• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:

    公开号:NL2013249A
    申请号:NL2013249
    申请日:2014-07-24
    公开日:2015-02-23
    发明作者:Emil Schmitt-Weaver；Wolfgang Henke；Thomas Hoogenboom；Pavel Izikson；Paul Luehrmann；Daan Slotboom；Jens Staecker；Alexander Ypma
    申请人:Asml Netherlands Bv；
    IPC主号:
    专利说明:

    LITHOGRAPHY SYSTEM AND A MACHINE LEARNING CONTROLLERFOR SUCH A LITHOGRAPHY SYSTEM
    Background [001] The present description relates a lithography system configured to apply apattern to a substrate, the system comprising, e.g.,: a track unit configured to apply alayer on the substrate for lithographic exposure, a lithography apparatus configured toexpose the layer according to the pattern, a metrology unit configured to measure aproperty of the exposed pattern in the layer, a control unit configured to control anautomatic substrate flow among the track unit, the lithography apparatus, and themetrology unit, and a machine learning controller configured to control the system tooptimize the property of the pattern.
    [002] An example of such a lithography system is known from U.S. Patent No. US6,979,522. In the system of U.S. Patent No. US 6,979,522, a lot (also called a “batch”)of substrates is exposed after alignment in a lithography apparatus wherein alignmentparameters are determined. Using a formula with tool specific coefficients, the overlayaccuracy can be calculated (predicted) from these alignment parameters in advance.Next, the exposure tool-offset can be adjusted on a substrate-to-substrate basis tocorrect for the derived overlay inaccuracy. Moreover, the alignment parameters for aspecific substrate can be used to change the tool-offset for the same substrate prior toexposure.
    Summary [003] A potential disadvantage of the foregoing lithography system is that the machinelearning controller may not be effective in correcting system drift. The known systemhas a controller that is non-optimal in keeping system drift from interfering with dataused to calculate accuracy. A linear formula with tool specific coefficients may not bediverse enough to properly characterize all external and internal elements that affectaccuracy (for example, the substrate chuck used for exposure, dynamic behavior of thesystem over time, etc.). It predicts overlay for an individual substrate on the basis ofalignment parameters (pre-exposure alignment data) and applies an overlay offset correction on the basis of that prediction. The prediction may not be reliable becausethere is a mix up of real-time (random, substrate-to-substrate varying) parameters withsystematic drift parameters.
    [004] It is an objective, for example, to alleviate at least part of a disadvantage of aknown lithography system.
    [005] According to an embodiment, there is provided a lithography system configuredto apply a pattern to a substrate, comprising: a track unit configured to apply a layer onthe substrate for lithographic exposure; a lithography apparatus configured to exposethe layer according to the pattern; a metrology unit configured to measure a property ofthe exposed pattern in the layer; a control unit configured to control an automaticsubstrate flow among the track unit, the lithography apparatus, and the metrology unit;and a machine learning controller configured to control the lithography system tooptimize a property of the pattern, the machine learning controller configured to betrained on the basis of the measured property and to correct lithography system drift byadjusting one or more selected from: the lithography apparatus, the track unit and/or thecontrol unit.
    [006] According to an embodiment, there is provided a lithography system having amachine learning controller configured to be trained on the basis of a measuredproperty and to correct lithography system drift by adjusting at least one selected from:the lithography apparatus, the track unit and/or the control unit. In this way thecorrection of the system is based on an accurate measurement of the pattern which isthe desired output. The machine learning is based on accurate post-exposuremeasurement data of the desired output (the pattern, for example overlay, imagingparameters such as a critical dimension) which is a good basis for correcting systemdrift.
    [007] In an embodiment, the machine learning controller is configured to correctlithography system drift on the basis of measured properties of substrates of at least twolots (or more lots) of substrates. A lot comprises at least two substrates, but generallyten or more substrates. Such a sequence may allow for an accurate monitoring ofsystem drift.
    [008] According to an embodiment, the machine learning controller comprises a firstcontroller configured to generate an overlay control signal to control overlay betweenpattern layers. The machine learning controller may correct overlay system drift byadjusting the lithography apparatus, the track unit and/or the control unit on the basis ofthe overlay control signal. This may yield a lithography system having an optimal controlof overlay.
    [009] According to an embodiment, the machine learning controller comprises asecond controller configured to generate a critical dimension control signal to control acritical dimension of the pattern. The machine learning controller may correct criticaldimension system drift by adjusting the lithography apparatus, the track unit and/or thecontrol unit on the basis of the critical dimension control signal. This may yield alithography system having an optimal control of critical dimension.
    [0010] According to an embodiment, the machine learning controller comprises both thefirst controller and the second controller. The machine learning controller may generatean edge-to-edge control signal which is a combination of the overlay control signal andthe critical dimension control signal. Accurate edge-to-edge control is desirable incomplementary lithography where one-dimensional grating lines are exposed withimmersion lithography and the two-dimensional shape is generated by putting so-called“cuts” on these lines to define line ends. These “cuts” may be exposed using extremeultraviolet (EUV) radiation lithography or with multiple exposures using immersionlithography (generally using deep ultraviolet radiation). It is noted that edge-to-edgevariation is not controlled directly. The edge-to-edge error is the result of an overlayerror and a critical dimension error. Thus, edge-to-edge error control involves acombination of the mentioned overlay and critical dimension (or another imagingparameter, e.g., a focus parameter) control signals.
