Patent Application
20250132206
The present disclosure relates to methods including adjusting a position a source die or a destination site to compensate for overlay error between a previously bonded dummy source die and a different destination site.
Advanced packaging technologies demand high throughput and precise placement of dies. Hybrid bonding can be particularly challenging with small misalignment tolerances. An apparatus can include a bridge and a base, where a bonding head, alignment hardware, and an alignment reference can be coupled to the bridge, and a carriage is coupled to the base. A destination substrate chuck and alignment hardware can be coupled to the carriage. During a hybrid bonding sequence, a source die is held by the bonding head, and a destination substrate is held by the destination substrate chuck. The amount of misalignment can be measured for the source die separately from the destination substrate chuck. Some compensation can be made for the misalignment but there is no end-point sensing, and overlay error may be greater than desired. A need exists for better overlay error measurement.
A method can include bonding a first dummy source die to a first destination site of a destination substrate, wherein the first dummy source die has a first dummy source metrology pattern, and the first destination site has a first destination metrology pattern. The method can further include collecting first radiation data regarding at least portions of the first dummy source metrology pattern and the first destination metrology pattern within a first radiation area, wherein collecting the first radiation data is performed after bonding the first dummy source die to the first destination site; analyzing the first radiation data to determine a first overlay error between the first dummy source die and the first destination site; and adjusting a position of a known good source die or a second destination site of the destination substrate to compensate for the first overlay error between the first dummy source die and the first destination site.
In another implementation, the method further includes irradiating the first radiation area with a radiation. Within the first radiation area, the first dummy source metrology pattern includes a relatively transmissive portion with respect to the radiation and a relatively non-transmissive portion with respect to the radiation. The relatively transmissive portion of the first dummy source metrology pattern includes a substrate, and the relatively non-transmissive portion of the first dummy source metrology pattern includes the substrate and dummy source alignment marks, wherein at least one of the dummy source alignment marks is disposed along a bonding surface of the first dummy source die. The dummy source alignment marks are only within the relatively non-transmissive portion, and no dummy source alignment mark of the first dummy source metrology pattern is within the relatively transmissive portion.
In a particular implementation, the first destination metrology pattern comprises destination alignment marks. During irradiating, no more than 25% of the radiation is transmitted through the destination alignment marks.
In a more particular implementation, during bonding, at least one of the dummy source alignment marks does not contact any of the destination alignment marks.
In another implementation, the method further includes irradiating the first radiation area with a radiation. The first radiation area includes at least part of each of the first dummy source metrology pattern and the first destination metrology pattern. Within the first radiation area, the first dummy source metrology pattern includes a relatively transmissive portion with respect to the radiation and a relatively non-transmissive portion with respect to the radiation. With respect to the radiation, a difference in transmission percentages between the relatively transmissive portion and the relatively non-transmissive portion is at least 25%.
In still another implementation, the method further includes irradiating the first radiation area with a radiation. Within the first radiation area, the first dummy source metrology pattern includes a relatively transmissive portion with respect to the radiation and a relatively non-transmissive portion with respect to the radiation. During irradiating, no more than 25% of the radiation is transmitted through the relatively non-transmissive portion, and at least 75% of the radiation is transmitted through all of the relatively transmissive portion.
In yet another implementation, the method further includes irradiating the first radiation area with infrared radiation.
In a further implementation, collecting the first radiation data is performed using a radiation and a radiation detector, wherein the radiation detector has a depth of focus of at most 9 μm with respect to the radiation.
In another implementation, the destination substrate includes a known good destination die and a bad destination die, and the first destination site is at least part of the bad destination die.
In a particular implementation, the method further includes bonding the known good source die to the second destination site of the destination substrate after adjusting the position of the known good source die or the second destination site, wherein the second destination site is at least part of the known good destination die.
In another particular implementation, the method further includes bonding a second dummy source die to a third destination site of the destination substrate, wherein the second dummy source die has a second dummy source metrology pattern, and the third destination site die has a second destination metrology pattern. The method further includes collecting second radiation data regarding at least portions of the second dummy source metrology pattern and the second dummy source metrology pattern within a second radiation area; and analyzing the second radiation data to determine a second overlay error between the second dummy source die and the third destination site. Bonding the second dummy source die, collecting the second radiation data, and analyzing the second radiation data are performed after adjusting the position of the known good source die or the second destination site.
In another implementation, the first dummy source metrology pattern comprises a first dummy source alignment mark.
In a particular implementation, the first destination metrology pattern comprises a first destination alignment mark.
In another particular implementation, the first destination metrology pattern comprises a set of contact pads that is not a part of an alignment mark.
In a further implementation, the first dummy source metrology pattern and the first destination metrology pattern comprise areas including arrays of corresponding contact pads.
In a particular implementation,
the array of contact pads of the first dummy source metrology pattern is phase shifted by an angle in a range from 90° to 270° relative to the array of contact pads of the first destination metrology pattern.
In another particular implementation, the array of contact pads of the first dummy source metrology pattern is phase shifted by an angle in a range of 150° to 210° relative to the array of contact pads of the first destination metrology pattern.
In another implementation, analyzing comprises analyzing the first radiation data using spatial frequency analysis.
In a particular implementation, the first dummy source metrology pattern, the first destination metrology pattern, or each of the first dummy source metrology pattern and the first destination metrology pattern has a pitch and size of a feature corresponding to the pitch. The first dummy source metrology pattern, the first destination metrology pattern, or each of the first dummy source metrology pattern and the first destination metrology pattern has a pitch and a size of a feature corresponding to the pitch corresponds to a value of a spatial frequency. A quotient is the pitch divided by the size of the feature corresponding to the pitch, and the value is the quotient+/−10% of the quotient and is in a range from 1.5 to 9.0.
