1. A method of determining a correction (750) for a process parameter associated with a lithographic process on a substrate, the lithographic process comprising a plurality of runs during each of which a pattern is to be applied to one or more substrates, the method comprising:

obtaining pre-exposure metrology data relating to a property of the substrate (725);

obtaining post-exposure metrology data (775) comprising one or more measurements of a process parameter that have been performed by an equivalent lithographic process on one or more previously exposed substrates of the lithographic process;

assigning a panelist relationship status from one or more lots (715) to the substrate based on the pre-exposure metrology data (725), wherein the one or more lots (715) have been determined using a classifier (710) trained to classify pre-exposure metrology data associated with the one or more previously exposed substrates (770) according to characteristics of corresponding post-exposure metrology data (775) of the one or more previously exposed substrates (770); and

determining a correction (750) for the process parameter based on the panelist relationship status and the post-exposure metrology data (775), wherein each of the one or more lots (715) has an associated process correction.

2. The method of claim 1, wherein the pre-exposure metrology data (725) includes data describing a shape of the substrate in a direction perpendicular to the substrate plane.

3. The method of claim 1, wherein the pre-exposure metrology data (725) includes leveling data.

4. The method of claim 1, further comprising the step of identifying a most representative substrate of the group for a particular group.

5. The method of claim 4, comprising: measuring the most representative of the substrates of the set; and updating corrections (780a, 780b, 780c) to the process parameters corresponding to the set based on the measurements.

6. The method of claim 1, wherein the group membership status is a degree that enables assignment of membership to one or more of the groups to each substrate.

7. The method of claim 6, wherein the step of determining the correction of the process parameter is based on a weighting of corrections associated with one or more groups based on a degree of membership of the substrate to the one or more groups.

8. The method of claim 7, wherein the degree of membership to a group is based on a classification score representing a metric defining the presence of a fingerprint of the group in the pre-exposure metrology data (725).

9. The method of claim 8, comprising an initial training phase (700) in which the one or more groups (715) are determined from a principal component analysis performed on the pre-exposure metrology data (725), each group being defined by an intrinsic fingerprint determined from the principal component analysis.

10. The method of claim 1, comprising an initial training phase (700) in which the one or more groups (715) are determined.

11. The method of claim 10, wherein the initial training phase (700) comprises a validation step comprising: assigning substrates to the one or more lots based on historical data and simulating an effect on the post-exposure metrology data.

12. The method of claim 1, wherein the post-exposure metrology data (775) comprises overlay data.

13. The method of claim 1, further comprising: the method is performed for each substrate in a run during which a pattern is applied to the substrate.

14. A computer program product including one or more sequences of machine-readable instructions for implementing the steps of the method of any preceding claim.

Background

A lithographic process is one in which a lithographic apparatus applies a desired pattern onto a substrate (usually onto a target portion of the substrate), and then various chemical and/or physical processing steps complete the pattern to produce functional features of a complex product. Accurate placement of patterns on substrates is a major challenge for reducing the size of circuit components and other products that can be produced by photolithography. In particular, a challenge in accurately measuring features already placed on a substrate is the critical step of being able to position successive layers of stacked features with sufficient accuracy to produce a working device at high production rates. Generally, the so-called overlay should be achieved within tens of nanometers in today's submicron semiconductor devices, down to a few nanometers in the most critical layers.

Thus, modern lithographic apparatus involve a centralized measurement or "mapping" operation prior to the step of actually exposing or otherwise patterning the substrate at the target site. So-called advanced alignment models have been and continue to be developed to more accurately model and correct for the non-linear distortions of the wafer "grid" caused by the processing steps and/or the lithographic apparatus itself. However, not all deformations are correctable, and it is still important to track as many causes of such deformations as possible and to eliminate them.

Modern lithographic processes and products are so complex that problems due to processing are difficult to explore for root cause. The overlay and alignment residuals typically show the pattern on the wafer (of the process and/or lithography tool). This can be interpreted as an uncorrectable amount with respect to a predetermined model, while a visual inspection and detailed analysis of the fingerprint can give an indication of the cause and correction strategy. The spatial pattern in a fingerprint is not used to quantify the fingerprint, and observation that multiple causes may appear in an apparent fingerprint simultaneously is not used to quantify the fingerprint. Overlay measurements are generally not applicable to each individual wafer and the relationship to process history and context is generally not known or used. Furthermore, it is difficult and time consuming to list all possible sources of spatial variation for nearby machines and processes.

In addition to the problem of identifying the cause of processing errors, process performance monitoring systems have been implemented that allow for the measurement of performance parameters from the product being processed, which are then used to calculate corrections for use in processing subsequent products. A limitation on current performance monitoring systems is that there is a tradeoff between the amount of time and the instrumentation dedicated to performance monitoring and the speed and accuracy with which corrections can be implemented. In a "run-to-run" control strategy, historical performance measurements are fed back to calculate new process corrections using (e.g., online) measurements performed between or during "runs," which may include one or more batches. In previous continuous run control strategies, each run included a "batch" of typically 25 substrates. Improved lithographic apparatus hardware has enabled wafer level control whereby operations may include a single substrate. However, making a complete overlay measurement for each substrate to take advantage of this wafer level control would be prohibitive in terms of time and throughput.

Disclosure of Invention

The present invention is directed to improving systems for controlling the performance of parameters, such as overlay, in a lithographic process.

In another aspect, the present invention is directed to optimizing a continuous operation control strategy during high capacity manufacturing.

According to a first aspect of the present invention, there is provided a method of determining a correction for a process parameter relating to a lithographic process on a substrate, the lithographic process comprising a plurality of runs during each of which a pattern is applied to one or more substrates, the method comprising: obtaining pre-exposure parameter data related to attributes of the substrate; obtaining post-exposure metrology data comprising one or more measurements of a process parameter that have been performed by an equivalent lithographic process on one or more previously exposed substrates of the lithographic process; assigning a group membership status from one or more groups to the substrate based on the pre-exposure parameter data; and determining a correction for the process parameter based on the panelist relationship status and the post-exposure metrology data.

The present invention still further provides a method of manufacturing a device wherein device features are formed on a series of substrates by a patterning process, wherein a correction to a process parameter of the lithographic process is determined by performing the method of the first aspect.

The invention still further provides a control system for a lithographic apparatus, the control system exposing: a memory for receiving pre-exposure parameter data and post-exposure metrology data relating to an attribute of a substrate, the post-exposure metrology data comprising one or more measurements of a process parameter that have been performed on one or more previous substrates; and a processor operable to assign a panelist status from one or more of the groups to the substrate based on the pre-exposure parameter data; and determining a correction to the process parameter based on the panelist relationship status and the post-exposure metrology data.

The invention still further provides a lithographic apparatus comprising a control system according to an aspect of the invention set forth above.

The present invention still further provides a method of dynamically updating one or more groups and/or corrections to a process parameter related to a lithographic process on a substrate, wherein one of a plurality of corrections is applied to the process parameter for each substrate based on a group membership status assigned to the substrate, the method comprising: obtaining post-exposure metrology data describing performance parameters of the substrate; and dynamically updating the one or more and/or more corrections in the set based on the post-exposure metrology data.

The invention still further provides a computer program product comprising one or more sequences of machine-readable instructions for carrying out the computing steps in the method of any aspect of the invention set out above.

These and other aspects and advantages of the apparatus and methods disclosed herein will become apparent by consideration of the following description of exemplary embodiments and accompanying drawings.

Drawings

Embodiments of the invention will now be described, by way of example only, with reference to the accompanying schematic drawings in which corresponding reference symbols indicate corresponding parts, and in which:

FIG. 1 depicts a lithographic apparatus suitable for use in embodiments of the invention;

FIG. 2 depicts a lithographic cell or cluster in which an inspection apparatus according to the present invention may be used;

FIG. 3 schematically illustrates a measurement process and an exposure process in the apparatus of FIG. 1, according to known practice;

FIG. 4 is a schematic diagram of an advanced process control method for controlling the apparatus of FIG. 1 in accordance with known practice;

FIG. 5 is a flow chart depicting an initial setup phase according to an embodiment of the present invention;

FIG. 6 is a plot of scores for each substrate belonging to a lot against a first intrinsic wafer fingerprint;

FIG. 7 is a graph of scores (y-axis) for an intrinsic wafer belonging to measured overlay fingerprints versus measured alignment fingerprints (x-axis) for a batch of (25) substrates, where each point represents a different substrate;

FIG. 8 is a flow chart conceptually illustrating a method according to a first embodiment of the present invention;

FIG. 9 conceptually illustrates an exemplary grouping of substrates for pre-exposure metrology data and post-exposure metrology data; and

figure 10 is a flow chart conceptually illustrating a method according to a second embodiment of the present invention.