    [0011] According to an embodiment, the first and/or second controller comprises (a)sub-controller(s) configured to generate (a) drift control signal(s) and/or (a) sub-controllers) configured to generate (a) real-time (relating to a random or substrate-to-substrate varying parameter) control signal(s). In an embodiment, the training signal forthe sub-controller for the real-time control signal may have the theoretical computed influence of systematic drift removed from the training signal before training the realtime controller.
    [0012] The drift control signal(s) may be based on a measured property (post-exposureinformation). In an embodiment, the drift control signal(s) may also be based on pre¬exposure information such as lithography apparatus information, substrate processinformation and/or plant information. Similarly, the real-time control signal(s) may bebased on pre-exposure information.
    [0013] The lithography apparatus information may include at least one selected from:information about a substrate chuck of the lithography apparatus used for exposure,information about the dynamics of a patterning device support of the lithographyapparatus, information about the dynamics of a substrate stage of the lithographyapparatus, information about the substrate alignment, information about the substrateleveling, information about an optical property of a projection system of the lithographyapparatus, and/or information about a parameter or property associated with exposingthe pattern of the patterning device onto a substrate.
    [0014] The substrate process information may include at least one selected from: spincoating information, baking information, etching information and/or the sequence of thesubstrate in the lot of substrates.
    [0015] The plant information may include environmental data comprising at least oneselected from: temperature in the plant and/or humidity in the plant.
    [0016] According to an embodiment, the machine learning controller generates an edge-to-edge control signal which is a combination of (an) overlay control signal(s) [desirablyincluding both a drift and real-time overlay control signal] and (a) critical dimensioncontrol signal(s) [desirably including both a drift and real-time critical dimension controlsignal]. The machine learning controller may correct system drift by adjusting at leastselected from: the lithography apparatus, the track unit and/or the control unit, on thebasis of the edge-to-edge control signal.
    [0017] According to an embodiment, there is provided a machine learning controller anda machine learning computer algorithm configured to control the lithography system.
    [0018] According to an embodiment, there is provided a machine learning controllerconfigured to control a lithography system to optimize a property of a pattern to be applied to a substrate, the machine learning controller configured to be trained on thebasis of the property measured by a metrology unit configured to measure the propertyof the exposed pattern in the layer, and to correct lithography system drift by adjustingone or more selected from: a lithography apparatus configured to expose a layer of thesubstrate according to the pattern, a track unit configured to apply the layer on thesubstrate for lithographic exposure, and/or a control unit configured to control anautomatic substrate flow among the track unit, the lithography apparatus, and themetrology unit.
    [0019] According to an embodiment, there is provided a method, comprising: exposing alayer of a substrate according to a pattern using a lithography apparatus of a lithographysystem; measuring a property of the exposed pattern in the layer using a metrology unitof the lithography system; training a machine learning controller on the basis of themeasured property; and controlling the lithography system to optimize a property of thepattern using the machine learning controller by correcting lithography system drift byadjusting one or more selected from: the lithography apparatus, a track unit configuredto apply the layer on the substrate for lithographic exposure, and/or a control unitconfigured to control an automatic substrate flow among the track unit, the lithographyapparatus, and the metrology unit.
    [0020] The machine learning controller may be provided with known artificialintelligence.
    Brief Description of the Figures [0021] Embodiments of the invention will now be described, by way of example only,with reference to the accompanying drawings in which: [0022] Figure 1 depicts a lithographic apparatus which may be part of a lithographysystem according to an embodiment of the invention; [0023] Figure 2 is a schematic view of a lithography system comprising a machinelearning controller according to an embodiment of the invention; [0024] Figure 3A is a schematic view of a lithography system comprising a machinelearning controller according to an embodiment of the invention; and [0025] Figure 3B is a schematic view of a machine learning controller according to anembodiment of the invention for a lithography system.
    Detailed Description [0026] Figure 1 schematically depicts a lithographic apparatus (2) (that may be part ofthe lithography system (4)) that comprises: [0027] an illumination system (illuminator) IL configured to condition a radiation beam B(e.g. UV radiation or EUV radiation).
    [0028] a support structure (e.g. a mask table) MT constructed to support a patterningdevice (e.g. a mask) MA and connected to a first positioner PM configured to accuratelyposition the patterning device in accordance with certain parameters; [0029] a substrate table (e.g. a wafer table) WTa or WTb constructed to hold a substrate(e.g. a resist coated wafer) W and connected to a second positioner PWa or PWbconfigured to accurately position the substrate in accordance with certain parameters;and [0030] a projection system (e.g. a refractive projection lens system) PS configured toproject a pattern imparted to the radiation beam B by patterning device MA onto a targetportion C (e.g. comprising one or more dies) of the substrate W.
    [0031] The illumination system may include various types of optical components, suchas refractive, reflective, magnetic, electromagnetic, electrostatic or other types of opticalcomponents, or any combination thereof, for directing, shaping, or controlling radiation.