In a further implementation, analyzing the first radiation data comprises analyzing a pattern of a first image generated from the first radiation data as compared to a pattern of a standard image.
Implementations are illustrated by way of example and are not limited to the accompanying figures.
Skilled artisans appreciate that elements in the figures are illustrated for simplicity and clarity and have not necessarily been drawn to scale. For example, the dimensions of some of the elements in the figures can be exaggerated relative to other elements to help improve understanding of implementations of the invention.
The following description in combination with the figures is provided to assist in understanding the teachings disclosed herein. The following discussion will focus on specific implementations of the teachings. This focus is provided to assist in describing the teachings and should not be interpreted as a limitation on the scope or applicability of the teachings.
Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The materials, methods, and examples are illustrative only and not intended to be limiting. To the extent not described herein, many details regarding specific materials and processing acts are conventional and can be found in textbooks and other sources within the arts.
A dummy source die can be bonded to a first destination site of a destination substrate. The dummy source die and first destination site can include metrology patterns. Radiation data regarding at least portions of the metrology patterns can be collected and analyzed to determine an overlay error between the dummy source die and the first destination site. During a subsequent die bonding, a position of a known good source die or a second destination site of the destination substrate can be adjusted to compensate for the overlay error between the dummy source die and the first destination site.
The dummy source die does not have any electrical circuit element that is electrically connected to an electrical circuit element within the destination substrate. Accordingly, the dummy source die may not have interconnects or other features that could interfere when collecting radiation data after the dummy source die is bonded to the first destination site.
The analysis of radiation data has many options. The analysis may or may not include generating an image. When an image is to be generated, further refinements may or may not be used to improve the image. The imaging may be performed with a range of focus that can focus on the alignment marks within the dummy source die and the destination die and produce an image where electrical circuit elements and interconnects in the destination die are not in focus or are blurred out. Pattern recognition may be used so that electrical circuit elements and interconnects below the contact pads in the destination die are not considered when determining an overlay error. Further, spatial frequency analysis can also be used when imaging or when determining the overlay error.
If needed or desired, the metrology patterns can be designed so that alignment marks within one of the metrology patterns do not overlap alignment marks within the other metrology pattern. Ideally, the alignment marks within one of the metrology patterns is 180° out of phase from the alignment marks within the other metrology pattern. The overlay error may correspond to how far alignment marks within one of the metrology patterns are away from being 180° out of phase from the alignment marks within the other metrology pattern.
More than one dummy source die can be bonded to the destination substrate and provide information regarding drift of the apparatus, whether or not there is center-to-edge variation or other positional relationship regarding overlay error with respect to where on the destination substrate the source die is bonded, or the like.
The methods can be performed by a system that can be in the form of an apparatus or can include a combination of an apparatus and a die loading machine or another apparatus used in hybrid bonding dies. The systems and methods of using the systems are better understood with the description below in conjunction with the corresponding figures.
Referring to
In
The source substrate chuck 122 can be a vacuum chuck, pin-type chuck, a groove-type chuck, an electrostatic chuck, an electromagnetic chuck, or the like. The source substrate chuck 122 can be coupled to the bridge 120 by being attached to the bridge 120 directly or can be coupled to the bridge 120 via a carriage (not illustrated). The source substrate chuck 122 has a source holding surface that faces the base 140 or a component coupled to the base 140.
Each of the bonding heads within the array of bonding heads 124 can include a die chuck and a body disposed between the die chuck and the bridge 120. Each of die chucks can be a vacuum chuck, pin-type chuck, a groove-type chuck, an electrostatic chuck, an electromagnetic chuck, a Bernoulli chuck, or the like. Alternatively, the die chucks can be contactless die chucks, where the die chucks contact the lateral sides, and not the device side or back side, of a die. The device side is a side of the die where electrical components are formed, the back side of the die is opposite the device side, and the lateral sides are disposed between the device and back sides of the die. When a die includes a thru-substrate via (TSV), the TSV may be exposed along the back side of the die. Contactless chucks can help to reduce the likelihood that an activated surface for bonding contacts the bonding head. In an implementation, the device side, the back side, or both the device side and back side can have activated surface(s). The lateral sides may not be activated for bonding.
The bonding heads may be configured such that the die chucks have a limited range of motion relative to their corresponding bodies to provide better positioning when dies are transferred from the array of bonding heads 124 to a destination substrate (not illustrated in
Referring to
The optical component 130 is coupled to the bridge 120 and can be used to determine a pitch of the array of die transfer seats 144. The optical component 130 may also be used to confirm the presence of or identity of a die (for example, a part number or type of die) coupled to a die transfer seat within the array of die transfer seats 144 or a destination substrate coupled to the destination substrate chuck 148. If needed or desired, more than one optical component 130 may be coupled to the bridge 120. The optical component 130 may also be used to determine positions of destination sites of the destination substrate coupled to the destination substrate chuck 148.
The carriage 146 is coupled to the base 140. The carriage 146 can be a positioning stage and provide translating motion along the base 140 in the X-direction, Y-direction, or Z-direction or rotational motion about one or more of axes, such as rotation about a Z-axis and along a plane lying along the X-direction and Y-direction.
The array of die transfer seats 144 are coupled to the carriage 146. The array of die transfer seats 144 have bodies and die chucks. The bodies are coupled to the carriage 146. The bridge 120 or a component coupled to the bridge 120 is closer to the die chucks than to the bodies of die transfer seats within the array of die transfer seats 144. In an implementation, any one or more of the die transfer seats can have a die chuck that can be of any type described with respect to the array of bonding heads 124. The die transfer seats and the bonding heads may be of the same type or different types. In an implementation, the die transfer seats within the array of die transfer seats 144 can be pick-up heads.
