Patent Application
20080090312
The various features of embodiments of the present invention may be more readily understood with reference to the following detailed description taken in conjunction with the accompanying drawings, where like reference numerals designate like structural elements, and in which:
The embodiments of the present invention facilitate employing lithography to apply a pattern to a substrate (i.e., “patterning a substrate”). In some embodiments, the lithography comprises contact lithography involving a contact between a patterning tool and a substrate. In various embodiments, the present invention employs nanoscale displacement sensing and estimation (nDSE) to estimate and reduce and effect of a disturbance on an alignment associated with the lithography. The nDSE is image-based according to the present invention. In particular, the present invention employs images of aligned objects acquired before and after the disturbance. In some embodiments, the images are optical images. The disturbance is one or more of induced by a contact between the aligned objects, associated with differential vibration of aligned objects, produced by a temperature differential between and across aligned objects, and generated by a mechanical drift or slippage of a lithography system operating on the objects, according to various embodiments of the present invention.
As used herein, nanoscale displacement sensing and estimation (nDSE) is an image-based methodology for detecting and quantifying a relative displacement of or between one or more objects using a pair of images of the objects. In particular, as defined herein nDSE is essentially any methodology that employs a comparison of pixel-level data contained in the pair of images to estimate one or both of a magnitude and a direction of the relative displacement (i.e., a vector displacement). In some embodiments, nDSE comprises one or more of phase delay detection-based (PDD-based) nDSE, statistical image correlation-based nDSE, and feature extraction-based nDSE applied to the images. Using nDSE, sub-pixel resolution of an estimated relative displacement often can be achieved. For example, a statistical image correlation using nearest neighbor (N-cubed) algorithms can provide a displacement estimate resolution of less than or equal to about 1/100th of a pixel.
PDD-based nDSE models a relative displacement of objects in terms of a constant phase delay across frequencies present within a frequency-domain transformation of the pixel-level data of images of the object. As an algorithm, phase delay detection (PDD) is based on a shift property of frequency-domain transformations, such as the Fourier transform. To estimate displacement, PDD extracts the phase delay from a pair of transformed images by observing phases of the frequencies before and after a displacement of the objects as recorded by the images. A slope of the extracted phase delay provides the estimated displacement of the objects. PDD is generally considered to be a global method with a capacity to handle relatively large relative displacements with reasonable computational complexity. PDD also provides for an ability to compensate for colored noise sources that may be present in the images. PDD-based nDSE is described further by Gao et al., in U.S. Patent Application Publication, US 2006-0045314 A1, incorporated herein by reference.
Statistical image correlation-based nDSE employs variations of a nearest neighbor navigation (N-cubed) algorithm. The statistical image correlation-based nDSE statistically extracts a relative displacement of the objects represented in a pair of images (i.e., an image before and an image after displacement) by correlating one or both of features and patterns within the image. In particular, values of neighboring pixels in the images are correlated and a result is fit to a function. An extremum (i.e., a minimum or a maximum) of the fitted function then provides the relative displacement estimate. Statistical image correlation-based nDSE is generally considered to be a local method with an ability to handle relatively small displacements and often to provide much higher resolution than the PDD-based nDSE. While somewhat less able to deal with colored noise sources than PDD-based nDSE, statistical image correlation-based nDSE is generally more efficient to implement and maintain. In addition, statistical image correlation-based nDSE may be more robust when presented with white noise than the PDD-based nDSE. Statistical image correlation-based nDSE is described further by Gao et al., U.S. Patent Application Publication, US 2006-0045313 A1, incorporated herein by reference.
Feature extraction-based nDSE estimates a relative displacement using image features or patterns extracted from the pair of images. In particular, one or more image features are identified in a first image of the pair. A corresponding one or more image features is identified in a second image of the pair. For example, edge detection may be employed to identify the one or more features (e.g., edges of or within the objects). In other examples, any one of various object identification techniques is used to identify the features (e.g., specific elements of the objects). Locations of the identified features are then compared between the two images to estimate the relative displacement. For example, a middle or centroid of an identified object in the first and second images may be determined. A difference in a location of the respective determined middles is the measured. The measured difference provides an estimate of the relative displacement of the objects.