    [0032] The support structure, which may be a reticle table or reticle chuck, holds thepatterning 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 examplewhether or not the patterning device is held in a vacuum environment. The supportstructure can use mechanical, vacuum, electrostatic or other clamping techniques tohold the patterning device. The support structure may be a frame or a table, forexample, which may be fixed or movable as required. The support structure may ensurethat the patterning device is at a desired position, for example with respect to theprojection system.
    [0033] Any use of the terms “reticle” or “mask” herein may be considered synonymouswith the more general term “patterning device.” [0034] The term “patterning device” used herein should be broadly interpreted asreferring to any device that can be used to impart a radiation beam with a pattern in itscross-section such as to create a pattern in a target portion of the substrate. It should benoted that the pattern imparted to the radiation beam may not exactly correspond to thedesired pattern in the target portion of the substrate, for example if the pattern includesphase-shifting features or so called assist features. Generally, the pattern imparted tothe radiation beam will correspond to a particular functional layer in a device beingcreated in the target portion, such as an integrated circuit.
    [0035] The patterning device may be transmissive or reflective. Examples of patterningdevices include masks, programmable mirror arrays, and programmable LCD panels.Masks are well known in lithography, and include mask types such as binary, alternatingphase-shift, and attenuated phase-shift, as well as various hybrid mask types. Anexample of a programmable mirror array employs a matrix arrangement of smallmirrors, each of which can be individually tilted so as to reflect an incoming radiationbeam in different directions. The tilted mirrors impart a pattern in a radiation beam whichis reflected by the mirror matrix.
    [0036] The term “projection system” used herein should be broadly interpreted asencompassing any type of projection system, including refractive, reflective,catadioptric, magnetic, electromagnetic and electrostatic optical systems, or anycombination thereof, as appropriate for the exposure radiation being used, or for otherfactors such as the use of an immersion liquid or the use of a vacuum. Any use of theterm “projection lens” herein may be considered as synonymous with the more generalterm “projection system”.
    [0037] As here depicted, the apparatus is of a transmissive type (e.g. employing atransmissive mask). 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 areflective mask).
    [0038] The lithographic apparatus may be of a type having two (dual stage) or moresubstrate tables (and/or two or more patterning device tables). In such “multiple stage” machines the additional tables may be used in parallel, or preparatory steps may becarried out on one or more tables while one or more other tables are being used forexposure. The two substrate tables WTa and WTb in the example of Figure 1 are anillustration of this. An embodiment of the invention disclosed herein can be used in astand-alone fashion, but in particular it may provide additional functions in the pre¬exposure measurement stage of either single- or multi-stage apparatuses.
    [0039] The lithographic apparatus may also be of a type wherein at least a portion of thesubstrate 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. Animmersion liquid may also be applied to other spaces in the lithographic apparatus, forexample, between the mask and the projection system. Immersion techniques are wellknown in the art for increasing the numerical aperture of projection systems. The term“immersion” as used herein does not mean that a structure, such as a substrate, mustbe submerged in liquid, but rather only means that liquid is located between theprojection system and the substrate during exposure.
    [0040] Referring to Figure 1, the illuminator IL receives a radiation beam from aradiation source SO. The source and the lithographic apparatus may be separateentities, for example when the source is an excimer laser. In such cases, the source isnot considered to form part of the lithographic apparatus and the radiation beam ispassed from the source SO to the illuminator IL with the aid of a beam delivery systemBD comprising, for example, suitable directing mirrors and/or a beam expander. In othercases the source may be an integral part of the lithographic apparatus, for examplewhen the source is a mercury lamp. The source SO and the illuminator IL, together withthe beam delivery system BD if required, may be referred to as a radiation system. Theilluminator IL may comprise an adjuster AD for adjusting the angular intensitydistribution of the radiation beam. Generally, at least the outer and/or inner radial extent(commonly referred to as -outer and -inner, respectively) of the intensity distribution in apupil plane of the illuminator can be adjusted. In addition, the illuminator IL maycomprise various other components, such as an integrator IN and a condenser CO. Theilluminator may be used to condition the radiation beam, to have a desired uniformityand intensity distribution in its cross section.
    [0041] The radiation beam B is incident on the patterning device (e.g., mask) MA, whichis held on the support structure (e.g., mask table) MT, and is patterned by the patterningdevice. Having traversed the patterning device MA, the radiation beam B passesthrough the projection system PS, which focuses the beam onto a target portion C ofthe substrate W. With the aid of the second positioner PWa/PWb and position sensor IF(e.g. an interferometric device, linear encoder or capacitive sensor), the substrate tableWTa/WTb can be moved accurately, e.g. so as to position different target portions C inthe path of the radiation beam B. Similarly, the first positioner PM and another positionsensor (which is not explicitly depicted in Figure 1) can be used to accurately positionthe patterning device MA with respect to the path of the radiation beam B, e.g. aftermechanical retrieval from a mask library, or during a scan. In general, movement of thesupport structure MT may be realized with the aid of a long-stroke module (coarsepositioning) and a short-stroke module (fine positioning), which form part of the firstpositioner PM. Similarly, movement of the substrate table WTa/WTb may be realizedusing a long-stroke module and a short-stroke module, which form part of the secondpositioner PWa/PWb. In the case of a stepper (as opposed to a scanner) the supportstructure MT may be connected to a short-stroke actuator only, or may be fixed.Patterning device MA and substrate W may be aligned using patterning devicealignment marks M1, M2 and substrate alignment marks P1, P2. Although the substratealignment marks as illustrated occupy dedicated target portions, they may be located inspaces 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 deviceMA, the patterning device alignment marks may be located between the dies. Thedepicted apparatus could be used in at least one of the following modes: [0042] 1. In step mode, the support structure MT and the substrate table WTa/WTb arekept essentially stationary, while an entire pattern imparted to the radiation beam isprojected onto a target portion C at one time (i.e. a single static exposure). Thesubstrate table WTa/WTb is then shifted in the X and/or Y direction so that a differenttarget portion C can be exposed. In step mode, the maximum size of the exposure fieldlimits the size of the target portion C imaged in a single static exposure.