Detailed Description

Before describing embodiments of the present invention in detail, it is instructive to provide an exemplary environment in which embodiments of the present invention may be implemented.

FIG. 1 schematically depicts a lithographic apparatus LA. The apparatus comprises: an illumination system (illuminator) IL configured to condition a radiation beam B (e.g. UV radiation or DUV radiation); a patterning device support or support structure (e.g. a mask table) MT constructed to support a patterning device (e.g. a mask) MA and connected to a first positioner PM configured to accurately position the patterning device in accordance with certain parameters; two substrate tables (e.g. wafer tables) WTa and WTb, each constructed to hold a substrate (e.g. a resist-coated wafer) W and each connected to a second positioner PW configured to accurately position the substrate in accordance with certain parameters; and a projection system (e.g. a refractive projection lens system) PS configured to project a pattern imparted to the radiation beam B by patterning device MA onto a target portion C (e.g. comprising one or more dies) of the substrate W. The reference frame RF connects the various components and serves as a reference datum for setting and measuring the position of the patterning device and the substrate and features on the patterning device and the substrate.

The illumination system may include various types of optical components, such as refractive, reflective, magnetic, electromagnetic, electrostatic or other types of optical components, or any combination thereof, for directing, shaping, or controlling radiation. For example, in an apparatus using Extreme Ultraviolet (EUV) radiation, a reflective optical component will typically be used.

The patterning device support holds the patterning device in a manner that depends on the orientation of the patterning device, the design of the lithographic apparatus, and other conditions, such as for example whether or not the patterning device is held in a vacuum environment. The patterning device support may use mechanical, vacuum, electrostatic or other clamping techniques to hold the patterning device. The patterning device support MT may be a frame or a table, for example, which may be fixed or movable as required. The patterning device support may ensure that the patterning device is at a desired position, for example with respect to the projection system.

The term "patterning device" used herein should be broadly interpreted as referring to any device that can be used to impart a radiation beam with a pattern in its cross-section such as to create a pattern in a target portion of the substrate. It should be noted that the pattern imparted to the radiation beam may not exactly correspond to the desired pattern in the target portion of the substrate, for example if the pattern includes phase-shifting features or so called assist features. Generally, the pattern imparted to the radiation beam will correspond to a particular functional layer in a device being created in the target portion, such as an integrated circuit.

As here depicted, the apparatus is of a transmissive type (e.g. employing a transmissive patterning device). Alternatively, the apparatus may be of a reflective type (e.g. employing a programmable mirror array of a type as referred to above, or employing a reflective mask). Examples of patterning devices include masks, programmable mirror arrays, and programmable LCD panels. Any use of the terms "reticle" or "mask" herein may be considered synonymous with the more general term "patterning device". The term "patterning device" may also be construed to mean a device that stores, in digital form, pattern information used to control such a programmable patterning device.

The term "projection system" used herein should be broadly interpreted as encompassing any type of projection system, including refractive, reflective, catadioptric, magnetic, electromagnetic and electrostatic optical systems, or any combination thereof, as appropriate for the exposure radiation being used, or for other factors such as the use of an immersion liquid or the use of a vacuum. Any use of the term "projection lens" herein may be considered as synonymous with the more general term "projection system".

The lithographic apparatus may also be of a type: wherein at least a portion of the substrate may be covered by a liquid having a relatively high refractive index, e.g. water, so as to fill a space between the projection system and the substrate. Immersion liquids may also be applied to other spaces in the lithographic apparatus, for example, between the mask and the projection system. Immersion techniques are well known in the art for increasing the numerical aperture of projection systems.

In operation, the illuminator IL receives a radiation beam from a radiation source SO. For example, when the source is an excimer laser, the source and the lithographic apparatus may be separate entities. In such cases, the source is not considered to form part of the lithographic apparatus and the radiation beam is passed from the source SO to the illuminator IL with the aid of a beam delivery system BD comprising, for example, suitable directing mirrors and/or a beam expander. In other cases the source may be an integral part of the lithographic apparatus, for example when the source is a mercury lamp. The source SO and the illuminator IL, together with the beam delivery system BD if required, may be referred to as a radiation system.

The illuminator IL may, for example, comprise an adjuster AD for adjusting the angular intensity distribution of the radiation beam, an integrator IN, and a condenser CO. The illuminator may be used to condition the radiation beam, to have a desired uniformity and intensity distribution in its cross-section.

The radiation beam B is incident on the patterning device MA, which is held on the patterning device support MT, and is patterned by the patterning device. Having traversed the patterning device (e.g. mask) MA, the radiation beam B passes through the projection system PS, which focuses the beam onto a target portion C of the substrate W. With the aid of the second positioner PW and position sensor IF (e.g., an interferometric device, linear encoder, two-dimensional encoder or capacitive sensor), the substrate table WTa or WTb can be moved accurately, e.g., so as to position different target portions C in the path of the radiation beam B. Similarly, the first positioner PM and another position sensor (which is not explicitly depicted in fig. 1) can be used to accurately position the patterning device (e.g. mask) MA with respect to the path of the radiation beam B, e.g. after mechanical retrieval from a mask library, or during a scan.

Patterning device (e.g. mask) MA and substrate W may be aligned by using mask alignment marks M1, M2 and substrate alignment marks Pl, P2. Although the substrate alignment marks as shown occupy dedicated target portions, they may be located in spaces between target portions (these are known as scribe-lane alignment marks). Similarly, in situations in which more than one die is provided on the patterning device (e.g. mask) MA, the mask alignment marks may be located between the dies. Small alignment marks may also be included within the die among device features where it is desirable to identify features that are as small as possible and do not require any imaging or process conditions that differ from adjacent features. An alignment system for detecting alignment marks is described further below.

The depicted apparatus may be used in various modes. In scan mode, the patterning device support (e.g. mask table) MT and the substrate table WT are scanned synchronously while a pattern imparted to the radiation beam is projected onto a target portion C (i.e. a single dynamic exposure). The velocity and direction of the substrate table WT relative to the patterning device support (e.g. mask table) MT may be determined by the (de-) magnification and image reversal characteristics of the projection system PS. In scan mode, the maximum size of the exposure field limits the width (in the non-scanning direction) of the target portion in a single dynamic exposure, whereas the length of the scanning motion determines the height (in the scanning direction) of the target portion. Other types of lithographic apparatus and modes of operation are possible, as is well known in the art. For example, a step mode is known. In so-called "maskless" lithography, the programmable patterning device is held stationary, but with a changing pattern, and the substrate table WT is moved or scanned.

Combinations and/or variations on the above described modes of use or entirely different modes of use may also be employed.

The lithographic apparatus LA is of the so-called dual stage type, having two substrate tables WTa, WTb and two stations, an exposure station EXP and a measurement station MEA, between which the substrate tables can be exchanged. When one substrate on one substrate table is exposed at the exposure station, another substrate may be loaded onto another substrate table at the measurement station and various preparatory steps performed. This enables a substantial increase in the throughput of the device. On a single stage apparatus, the preparatory steps and the exposure steps need to be carried out sequentially on a single stage for each substrate. The preliminary steps may include mapping the surface height profile of the substrate using the level sensor LS and measuring the position of the alignment marker on the substrate using the alignment sensor AS. IF the position sensor IF is not capable of measuring the position of the substrate table at the measurement station and at the exposure station, a second position sensor may be provided to enable tracking of the position of the substrate table relative to the reference frame RF at both stations. Other arrangements are known and may be used in place of the dual platform arrangement shown. For example, other lithographic apparatus are known in which a substrate table and a measurement table are provided. The substrate table and the measurement table are docked together when performing the preliminary measurement and then undocked when the substrate table is subjected to the exposure.

As shown in FIG. 2, lithographic apparatus LA forms part of a lithographic cell LC and is sometimes referred to as a lithographic cell or cluster, which also includes apparatus for performing pre-exposure and post-exposure processes on a substrate. Conventionally, these devices comprise: a spin coater SC for depositing a resist layer, a developer DE for developing the exposed resist, a chill plate CH, and a bake plate BK. The substrate handler or robot RO picks up the substrate from the input/output ports I/O1, I/O2, moves the substrate between different process tools, and then transfers it to the load station LB of the lithographic apparatus. These devices are generally referred to collectively as tracks and are controlled by a track control unit TCU, which is itself controlled by a supervisory control system SCS, which also controls the lithographic apparatus via the lithographic control unit LACU. Thus, different equipment may be operated to maximize throughput and processing efficiency.