The array of die transfer seats 144 can be configured as a vector (a row or a column of pick-up heads), or as a matrix (at least two rows and at least two columns of pick-up heads), or as a staggered array. Regarding the matrix, the number of bonding heads within the array of die transfer seats 144 may be different between rows, between columns, or between rows and columns. Some array configurations can be 3×1, 6×1, 2×2, 2×3, 2×4, 4×2, 10×10, or another rectangular shape, where the first number corresponds to the number of pick-up heads along a row or column, and the second number corresponds to the number of pick-up heads along the other of the row or column.
In theory, dies from an entire source wafer may be transferred all at once. From a top view, for such a configuration, the array of die transfer seats 144 will have fewer pick-up heads along rows closer to the top and bottom of the array as compared to the row or the pair of rows closest to the center of the array, and the array of die transfer seats 144 will have fewer pick-up heads along columns closer to the left-side and right-side of the array as compared to the column or the pair of columns closest to the center of the array. After reading this specification, skilled artisans will be able to determine an array configuration for the array of die transfer seats 144 that meets the needs or desires for a particular application.
The array of die transfer seats 144 can be configured to have an adjustable pitch that can be reversibly changed between a source-matching pitch and the bonding head-matching pitch. The array of die transfer seats 144 or the carriage 146 can include motors, electrical components or the like that can be activated to move die transfer seats to achieve a desired pitch. In an implementation, the array of die transfer seats 144 can be at the source-matching pitch when picking up a set of dies coupled to the source substrate chuck 122 and at the bonding head-matching pitch when transferring the set of dies to the array of bonding heads 124. After the dies are transferred to the array of bonding heads 124, the pitch for the array of die transfer seats 144 can be changed back to the source-matching pitch before picking up more dies.
The terms “transfer operation” and “transfer cycle” are addressed to aid in understanding implementations as described herein. A transfer operation starts no later than loading a first set of dies on the array of die transfer seats 144 and ends with a last set of dies bonded to destination sites of a destination substrate overlying the destination substrate chuck 148. A transfer cycle starts no later than loading a set of dies onto the array of die transfer seats 144 until that same particular set of dies is bonded to the destination sites of the destination substrate that is coupled to the destination substrate chuck 148. A transfer operation can include one or more transfer cycles.
In an implementation, die transfer seats within the array of die transfer seats 144 do not need to be pick-up heads. The die transfer seats within the arrays of die transfer seats 144 may or may not be able to extend in the Z-direction (toward the bridge 120 or a component coupled to the bridge 120). Dies can be loaded onto die transfer seats within the arrays of die transfer seats 144 by the die loading machine 180. Alternatively, the dies can be loaded manually by a human operator.
The optical component 150 in
Referring to
Referring to
The die loading machine 180 can be a robot, a pick-and-place tool, or the like. Referring to
Referring to
Attention is directed to methods of using the system 100.
The method can include mounting a destination substrate onto a destination substrate chuck at block 522 and mounting a source substrate onto a source substrate chuck at block 524 and in
Referring to
Each of the destination dies 820 can be a known good die (KGD) 824 or a bad die 834. A KGD, whether a destination die, a source die, or another die, passed one or more tests and is acceptable for bonding to another die. A bad die failed at least one test, where the test can be for electrical opens, electrical shorts, other functional testing (e.g., a process operation may not have been performed (missing an implant, an insulating layer was not formed, an etch operation was missed, or the like), the process operation may have been inadvertently performed more than one time, or the like), or the die operates too slowly (e.g., operating frequency is too low, read access time is too long, or the like)) for a particular application.
Referring to
Referring to
Each die within the plurality of dies 724 may include a plurality of contact pads (such as contact structures or a conductive interconnect). The plurality of contacts may be arranged on a rectangular, square, oblique, or hexagonal lattice, in which each lattice point in a region of the die includes a contact pad. The bonding sites of the destination substrate 748 will have an identical set of contact pads to match the contact pads on the plurality of dies 724. The arrangement of the contact pads on each die may have a spatial frequency that is characterized by its periodicity in two directions and also the relative phases of the periodic arrangement in two directions. The phases may be relative to a fixed point on the die such as a corner of the die.
Any or all dies within the plurality of dies 724 can include a microprocessor, a microcontroller, a graphic processing unit, a digital signal processor, a memory die (for example, a Level 2 or Level 3 cache, a flash memory, or the like), a field programmable gate array (FPGA), a power transistor die, a power circuit die, a charge coupled-device (CCD), an image sensor, a semiconductor circuit element, or the like. All dies within the plurality of dies 724 can include identical circuit element(s) and, if any circuit is present, identical circuit(s).
The die has a device side, which has most or all of the electrical circuit elements of the die, and a back side opposite the device side. In the implementation as illustrated in
Before continuing with the method, a portion of a dummy source die 924 is described.
alignment marks 944. In an implementation, the alignment marks 944 can be an array of dummy contact pads. Radiation can be used during a later radiation analysis operation. With respect to any particular layer, feature (such as an alignment mark, an interconnect, a conductive plug, etc.) or material, radiation can be transmitted through, absorbed by, or reflected by such layer, feature, or material. As used herein, “non-transmissive” and its variants refer to radiation that is absorbed, reflected, or a combination of absorbed and reflected by such layer, feature, or material. The substrate 932 and the insulating layer 942 are relatively transmissive to radiation as compared to the alignment marks 944, and the alignment marks 944 are relatively non-transmissive with respect to the radiation as compared to the substrate 932 and the insulating layer 942. More details regarding metrology patterns is described in more detail later in this specification.
The substrate 932 can include a semiconductor base material (e.g., Si, SiC, a III-V semiconductor material, a II-VI semiconductor material, or the like) and may or may not include an insulating layer, a patterned polysilicon layer, or the like overlying the semiconductor base material. In an implementation, the semiconductor material can include Si, 4H-SiC, 6H-SiC, AlxGa(1-x)N where 0≤x≤1, or a combination of layers, wherein each layer within combination of layers includes one of the semiconductor materials. The insulating layer 942 can include one or more films of a silicon oxide, a silicon nitride, or a silicon oxynitride. The alignment marks 944 can include a metal or metal alloy, such as Al, Al−1 wt % Cu, Cu, Ag, Au, Pt, Co, W, TiW, Ti, Ta, or the like. More layers may be used for the dummy source die 924.