    [0043] 2. In scan mode, the support structure MT and the substrate table WTa/WTb arescanned synchronously while a pattern imparted to the radiation beam is projected ontoa target portion C (i.e. a single dynamic exposure). The velocity and direction of thesubstrate table WTa/WTb relative to the support structure MT may be determined by the(de-)magnification and image reversal characteristics of the projection system PS. Inscan mode, the maximum size of the exposure field limits the width (in the non-scanningdirection) of the target portion in a single dynamic exposure, whereas the length of thescanning motion determines the height (in the scanning direction) of the target portion.
    [0044] 3. In another mode, the support structure MT is kept essentially stationaryholding a programmable patterning device, and the substrate table WTa/WTb is movedor scanned while a pattern imparted to the radiation beam is projected onto a targetportion C. In this mode, generally a pulsed radiation source is employed and theprogrammable patterning device is updated as required after each movement of thesubstrate table WTa/WTb or in between successive radiation pulses during a scan. Thismode of operation can be readily applied to maskless lithography that utilizesprogrammable patterning device, such as a programmable mirror array of a type asreferred to above.
    [0045] Lithographic apparatus LA is of a so-called dual stage type which has two tablesWTa and WTb and two stations - an exposure station and a measurement station-between which the tables can be exchanged. In an embodiment, each of the two tablesWTa and WTb is a substrate table. In an embodiment, one of the tables WTa, WTb is asubstrate table and another table WTa, WTb is a measurement table which does nothold a substrate. While a substrate on a substrate table WTa, WTb is being exposed atthe exposure station, a measurement table WTa, WTb or another substrate on anothersubstrate table WTa, WTb is at the measurement station so that various preparatorysteps may be carried out. The preparatory steps may include mapping the surface of asubstrate using a level sensor LS, measuring the position of one or more alignmentmarkers on a substrate using an alignment sensor AS, and/or measuring a property ofthe projection system or projection radiation. This enables a substantial increase in thethroughput of the apparatus. If the position sensor IF is not capable of measuring theposition of a table while it is at the measurement station as well as at the exposure station, a second position sensor may be provided to enable the positions of the table tobe tracked at both stations.
    [0046] The apparatus may further include a lithographic apparatus control unit LACUwhich controls movements and measurements of the various actuators and sensorsdescribed. The control unit LACU may include signal processing and data processingcapacity to implement desired calculations relevant to the operation of the apparatus. Inpractice, control unit LACU may be realized as a system of many sub-units, eachhandling the real-time data acquisition, processing and control of a subsystem orcomponent within the apparatus. For example, one processing subsystem may bededicated to servo control of the positioner PWa/PWb. Separate units may handlecoarse and fine actuators, or different axes. A unit might be dedicated to the readout ofthe position sensor IF. Overall control of the apparatus may be controlled by a centralprocessing unit, communicating with these sub-system processing units, with operatorsand with other apparatuses involved in the lithographic manufacturing process.
    [0047] In Fig. 2 a lithography system 4 is shown for applying a pattern to a substrate.
    [0048] The system 4 comprises a track unit 6 configured to apply a layer on thesubstrate for lithographic exposure. The system 4 also comprises a lithographyapparatus 2 configured to expose the layer according to the pattern. In use, there is asubstrate flow 7 via the track unit 6 to the lithography apparatus 2 and then back to thetrack unit 6 (to apply a next layer or a final treatment before exit).
    [0049] The lithography apparatus 2 is described hereinbefore and is of the dual-stagetype in this example. The system 4 also comprises a metrology unit 8 configured tomeasure a property of the exposed pattern in the layer and a control unit 10 configuredto control the automatic substrate flow 7 among the track unit, the lithographyapparatus, and the metrology unit. For example, the metrology unit 8 is configured tomeasure the overlay between exposed patterns in successive layers on a certainsubstrate and/or an imaging parameter of an exposed pattern in a certain layer of acertain substrate. Examples of imaging parameters include the critical dimension of thepattern (so called “CDU” of the pattern), magnification (errors) and/or distortion.
    [0050] The lithography system 4 comprises a machine learning controller 12 configuredto control at least part of the system 4. In this particular example the machine learning controller 12 is configured to control the track unit 6, the lithography apparatus 2 and/orthe control unit 10 in order to optimize a property of the pattern. For this, the machinelearning controller 12 is trained on the basis of the measured property of patternsexposed on substrates of at least two so-called lots of substrates. A lot of substrates”comprises at least two substrates, and often between ten to thirty substrates. Themeasurements performed by the metrology unit 8 of the patterns of such a sequence ofsubstrates are fed to the machine learning controller via a measurement signal 14. Thismeasurement signal 14 allows the machine learning controller 12 to monitor system driftand to correct for this system drift. These measurements relate to “post-exposure” datawhich can be compared directly with the desired output result and which allowsmonitoring drift in an accurate way. This information may be used for an effective andaccurate control of the system 4.