In order to properly and consistently expose a substrate exposed by a lithographic apparatus, it is desirable to inspect the exposed substrate to measure properties such as overlay error, line thickness, Critical Dimension (CD), etc. between subsequent layers. Thus, the manufacturing facility in which the lithography cell LC is located also includes a metrology system MET that receives some or all of the substrates W that have been processed in the lithography cell. The measurement results are directly or indirectly provided to the supervisory control system SCS. If an error is detected, the exposure of subsequent substrates may be adjusted.

Within the metrology system MET, an inspection apparatus is used to determine properties of a substrate, and in particular to determine how properties of different substrates or different layers of the same substrate vary between different layers. The inspection apparatus may be integrated into the lithographic apparatus LA or the lithographic cell LC or may be a stand-alone device. To achieve the fastest measurements, it may be desirable for the inspection apparatus to measure properties in the exposed resist layer immediately after exposure. However, not all inspection devices have sufficient sensitivity to make useful measurements of the latent image. Thus, measurements can be taken after a post-exposure bake step (PEB), which is typically the first step performed on an exposed substrate and increases the contrast between exposed and unexposed portions of the resist. At this stage, the image in the resist may be referred to as a semi-latent image. Measurements can also be made of the developed resist image at points where either the exposed or unexposed portions of the resist have been removed. In addition, substrates that have been exposed may be stripped and reworked-to improve yield-or discarded, thereby avoiding further processing of substrates that are known to be defective. In the event that only some target portions of the substrate are defective, further exposures may be performed only on those target portions that are good.

The metrology step may also be performed using the metrology system MET after the resist pattern has been etched into the product layer. The latter possibility limits the possibility of defective substrate rework, but may provide additional information about the performance of the manufacturing process as a whole.

Fig. 3 shows a step for exposing a target portion (e.g. a die) on a substrate W in the dual stage apparatus of fig. 1. A process according to conventional practice will first be described. The present disclosure is in no way limited to the type of dual platform device shown. Those skilled in the art will recognize that similar operations are performed in other types of lithographic apparatus (e.g., lithographic apparatus having a single substrate table and a dual stage measurement table).

The steps performed at the measurement station MEA are on the left hand side within the dashed box, while the right hand side shows the steps performed at the exposure station EXP. Often, one of the substrate tables WTa, WTb will be at the exposure station and the other at the measurement station, as described above. For the purposes of this description, it is assumed that the substrate W has been loaded into the exposure station. At step 200, a new substrate W' is loaded into the apparatus by a mechanism not shown in the figures. The two substrates are processed in parallel in order to increase the throughput of the lithographic apparatus.

Referring initially to the newly loaded substrate W', this substrate may be a previously unprocessed substrate and is prepared with new photoresist for use in a first exposure in the apparatus. In general, however, the lithographic process described will be only one of a series of exposure and processing steps, such that the substrate W' has passed through this apparatus and/or other lithographic apparatus several times, and may also undergo subsequent processes. Particularly with respect to the problem of improving overlay performance, the task will be to ensure that the new pattern is accurately applied in the correct position on the substrate that has undergone one or more cycles of patterning and processing. Each patterning step may introduce positional deviations in the applied pattern, while subsequent processing steps gradually introduce distortions in the substrate and/or the pattern applied to the substrate, which must be measured and corrected to obtain satisfactory overlay performance.

The previous and/or subsequent patterning steps may be performed in other lithographic apparatus (as just mentioned), and may even be performed in different types of lithographic apparatus. For example, some layers in the device fabrication process that are very demanding in parameters such as resolution and overlay may be performed in more advanced lithography tools than other less demanding layers. Thus, some layers may be exposed in an immersion type lithographic tool, while other layers are exposed in a "dry type" tool. Some layers may be exposed in a tool operating at a DUV wavelength, while other layers are exposed using EUV wavelength radiation. Some layers may be patterned by steps that are an alternative or supplement to the exposure in the illustrated lithographic apparatus. These alternative and complementary techniques include, for example, imprint lithography, self-aligned multiple patterning, and directed self-assembly. Similarly, the other processing steps performed on each layer (e.g., CMP and etching) may be performed on different equipment for each layer.

At 202, alignment measurements using substrate mark P1 or the like and an image sensor (not shown in the figure) are used to measure and record the alignment of the substrate relative to the substrate table WTa/WTb. In addition, the alignment sensor AS will be used to measure several alignment marks across the substrate W'. These measurements are used in one embodiment to create a substrate model (sometimes referred to as a "wafer grid") that maps the distribution of marks across the substrate with great accuracy, including any distortions relative to a nominal rectangular grid.

At step 204, a wafer height (Z) map is also measured with respect to the X-Y position using the level sensor LS. Primarily, the height map is only used to achieve accurate focusing of the exposed pattern. In addition, height maps may be used for other purposes.

When the substrate W' is loaded, recipe data 206 is received that defines the exposures to be performed and also defines the wafer, the previously generated patterns and the attributes of the patterns to be generated on the wafer. When options for alignment marks on the substrate are present and when options for settings of the alignment sensors are present, these options are defined by the alignment recipes in the recipe data 206. Thus, the alignment matching scheme defines how to measure the positions of the alignment marks and how to mark the positions of the alignment marks.

At 210, wafers W 'and W are exchanged so that the measured substrate W' becomes substrate W, entering exposure station EXP. In the example apparatus of fig. 1, this exchange is performed by exchanging supports WTa and WTb within the apparatus so that substrate W, W' remains accurately clamped and positioned on those supports to maintain relative alignment between the substrate table and the substrate itself. Thus, once the stage has been switched, all that is necessary to control the exposure step using the measurement information 202 and 204 for the substrate W (formerly W') is to determine the relative position between the projection system PS and the substrate table WTb (formerly WTa). At step 212, reticle alignment is performed using mask alignment marks M1, M2. In steps 214, 216, 218, the scanning motion and radiation pulses are applied at successive target sites across the substrate W in order to complete the exposure of a number of patterns.

By using the alignment data and the height map obtained at the measurement station in performing the exposure step, the patterns are accurately aligned with respect to the desired locations, in particular with respect to features previously placed on the same substrate. At step 220, the exposed substrate, now labeled W ", unloaded from the apparatus undergoes an etch or other process in accordance with the exposed pattern.

Advanced process control using performance data

For best performance, historical performance data about the lithographic process is typically used in addition to measurements made while the current substrate is loaded into the lithographic apparatus. For this purpose, performance measurements are made with a metrology system MET (fig. 2). Different forms of advanced process control may be implemented. Fig. 4 illustrates only one example of implementing a known stability control method.

Fig. 4 depicts a stability module 300. This module is for example an application running on a processor, for example within the control unit LACU or the supervisory control system SCS of fig. 2. Three main process control loops are shown, labeled 1, 2, 3. The first loop provides local control of the lithographic apparatus using the stability module 300 and the monitor wafer. The monitor wafer 302 is shown being transferred from a lithography unit 304, which lithography unit 304 may be, for example, the lithography unit LC of fig. 2. The monitor wafer 304 has been exposed with a calibration pattern to set the "baseline" parameters for focus and overlay. Later, the metrology tool 306 reads the baseline parameters, which are then interpreted by the stability module 300 to calculate the stability correction 308 specific to the lithography unit. The performance data may be fed back to the lithography unit 304 and used when performing further exposures. The exposure of the monitor wafer may involve printing a pattern of marks on top of the reference marks. By measuring overlay error between the top and bottom marks, performance deviation of the lithographic apparatus can be measured even when the wafer has been removed from the apparatus and placed in a metrology tool.

The second (APC) control loop is based on measurements of performance parameters such as focus, dose and overlay on the actual production wafer. The exposed production wafer 320 is transferred to a metrology tool 322, which metrology tool 322 may be the same as or different from the metrology tool 306 in the first control loop. At 322, information regarding parameters such as critical dimension, sidewall angle, and overlap, for example, is determined and passed to an Advanced Process Control (APC) module 324. This data is also passed to the stability module 300. Process corrections 326 are calculated and used by a Supervisory Control System (SCS)328, which provides control of the lithography units 304 in communication with the stability module 300.

The third control loop allows for integration of metrology into the second (APC) control loop in, for example, double patterning applications. The etched wafer 330 is transferred to a metrology unit 332, which metrology unit 332 may also be the same as or different from the metrology tools 306, 322 used in the first control loop and/or the second control loop. The metrology tool 332 measures performance parameters read from the wafer, such as critical dimension, sidewall angle, and overlay. These parameters are passed to an Advanced Process Control (APC) module 324. This loop continues as the second loop.