Zones 954 correspond to the alignment marks 944, and the zones 952 correspond to regions between the alignment marks 944. The zones 952 and 954 extend through, above, and below the alignment marks 944. Any object or portion of the dummy source die 924 that is within a particular zone 954 and outside its corresponding alignment mark 944 is directly above or directly below the corresponding alignment mark 944, regardless whether or not such object or portion thereof contacts or is spaced apart from the corresponding alignment mark 944. For example, referring to
The method can further include moving a carriage to a dummy source die loading position at block 526 and placing a dummy source die on a die transfer seat at block 528 in
Referring to
In an implementation, the die chucks of the array of die transfer seats 144 do not contact the activated surfaces of the dies, such as the dummy source die 924. The dummy source die is drawn in
If the dummy source die 924 die is too thin to be held by its sides, a backing plate (not illustrated) can be coupled to the die. For example, a die may have a thickness less than 50 μm. A thickness of the backing plate or a combined thickness of the backing plate and die is sufficient to allow a pick-up head to pick up the backing plate or a combination of the backing plate and die without having an activated surface of the die contacting the pick-up head. The backing plate can have a thickness in a range from 100 μm to 500 μm.
The backing plate can be coupled to the dummy source die 924 using an adhesive compound. The backing plate may be removed at a later time or remain coupled to the die in the finished electrical device. After the die is bonded to the destination substrate 748, the backing plate may be removed. In an implementation, the adhesive compound may be deactivated by exposure to actinic radiation. The actinic radiation may be in a range from 100 nm to 1000 nm. In such an implementation, at least 70% of the actinic radiation to be transmitted through the backing plate. In another implementation, a solvent can be used to remove the adhesive compound from between the dummy source die and the backing plate.
In a further implementation, the die chucks for the array of die transfer seats 144 can have a design where the die chucks contact the bottom-facing surface of the dummy source die 924. After reading this specification in its entirety, skilled artisans will be able to determine whether the array of die transfer seats 144 should or should not contact the dummy source die 924 for a particular application.
In another implementation, the dummy source die 924 and production source dies within the plurality of dies 724 can be attached to the source substrate 722. The dummy source die 924 can be picked up from the source substrate 722. The method of picking up die from the source substrate 722 is described in more detail with the process flow diagram in
The method can further include transferring the dummy source die from the die transfer seat to a bonding head at block 542 in
The die chuck for the die transfer seat within the array of die transfer seats 144 that is holding the dummy source die 924 can be extended toward a bonding head within the array of bonding heads 124, the die chuck for the bonding head can be extended toward the die transfer seat that holds the dummy source die, or both.
The method can include measuring alignment of the dummy source die using an optical component at block 544 in
The controller 160 or a local controller can use the information to determine an alignment error and an amount of positioning of the dummy source die 924 so that the dummy source die 924 will be more closely aligned to its corresponding destination site of the destination substrate 748. The controller 160 or a local controller can transmit a signal, so that the position of the dummy source die 924 is adjusted by moving the die chuck of the bonding head using the limited range of motion of the die chuck relative to its corresponding body. Thus, in an implementation, moving the die chuck allows the position of the dummy source die 924 to be adjusted relative to the destination substrate 748 that is held by the destination substrate chuck 148. Positioning and measuring alignment error may be performed iteratively until the alignment error is zero or an acceptably low value.
The method can include moving the carriage such that the bonding head is over the destination substrate at block 622 in
Before continuing with the method, an exemplary construction for part of a destination site is described with respect to
The insulating layer 1702 can include a plurality of insulating layers each including an oxide, a nitride, or an oxynitride. The interconnects 1704 can be made of a metal or metal alloy, such as Al, Al−1 wt % Cu, Cu, Ag, Au, Pt, Co, W, TiW, Ti, Ta, or the like. The conductive plugs 1706 can include any of the materials described with respect to the interconnects 1704. The silicon oxide layer 1724 can include a silicon oxide, a silicon nitride, or a silicon oxynitride. The alignment marks 1744 can include any of the materials as described with respect to the interconnects 1704. The alignment marks 1744 and the interconnects 1704 can have the same material or different materials.
Referring to
The method can further include bonding the dummy source die to a corresponding destination site of the destination substrate at block 624 in
The method can include irradiating metrology patterns within a radiation area with radiation at block 642 and collecting radiation data corresponding to the metrology patterns within the radiation area at block 644 in
Radiation emitted by the radiation source 474 can be reflected by some components within metrology patterns within the dummy source die 924 and the destination site of the destination substrate 748. The reflected radiation can be received by the radiation detector 478. In the same or different implementation, the radiation source 474 may be absorbed and cause some components within the dummy source die 924 or the destination site of the destination substrate 748 to fluoresce, phosphoresce, or produce ionized radiation, and the radiation detector 478 can detect the fluorescence, the phosphorescence, or the ionized radiation emitted by such components. Thus, the radiation detector 478 may detect the same or a different form of radiation as emitted by the radiation source 474. For the purposes of this specification, a component within the dummy source die 924 or the destination site of the destination substrate 748 that fluoresces, phosphoresces, or produces ionized radiation in response to being irradiated by the radiation is a type of non-transmissive component.
In another implementation, radiation can be transmitted through the dummy source die 924 and the destination substrate 748 may be detected by a radiation detector within the destination substrate chuck 148, the carriage 146, the base 140, or another suitable portion of the system 100. Similar to the radiation detector 478, the radiation detector for the transmitted radiation may detect the same type of radiation emitted by radiation source 474 or another form of radiation, such as fluorescence, phosphorescence, or ionized radiation.