Estimates of a relative displacement are provided by nDSE, according to the present invention. As used herein, nDSE generally does not track an actual position of the object or objects. The relative displacement (i.e., offset) may comprise a displacement of an object(s) relative to a reference frame, in some embodiments. For example the relative displacement may be a displacement of the object relative to a camera producing the images. In other embodiments the relative displacement estimate represents relative displacement of an object with respect to another object or objects being imaged. For example, the relative displacement may be between a patterning tool (e.g., mask, mold, etc.) and a substrate using features visible separately on each of the patterning tool and the substrate. Such relative displacement estimates are readily produced by nDSE. Moreover, once aligned, nDSE can detect and quantify a change in an alignment between objects (e.g., the patterning tool and the substrate). Estimating relative displacements using nDSE is further described by Picciotto et al., U.S. Pat. No. 7,085,673, as well as by Picciotto et al., U.S. Patent Application Publication, US 2006-0047462 A1, both of which are incorporated herein by reference, in their entireties.
The displacement estimate produced by nDSE does not employ specific image information (e.g., an alignment mark or pattern) but instead employs arbitrary image features and patterns encoded in the pixel-level data. The use of arbitrary image features and patterns in the objects distinguishes nDSE used for object alignment from conventional alignment approaches that require a high-precision alignment mark to detect and quantify misalignments and displacements. Furthermore, the image features and patterns employed by nDSE are generally not optically colocated (i.e., not visually aligned on top of one another) as required for alignment marks employed in conventional alignment. Instead, nDSE employs image features and patterns that are optically nearby but not visually occluding one another. For example, an arbitrary image feature on a patterning tool employed by nDSE may be visually adjacent to another arbitrary image feature on the substrate when viewed from a perspective of an optical sensor (e.g., camera). However, while generally not employing image features and patterns that are optically colocated, nDSE is capable of employing optically colocated image features and patterns instead of or in addition to those that are not optically colocated.
Further herein, the term ‘contact lithography’ generally refers to essentially any lithographic methodology that employs a direct or physical contact between means for providing a pattern or the patterning tool and means for receiving the pattern or the substrate, including a substrate having a pattern receiving layer, without limitation. Specifically, ‘contact lithography’ as used herein includes, but is not limited to, various forms of photographic contact lithography, X-ray contact lithography, and imprint lithography. Imprint lithography includes, but is not limited to, micro-imprint lithography and nano-imprint or nanoscale imprint lithography (NIL).
For example, in photographic contact lithography, a physical contact is established between a photomask (i.e., the patterning tool) and a photosensitive resist layer on the substrate (i.e., the pattern receiving means). During the physical contact, visible light, ultraviolet (UV) light, or another form of radiation passing through the photomask exposes the photoresist. As a result, a pattern of the photomask is transferred to the substrate. In imprint lithography, a mold (i.e., the patterning tool) transfers a pattern to the substrate through an imprinting process, for example. In some embodiments, a physical contact between the mold and a layer of formable or imprintable material on the substrate (i.e., the pattern receiving means), transfers the pattern to the substrate.
For simplicity herein, no distinction is made between the substrate and any layer or structure on the substrate (e.g., photoresist layer or imprintable material layer) unless such a distinction is necessary for proper understanding. As such, the pattern receiving means is generally referred to herein as a ‘substrate’ irrespective of whether a resist layer or a formable layer may be employed on the substrate to receive the pattern. Moreover, the patterning tool (e.g., photomask, X-ray mask, imprint mold, template, etc.) is also referred to herein as either a ‘mold’ or a ‘mask’ for simplicity of discussion and not by way of limitation. Examples described herein are provided for illustrative purposes only and not by way of limitation.
In other embodiments, establishing 110 an initial alignment comprises aligning separate ones of the objects with respect to one another. In particular, establishing 110 an initial alignment establishes a relative orientation of a plurality of objects. For example, establishing 110 an initial alignment may align a patterning tool (or a pattern thereon) with a substrate being patterned, each of the patterning tool and substrate representing an object of the plurality. Establishing 110 establishes a relative positioning of the patterning tool and the substrate, according to the embodiment. In general, the established 110 initial alignment may be one or both of a lateral alignment (e.g., x-y) and an angular alignment (e.g., ω alignment) of the patterning tool relative to the substrate.