    [0051] In the example of Fig. 2 the machine learning controller 12 is configured togenerate drift control signals (16.1 and/or 16.2) on the basis of the measurement signal14. Via drift control signal 16.1 system drift in the lithography apparatus 2 can becorrected and via drift control signal 16.2 system drift in the track unit 6 can becorrected. In a similar way a drift control signal 16.3 can be fed to the control unit 10.
    [0052] In this example the machine learning controller 12 may comprise artificialintelligence for learning from data. This artificial intelligence may comprise knownmachine learning and data mining techniques. The machine learning controller 12 maycomprise at least one algorithm selected from the following: Time Series, NeuralNetworks, Support Vector Machines, Principal Component Analysis, GeneticProgramming, Association Rule Learning, Decision Tree Learning, and/or InductiveLogic Programming.
    [0053] In an embodiment, the measurement control signal 14 may represent a propertyor parameter other than a measured property of a pattern exposed on a substrate. Forexample, measurement control signal 14 may represent a parameter or propertyassociated with exposing the pattern of the patterning device onto a substrate. In anembodiment, the parameter or property associated with exposing the pattern of thepatterning device onto a substrate may be a focus parameter or property, or a doseparameter or property, or both. Thus, in an embodiment, the machine learning controller 12 is configured to generate drift control signals (16.1 and/or 16.2) on the basis of ameasurement signal 14 that represents a focus and/or dose parameter or property thatrelates to the exposure radiation used by the lithography apparatus to expose thesubstrate according to the patterning device pattern.
    [0054] In an embodiment, the metrology unit 8 may comprise a focus sensor to measurea focal property or determine a focal parameter associated with exposing the pattern ofthe patterning device onto a substrate. Additionally or alternatively, the metrology unit 8may comprise a dose sensor to measure a dose property or determine a doseparameter associated with exposing the pattern of the patterning device onto asubstrate. In an embodiment, the focus sensor and/or dose sensor is part of thelithography apparatus 2.
    [0055] A further embodiment of a lithography system 4 is shown in Fig. 3A. Themachine learning controller 12 is configured to generate the drift control signal(s) 16.1,16.2 (and probably also a drift control signal 16.3, not depicted in the Figure, forcontrolling the control unit 10) on the basis of the measurement signal 14. In thisexample the drift control signals 16.1, 16.2 (probably 16.3) are also based onlithography apparatus information 18. The lithography apparatus information includes atleast one selected from: information about which substrate chuck of the dual stagelithography apparatus was used for the particular substrate [where the dual stagelithography apparatus comprises two chucks for simultaneously supporting two differentsubstrates in a measurement phase and exposure phase], information about thedynamics of the patterning device support of the lithography apparatus, informationabout the dynamics of the substrate stage of the lithography apparatus, informationabout the substrate alignment, information about the substrate leveling, informationabout an optical property of the projection system of the lithography apparatus and/orinformation about a parameter or property associated with exposing the pattern of thepatterning device onto a substrate such as focus, dose and/or another parameter (inwhich case the measurement signal 14 may represent a measured property of patternsexposed on substrates). The mechanical (stage dynamics) and optical (lens elements)parts of the lithography suffer from wear. Generally, this occurs gradually in time andcan be compensated for on the basis of a machine learning drift control signal 16.1.
    [0056] In an embodiment, the machine learning controller 12 may be configured togenerate the drift control signal(s) (16.1,16.2 and/or 16.3)) on the basis of substrateprocess information 20. The substrate process information 20 may comprise track unitinformation including at least one selected from: spin coating information, bakinginformation and/or the sequence of the substrate in the lot of substrates. In a lithographysystem 4 having more than one (parallel) track units 6, information about which trackunit 6 treated a certain substrate with a new layer may be part of the track unitinformation.
    [0057] In an embodiment, the machine learning controller 12 may be configured togenerate the drift control signal(s) (16.1,16.2 and/or 16.3) on the basis of plantinformation 22 regarding the plant where the lithographic system is housed. Plantinformation 22 may include environmental data comprising at least one selected from:temperature in the plant, humidity in the plant, and/or external information such asetching information.
    [0058] The lithography apparatus information 18, the substrate process information 20and the plant information 22 can be seen as “pre-exposure” information in the sensethat use of this information can be made before exposure of the actual substrate.However, this information also can be used after the exposure of a substrate forsubstrates later in the sequence (in a “post-exposure” way) which may also be useful fordrift control.