Current process correction strategies in High Volume Manufacturing (HVM) environments are typically performed on a per-chuck and per-batch basis. However, correction of each substrate is recently considered. It is then possible to define per-substrate process corrections rather than per-batch process corrections. A practical strategy needs to be designed for utilizing per-substrate corrections for process control at the per-substrate level (referred to herein as wafer level control or WLC). Performing overlay measurements on each processed substrate is expensive (particularly in terms of time and throughput). Instead, every substrate prediction of a "process fingerprint" can be made. A process fingerprint (or signature) describes the deformation or other deformation that a particular process step and/or process tool imposes on a substrate. Such predictions may be based on exposure order (known in advance) or context/processing history. However, this has some disadvantages. First, tracking and managing all historical process steps, especially for higher layers, requires a significant amount of effort. Second, it can be difficult to establish an unambiguous relationship between the process tool and the impact on the overlap.

Using metrology data that is typically generated on a more regular basis, such as alignment data or leveling data generated on a per substrate basis, is an alternative to reducing substrate-to-substrate variation. However, considering the example of alignment data in particular, the correction capability is limited by: to avoid loss of throughput, only a limited number of alignment marks can be measured; the alignment model is typically limited to a global (inter-field) model; and often the alignment marks suffer from process induced mark damage resulting in less reliable measurements.

It is proposed to group the substrates together and determine corrections based on the group of substrates in a continuously running wafer-level control strategy. In an HVM environment, cluster allocation may be performed according to a context history of a substrate. However, as already described, tracking the context history is an undesirable burden. Instead, it is proposed to group substrates according to pre-exposure parameter data (e.g., pre-exposure metrology data) that is related to the post-exposure performance parameter being controlled (e.g., overlay). By grouping the substrates in this manner, accuracy close to "per substrate" may be achieved while benefiting from a relatively large average across each group of substrates.

In this context, pre-exposure metrology data includes metrology data from measurements performed prior to exposure of a layer controlling a performance parameter, i.e. the term "pre-exposure" is exposure relative to the next layer. Thus, the pre-exposure metrology data may comprise measurements performed on a substrate on which a previous layer has been exposed for controlling exposure of another layer on the substrate.

The pre-exposure data may include data from measurements performed before loading onto the lithographic apparatus (scanner) for exposing the current layer, or after loading onto the lithographic apparatus (scanner) for exposing the current layer. In the latter example, the pre-exposure data may include a preparatory metrology technique for exposing the layer. In one embodiment, the pre-exposure metrology data may include alignment data. The alignment data may include measurements performed in preparation for exposing the current layer after loading the substrate. Alternatively, or in combination, the alignment data may include measurements performed in preparation for exposing a previous layer, i.e., measurements prior to loading the substrate for measuring and exposing a current layer. Alternatively, or in combination, the pre-exposure metrology data may include leveling data describing the shape of the substrate. As with the alignment data, the leveling data may be from measurements performed in preparation for exposing the current layer or previous layers. Alternatively, or in combination, the pre-exposure metrology data may include wafer geometry data and/or in-plane deformation data.

Considering an example in which the pre-exposure metrology data comprises alignment data, the alignment data may be measured across the substrate at a metrology station of the lithographic tool. The alignment data may include a plurality of vectors across the substrate, each vector representing the position and displacement of a mark position measured by the alignment sensor AS relative to a nominal position (e.g., positional deviation) for a particular mark on the substrate. All substrates may have the same spatial distribution of measurements and markings, but the actual deviations are usually unique for each substrate. Analysis of pre-exposure metrology data (alignment measurements) may be performed on a cluster of substrates in order to reveal various "fingerprints" that may be hidden in the data. Similarly, a fingerprint may be obtained from substrate topography or shape measurements, for example measured using a level sensor LS. It is known that any one of the different steps in the production of a processed substrate can have its own fingerprint contribute to the distribution of positional errors across the substrate. Considering that a real product may go through tens of process steps, including many cycles of patterning and processing in different devices and types of devices, it is difficult to know which type of device (let alone which individual devices) contributes to errors occurring in the finished product.

The proposed method may comprise two stages. An initial setup or training phase is performed to classify a group of substrates into a plurality of groups. The setup phase may include training a classifier to classify pre-exposure metrology data (input objects) according to (e.g., labeled by) characteristics of the performance parameters (outputs). Hard or soft classification of the data may be performed using any suitable (e.g., supervised, semi-supervised, or unsupervised) machine learning technique, such as linear discriminant analysis, logistic regression, support vector classifiers, or Principal Component Analysis (PCA). Other suitable classification methods are described in WO2015049087, which is incorporated herein by reference. This describes the following method: alignment data or other measurements are made at a stage during the performance of the lithographic process to obtain object data representing other parameters or positional deviations measured at points spatially distributed across each wafer. Such object data is used to obtain diagnostic information by performing multivariate analysis to decompose a set of the vectors representing the wafer in the multidimensional space into one or more component vectors. Diagnostic information about the industrial process is extracted using the component vectors. The performance of the lithographic process of subsequent product units may be controlled based on the extracted diagnostic information.

A training phase may be performed on historical data from multiple substrates for which pre-exposure metrology data and post-exposure metrology data (measurements of performance parameters) are applicable. For the specific example already mentioned, the alignment fingerprints (which describe the substrate grid deformation in the substrate plane) or the substrate shapes or topography (which describe the substrate deformation in the direction perpendicular to the substrate plane) are classified according to the respective overlay measured properties (e.g. overlay fingerprint properties). The results of the training phase may include a plurality of substrate sets, each labeled with a common fingerprint or topographical property and corresponding coefficients. Each performance parameter characteristic will have an associated process correction (e.g., best correction matching scheme). In one embodiment, the setup phase may be consistent with normal production (control phase, based on lot-based process corrections).

FIG. 5 is a flow chart describing a specific example of a training phase for predictive substrate classification. The setup phase may include using a supervised learning algorithm on training data 400 (e.g., historical) including pre-exposure metrology data for a plurality of substrates labeled with post-exposure metrology data describing performance parameters of the substrates. Since there will typically be much more pre-exposure metrology data available than post-exposure metrology data, the setup phase may include using a semi-supervised learning algorithm, where only a small number of substrates are labeled (post-exposure measured substrates), while only pre-exposure measured substrates are unlabeled. Such semi-supervised classification may use label diffusion methods, for example. In a particular example, the pre-exposure metrology data can include an alignment fingerprint or substrate topography, and the post-exposure metrology data can include an overlay, critical dimension, or focus fingerprint. First, an intra-batch clustering (or clustering) step 410 is performed to identify clusters/groups per batch. Thereafter, an inter-batch clustering step 420 is performed to identify similar fingerprints/topologies between batches. A classification step 430 is then performed to train the classifier using the training data. During this step, a pre-exposure metrology data set (i.e., a substrate set) is defined. A verification step 440 is then performed. In this verification step 440, substrates are assigned to groups of substrates based on historical data. Then, shadow pattern simulations of performance parameter benefits are performed for each process correction thread.

In a second or control phase, pre-exposure metrology data of the substrate is obtained, for example by performing alignment and/or leveling metrology on the substrate. The metrology may be performed in the lithographic apparatus as part of an alignment and substrate measurement process; for example, the alignment sensor AS and the level sensor LS of fig. 1 are respectively used by the above-described method. Assigning a panelist relationship status to a substrate based on the pre-exposure metrology data. The group membership status may include assigning each substrate a single group membership (hard classification) or partial membership of some or all groups (soft classification; e.g., using a flexible max function). In either case, these groups will be one of the substrate groups determined during the setup phase (or possibly determined during the control phase, as will be described later). Pre-exposure metrology data may include measuring any physical property common to all substrates that is sensitive to substrate process variations; for example: alignment fingerprint or substrate topography or flatness measurements. Based on the panelist relationship status, an associated process correction will be identified. This associated process correction can then be used during the exposure in which the layer is exposed.

In one embodiment, pre-exposure metrology on a substrate and subsequent substrate classification, identification of an associated correction, and exposure using the corrected layer are performed by a lithographic apparatus. This would mean that the calibration loop would be short (pre-exposure metrology data is used directly for subsequent exposure steps). No additional tools are required to measure the substrate.

In an embodiment, in case the lithographic apparatus comprises more than one support (more than one chuck), as shown in fig. 1, the chuck assignment may be considered when identifying the correction based on the substrate classification. Thus, the correction can be based on each chuck being associated with a particular substrate group.

A performance parameter (e.g., overlay) is measured after exposure on some or all of the substrates. The resulting metrology data can then be modeled and the parameters used to update or replace process corrections associated with the substrate lot for the measured substrate. The process correction update may be implemented using a temporal filter and/or averaging (e.g., using a moving average). Each substrate group can be modeled at a time. Alternatively, the modeling may include modeling all parameters using the category-specific parameters and the shared parameters at once.