In some embodiments, establishing 110 an initial alignment comprises manually adjusting one or both of a position and an orientation of the objects. For example, an operator may observe the objects and manually command a positioning system to move the objects (e.g., one or more of the objects may be moved) until a desired alignment is achieved. An alignment operation of a conventional mask aligner system is an example of establishing 110 an initial alignment of an exemplary patterning tool and substrate.
In other embodiments, establishing 110 an initial alignment comprises automatically aligning the objects (e.g., a relative position of the patterning tool and the substrate). An automated alignment system compares a current position of an object or objects relative to a desired position or positions and provides an input to the positioning system that adjusts positions or relative positions of the objects to correspond to desired positions, for example. In some embodiments of establishing 110 an initial alignment automatically, nDSE is employed to measure a displacement error between the current position and the desired position. Once the displacement error is determined, one or more of the objects are moved by the positioning system to the desired position or positions. Generally, establishing 110 an initial alignment is performed before or at a beginning of the lithography process.
The method 100 of maintaining an alignment further comprises employing 120 nanoscale displacement sensing and estimation (nDSE) based feedback control of relative positions of one or more of the objects. In particular, employing 120 nDSE-based feedback control uses nDSE to quantify a relative displacement of the objects from the established 110 initial position. Employing 120 nDSE-based feedback control maintains the alignment during the lithography process. Specifically, the quantified relative displacement is reduced and, in some embodiments, minimized by nDSE-based feedback control of the position. The nDSE-based feedback control of the position accounts and corrects for disturbances of the object after the initial alignment is established 110.
The disturbance of the objects after initial alignment represents a change in the alignment (i.e., one or both of a location and an orientation) of the objects relative to the initial alignment. The change degrades the alignment. In general, the disturbance may be one or both of a periodic disturbance and an aperiodic disturbance. A periodic disturbance is a change in the alignment of the objects that is repeating and is characterized by a disturbance frequency. A periodic disturbance may have multiple, discrete disturbance frequencies. An aperiodic disturbance is generally non-repeating and may be represented by a continuous range of disturbance frequencies. Whether periodic, aperiodic, or both, the disturbance may be represented by a frequency spectrum. A Nyquist rate or frequency is defined as two (2) times a highest frequency of interest in the frequency spectrum. In some embodiments, a highest frequency of interest is determined by a frequency in the frequency spectrum above which a power of the spectrum is sufficiently small to be ignored. In some embodiments, the highest frequency of interest is a point in the frequency spectrum above which the disturbance essentially does not interfere with an alignment of the objects as defined by a predetermined alignment accuracy.
In some instances, the disturbance may be due to or be a result of a vibration in the environment. Such environmental vibrations may be mechanical vibrations transmitted through a table or bench that supports a lithography system that performs the lithography. Examples of mechanical vibrations include, but are not limited to, people walking in a vicinity of the lithography system or equipment being moved in the vicinity of the lithography system. Such mechanical vibrations are often present even when using one or both of a passive vibration isolation system and an active vibration isolation system, for example. Such mechanical vibrations can be both periodic and aperiodic. Other environmental vibrations may be acoustic vibrations. Acoustic vibrations are often associated with an air ventilation system or other clean room equipment, for example. Acoustic vibrations are generally coupled to the lithography system through the air surrounding the system. Acoustic vibrations of significance are largely periodic.
In some instances, the disturbance may be a result of a temperature difference within or between the objects or within an environment through which the objects are moved. For example, a thermal expansion or contraction that introduces a relative displacement between the patterning tool and the substrate may occur during lithography. Temperature differences are typically aperiodic and characterized by mostly low frequency spectral components.
In other instances, the disturbance may be associated with the objects themselves or with the lithography system. For example, the disturbance may result from a physical contact between one object and another object. When the type of lithography is contact lithography, a contact between contacting surfaces of the patterning tool and the substrate may induce a drift or a shift in a relative alignment therebetween, for example. In another example, the disturbance is due to a mechanical drift or slippage of a mechanism of the contact lithography system as the patterning tool is moved into contact with the substrate. As used herein, mechanical drift or slippage is defined as an unwanted or unintended motion of the lithography system.