The radiation can be selected so that the dummy source die 924 includes relatively non-transmissive portions and relatively transmissive portions. As previously described, the alignment marks 944 can include a metal or metal alloy that transmits absorbs, reflects, or absorbs and reflects the radiation more than the insulating layer 942 and the substrate 932, where the substrate 932 can include a semiconductor material, an insulating layer, or a combination thereof. The non-transmissive portions can be the alignment marks 944, and the relatively transmissive portions can include the insulating layer 942 and the substrate 932. In an implementation, the radiation can be infrared radiation (IR) and have a wavelength in a range from 0.7 μm to 1000 μm and, in a particular implementation, from 1 μm to 10 μm or from 1 μm to 2 μm.
The radiation detector 478, the radiation detector within the destination substrate chuck 148, the carriage 146, the base 140, or another suitable portion of the system 100, or another component coupled to the radiation tool 272 can collect at least some radiation data. The radiation data can include the measurements of intensities of radiation detected and locations where such intensities were detected. If the radiation detector 478 does not collect location information, such location information may be obtained from the carriage 146 or another suitable component within the system 100.
Metrology patterns are described with respect to
In any of the foregoing implementations, the metrology pattern 2440 may be an array of contact pads (dummy or active) that lie along the bonding surface of the destination die or may be an array of metal or metal alloy features that are formed at a lower elevation within the destination site. A destination die can include a lowest level of interconnects that may include gate electrodes and word lines, another level that includes bit lines and interconnects that are electrically connected to source regions, drain regions, gate electrodes, or a combination thereof, still another level interconnects, a further level of interconnects, and finally a level including contact pads. Due to electromagnetic radiation, parasitic coupling, or the like, the alignment marks 1744 within the destination site may be formed with a lower level of interconnects as opposed to the contact pad level. Thus, the metrology pattern 2440 may be spaced apart from and not lie along the bonding surface.
Referring to
The difference in transmission percentages between the relatively transmissive portions and the relatively non-transmissive portions is at least 25%. In the same or different implementation, (1) no more than 25% of the radiation is transmitted through the relatively non-transmissive portions, (2) at least 75% of the radiation is transmitted through all of the relatively transmissive portions, or (3) no more than 25% of the radiation is transmitted through the relatively non-transmissive portions and at least 75% of the radiation is transmitted through all of the relatively transmissive portions.
The metrology patterns 2340 and 2440 are designed such that the overlay error is zero when the alignment marks (such as the arrays of dummy contact pads for the particular example) within the metrology patterns 2340 and 2440 are phase shifted by an angle of 180°. In an implementation, the phase shift can be in a range from 90° to 270°, in a range from 150° to 210°, or another suitable range that can be useful for determining overlay error. The alignment marks 944 do not overlie any of the alignment marks 1744. The phase shifting can be in one direction or in different directions. Referring to
The method can include analyzing the radiation data to determine an overlay error between the dummy source die and the corresponding destination site at block 646 in
An image can be generated from the radiation data. Imaging can be performed using radiation data to generate an image corresponding to the radiation area 2540. The radiation area may include all of the metrology pattern 2340, the metrology pattern 2440, or both or may include a portion, and not all, of either or both of the metrology patterns 2340 and 2440. In
Imaging of the radiation area 2540 can be aided in view of some considerations. Imaging may be performed with the radiation detector having a depth of focus that can focus on the alignment marks within the metrology patterns 2340 and 2440. The depth of focus may be at most 9 μm, at most 7 μm, or at most 5 μm. In a particular implementation, alignment marks 944 and 1744 can be arrays of dummy contact pads that lie at the bonding surfaces of the dummy source die 924 and the corresponding destination site. The thickness of the dummy contact pads may be in a range from 1 μm to 4 μm. The limit on the depth of focus can reduce the likelihood of other metal or metal alloy components lying at lower elevations within the destination site, such as the interconnects 1704 and conductive plugs 1706, from being in focus. Such components may not be seen or may be blurred. The depth of focus may be adjusted to the particular thicknesses used for the alignment marks 944 and 1744. Thus, the depth of focus may be significantly less than 9 μm or could be greater than 9 μm when no components adjacent to the metrology pattern 2440 would interfere with imaging. In the same or different implementation, the other metal or metal alloy components lying at lower elevations within the destination site may be blurred.
After an image has been generated, the controller 160, a local controller, or a computer external to the system 100 can determine an X-direction offset, a Y-direction offset, a rotational offset, or a combination thereof between the metrology patterns 2340 and 2440 within the radiation area 2540. Alternatively, a human operator may direct the controller 160, a local controller, or a computer external to the system 100 to use the image and measure the X-direction offset, the Y-direction offset, the rotational offset, or a combination thereof between the metrology patterns 2340 and 2440 within the radiation area 2540.
Analysis of the radiation data can be performed without generating an image. The radiation data may be collected by scanning many rows in the X-direction or many columns in the Y-direction of the radiation area 2540. The alignment marks 944 and 1744 absorb, reflect, or a combination of absorb and reflect substantially more radiation as compared to locations between the alignment marks 944 and 1744, and substantially more radiation can be transmitted between the alignment marks 944 and 1744 as compared to through the alignment marks 944 and 1744.
