as needed. Thus, the system can use machine learning algorithms to actively compensate for potential errors before they occur, rather than correcting them in real time as they occur.

Using the above method, lithography system behavior can be accurately modeled and calibrated to estimate positional disturbances that occur during the stabilization period, which enhances overall pitch and overlay correction repeatability. The model is then used to correct overlay and total pitch errors in real time by adjusting the digital mask during a subsequent settling period of the system. Furthermore, if the system can use large amounts of data, the system can actively compensate for potential positional disturbances or errors before they occur using a machine learning model with model guidance.

Using the model for dynamic position correction may eliminate or reduce expensive hardware solutions. Since position correction is applied to the digital mask, the model can be easily used for dynamic position correction. Furthermore, because the model is a software-based solution, new model forms can be developed to include new effects that were not previously included or covered, or to include additional sensors that were not originally available. Thus, the lithography system can be accurately used for exposure during a stabilization period of the plate or substrate.

While the foregoing is directed to embodiments of the present disclosure, other and further embodiments of the disclosure may be devised without departing from the basic scope thereof, and the scope thereof is determined by the claims that follow.