In one embodiment, during the control phase, it may be determined from the index that the pre-exposure metrology data for a particular substrate incorrectly belongs to any of the substrate lots identified in the training phase. For example, the indicator may be a distance indicator, and if the distance indicator to the nearest substrate group is above a threshold value, the particular substrate may be deemed to have incorrectly belonged to any of the substrate groups. In a particular example, the distance index may represent a distance between a measured alignment fingerprint (or other pre-exposure metrology data) of the substrate and a metrology fingerprint defining a closest set. In such embodiments, the method can include updating the substrate lots by updating the respective characteristics characterizing one or more substrate lots such that one of the substrate lots now contains characteristics of the pre-exposure metrology data for that substrate. In this way, the characteristics of each substrate group can be updated while maintaining group number consistency. Alternatively, embodiments may include adding a new set of characteristics corresponding to pre-exposure measurement data for an unclassified substrate. A performance parameter (e.g., overlay) of this substrate is then measured after exposure and used to label pre-exposure metrology data for unclassified substrates. Furthermore, a corresponding correction for the new substrate group should be determined (e.g., by modeling), which can then be used to correct subsequent substrates categorized in the group. Alternatively, substrates that do not fit into any group can be reworked and set aside.

The selection of the substrate for post-exposure measurements can be optimized during the control phase. Such optimization may include selecting the most representative substrate identified as its corresponding group of substrates. This may include selecting a substrate for which the distance index associated therewith used in its classification is less than the distance indexes of the other substrates in the group. However, selection based on only a representative substrate may result in some groups being updated more frequently than others. Thus, in another embodiment, the substrate may be selected based on a combination of the representation and how recent the corresponding correction set was last updated.

As already described, at least in the case of hard sorting, the substrate is "binned" in a predetermined group, and then group-based correction is applied uniformly to each member of the group. In such embodiments, the classification/clustering may be performed by a statistical tool such as Principal Component Analysis (PCA) using online or offline data from a scanner or other metrology tool. However, in practice, the distinction between groups is usually not trivial and thus grading is not preferred. This is illustrated by the example plot of fig. 6. This is a plot of the PCA score on the y-axis relative to a particular intrinsic wafer (intrinsic fingerprint or principal component) versus substrate (x-axis) for a batch of 25 substrates. To obtain the map, PCA was performed on a batch (25 substrates) of substrate alignment data (pre-exposure metrology data). The most prominent fingerprint (the first intrinsic wafer or first intrinsic fingerprint) is then identified, and the graph plots the score value for that intrinsic wafer for each measured substrate (i.e., the measurement for which an intrinsic fingerprint is present in each substrate). As can be seen in fig. 6, the substrate has no significant clustering; the dispersion of the substrates is more continuous than this. This may result in substantially arbitrary sorting of certain substrates.

Thus, in this embodiment, it is proposed to use a weighted classification based on the classification score to improve the correction. Weighted classification improves the correction for each substrate by weighting each correction using the score values of the substrate found by PCA. In such embodiments, each intrinsic wafer identified may represent a different group in the classification. In one embodiment, not all of the identified intrinsic wafers define a separate group. For example, one or more of the least significant intrinsic wafers (intrinsic fingerprints/principal components) may be ignored in defining the group. The setup phase may be largely as already described, but in particular uses a classification technique (an example of which is PCA) which provides a score to each substrate according to a set (eigenwafer), for example a measure of the degree of membership within the set (which may be positive or negative). Other examples of suitable statistical classification methods to assign weights or scores to members of a group include: random forest, Bayesian network, neural network, linear discriminant analysis. The pre-exposure metrology data and post-exposure metrology data may include data already described in other embodiments. In one embodiment, a scoring-based weighting may be applied (e.g., multiplied) to the overlapping (or other process parameter) fingerprint correction to provide a weighted correction.

Fig. 7 is a graph of scores for post-exposure metrology data (y-axis), such as measured overlay fingerprints, versus pre-exposure metrology data (x-axis), such as measured alignment fingerprints, for a particular set or intrinsic wafer. Each dot represents a different substrate in a batch of (25) substrates. It is apparent that there is a correlation between the scores of the post-exposure metrology data and the pre-exposure metrology data. This correlation indicates that scoring attributed to pre-exposure metrology data (e.g., alignment data) can be used to determine the best correction for wafer-level control. It should be noted that the better this correlation, the better the predicted overlay correction will be.

It should be noted that in such an embodiment, only one group needs to be defined. Although technically this may be true for hard sorting, the result is essentially meaningless, since each substrate can only be assigned to the single group and therefore cannot be distinguished. However, in this embodiment, weighting based on the scoring values would mean that different corrections may be applied based on pre-exposure metrology data, whether only one lot (e.g., a single, most dominant, intrinsic wafer) is present or more than one lot is defined. Where more than one group is defined, the correction applied to a substrate may be a correction applicable to the group to which the substrate is assigned, the correction being weighted according to the group for which the substrate's score is relevant. In one embodiment, the actual classification may be a hard classification, where each substrate is assigned to a single group, and the corresponding corrected scores are weighted according to their scores (e.g., the degree of membership thereof) within that group. In an alternative embodiment, each substrate may optionally be assigned a partial membership of a plurality of groups, with the scoring values used to weight corrections between the groups. In the latter example, some substrates may only be classified in a single group if the scoring value is particularly high (in absolute terms) for that group.

FIG. 8 illustrates a flow diagram that conceptually describes a method in accordance with certain embodiments. Setup phase 700 includes obtaining historical training data for a plurality of substrates 705 and performing clustering and classification step 710 to obtain a plurality of substrate groups 715. Within the lithographic apparatus 720, pre-exposure metrology 725 is performed. The resulting pre-exposure metrology data is processed 730 within the processor 735 to assign each substrate to a substrate lot 740 (which may include partially assigning substrates to more than one substrate lot as appropriate). The steps may also optionally include storing a score (or scores) associated with each substrate that indicates the extent to which the substrate belongs to its assigned group (or groups). The processor 735 then assigns 745 the process corrections 750 to each substrate 748 based on the assigned category(s) and/or score(s). The processor 735 may be integral with (e.g., form part of) the lithographic apparatus 720, or may be a stand-alone processing module. At step 760, each assigned process correction 750 is used for wafer level control when exposing the next layer on that substrate. Within a metrology apparatus (e.g., a scatterometer) 765, one or more already classified substrates 770 (e.g., substrates that have been exposed before the plurality of substrates 705 were introduced to the lithographic apparatus 720) are measured post-exposure 775. For each substrate group, correction updates 780a, 780b are determined and fed to the processor 735 to be used in the algorithmic allocation process corrections 745. The correction update may be a weighted correction update based on the group membership status of each substrate, particularly if a soft distribution method (e.g., a flexible max method) is used. The method may further include an optimization step (not shown) that optimizes the updated process corrections 750 such that pre-optimized process corrections (one per category) are selected when assigning process corrections at step 745.

In this way, wafer-level control for continuous operation of the lithographic process uses pre-exposure metrology data in an automated solution without the need to track process history information and without the need to take off-line measurements.

It may be desirable to evaluate the clustering/classification of the methods described herein. In particular, it is useful to assess how well a substrate is correlated to a certain set of substrates, and whether, for example, pre-exposure metrology data is representative/used as a basis for an initial cluster of substrates (e.g., how well pre-exposure metrology data correlates to post-exposure metrology data correlated to a substrate).

Such evaluation may be performed as part of a training phase, such as the setup phase 700, more specifically the clustering and classification step 710 and/or the clustering steps 410, 420 and/or the classification step 430. The evaluation may include applying a clustering algorithm (e.g., k-means, gaussian mixture model, etc.) to determine k groups of substrates based on the post-exposure metrology data and j groups of substrates based on the pre-exposure metrology data, respectively.

In one embodiment, the optimal numbers k and j may be determined automatically by using Bayesian information criteria or similar model selection techniques. In such embodiments, this may include finding a minimum of the bayesian information criterion BIC, which may take the form:

whereinIs the maximum value of the likelihood function of the cluster model used, k is the number of model parameters, and n is the number of samples.

In a particular example, the cluster model used on the pre-exposure metrology data and the post-exposure metrology data may be a gaussian mixture model, e.g., a weighted sum of gaussians multiplied by a prior probability. In particular embodiments, this model p (x) may include:

whereinx is the clustered data, k is the number of elements (clusters), μ_iIs the mean value, Σ_iIs the covariance of element i.

