JP6473116B2 - Alignment apparatus and alignment method - Google Patents

Description

  The present invention relates to an alignment apparatus and an alignment method for aligning two objects.

  There are techniques for aligning two objects. For example, Patent Document 1 describes a mounting device (joining device) that joins a component held on a head and a substrate held on a stage after aligning and mounting the component on the substrate. Yes.

  In such an apparatus, first, the positions of two objects to be joined (specifically, horizontal positions) are detected in a non-contact state. And based on the detection result of the horizontal position of both to-be-joined objects, alignment (alignment) is performed so that the relative position shift in the horizontal direction of the to-be-joined objects may be eliminated. The two objects to be joined are joined by approaching and further contacting.

  However, the above-described technique has a problem that misalignment occurs due to various factors when both non-contacted objects are in contact with each other. For example, when the objects to be bonded come into contact with each other, the horizontal position of the objects to be bonded is slightly changed due to warpage of the wafer, inclination between the head and the stage, and / or physical impact force. However, it may shift.

  In recent years, miniaturization of processing technology has progressed, and a situation in which such positional deviation is not allowed is also emerging.

  On the other hand, Patent Document 2 proposes the following technique. Specifically, the alignment of both the objects to be bonded (also referred to as the first alignment) was performed in a state where the objects to be bonded of the first object to be bonded and the second object to be bonded face each other in a non-contact manner. After that, both the objects to be bonded are brought into contact with each other, and a positional shift (also referred to as a second positional shift) (for example, a rightward shift) of both the bonded objects in the contact state is measured. Then, after the contact state of both the objects to be bonded is once released, both objects to be bonded are moved so as to eliminate the second positional shift, and the objects to be bonded are brought into contact again.

  More specifically, when the contact state of both the objects to be bonded is once released, the positional deviation caused by the contact is naturally canceled by the contact release operation, and ideally the first The state returns to the state immediately after the alignment (for example, the state in which the shift amount is zero). Then, immediately after both the objects to be joined are relatively moved so as to eliminate the second misalignment from this state, for example, the alignment marks do not coincide with each other, rather than the second misalignment. Transition is made to a state in which a deviation in direction (for example, a deviation in the left direction) occurs. Then, when both the objects to be joined come into contact again thereafter, a deviation (similar to the second positional deviation) (for example, a rightward deviation) due to the re-contact occurs again. In a state where a shift opposite to the second position shift (for example, a leftward shift) is generated, a shift similar to the second position shift (for example, a right shift) occurs, resulting in a re-contact state. It is possible to make the positional deviation of both objects to be close to zero (e.g., to match both alignment marks (ideally) exactly).

  According to such a technique (described in detail later), it is possible to correct the positional deviation caused by the contact better.

JP 2008-85322 A JP 2011-66287 A

  However, the inventor of the present application may not always sufficiently reduce the deviation that occurs when both objects (both objects to be joined) contact each other due to various circumstances. Was elucidated.

  Examples of the circumstances include the following.

  For example, when the contact between both objects to be bonded is temporarily released (in other words, when both objects to be peeled are peeled off), a holding member (adsorption member) that holds the objects to be bonded by an impact associated with the contact release There exists a situation that the said to-be-joined object may mutually shift | deviate and the positional relationship between a holding member and a to-be-joined object may shift | deviate. Other circumstances will be described later.

  Accordingly, an object of the present invention is to provide an alignment technique capable of aligning two objects with higher accuracy.