Referring to
The controller 160, a local controller, or a computer external to the system 100 can use the difference in signal amplitudes and location information to determine the sizes and locations of the alignment marks 944 and 1744 and the pitch corresponding to the metrology patterns 2340 and 2440 within the radiation area 2540 in
The Y-direction dimension may be obtained by identifying to which row the scan belongs. A scan may refer to one or more lines of an image obtained by a CCD or one or more lines obtained by scanning using an optical component. As a further example, Scan 0 may be along the upper border (closest to top of
Referring to
Spatial frequency analysis may be performed on the metrology patterns 2340 and 2440 for the radiation data collected from the radiation area 2540. The spatial frequency analysis can be performed in the spatial frequency domain with or without an image of the radiation area 2540. The spatial frequency analysis may be performed using a value that is based on a detected pitch and feature size. The spatial frequency may be set to exclude features, such as word lines, bit lines, or other interconnects, and isolated features. In view of the foregoing, the pitch divided by the feature size is a quotient and the value for spatial frequency analysis can be the quotient+/−10% to account for manufacturing or other variances. In an alternative implementation, optical techniques may be used to blur out features that are at a plane that is not of interest. The value may be in a range from 1.5 to 9.0.
The word lines or the bit lines may not have a pitch in one of the directions, or, when present, segmented word lines or segmented bit lines can have relatively large lengths in a direction and have relatively small spaces between the segments in that same direction. Thus, for a segmented bit line or a segmented word line, the quotient (the pitch divided by the feature size) will be less than 1.1.
Isolated features may not have a pitch or have a relatively small feature size and a relatively large spaces between the isolated features. If the isolated features have a pitch, the pitch divided by the feature size may exceed 10, 50, or more. Alternatively, the isolated features may vary greatly in size. An X-direction dimension, a Y-direction dimension, or both may be as little as 1.0 μm or as large as 10 μm, 50 μm, or more. The spatial frequency analysis may include a lower limit, an upper limit, or both for the X-direction dimension, the Y-direction dimension, or both. A non-limiting example, each alignment marks 944 and 1744 may be designed to have an X-direction dimension of 1.5 μm and a Y-direction dimension of 1.5 μm. In practice, the X-direction dimension and Y-direction dimension may vary by up to 0.2 μm due to manufacturing or other variations. Thus, the features will have X-direction dimensions and Y-direction dimensions that are in a range from 1.3 μm to 1.7 μm.
The spatial analysis may be performed only on features that meet one or more of (1) the value associated with the quotient (pitch divided by feature size), (2) feature size (X-direction dimension, Y-direction dimension, or both)), another suitable criterion, or a combination thereof.
Analysis of the radiation data can include comparing an image corresponding the radiation area 2540 to a standard image corresponding to the radiation area 2540. The standard image may be an image from a different bonded dummy source die bonded to a different destination site of a different destination substrate, where the bonded different die and different destination site has no or a relatively low overlay error. In the same or another implementation, the standard image may be a computer-generated image corresponding to zero overlay error. The computer-generated image may be based on theoretical data and not include actual radiation data from bonded dies.
Pattern recognition can be useful for both reflective and transmission radiation modes and can be particularly useful when radiation is transmitted through the destination substrate 748. The metrology pattern 2440 may have other features that are close to or underlie the metrology pattern 2440 of the destination site, such as the interconnects 1204 and conductive plugs 1206. Pattern recognition can determine that the other features are not part of the metrology pattern 2440, and such other features may be excluded from the analysis when determining overlay error.
As previously described, different techniques can be performed when analyzing the radiation data or a derivative thereof (for example, an image generated from the radiation data). The analysis can be performed using any particular technique or a combination of the techniques may be used. The analysis can continue further to determine an overlay error between the dummy source die 924 and its corresponding destination site.
As illustrated in
Referring to
The metrology patterns 2340 and 2440 are designed such that the alignment marks 944 and 1744 are 180° out of phase of each other in each of the X-direction and the Y-direction when there is no overlay error. Accordingly, when the alignment marks 944 and 1744 are 180° out of phase compared to each other, along a row of the alignment marks 944 (X-direction), the centers of the alignment marks 944 are 1.5 μm from each of lines defined centers of the alignment marks 1744 in immediately adjacent columns of the alignment marks 1744. Along a column of the alignment marks 944 (Y-direction), the centers of the alignment marks 944 are 2.0 μm from each of lines defined centers of the alignment marks 1744 in immediately adjacent rows of the alignment marks 1744.
In the particular example, the alignment mark 2644 is 1.2 μm from the line defined by centers of the alignment marks 1744 in the column just to the left of the alignment mark 2644, and 1.8 μm from the line defined by centers of the alignment marks 1744 in the column just to the right of the alignment mark 2644. For zero overlay error, both distances would be 1.5 μm. The alignment mark 2644 is 0.3 μm closer to the left-hand side of
Continuing with the same example, the alignment marks 2644 is 2.6 μm from the line defined by centers of the alignment marks 1744 in the row just above the alignment mark 2644, and 1.4 μm from the line defined by centers of the alignment marks 1744 in the row just below the particular alignment mark 944. For zero overlay error, both distances would be 2.0 μm. The alignment mark 2644 is 0.6 μm closer to the bottom of
The overlay error may or may not have a rotational component. For example, the line defined by the centers of a row or column of the alignment marks 944 should be parallel to lines defined by the centers of immediately adjacent rows or columns of the alignment marks 1744. When rotational overlay error is present, a line corresponding to the alignment marks 944 intersects a line corresponding to the alignment marks 1744. The intersection of the lines may occur within the radiation area 2540 or outside the radiation area 2540. For this particular example, there is no rotational overlay error.