Employing 200 feedback control further comprises estimating 220 a relative displacement of the objects using nDSE applied to the image. In particular, nDSE is applied to the acquired 210 image and a reference image. In some embodiments, the reference image represents the objects prior to the disturbance. Estimating 220 a relative displacement produces an estimate of the relative displacement by sensing a difference evident between the acquired 210 image and the reference image according to nDSE. The estimated 220 relative displacement represents a position error of the objects due to the disturbance.
In some embodiments, estimating 220 a relative displacement uses statistical image correlation-based nDSE. In other embodiments, phase delay detection-based (PDD-based) nDSE is used for estimating 220 a relative displacement. In yet other embodiments, estimating 220 a relative displacement uses both PDD-based nDSE and statistical image correlation-based nDSE. In still other embodiments, other nDSE methods such as feature extraction-based nDSE are used for estimating 220 a relative displacement instead of or in addition to one or both of statistical image correlation-based nDSE and PDD-based nDSE.
Employing 200 feedback control further comprises adjusting 230 relative positions of one or more of the objects to reduce the relative displacement. For example, a position of one of the objects may be adjusted 230 by instructing a positioning system of the lithography system to move the object relative to others of the objects in a manner that reduces the relative displacement. In terms of the position error, the position system moves the object by (−Δx, −Δy) when the position error is estimated by nDSE to be (Δx, Δy). After the object is moved, the position error is reduced and, in some embodiments, the position error is minimized. As such, the employed 200 feedback control applies negative feedback of the estimated relative displacement to reposition the objects after the displacement, in some embodiments. In some embodiments, acquiring 210 an image, estimating 220 a relative displacement, and adjusting 230 relative positions of one or more of the objects are repeated at a rate that exceeds a Nyquist frequency of the disturbance to facilitate real-time feedback control during lithography.
In some embodiments (not illustrated), the method 100 of maintaining an alignment further comprises acquiring a reference image of the objects. The reference image is acquired after the initial alignment is established 110. For example, the reference image may be acquired by the same optical sensor used in acquiring 210 an image described above. In some embodiments, the reference image is acquired before the disturbance degrades the alignment of the objects. In some embodiments, acquiring the reference image provides the reference image used in estimating 220 a relative displacement described above with respect to employing 200 feedback control. In some embodiments, the reference image is acquired after adjusting 230 relative positions of one or more of the objects. In particular, acquiring the reference image after adjusting 230 relative positions provides a new reference image to replace a previously acquired reference image, according to some embodiments.
As illustrated in
As illustrated in
In some embodiments, determining 330′ a phase-frequency slope comprises identifying at least one frequency of the identified 320′ phases and corresponding frequencies, the at least one frequency having a known property. In these embodiments, determining 330′ a phase-frequency slope further comprises assigning weights to the identified 320′ phases, the weights being dependent on the at least one frequency of the identified 320′ phases. For example, the known property and the at least one frequency associated therewith may be frequencies that are expected to be noisy based on known characteristics of the lithography system. In some embodiments, a zero weight is assigned. The assigned weights are used in determining 330′ a phase-frequency slope, in some of these embodiments. Further details of PDD-based nDSE are described further in Gao et al., in U.S. Patent Application Publication, US 2006-0045314 A1, referenced above.
As illustrated in
The method 400 of disturbance compensation further comprises acquiring 420 a second image of the patterning tool and the substrate after a disturbance to the alignment that degrades the alignment. In particular, the second image is acquired 420 either after a disturbance is detected or after a disturbance is suspected. For example, the second image may be acquired 420 when a sensor detects a disturbance. Alternatively, the second image is acquired 420 periodically during contact lithography expecting or assuming that a disturbance has happened.
The method 400 of disturbance compensation further comprises estimating 430 an alignment error induced by the disturbance. In particular, estimating 430 an alignment error is performed using nDSE applied to the first image and the second image. In various embodiments, using nDSE may comprise using one or more of PDD-based nDSE, statistical image correlation-based nDSE, and feature extraction-based nDSE. For example, using nDSE applied to the images may comprise using 300, 300′ nDSE described above with respect to
The method 400 of disturbance compensation further comprises adjusting 440 a relative position of the patterning tool and the substrate to reduce the alignment error. In particular, the estimated 430 alignment error is provided to a positioning system of the contact lithography system. The positioning system moves one or both of the patterning tool and the substrate in a manner responsive to the provide alignment error. The movement reduces the alignment error. In some embodiments, acquiring 420 a second image, estimating 430 an alignment error, and adjusting 440 a relative position are repeated to further reduce and in some embodiments, are repeated to minimize the alignment error. In some embodiments, acquiring 420 a second image, estimating 430 an alignment error, and adjusting 440 a relative position are repeated at a rate that is equal to or greater than a Nyquist frequency of the disturbance to facilitate real-time disturbance compensation during contact lithography.