  In order to solve the above problem, the invention of claim 1 is an alignment method for aligning both objects of a first object and a second object, and a) the first object and In a state where the second object is opposed to the second object in a non-contact manner, the first object displacement, which is a position displacement between the objects and in a plane perpendicular to a predetermined direction, is measured. Correcting the positional deviation of 1 and aligning the two objects; b) after the step a), the first object and the second object are relative to each other in the predetermined direction. Moving the first object and the second object into contact with each other, and c) the positions of the two objects in the contact state between the first object and the second object. Misalignment and misalignment in a plane perpendicular to the predetermined direction. And d) after the second misalignment measurement operation, the two objects are relatively moved in the predetermined direction to temporarily change the contact state between the objects. A step of canceling, e) a step of estimating a contact-induced shift that is a positional shift caused by the contact between the two objects based on the second positional shift, and f) a contact cancellation of the two objects. Measuring a third positional shift which is a relative positional relationship between the two objects in a state and is a relative positional relationship in a plane perpendicular to the predetermined direction; and g) a reverse direction of the contact-induced shift. Is set as a target state, and in the contact release state of both objects, the two objects are relatively moved in a plane perpendicular to the predetermined direction, and the target state is set. Said to achieve And adjusting the positional deviation of 3, h) after said step g), characterized by comprising the steps of: performing a re-contacting operation of contacting the two object again.

  According to a second aspect of the present invention, in the alignment method according to the first aspect of the present invention, the step f) includes: f-1) a first mark attached to the first object and the second object. Reading the attached second mark, and f-2) based on the positional deviation between the two marks measured using both the first mark and the second mark, Measuring the third misregistration.

  According to a third aspect of the present invention, in the alignment method according to the second aspect of the present invention, in the step f-1), the first mark attached to the first object and the second object are attached to the first object. A captured image including both of the second mark and the attached second mark is acquired, and in step f-2), a positional deviation between the marks is measured based on the captured image. To do.

  According to a fourth aspect of the present invention, in the alignment method according to the second aspect of the present invention, the step f-1) includes: f-1-1) a step of reading the first mark attached to the first object. And f-1-2) reading the second mark attached to the second object at a timing different from the step f-1-1).

  According to a fifth aspect of the present invention, in the alignment method according to the first aspect of the invention, the step f) includes: f-1) a first mark attached to the first object and the second object. A step of reading only one of the two marks with the attached second mark, and f-2) a position measured before step f) with respect to the other of the two marks. Obtaining a relative positional relationship between the two marks based on the position measured using the one mark read in step f-1).

  According to a sixth aspect of the present invention, in the alignment method according to any one of the first to fifth aspects of the present invention, in the step a), the first and second misalignments are kept within a first permissible range. The alignment of the object is performed, and the residual deviation within the first allowable range is obtained. In step e), the residual deviation which is the deviation remaining in step b) is also based on the residual deviation. The contact-induced deviation is estimated.

  According to a seventh aspect of the present invention, in the alignment method according to any one of the first to sixth aspects, the step c) is executed again after the step h), and the second position is determined in the step c). Steps d) to h) and step c) are repeatedly executed in this order until it is determined that the deviation falls within the second allowable range.

  According to an eighth aspect of the present invention, in the alignment method according to the seventh aspect of the invention, in the step e), the non-detection of the two objects finally detected in the step g) immediately before the step e) is performed. The contact-induced deviation is estimated based on the positional deviation in the contact state and the second positional deviation obtained in step c) immediately before step e).

  According to a ninth aspect of the present invention, in the alignment method according to any one of the first to eighth aspects of the invention, in the step e), the contact is based on a plurality of past estimation results relating to the contact-induced deviation. The cause deviation is estimated.

  According to a tenth aspect of the present invention, in the alignment method according to any one of the first to ninth aspects, the third step in step f) until the deviation from the target state falls within a third allowable range. The positional deviation measuring operation and the relative movement operation of the two objects in step g) are repeatedly performed.

  The invention of claim 11 is the alignment method according to any one of claims 1 to 10, wherein each of the first and second objects is a substrate, and at least one of the steps b). The contact between the two objects is started in a state where the objects are bent and the two objects are partially in contact with each other.