The method can include determining an overlay error compensation based at least in part on the overlay error at block 648 in
When positioning the next source die or its corresponding destination site of the destination substrate 748, the next source die should be 0.3 μm closer to the right-hand side of
The method can continue with other dies to be bonded to other destination sites of the destination substrate 148. This part of the method corresponds to the process flow diagram in
The method includes performing registration and metrology with respect to a plurality of dies on the source substrate and a plurality of die transfer seats at block 2722 in
The method can further include changing a pitch of the array of die transfer seats to a source-matching pitch at block 2724 in
The method can include loading a set of dies onto the array of die transfer seats at block 2726 in
In an implementation, the die chucks of the array of die transfer seats 144 do not contact the activated surfaces of the set of dies 3024. The die chucks of the array of die transfer seats 144 may be Bernoulli chucks. Although the set of dies 3024 are held by the die chucks, the set of dies 3024 are drawn in
If a die is too thin to be held by its sides, a backing plate can be coupled to the die. For example, a die may have a thickness of less than 50 μm. A thickness of the backing plate or a combined thickness of the backing plate and die is sufficient to allow a pick-up head to pick up the backing plate or a combination of the backing plate and die without having an activated surface of the die contacting the pick-up head. The backing plate can have a thickness in a range from 100 μm to 500 μm.
The backing plate can be coupled to the die using an adhesive compound. The backing plate may be removed at a later time or remain coupled to the die in the finished electrical device. After the die is bonded to the destination substrate 748, the backing plate may be removed. In an implementation, the adhesive compound may be deactivated by exposure to actinic radiation. The actinic radiation may be in a range from 100 nm to 1000 nm. In such an implementation, at least 70% of the actinic radiation is to be transmitted through the backing plate. In another implementation, a solvent can be used to remove the adhesive compound from between the die and the backing plate.
In another implementation, a die may not have an activated surface but has a relatively fragile component along a surface that will be bonded to the destination substrate 748, and such surface should not contact a die transfer seat within the array of die transfer seats 144. A die transfer seat as described with respect to the die having the activated surface can be used for the die with a fragile component along the surface facing the die transfer seat.
In a further implementation, the die chucks for the array of die transfer seats 144 can have a design where the die chucks contact the bottom-facing surfaces of the set of dies 3024. After reading this specification in its entirety, skilled artisans will be able to determine whether the array of die transfer seats 144 should or should not contact device sides or back sides of the dies and determine a design that meets the needs or desires for a particular application.
The method can further include changing a pitch for the array of die transfer seats to a bonding head-matching pitch at block 2742 in
The method can include moving the carriage to an alignment position at block 2744 in
The method can further include moving the carriage 146 so that the set of dies 3024 can be transferred from the array of die transfer seats 144 to the array of bonding heads 124. In an implementation as illustrated in
The method can further include transferring the set of dies from the array of die transfer seats to the array of bonding heads at block 2822 in
The method can include measuring alignment of the set of dies using an optical component at block 2824 in
The method can also include adjusting a position of the set of dies or the destination site to compensate for overlay error at block 2842 in
With respect to positioning of dies, the overlay error compensation will be in the direction opposite of the overlay error. Regarding the particular example described with respect to
In an alternative implementation, the destination substrate 748 may be positioned instead of positioning the set of dies 3024. With respect to positioning of the destination substrate 748, the overlay error compensation will be in the same direction as the overlay error. When positioning the destination substrate 748, the X-direction overlay error compensation is −0.3 μm, and the Y-direction overlay error compensation is −0.6 μm. Before bonding, the position of the destination substrate 748 can be adjusted by moving the set of dies 3024 by −0.3 μm in the X-direction and −0.6 μm in the Y-direction. Referring briefly to
The overlay error may include a rotational component. In the example, the overlay does not include a rotational component. However, in another implementation, the overlay error may include a rotational component having an angle θ in a counterclockwise direction (−θ). When positioning the set of dies 3024, the overlay error compensation can rotate the set of dies 3024 by an angle θ in a clockwise direction (+θ). When positioning the destination substrate 748, the overlay error compensation can rotate the destination substrate 748 by an angle θ in a counterclockwise direction (−θ).
The method can be used to compensate for X-direction overlay error, Y-direction overlay error, rotational overlay error, or a combination thereof. Compensating for overlay error allows for a more accurate bonding of active contact pads between source die and destination sites of the destination substrate. Compensating for the overlay error may be performed before or after adjusting the position of the set of dies 3024 regarding misalignment (block 2824 in
The method can include moving the carriage such that the array of bonding heads are over the destination substrate at block 2844 in
The method can further include bonding the set of dies to the corresponding destination sites of the destination substrate at block 2846 in
A determination is made whether another set of dies is to be transferred to the destination substrate at decision diamond 2848 in
If needed or desired, another dummy source die may be bonded to the destination substrate 748 and used to determine another overlay error. The method can be any of the methods as previously described with respect to
The overlay errors for the dummy source dies 924 and 3824 may be different. The information provided by the overlay error regarding the dummy source die 3824 may depend on when the dummy source die 3824 was bonded as compared to bonding sets of source KGDs.
The bonding of the dummy source die 3824 may be performed before bonding the set of dies 3024 to destination sites the KGDs 824. Radiation data regarding the dummy source die 3824 and its corresponding destination site collected before bonding any source KGD can be helpful to confirm that the overlay error compensation regarding the dummy source die 924 was sufficient or if the overlay error compensation should be changed to reduce the amount of undercompensation or overcompensation associated with the overlay error compensation regarding the dummy source die 924.
The bonding of the dummy source die 3824 may be performed (1) after bonding the set of dies 3024 to destination sites KGDs 824 and before bonding another set of source dies to other destination sites of the destination substrate 748 (hereinafter “intermediate bonding”) or (2) after all destination sites of KGDs 824 have been bonded with source KGDs from the plurality of dies 724 (hereinafter “last bonding”). Radiation data regarding the dummy source die 3824 and its corresponding destination site collected for intermediate bonding or last bonding can be helpful to whether the overlay error changed over time since the time of bonding the dummy source die 924. The change may be due to vibration or other movement of any one or combination of components within the apparatus 110, temperature change within the apparatus 110, or another environmental condition change within the apparatus 110.