    [0059] In the example of Fig. 3A the machine learning controller 12 is configured togenerate a real-time control signal (24.1,24.2 and/or 24.3) on the basis of lithographyapparatus information. The real-time control signal 24.1 may be used to control thelithography apparatus 2 and the real-time control signal 24.2 may be used to control thetrack unit 6 (the real-time control signal 24.3 may be generated to control the unit 10).The real-time control signal 24.1 may be effective in correcting, for example, substrate-to-substrate variation: for example an alignment error for a specific substrate can becorrected before exposure in the lithography apparatus 2. The real-time control signal24.2 may be effective for correcting, for example, a specific characteristic in the trackunit 6 (as an example: when more than one substrate flow through a track unit, orthrough track units in parallel, is used in the lithography system 4, relevant information includes information about which track unit has been used for a substrate and whichcharacteristic that yields to the layer in order to correct the lithography apparatus 2 forthat specific substrate).
    [0060] The real-time control signal (24.1,24.2, 24.3) may be based, in addition to thelithography information, also on the measured property by the metrology unit 8(measurement signal 14), the substrate process information 20 and/or plant information22.
    [0061] In Fig. 3B a machine learning controller 12 is shown which can be used in alithography system 4. In this particular example, an accurate edge-to-edge control isdescribed which is desirable in complementary lithography where one-dimensionalgrating lines are exposed with immersion lithography and the two-dimensional shape isgenerated by putting so-called “cuts” on these lines to define line ends.
    [0062] In the example of Fig. 3B, the machine learning controller 12 is configured togenerate edge-to-edge control signals (16.1,24.1, 16.2, 24.2) which are combinationsof the overlay and critical dimension control signals generated by respectively a firstcontroller (26.1) and a second controller (26.2). The lithography apparatus 2 iscontrolled with the edge-to-edge control signals (16.1,24.1) and the track unit 6 iscontrolled with the edge-to-edge control signals (16.2, 24.2). The machine learningcontroller 12 is also capable of controlling the control unit 14 with edge-to-edge controlsignals (16.3, 24.3).
    [0063] The signals 16.1, 16.2 (16.3) are drift control signal and the signals 24.1, 24.2(24.3) are real-time control signals. The generated control signals are based onmeasurement signal 14, lithography information 18, substrate process information 20and/or plant information 22.
    [0064] The first controller 26.1 comprises a first sub-controller 28.1 configured togenerate a drift overlay control signal and a second sub-controller 28.2 configured togenerate a real-time overlay control signal. The second sub-controller 26.2 comprises athird sub-controller 30.1 configured to generate a drift critical dimension control signaland a fourth sub-controller 30.2 configured to generate a real-time critical dimensioncontrol signal.
    [0065] Furthermore, the machine learning controller comprises an estimation unit 28.3configured to generate an overlay prediction signal and an estimation unit 30.3configured to generate a critical dimension prediction signal. The overlay predictionsignal and the critical dimension signal are fed to the edge-to-edge controller 26.3 forgenerating edge-to-edge control signals (16.1,24.1,16.2, 24.2) as describedhereinbefore (thus in this example there are both a drift edge-to-edge control signal andreal-time edge-to-edge control signal).
    [0066] In an embodiment, the edge-to-edge control signals (16.1,24.1, 16.2, 24.2) mayinclude another signal representing another parameter. For example, the otherparameter may be another imaging parameter, such as a focus parameter. In anembodiment, the edge-to-edge error control signal may involve a different combinationof imaging parameters. For example, the edge-to-edge control signal may comprise acombination of the mentioned overlay control signal and another imaging parametersuch as a focus parameter. Or, the edge-to-edge control signal may comprise acombination of the mentioned critical dimension signal and another imaging parametersuch as a focus parameter.
    [0067] In an embodiment, reference herein to machine learning controller 12 beingconfigured to control the track unit 6, the lithography apparatus 2 and/or the control unit10 may include control of one or more specific devices within the track unit 6, thelithography apparatus 2 and/or the control unit 10, and/or include control of one or morespecific devices in the lithography system 4 that are outside, but associated with, thetrack unit 6, the lithography apparatus 2 and/or the control unit 10, or both. For example,the machine learning controller 12 may adjust the radiation source SO in, or associatedwith, the lithography apparatus 2. Thus, in an embodiment, the machine learningcontroller 12 adjusts radiation source SO used to generate the exposure radiation,which adjustment may include adjusting the focus and/or dose of the exposureradiation. Thus, in an embodiment, the radiation source SO (whether included in thelithography apparatus or associated with the lithography apparatus) may have themachine learning controller 12 or have a controller operable with the machine learningcontroller 12.
    [0068] The controllers described herein may each or in combination be operable whenone or more computer programs are read by one or more computer processors locatedwithin at least one component of the lithographic system. The controllers may each or incombination have any suitable configuration for receiving, processing, and sendingsignals. One or more processors are configured to communicate with the at least one ofthe controllers. For example, each controller may include one or more processors forexecuting computer programs that include machine-readable instructions for one ormore of the methods described above. The controllers may include data storagemedium configured to store such computer programs, and/or hardware to receive sucha medium. So the controller(s) may operate according to the machine readableinstructions of one or more computer programs.
    [0069] An embodiment may take the form of a computer program containing one ormore sequences of machine-readable instructions describing a method as disclosedabove, or a data storage medium (e.g. semiconductor memory, magnetic or optical disk)having such a computer program stored therein. Other aspects of the invention are set-out as in the following numbered clauses.