  According to a twelfth aspect of the present invention, in the alignment method according to the eleventh aspect of the invention, in the step f), the contact between the two objects is released, and the deflection of the at least one object is returned. The third positional deviation is measured in a state where the at least one object is flat.

  The invention of claim 13 is an alignment apparatus for aligning both objects of a first object and a second object, wherein the first object and the second object are non-aligned. In a state of being opposed by contact, a measuring means for measuring a first positional deviation, which is a positional deviation between the two objects in a plane perpendicular to a predetermined direction, and correcting the first positional deviation. An alignment means for aligning the two objects, and after the correction of the first displacement, the first object and the second object are relatively moved in the predetermined direction. Relative movement means for bringing the first object and the second object into contact with each other, and the measuring means is configured to move the first object and the second object by a relative movement operation by the relative movement means. In the state where the object of the A second positional deviation that is a positional deviation of the object in a plane perpendicular to the predetermined direction is measured, and the relative movement means is configured to measure the two objects after the second positional deviation measuring operation. The object is relatively moved in the predetermined direction to temporarily release the contact state of the two objects, and the alignment means detects a contact-induced deviation that is a positional deviation caused by the contact of the two objects. 2 based on the positional deviation of the two objects, and the measuring means is a relative positional relation between the two objects in a released state, and a relative positional relation in a plane perpendicular to the predetermined direction. And the alignment means sets a state in which a position shift opposite to the contact-induced shift occurs as a target state, and in the contact release state of both objects, the predetermined method The relative displacement means adjusts the third positional shift so as to achieve the target state by relatively moving the objects in a plane perpendicular to the object, and the relative movement means is configured to adjust the contact between the objects. After the state is once released and the both objects are relatively moved, a re-contact operation is performed in which both the objects are brought into contact again.

  According to a fourteenth aspect of the present invention, in the alignment apparatus according to the thirteenth aspect of the invention, when the measuring means measures the third positional deviation, the first mark attached to the first object and the first mark A position between the two marks measured by using both the first mark and the second mark while reading both the marks with the second mark attached to the second object The third positional deviation is measured based on the deviation.

  According to a fifteenth aspect of the present invention, in the alignment apparatus according to the fourteenth aspect of the present invention, the measurement means measures the first mark attached to the first object when measuring the third positional deviation. A captured image including both of the second mark and the second mark attached to the second object is acquired, and a positional deviation between the marks is measured based on the captured image.

  According to a sixteenth aspect of the present invention, in the alignment apparatus according to the fourteenth aspect of the invention, the measuring means measures the first mark attached to the first object when measuring the third positional deviation. In addition to reading, the second mark attached to the second object is read at a timing different from the reading timing of the first mark.

  According to a seventeenth aspect of the present invention, in the alignment apparatus according to the thirteenth aspect of the invention, the measuring means measures the first misalignment and measures the first mark attached to the first object and the first mark. Only one of the two marks with the second mark attached to the second object is read, and the position measured using the read one mark and the other of the two marks The relative positional relationship between the two marks is acquired based on the position measured before the third misalignment measurement time for the mark.

  According to an eighteenth aspect of the present invention, in the alignment apparatus according to any one of the thirteenth to seventeenth aspects of the invention, the alignment means is configured so that the both objects are in the first position until the first displacement is within a first allowable range. The measurement means obtains a residual deviation within the first allowable range still remaining after the alignment, and the alignment means determines the contact-induced deviation based on the residual deviation. Is estimated.

  According to a nineteenth aspect of the present invention, in the alignment apparatus according to any one of the thirteenth to eighteenth aspects of the present invention, the contact state between the two objects is maintained until the second positional deviation falls within a second allowable range. A contact releasing operation for once releasing, an operation for estimating the contact-induced deviation, a measuring operation for the third positional deviation, an adjusting operation for adjusting the third positional deviation to achieve the target state, The re-contact operation for bringing both objects into contact again and the second position shift measurement operation are repeatedly performed.