It is proposed to apply a matching algorithm to match k bins in the post-exposure metrology data with j bins in the pre-exposure metrology data. This may include optimizing one or more match indicators or Key Performance Indicators (KPIs). Possible KPIs may include, for example, matching accuracy or purity. Evaluating matching accuracy may include determining correlation and correlation on a plot of true positive rate versus false positive rate for different discrimination thresholds from a Receiver Operating Characteristic (ROC) curveAnd/or area under the curve. Purity is a measure of the similarity (e.g., in terms of labeling after classification) of the samples within a group. More specifically, for a set of Ω ═ { ω₁,ω₂,…,ω_KAnd a set of labelsPurity is thenMay be defined as:

thus, the purity of the pre-exposure metrology data set may include the uniformity of the pre-exposure metrology data within each cluster relative to the cluster(s) of its one or more matching post-exposure metrology data (e.g., it is the case that all or most of the members of the pre-exposure cluster are from only the post-exposure cluster(s) that match the pre-exposure cluster, and vice versa).

It is possible to envision that the number of k sets in the post-exposure metrology data and the number of j sets in the post-exposure metrology data will be the same; i.e., j ═ k. This would mean that the pre-exposure measurement data results in the same set as the post-exposure metrology data. However, there are many reasons why this may not be the case, in practice j ≠ k. In the described embodiment, the matching algorithm matches groups even when j ≠ k.

FIG. 9 shows that clustering has revealed four pre-exposure metrology data sets G_a1-G_a4 and three post-exposure metrology data sets G_b1-G_bExample 3. FIG. 9 also shows a set G of pre-exposure metrology data_a1 and G_a2 and set G of post-exposure metrology data_b1 matching, set G of pre-exposure metrology data_a3 and set G of post-exposure metrology data_b2 match, and set G of pre-exposure metrology data_a4 and set G of post-exposure metrology data_bAnd 3, matching.

In one embodiment, samples with uncertain panelist relationships (e.g., samples and/or outliers near a group or decision boundary) may be excluded from the grouping. For example, any samples within a certain distance (tolerance) from the decision boundary may be excluded. In a particular example, where w is a vector orthogonal to the decision boundary and b is a scalar offset term, then the decision boundary can be written as:

w^Tx+b＝0

the tolerance may be defined as any position within the following inequality:

(w^Tx_i+b)y_i>c

where c determines the size of the margin on either side of the decision boundary. In another embodiment, where weighted classification is used (as already described), then the weights assigned to the substrates may be used to determine uncertain panelist relationships and thus whether a particular substrate may be excluded. FIG. 9 shows an example including two substrates W23, W15 having indeterminate group states. Due to the uncertainty of their lot status, they are incorrectly assigned to the incorrect post-exposure metrology lot: the substrate W23 has been assigned to group G_b2 instead of G_b3, and the substrate W15 has been assigned to group G_b3 instead of G_b2. It is proposed that these substrates can be excluded altogether or given less weight based on their indeterminate state.

In one embodiment, an initial step of removing irrelevant or less relevant features is performed such that clustering is performed only on the pre-exposure features that are relevant to post-exposure metrology changes. Uncorrelated features will increase the number of sets required and will result in a low match quality that may not be significantly better than random. In one embodiment, the dimension of the pre-exposure metrology data may be limited to two. For example, having only two-dimensional data typically means that far fewer sets will be needed. By way of specific example, only three groups are required to display two-dimensional data on a test dataset, while eight groups are required to add another dimension. It will be appreciated that the actual number of groups required in each case will depend on the data set.

In a verification phase, KPIs describing the quality of the match may be determined. This may include determining the statistical significance of the packet performance. For example, a p-value may be calculated that indicates whether the quality of the group match is significantly better (e.g., better than a threshold tolerance) than random. If it is determined that the quality of the set match is not significantly better than random, this may indicate that the pre-exposure metrology data does not adequately account for the substrate-to-substrate variations observed in the post-exposure metrology data. If this is the case, the foregoing steps may be repeated using different types of pre-exposure metrology data or different types of pre-exposure parameter data (e.g., alignment, leveling, processing history, etc.). Another reason that the group match quality may not be significantly better than random is that the clustering algorithm may not work effectively on the data set. If neither the available clustering algorithms nor the pre-exposure metrics result in statistically significant group matching performance, it can be concluded that cluster-based control should not be used in the current substrate production scenario. On the other hand, when KPIs indicate good, statistically significant packet performance, then cluster-based control strategies may be enabled in production.

In another embodiment, the concepts described herein may be used for calibration between patterning steps of a multiple patterning process, such as a LELE (photo-etch-photo-etch) process. In such embodiments, the pre-exposure parameter data may include contextual data relating to the processing context used.

In a practical sense, a large number of different context variables (context parameters) may be involved. For example, each process tool, process chamber, and process recipe can be considered a separate context. Thus, the number of context combinations may be very large. The logic that monitors each unique combination of contexts is not always practical.

Thus, in a particular embodiment, the context data on which control is proposed to be based is limited to an etch chamber used in an etch step immediately preceding a previous lithography step. An etch tool may have multiple chambers (typically up to 4), resulting in a limited set of unique context values (corresponding to the number of etch chambers). By tracking which etch chamber is used to process each substrate, each batch of substrates may be sorted into multiple (e.g., four) groups. Then, for each group, a separate WLC control can be determined. These WLC corrections can be added to the "normal" (APC) corrections applied in the inter-batch control. Generally, in multiple patterning applications, it is typically recommended to use the same correction for each patterning (photo-etching) step. The proposal described in this embodiment improves intra-layer "overlap" (between each lithography step of a layer) by proposing context-based wafer-level control for the difference between the layer positions of two layers.

Fig. 10 is a flow chart illustrating such an embodiment. The examples herein show a double patterning LELE process, where each layer is exposed in two separate lithographic etching steps. However, the embodiments are more generally applicable to multiple patterning processes, including processes having more than two separate lithographic etching steps. In the initial lot (lot N), a first patterning step L1 is performed_N. As already described, the first patterning step L1 may be performed with corrections determined using the APC control loop based on measurements of one or more previous lots_N. After the first patterning step, a first etching step E1 is performed_N. In the first etching step E1_NIn, for example, four etch chambers EC_a-EC_dEach substrate is etched in one of the chambers. The etch chamber used to etch each substrate is recorded (as a relevant context) and each substrate is assigned to the group corresponding to the etch chamber used. In addition, the measurement in the patterning step L1 was carried out in a first measurement step MET1 using a metrology device_NFollowed by one or more patterned substrates. In one particular embodiment, at least two substrates are measured, one substrate per chuck, to obtain measurements for each chuck. The first measuring step may comprise, for example, measuring the first overlay fingerprint OV 1. In a first patterning step L1_NAnd a first etching step E1_NThereafter, the first patterning step L1 is performed_NCorrection performed the same APC correction is performed to perform the second patterning step L2_N. In a second patterning step L2_NThen, a second measurement is performed using the measurement deviceQuantitative step MET 2. In a second metrology step MET2, at least one substrate in each set (and optionally also at least one substrate per chuck) is measured. For example, the second measurement step may include measurement of the second overlay fingerprint OV2 for each set (or class/cartridge combination). The difference between the second overlay fingerprint OV2 and the first overlay fingerprint OV1 may then be calculated and used to determine the intra-layer correction cor for each substrate group_a-cor_d(second correction). In-layer correction cor_a-cor_dTheir application may be made to minimize the difference between the second overlay fingerprint OV2 and the first overlay fingerprint OV 1.

In the subsequent batch (e.g., batch N +1), the first patterning step L1_N+1And a first etching step E1_N+1It is performed in a similar manner as batch N, using "standard" APC corrections (first correction) as appropriate. This may include an exponentially weighted moving average EWMA of previous measurements. As before, trace etch step E1_N+1The etch chamber (scenario) used in (1). Selecting a correction cor corresponding to the group of the context based on the context_a-cor_dAn appropriate one of them. When the second patterning step L2 is performed_N+1This (second) correction is used together with the APC (first) correction, after which a second etching step (not shown) will be performed. Thus, from an overlay perspective, the final double-patterned (LELE) layer will appear more like a single exposure.

It will be appreciated that in principle this concept can be extended to more complex context threads than shown (etch chamber) and parameters other than overlay (e.g. CD or Edge Placement Error (EPE)).

Other embodiments of the invention are disclosed in the following list of numbered embodiments:

1. a method of determining a correction for a process parameter associated with a lithographic process on a substrate, the lithographic process comprising a plurality of runs during each of which a pattern is applied to one or more substrates, the method comprising:

obtaining pre-exposure parameter data related to attributes of the substrate;

obtaining post-exposure metrology data comprising one or more measurements of a process parameter that have been performed by an equivalent lithographic process on one or more previously exposed substrates of the lithographic process;

assigning a group membership status from one or more groups to the substrate based on the pre-exposure parameter data; and

determining a correction to the process parameter based on the panelist relationship status and the post-exposure metrology data.