    Clauses 1. A lithography system configured to apply a pattern to a substrate, comprising: a track unit configured to apply a layer on the substrate for lithographic exposure;a lithography apparatus configured to expose the layer according to the pattern;a metrology unit configured to measure a property of the exposed pattern in thelayer and/or measure a property associated with exposing the pattern onto thesubstrate; a control unit configured to control an automatic substrate flow among the trackunit, the lithography apparatus, and the metrology unit; and a machine learning controller configured to control the lithography system tooptimize a property of the pattern, the machine learning controller configured to betrained on the basis of the measured property and to correct lithography system drift byadjusting one or more selected from: the lithography apparatus, the track unit and/or thecontrol unit. 2. A lithography system configured to apply a pattern to a substrate, the systemcomprising: a lithography apparatus configured to expose a layer of the substrate accordingto the pattern; and a machine learning controller configured to control the lithography system tooptimize a property of the pattern, the machine learning controller configured to betrained on the basis of a property measured by a metrology unit configured to measurea property of the exposed pattern in the layer and/or a property associated withexposing the pattern onto the substrate, and to correct lithography system drift byadjusting one or more selected from: the lithography apparatus, a track unit configuredto apply the layer on the substrate for lithographic exposure, and/or a control unitconfigured to control an automatic substrate flow among the track unit, the lithographyapparatus, and the metrology unit. 3. The lithography system according to clause 2, further comprising: the track unit configured to apply the layer on the substrate for lithographicexposure; the metrology unit configured to measure the property of the exposed pattern in thelayer; and the control unit configured to control the automatic substrate flow among the track unit,the lithography apparatus, and the metrology unit. 4. The lithography system according to any of clauses 1 to 3, wherein thelithography system drift comprises at least lithography apparatus drift, track unit drift,control unit drift, and/or metrology unit drift. 5. The lithography system according to any of clauses 1 -4, wherein the machinelearning controller is configured to be trained on the basis of at least two lots ofsubstrates. 6. The lithography system according to any of clauses 1 -5, wherein the machinelearning controller comprises a first controller configured to control overlay betweenpattern layers and/or a second controller configured to control a critical dimension of thepattern. 7. The lithography system according to clause 6, wherein the machine learningcontroller comprises both the first and second controller, wherein the first controllercomprises a first sub-controller configured to generate a first drift control signal for apattern overlay, wherein the second controller comprises a third sub-controllerconfigured to generate a second drift control signal for a critical dimension of thepattern, and wherein the machine learning controller is configured to generate an edge-to-edge placement signal which is a combination of the first and second drift controlsignals. 8. The lithography system according to clause 7, wherein the edge-to-edgeplacement signal is a warning signal, and wherein the machine learning controller is configured to control the lithography apparatus, track unit and/or control unit on thebasis of the warning signal if it exceeds a threshold level. 9. The lithography system according to clause 8, wherein the edge-to-edgeplacement signal is a third drift control signal, wherein the machine learning controller isconfigured to control the lithography apparatus, track unit and/or control unit with thethird drift control signal. 10. The lithography system according to any of clauses 7-9, wherein the machinelearning controller is configured to deduce at least one of the drift control signals fromthe measured property and to correct the lithography system for drift by adjusting thelithography apparatus, track unit and/or control unit with the deduced drift control signal. 11. The lithography system according to clause 10, wherein the machine learningcontroller is configured to deduce the at least one drift control signal also on the basis oflithography apparatus information, the lithography apparatus information including atleast one selected from: information about a substrate chuck of the lithographyapparatus used for exposure, information about the dynamics of a patterning devicesupport of the lithography apparatus, information about the dynamics of a substratestage of the lithography apparatus, information about substrate alignment, informationabout substrate leveling, information about an optical property of a projection system ofthe lithography apparatus, and/or information about a parameter or property associatedwith exposing the pattern of the patterning device onto a substrate. 12. The lithography system according to clause 10 or clause 11, wherein themachine learning controller is configured to deduce the at least one drift control signalalso on the basis of substrate process information, the substrate process informationincluding at least one selected from: spin coating information, baking information,etching information and/or the sequence of the substrate in the lot of substrates. 13. The lithography system according to any of clauses 10-12, wherein the machinelearning controller is configured to deduce the at least one drift control signal also on thebasis of plant information regarding the plant housing the lithography system, the plantinformation including environmental data comprising at least one selected from:temperature in the plant and/or humidity in the plant. 14. The lithography system according to any of clauses 10-13, wherein the firstcontroller comprises a second sub-controller configured to deduce a real-time overlaycontrol signal and the second controller comprises a fourth sub-controller configured todeduce a real-time critical dimension control signal, wherein the real-time control signalsare based on lithography apparatus information including at least one selected from:information about a substrate chuck of the lithography apparatus used for exposure,information about the dynamics of a patterning device support of the lithographyapparatus, information about the dynamics of a substrate stage of the lithographyapparatus, information about substrate alignment, information about substrate leveling,information about an optical property of a projection system of the lithography apparatusand/or information about a parameter or property associated with exposing the patternof the patterning device onto a substrate, wherein the real-time control signalscorrespond to substrate-to-substrate lithography system variation, and wherein themachine learning controller is configured to correct the lithography apparatus, track unitand/or control unit for the substrate-to-substrate lithography system variation with thereal-time control signals, and wherein the machine learning controller is configured todeduce the real-time control signals also on the basis of the measured property. 15. A machine learning controller for use in a lithography system according to any ofclauses 1-14.