  According to a twentieth aspect of the present invention, in the alignment apparatus according to the nineteenth aspect of the invention, the alignment means is the two objects finally detected in the third misalignment adjusting operation immediately before the recontact operation. Estimating the contact-induced deviation based on a positional deviation in a non-contact state of the object and the second positional deviation obtained by the second positional deviation measurement operation immediately after the re-contact operation. It is characterized by.

  According to a twenty-first aspect of the invention, in the alignment apparatus according to any of the thirteenth to twentieth aspects of the invention, the alignment means is configured to perform the contact-induced deviation based on a plurality of past estimation results related to the contact-induced deviation. Is estimated.

  According to a twenty-second aspect of the present invention, in the alignment apparatus according to any of the thirteenth to twenty-first aspects, the third positional deviation is measured until the deviation from the target state falls within a third allowable range. The operation and the adjustment operation for adjusting the third misalignment so as to achieve the target state are repeatedly performed.

  According to a twenty-third aspect of the present invention, in the alignment apparatus according to any of the thirteenth to twenty-second aspects, the first and second objects are substrates, respectively, and the relative movement means is at least one of the substrates. The object is bent, and the contact between the two objects is started from a state where the objects are partially in contact with each other.

  According to a twenty-fourth aspect of the present invention, in the alignment apparatus according to the twenty-third aspect of the invention, the measuring means is in a state in which the contact between the two objects is released, and the bending of the at least one object is returned and the at least one The third positional deviation is measured in a state where one object is flat.

  The invention of claim 25 is an alignment method for aligning both objects of a first object and a second object, and a) the first object and the second object In a state where the two are opposed to each other in a non-contact state, the first positional deviation, which is a positional deviation between the two objects and in a plane perpendicular to a predetermined direction, is measured, and the first positional deviation is corrected. And b) after the step a), the first object and the second object are relatively moved in the predetermined direction, and the both objects Forming a high-pressure contact state where the objects are in contact with each other and a pressure greater than a certain value is applied between the objects; c) the first object and the second object In a high-pressure contact state with Measuring a second positional deviation which is a positional deviation in a plane perpendicular to the predetermined direction; and d) after the second positional deviation measuring operation, the two objects are relatively moved in the predetermined direction. The two objects are moved from the high-pressure contact state to a low-pressure contact state in which a pressure smaller than the predetermined value is applied between the two objects and the two objects are in contact with each other. E) a step of estimating a high-pressure contact-induced shift, which is a positional shift caused by the formation of a high-pressure contact state regarding both objects, based on the second positional shift; and f) the above Measuring a third positional shift that is a relative positional relationship between the two objects in a low-pressure contact state between the two objects and in a plane perpendicular to the predetermined direction; g) High pressure A state in which a position displacement opposite to the touch-induced displacement occurs is set as a target state, and both the objects are relatively moved in a plane perpendicular to the predetermined direction in a low-pressure contact state of the both objects. Then, after the step of adjusting the third misalignment so as to achieve the target state, and h) after the step g), a re-pressurizing operation is performed to change both the objects to the high-pressure contact state again. And a step.