2. The method of embodiment 1, comprising performing a lithographic process on the substrate using the correction.

3. The method of embodiment 1 or 2, wherein the pre-exposure parameter data comprises pre-exposure metrology data.

4. The method of embodiment 3, comprising performing a pre-exposure metrology step on the substrate to obtain the pre-exposure metrology data.

5. The method of embodiment 4, wherein the pre-exposure metrology step and the subsequent step of performing a lithographic process on the substrate using the correction are performed by the same lithographic apparatus.

6. The method of embodiment 5, wherein the steps of assigning a panelist status and determining a correction to the process parameter are also performed by the same lithographic apparatus.

7. The method of any of embodiments 3-6, wherein the pre-exposure metrology data comprises alignment data describing a cross-substrate grid deformation in a substrate plane.

8. The method of any of embodiments 3-7, wherein the pre-exposure metrology data comprises data describing a shape of the substrate in a direction perpendicular to a substrate plane.

9. The method of any of embodiments 3-8, wherein the pre-exposure metrology data comprises leveling data.

10. The method as in any one of embodiments 3-9 comprising the step of identifying, for a particular group, the most representative substrate in the group.

11. The method of embodiment 10, comprising measuring the substrate most representative of the group; and

updating a correction to the process parameter corresponding to the set based on the measurement.

12. The method of any of embodiments 3-11, wherein the group membership status is a degree such that each substrate may be assigned a membership to one or more of the groups.

13. The method of embodiment 12 wherein the step of determining the correction of the process parameter is based on weighting the corrections associated with one or more groups based on the degree of membership of the substrate to the one or more groups.

14. The method of embodiment 13, wherein the degree of membership for a group is based on a classification score representing a metric defining the presence of a fingerprint of the group in the pre-exposure metrology data.

15. The method of embodiment 14 wherein the assigning step comprises performing principal component analysis on the pre-exposure metrology data to identify one or more intrinsic fingerprints present in the pre-exposure metrology data and the classification score representing a metric of the corresponding intrinsic fingerprint present in the pre-exposure metrology data.

16. The method of embodiment 14 or 15, comprising an initial training phase in which the one or more groups are determined from a principal component analysis performed on the pre-exposure metrology data, each group being defined by an intrinsic fingerprint determined from the principal component analysis.

17. The method as in any one of embodiments 3-14 comprising an initial training phase wherein the one or more groups are determined.

18. The method of embodiment 17, wherein the one or more lots are determined from a plurality of labeled sets of the pre-exposure metrology data, each labeled set being associated with a different substrate of a plurality of substrates and labeled by the post-exposure metrology data associated with the substrate.

19. The method of embodiment 18, wherein the initial training phase comprises executing a supervised or semi-supervised classification algorithm that trains classifiers for at least the labeled set of pre-exposure metrology data to define the one or more groups.

20. The method as in any one of embodiments 17-19 wherein the initial training phase comprises the step of identifying pre-exposure groups in the pre-exposure metrology data.

21. The method of embodiment 20, wherein the step of identifying the pre-exposure groups comprises identifying groups within the batch and identifying groups between batches.

22. The method of embodiment 20 or 21, wherein the initial training phase comprises a step of identifying exposed groups in the exposed metrology data.

23. The method of embodiment 22 comprising the step of matching the post-exposure and pre-exposure groups by optimizing at least one matching metric, the at least one matching metric being related to the matching quality.

24. The method of embodiment 23, wherein the match metric comprises one or more of:

purity of the pre-exposure and/or post-exposure groups in terms of homogeneity across each group; and

correlations and/or areas under the curve of the set describing different discrimination thresholds determined from the receiver operating characteristic curve.

25. The method of embodiment 23 or 24, wherein it is determined whether the statistical significance of the matched group described by the match indicator is significantly greater than random.

26. The method of embodiment 25 wherein the training phase is repeated using different types of pre-exposure metrology data where it is determined that the statistical significance of the matched set described by the match indicator is not significantly greater than random.

27. The method as in any one of embodiments 23-26 wherein substrates with uncertain membership status are excluded from the matching step or given less weight in the matching step.

28. The method as in any one of embodiments 23-27 wherein the matching step is performed only on pre-exposure features within pre-exposure data associated with post-exposure metrology changes.

29. The method of any of embodiments 22-28, comprising separately optimizing a number of pre-exposure groups and a number of post-exposure groups.

30. The method as in any one of embodiments 17-29 wherein the initial training phase comprises a verification step comprising assigning substrates to the one or more lots based on historical data and simulating an effect on the post-exposure metrology data.

31. The method as in any one of embodiments 17-30, wherein the post-exposure metrology data comprises overlay data.

32. The method of any of embodiments 1-11, wherein the team membership status is such that each substrate is always assigned to a single team, otherwise unclassified.

33. The method of any preceding embodiment, wherein the pre-exposure metrology data comprises data relating to previously exposed layers on the substrate.

34. The method of any of embodiments 1-32, wherein the pre-exposure metrology data comprises data relating to a layer to be exposed on the substrate in a subsequent step of the lithographic process.

35. The method of embodiment 1 or 2, wherein the pre-exposure parameter data comprises context data relating to a particular processing step.

36. The method of embodiment 35 wherein the contextual data relates to a tool used in a processing step to process the substrate.

37. The method of embodiment 36, wherein the context data relates to specific etch chambers used during an etch step, each group corresponding to one of the etch chambers.

38. The method of embodiment 35 or 36 or 37 wherein the lithographic process comprises a multiple patterning process with at least a first patterning and etching step and a second patterning and etching step per layer.

39. The method of embodiment 38, wherein the correction comprises a second correction relative to the first correction, the method comprising the steps of:

performing a first patterning and etching step using the first correction;

determining the context applicable to the first patterning and etching step;

assigning a panelist relationship status to the substrate based on the determination of the context; and

determining the second correction for the second patterning and etching step based on the panelist relationship status and the first correction.

40. The method of embodiment 39 comprising performing the following initial steps for each group:

obtaining first process parameter data relating to a first measurement of the process parameter between the first patterning and etching step and the second patterning and etching step;

obtaining second process parameter data relating to a second measurement of the process parameter after the second patterning and etching step; and

calculating the second correction based on a difference between the first process parameter data and the second process parameter data.

41. The method of embodiment 40 wherein the second correction is calculated to minimize the difference between the first and second process parameter data for each set.

42. The method of any preceding embodiment, wherein the process parameter comprises an overlap.

43. The method as in any one of embodiments 1-41, wherein the process parameter comprises one of a critical dimension and an edge placement error.

44. The method of any preceding embodiment, wherein the step of determining a correction for the process parameter based on the membership status further determines a correction for the process parameter based on which chuck the substrate is mounted to during the lithographic process.

45. The method of any preceding embodiment, comprising measuring the substrate after exposure to obtain a post-exposure measurement of the substrate; and

updating a correction of the process parameter corresponding to the membership status assigned to the substrate using the post-exposure measurement of the substrate.

46. The method of any preceding embodiment, wherein, in the event that it is determined from the index that the pre-exposure metrology data does not sufficiently conform to any of the one or more lots, the method comprises updating the one or more lots such that the pre-exposure metrology data can be classified.

47. The method of embodiment 46 wherein the step of updating the one or more lots includes maintaining a same number of lots and updating data characteristics defining one or more of the lots such that the pre-exposure metrology data for the substrate substantially conforms to at least one of the lots according to the index.

48. The method of embodiment 46 wherein the step of updating the one or more groups comprises adding a new group defined by data characteristics having improved compliance with the substrate relative to other groups.

49. The method of any preceding embodiment, comprising performing the method for each substrate in a run.

50. The method of embodiment 49, wherein the post-exposure metrology data comprises one or more measurements of process parameters that have been performed by an equivalent lithographic process on one or more previously exposed substrates in the same run of the lithographic process.

51. The method of embodiment 49, wherein the post-exposure metrology data comprises one or more measurements of process parameters that have been performed by an equivalent lithographic process on one or more previously exposed substrates in a previous run of the lithographic process.

52. A method of manufacturing a device, wherein device features are formed on a series of substrates by a patterning process, wherein a correction to a process parameter of the patterning process is determined by performing the method of any of embodiments 1-51 and 56-64.

53. A control system for a lithographic apparatus, the control system comprising:

a memory for receiving pre-exposure parameter data and post-exposure metrology data relating to an attribute of a substrate, the post-exposure metrology data comprising one or more measurements of a process parameter that have been performed on one or more previous substrates; and

a processor operable to:

assigning a group membership status from one or more groups to the substrate based on the pre-exposure parameter data; and

determining a correction for the process parameter based on the panelist relationship status and the post-exposure metrology data.

54. The control system of embodiment 53 wherein the pre-exposure parameter data comprises pre-exposure metrology data.

55. The control system of embodiment 54 wherein the pre-exposure metrology data comprises alignment data describing a deformation of a cross-substrate grid in a substrate plane.