In some embodiments, one or both of the method 100 of maintaining an alignment and the method 400 of disturbance compensation are employed in a contact lithography alignment system. For example, the method 100 of maintaining may be implemented as instructions of a computer program stored in a memory of the contact lithography alignment system. A processor of the contact lithography system may execute the instructions to apply the method 100 to maintain an alignment. Similarly, the method 400 of disturbance compensation may be implemented as instructions of a computer program stored in a memory of the alignment system, and executed by the processor, such that the method 400 may be applied to the contact lithography alignment system to provide disturbance compensation.
As illustrated in
The contact lithography alignment system 500 further comprises a feedback processor 530 providing nDSE. The feedback processor 530 receives the image or data representing the image from the optical sensor 520. The feedback processor 530 determines an alignment error from the image using nDSE. The provided nDSE comprises one or more of statistical image correlation-based nDSE, PDD-based nDSE, and feature extraction-based nDSE, in various embodiments. The feedback processor 530 produces an output representing the determined alignment error, in some embodiments. The alignment error represents a positional deviation from an initial alignment of the objects 510.
The contact lithography alignment system 500 further comprises a position controller 540. The alignment error from the feedback processor 530 is communicated to the position controller 540. The position controller 540 receives the determined alignment error (e.g., feedback processor output signal) and controls positioning of the objects 510. In particular, the position controller 540 adjusts relative positions of the objects 510 to reduce the alignment error.
In some embodiments, the position controller 540 provides an input to a stage 550 upon which is mounted one or more of the objects 510 (e.g., a substrate). The stage 550 enables controlled movement of the stage-mounted objects 510. The provided input from the position controller 540 instructs the stage 550 to move the objects 510. In some embodiments, the stage 550 is a precision stage 550. For example, the precision stage 550 may be an N-point XY200Z20A-A nano-positioning stage manufactured by nPoint Incorporated, Madison, Wis. In an example, only the substrate 514 may be mounted to the precision stage 550 and the patterning tool 512 is fixed. In this example, a position of the substrate 514 is adjusted relative to a patterning tool 512 position. In another example, the patterning tool 512 is moved (e.g., by another stage or similar positioning system, not illustrated) instead of or in addition to the stage-mounted substrate 514 to reduce the alignment error.
In some embodiments, the reduced alignment error compensates for a disturbance of the objects 510 during contact lithography. In some embodiments, the initial alignment of the objects 510 establishes a desired relative position of the objects 510. For example, the desired relative position may represent an alignment of a pattern on the patterning tool 512 with a receiving surface of the substrate 514. In some embodiments, the reference image is produced following the initial alignment and prior to the disturbance.
As illustrated in
In some embodiments, the instructions of the computer program 630 further implement statistical image correlation-based nDSE comprising fitting a function to correlation data for the image and the reference image and identifying an extremum of the function after fitting. The extremum, a minimum or a maximum of the function, provides the estimate. In some embodiments, the instructions of the computer program 630 further implement phase delay detection-based (PDD-based) nDSE comprising performing a frequency transform of image data representing the image and the reference image to produce frequency domain data. The instruction further implement finding a slope associated with a phase of the frequency domain data. The slope provides the estimate of the relative displacement. In some embodiments, the instructions implement feature extraction-based nDSE comprising identifying, locating and comparing corresponding image features and patterns in the image and reference image. The comparison determines the estimate of relative displacement.
Thus, there have been described embodiments of a contact lithography alignment system and methods that employ nDSE to one or both of maintain an alignment and provide disturbance compensation during lithography. It should be understood that the above-described embodiments are merely illustrative of some of the many specific embodiments that represent the principles of the present invention. Clearly, numerous other arrangements can be devised without departing from the scope of the present invention as defined by the following claims.