The difference in overlay errors with the dummy source dies 924 and 3824 may provide other information in addition or in place of the temporal information. Referring to
Continuing with the particular example that was previously described, the dummy source die 924 has an X-direction overlay error of −0.3 μm and a Y-direction overlay error of −0.6 μm, where X-direction positive values are closer to the right-hand side of
The distance from the center of the destination site of the destination die 3924 to the center of the destination substrate 748 may have a value that is halfway between the values for the distances between the centers of destination sites of the bad dies 834 that are bonded to the dummy source dies 924 and 3824. Thus, the overlay error compensations for the source KGD for each direction may be
OECKGD=−((OE924+OE3824)/2), where:
When the source KGD for the destination die 3924 is moved, the overlay error compensation is in a opposite direction as compared to the overlay errors, and thus, a minus precedes the overlay errors for the dummy source dies 924 and 3824.
Using the values above, the overlay error compensation for the source KGD when positioning it relative to the destination site for the destination die 3924 is:
OECKGD=−((−0.3 μm−0.1 μm)/2), or +0.1 μm, and
OECKGD=−((−0.6 μm−0.2 μm)/2), or +0.2 μm.
At least part of the positioning can include moving the source KGD 0.1 μm closer to the right-hand side of
When the destination substrate 748 is moved for the source KGD, the movement will be in the same direction as compared to the overlay error. The overlay error compensation when moving the destination substrate 748 is −0.1 μm in the X-direction (from a top view of the destination substrate 748, move the destination substrate 748 0.1 μm closer to the left-hand side) and −0.2 μm in the Y-direction (from a top view of the destination substrate 748, move the destination substrate 748 0.2 μm closer to the bottom).
The prior example with specific numbers is provided to illustrate and improve understanding how overlay errors can be used to determine overlay error compensations and not to limit the scope of the present invention as defined in the claims.
After all desired source dies (dummy source dies and source KGDs) are bonded to the destination substrate 748, a hybrid bonding process can continue. The hybrid bonding process can include three steps that include a bonding operation, a first anneal to cause the metal within the dies and at the destination sites to expand and contact each other, and a second anneal to cause metal atoms to cross the metal-metal interface and reduce contact resistance. The method previously described can correspond to the bonding operation of a hybrid bonding process. The destination substrate 748 can be removed from the apparatus 110 or moved to a different portion of the apparatus 110 or a different tool to perform the anneal operations.
The methods as described with respect to the process flow diagrams in
Many of the implementations described herein can help with determining an overlay error with respect to a die and a destination site of a destination substrate. A dummy source die can be bonded to a destination site of a destination substrate. The dummy source die does not have any electrical circuit element that is electrically connected to an electrical circuit element within the destination substrate. Accordingly, the dummy source die may not have interconnects or other features that interfere when collecting radiation data after dummy source die is bonded to a destination site. In an implementation, alignment marks within the metrology patterns of the dummy source die may reflect substantially more of the radiation as compared to other parts of the metrology pattern or may transmit substantially less of the radiation as compared to other parts of the metrology pattern.
The radiation can be collected and analyzed. The analysis may or may not include generating an image. When an image is to be generated, further refinements may or may not be used to improve the image. The imaging may be performed with a range of focus that can focus on the alignment marks within the dummy source die and the destination die and produce an image where electrical circuit elements and interconnects in the destination die are not in focus or are blurred out. Pattern recognition may be used so that electrical circuit elements and interconnects below the contact pads in the destination die are not considered when determining an overlay error. Furthermore, spatial frequency analysis can also be used when imaging or when determining the overlay error.
The metrology patterns for the dummy source die and its corresponding destination site can be significantly out of phase to produce a more accurate determination of overlay error. Referring to
The overlay error obtained for the dummy source die and its corresponding destination site may include an X-direction overlay error, a Y-direction overlay error, a rotational overlay error, or any combination thereof. An overlay error compensation can be used for subsequent source dies, such as source KGDs, that will be bonded to the destination substrate, wherein the overlay error compensation can help to reduce the overlay error for the subsequent dies. The overlay error compensation can allow for a greater overlap between the active contact pads of source KGDs to the active contact pads at the destination sites of destination KGDs of the destination substrate. The increase in overlap between active contact pads helps to reduce contact resistance.
More than one dummy source die can be bonded to the destination substrate and provide information regarding drift of the apparatus 110, determine if the overlay error depends on the position where source dies are bonded to destination site of the destination substrate (for example, center-to-edge variation with respect to the destination substrate), or the like.
Note that not all of the activities described above in the general description or the examples are required, that a portion of a specific activity may not be required, and that one or more further activities can be performed in addition to those described. Still further, the order in which activities are listed is not necessarily the order in which they are performed.
Benefits, other advantages, and solutions to problems have been described above with regard to specific implementations. However, the benefits, advantages, solutions to problems, and any feature(s) that can cause any benefit, advantage, or solution to occur or become more pronounced are not to be construed as a critical, required, or essential feature of any or all the claims.
The specification and illustrations of the implementations described herein are intended to provide a general understanding of the structure of the various implementations. The specification and illustrations are not intended to serve as an exhaustive and comprehensive description of all of the elements and features of apparatus and systems that use the structures or methods described herein. Separate implementations can also be provided in combination in a single implementation, and conversely, various features that are, for brevity, described in the context of a single implementation, can also be provided separately or in any subcombination. Further, reference to values stated in ranges includes each and every value within that range. Many other implementations can be apparent to skilled artisans only after reading this specification. Other implementations can be used and derived from the disclosure, such that a structural substitution, logical substitution, or another change can be made without departing from the scope of the disclosure. Accordingly, the disclosure is to be regarded as illustrative rather than restrictive.
This application is related to U.S. patent application No. ______ entitled “Method Including Positioning a Source Die or a Destination Site to Compensate for Overlay Error” by Choi filed of even date (Attorney Docket No. CNT C1159-US), which is assigned to the current assignee hereof and incorporated herein by reference in its entirety.