56. The control system of embodiment 54 or 55, wherein the pre-exposure metrology data comprises data describing a shape of the substrate in a direction perpendicular to a substrate plane.

57. The control system of any of embodiments 54-56, wherein the pre-exposure metrology data comprises leveling data.

58. The control system of any of embodiments 54 to 57, wherein the processor is operable to assign the group membership status such that each substrate may be assigned a degree of membership to one or more of the groups.

59. The control system of embodiment 58 wherein the corrections to the process parameters are determined based on weighting corrections associated with one or more groups based on a degree of membership of the substrate to the one or more groups.

60. The control system of embodiment 59, wherein the degree of membership to a lot is based on a classification score representing a measure of the presence of a fingerprint defining the lot in the pre-exposure metrology data.

61. The control system of embodiment 60 wherein the processor is operable to perform principal component analysis on the pre-exposure metrology data to identify one or more intrinsic fingerprints present in the pre-exposure metrology data and the classification score representing a measure of the presence of a corresponding intrinsic fingerprint in the pre-exposure metrology data.

62. The control system of any of embodiments 54-61, wherein the processor is operable to determine, for a particular group, a most representative substrate in the group.

63. The control system of embodiment 62 wherein the processor is operable to update a correction to the process parameter corresponding to the group based on a measurement of the substrate most representative in the group.

64. The control system of any of embodiments 53-57, wherein the processor is operable to assign the group membership status such that each substrate is always assigned to a single group and is otherwise unclassified.

65. The control system of any of embodiments 53-64, wherein the pre-exposure metrology data comprises data relating to a previously exposed layer on the substrate.

66. The control system of any of embodiments 53-64, wherein the pre-exposure metrology data comprises data relating to a layer to be exposed.

67. The control system of embodiment 53 wherein the pre-exposure parameter data comprises contextual data relating to a particular process step.

68. The control system of embodiment 67 wherein the contextual data relates to a tool that has been used in processing the substrate.

69. The control system of embodiment 67 or 68, being operable to control the lithographic apparatus to perform a multiple patterning process with at least a first patterning step and a second patterning step per layer.

70. The control system of embodiment 69, wherein the context data relates to particular etch chambers that have been used to etch the substrate between the first and second patterning steps, each group corresponding to one of the etch chambers.

71. The control system of embodiment 70, wherein the correction comprises a second correction relative to the first correction, the control system operable to control the lithographic apparatus to:

performing a first patterning step using a first correction;

determining the context applicable to the first patterning step;

assigning a panelist relationship status to the substrate based on the determination of the context; and

determining the second correction for the second patterning step based on the panelist relationship status and the first correction.

72. The control system of embodiment 71, operable to control the lithographic apparatus to, for each class:

obtaining first process parameter data relating to a first measurement of the process parameter between the first patterning step and the second patterning step;

obtaining second process parameter data relating to a second measurement of the process parameter after the second patterning step; and

calculating the second correction based on a difference between the first process parameter data and the second process parameter data.

73. The control system of embodiment 72 being operable such that the second corrections are calculated to minimize a difference between the first and second process parameter data for each set.

74. The control system of any of embodiments 53-73 wherein the process parameter comprises an overlap.

75. The control system of any of embodiments 53-74, wherein the process parameter comprises one of a critical dimension and an edge placement error.

76. The control system as in any one of embodiments 53-75, wherein the correction to the process parameter determined based on the membership status is also based on which chuck the substrate is mounted to during the lithographic process.

77. The control system of any of embodiments 53-76, wherein the processor is operable to update a correction to the process parameter corresponding to the state of membership assigned to the substrate using the post-exposure measurement of the substrate.

78. The control system of any of embodiments 53-77, wherein, in the event that the processor determines from the indicator that the pre-exposure metrology data is not sufficiently compliant with any of the one or more lots, the processor is operable to update the one or more lots such that the pre-exposure metrology data can be classified.

79. The control system of embodiment 78, wherein updating the one or more lots includes maintaining a same number of lots and updating data characteristics defining one or more of the lots such that the pre-exposure metrology data for the substrate substantially conforms to at least one of the lots according to the index.

80. The control system of embodiment 78 wherein said updating said one or more groups includes adding a new group defined by data characteristics having improved compliance with said substrate relative to other groups.

81. A control system operable to control a suitable apparatus to perform the method of any of embodiments 1 to 52 and 86 to 94.

82. A lithographic apparatus comprising the control system according to any one of embodiments 43 to 81.

83. The lithographic apparatus of embodiment 82, comprising a measurement system operable to perform pre-exposure metrology on the substrate to obtain the pre-exposure metrology data, a patterning system operable to form device features on the substrate in a lithographic process using the correction to process parameters of the patterning process, and a control system.

84. A computer program product comprising one or more sequences of machine-readable instructions for carrying out the steps of the method of any one of embodiments 1 to 52 and 86 to 94.

85. A computer program product containing one or more sequences of computer readable instructions for causing a processing device or a system of processing devices to implement the control system of any one of embodiments 53 to 81.

86. A method of dynamically updating one or more groups and/or corrections to a process parameter related to a lithographic process on a substrate, wherein one of a plurality of corrections is applied to the process parameter for each substrate based on a group membership status assigned to the substrate, the method comprising:

obtaining post-exposure metrology data describing performance parameters of the substrate; and

dynamically updating the one or more and/or more corrections in the set based on the post-exposure metrology data.

87. The method of embodiment 86 wherein the post-exposure metrology data comprises overlay data.

88. The method of embodiment 86 or 87, wherein the dynamically updating step comprises dynamically updating the correction to the process parameter corresponding to the group based on the measurement of the substrate determined to be the most representative of the group.

89. The method of any of embodiments 86-88, wherein the dynamically updating step comprises updating one or more of the plurality of corrections corresponding to the group membership status of the substrate.

90. The method as in any one of embodiments 86-89 wherein the dynamically updating step comprises applying a weighted update to the plurality of corrections based on the group membership status of the substrate.

91. The method of any one of embodiments 86-90, comprising the step of obtaining pre-exposure metrology data describing a property of each substrate;

assigning a panelist status from one or more lots to the substrate based on the pre-exposure metrology data; and

determining a correction to the process parameter based on the panelist relationship status.

92. The method of embodiment 91, wherein, in the event that it is determined from the index that the pre-exposure metrology data does not sufficiently conform to any of the one or more lots, the dynamically updating step comprises dynamically updating the one or more lots such that the pre-exposure metrology data can be classified.

93. The method of embodiment 92 wherein dynamically updating the one or more lots comprises maintaining a same number of lots and updating data characteristics defining one or more of the lots such that the pre-exposure metrology data for the substrate substantially conforms to at least one of the lots according to the index.

94. The method of embodiment 92 wherein dynamically updating the one or more groups comprises adding a new group defined by data characteristics having improved compliance with the substrate relative to other groups.

In association with the hardware of the lithographic apparatus and the lithographic cell LC, embodiments may include a computer program containing one or more sequences of machine-readable instructions for causing a processor of a lithographic manufacturing system to implement the method of mapping and controlling a model as described above. Such a computer program may be executed, for example, in a separate computer system for the image calculation/control process. Alternatively, the calculation steps may be performed entirely or partially within the processor, the metrology tool and/or the control unit LACU and/or the supervisory control system SCS of fig. 1 and 2. A data storage medium (e.g. semiconductor memory, magnetic or optical disk) may also be provided in which such a computer program is stored in a non-transitory form.

Although specific reference may have been made above to the use of embodiments of the invention in the context of optical lithography, it will be appreciated that the invention may be used in other patterning applications, for example imprint lithography. In imprint lithography, a topography in a patterning device defines the pattern created on a substrate. The topography of the patterning device may be pressed into a layer of resist provided to the substrate whereupon the resist is cured by applying electromagnetic radiation, heat, pressure or a combination thereof. After the resist is cured, the patterning device is moved out of the resist, leaving a pattern therein.

The foregoing description of the specific embodiments will so fully reveal the general nature of the invention that others can, by applying knowledge within the skill of the art, readily modify and/or adapt for various applications such specific embodiments, without undue experimentation, without departing from the general concept of the present invention. Therefore, based on the teachings and guidance provided herein, such adaptations and modifications are intended to be within the meaning and range of equivalents of the disclosed embodiments. It is to be understood that the phraseology or terminology herein is for the purpose of description by way of example and not of limitation, such that the terminology or phraseology of the present specification is to be interpreted by the skilled artisan in light of the teachings and guidance.

The breadth and scope of the present invention should not be limited by any of the above-described exemplary embodiments, but should be defined only in accordance with the following claims and their equivalents.