    权利要求:
    Claims (1)
    [1]
    A lithography device comprising: an illumination device adapted to deliver a radiation beam; a carrier constructed to support a patterning device, the patterning device being capable of applying a pattern in cross-section of the radiation beam to form a patterned radiation beam; a substrate table constructed to support a substrate; and a projection device adapted to project the patterned radiation beam onto a target area of the substrate, characterized in that the substrate table is adapted to position the target area of the substrate in a focal plane of the projection device.
    类似技术:
    公开号 | 公开日 | 专利标题
    NL2013249A|2015-02-23|Lithography system and a machine learning controller for such a lithography system.
    US20220011681A1|2022-01-13|Lithographic method and lithographic apparatus
    TWI388944B|2013-03-11|Lithographic apparatus and device manufacturing method
    US8706442B2|2014-04-22|Alignment system, lithographic system and method
    TWI646400B|2019-01-01|Photolithography device, component manufacturing method and related data processing device and computer program product
    KR100869308B1|2008-11-18|Lithographic apparatus and device manufacturing method
    US8717536B2|2014-05-06|Lithographic apparatus, device manufacturing method and associated data processing apparatus and computer program product
    WO2005053007A1|2005-06-09|Exposure method, device manufacturing method, exposure device, and program
    US10846457B2|2020-11-24|Lithography system, simulation apparatus, and pattern forming method
    US10078272B2|2018-09-18|Lithographic method and apparatus
    KR101905632B1|2018-10-10|Object positioning in lithography
    US20160041478A1|2016-02-11|Lithography Cluster, Method and Control Unit for Automatic Rework of Exposed Substrates
    JP2020513113A|2020-04-30|Lithographic method and apparatus
    JP4902685B2|2012-03-21|Stage system, lithographic apparatus including the stage system, and device manufacturing method
    US11022895B2|2021-06-01|Lithographic apparatus adjustment method
    同族专利:
    公开号 | 公开日
    IL233914A|2019-06-30|
    US20160170311A1|2016-06-16|
    WO2015024783A1|2015-02-26|
    TW201514638A|2015-04-16|
    TWI600976B|2017-10-01|
    US9971251B2|2018-05-15|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    US6368883B1|1999-08-10|2002-04-09|Advanced Micro Devices, Inc.|Method for identifying and controlling impact of ambient conditions on photolithography processes|
    EP1235109A1|2001-02-23|2002-08-28|Infineon Technologies AG|Method for exposing at least one or at least two semiconductor wafers|
    US7047099B2|2001-06-19|2006-05-16|Applied Materials Inc.|Integrating tool, module, and fab level control|
    US6737208B1|2001-12-17|2004-05-18|Advanced Micro Devices, Inc.|Method and apparatus for controlling photolithography overlay registration incorporating feedforward overlay information|
    US7076320B1|2004-05-04|2006-07-11|Advanced Micro Devices, Inc.|Scatterometry monitor in cluster process tool environment for advanced process control |
    US7239368B2|2004-11-29|2007-07-03|Asml Netherlands B.V.|Using unflatness information of the substrate table or mask table for decreasing overlay|
    US7875851B1|2006-05-01|2011-01-25|Advanced Micro Devices, Inc.|Advanced process control framework using two-dimensional image analysis|KR20180036239A|2016-09-30|2018-04-09|삼성전자주식회사|Method of optimizing a mask using pixel based learning|
    DE102016225899A1|2016-12-21|2018-06-21|Carl Zeiss Smt Gmbh|Method and apparatus for modifying imaging properties of an optical system for microlithography|
    US10409171B2|2017-01-25|2019-09-10|Kla-Tencor Corporation|Overlay control with non-zero offset prediction|
    US10444643B2|2017-03-24|2019-10-15|Nikon Research Corporation Of America|Lithographic thermal distortion compensation with the use of machine learning|
    EP3382606A1|2017-03-27|2018-10-03|ASML Netherlands B.V.|Optimizing an apparatus for multi-stage processing of product units|
    US10733744B2|2017-05-11|2020-08-04|Kla-Tencor Corp.|Learning based approach for aligning images acquired with different modalities|
    US10497105B2|2017-11-01|2019-12-03|Google Llc|Digital image auto exposure adjustment|
    EP3518040A1|2018-01-30|2019-07-31|ASML Netherlands B.V.|A measurement apparatus and a method for determining a substrate grid|
    WO2021063728A1|2019-10-02|2021-04-08|Asml Netherlands B.V.|Process monitoring and tuning using prediction models|
    WO2022028805A1|2020-08-06|2022-02-10|Asml Netherlands B.V.|Method and apparatus for concept drift mitigation|
    法律状态:
    2015-04-15| WDAP| Patent application withdrawn|Effective date: 20150402 |
    2015-04-22| WDAP| Patent application withdrawn|Effective date: 20150402 |
    优先权:
    申请号 | 申请日 | 专利标题
    US201361867992P| true| 2013-08-20|2013-08-20|
    US201361867992|2013-08-20|
    US201361891254P| true| 2013-10-15|2013-10-15|
    US201361891254|2013-10-15|
    [返回顶部]