  The invention of claim 26 is an alignment apparatus for aligning both objects of a first object and a second object, wherein the first object and the second object are non-aligned. In a state of being opposed by contact, a measuring means for measuring a first positional deviation, which is a positional deviation between the two objects in a plane perpendicular to a predetermined direction, and correcting the first positional deviation. An alignment means for aligning the two objects, and after the correction of the first displacement, the first object and the second object are relatively moved in the predetermined direction, A relative moving means that forms a high-pressure contact state in which the two objects are in contact with each other and a pressure of a certain value or more is applied between the two objects, and the measuring means includes the relative The relative movement by the moving means causes the first movement. In a state where the high-pressure contact state between the target object and the second target object is formed, the second target object is a position shift in the plane perpendicular to the predetermined direction. The relative displacement means measures the positional deviation, and after the second positional deviation measurement operation, the relative movement means moves the both objects relative to each other in the predetermined direction, and moves both the objects from the high-pressure contact state. , A pressure smaller than the predetermined value is applied between the two objects, and a transition is made to a low pressure contact state in which the two objects are in contact with each other. A high-pressure contact-induced shift, which is a positional shift that occurs with the formation of a state, is estimated based on the second positional shift, and the measuring means is configured to detect the two objects in a low-pressure contact state between the two objects. Relative position A third positional shift that is a relative positional relationship in a plane perpendicular to the predetermined direction is measured, and the alignment means targets a state in which a positional shift in a direction opposite to the high-pressure contact-induced shift occurs. The third position is set so as to achieve the target state by relatively moving the two objects in a plane perpendicular to the predetermined direction in the low pressure contact state of the two objects. The displacement is adjusted, and the relative movement means performs a re-pressurization operation for changing both the objects to the high-pressure contact state again after the third positional displacement is adjusted.

  According to the invention described in claims 1 to 26, it is possible to align two objects with higher accuracy.

It is a side view which shows the internal structure of a joining apparatus. It is a side view which shows the internal structure of a joining apparatus. It is a schematic perspective view which shows the stage and head vicinity. It is a figure which shows the two alignment marks attached | subjected to one to-be-joined object. It is a figure which shows the two alignment marks attached | subjected to the other to-be-joined thing. It is a figure which shows the picked-up image regarding both to-be-joined objects. It is a figure which shows the state which 1 set of marks have mutually shifted | deviated. It is a top view which shows the position shift of 2 sets of marks. It is a figure explaining calculation of deviation deltatheta. It is a figure explaining calculation of deviation deltatheta. It is a flowchart which shows operation | movement of the alignment apparatus which concerns on this embodiment. It is a figure which shows the state (non-contact state) where both the objects were correctly positioned in the horizontal direction. It is a figure which shows the state (contact state) in which both the objects were correctly positioned in the horizontal direction. It is a figure which shows the state (contact state) which has shifted | deviated from the correct positional relationship. It is a figure which shows the mode immediately after a 1st alignment operation | movement (non-contact state) (1st operation example). It is a figure which shows the contact state (state which has 2nd position shift) of both target objects. It is a figure which shows a contact cancellation | release state (ideal state which returned to the same state as FIG. 15). It is a figure which shows the movement operation for eliminating the contact-induced deviation in the non-contact state. It is a figure which shows the state (ideal state) immediately after re-contact. It is a figure which shows the state (state in which the shift | offset | difference (alpha) exists) immediately after re-contact. It is a figure which shows a contact cancellation | release state (state in which deviation | shift exists). It is a figure which shows the movement operation for eliminating the contact-induced deviation in the non-contact state. It is a figure which shows a mode that residual shift | offset | difference exists immediately after a 1st alignment operation | movement (non-contact state) (2nd operation example). It is a figure which shows the contact state (state which has 2nd position shift) of both target objects. It is a figure which shows a contact cancellation | release state (ideal state which returned to the same state as FIG. 23). It is a figure which shows the movement operation for eliminating the contact-induced deviation in the non-contact state. It is a figure which shows the state (ideal state) immediately after re-contact. It is a figure which shows the state (state in which the shift | offset | difference (alpha) exists) immediately after re-contact. It is a figure which shows both the objects which have a non-contact opposing state (2nd Embodiment). It is a figure which shows a mode that the upper side object was bent in the non-contact opposing state. It is a figure which shows a mode that the upper side object was bent and contact was started. It is a figure which shows a mode that both target objects are contacting in the state where an upper side target object is flat. It is a figure which shows the holding mechanism of a vacuum suction system. It is a figure which shows the head etc. which concern on a modification. It is a figure which shows position alignment (nanoimprint process) with a board | substrate and a mold (non-contact opposing state). It is a figure which shows position alignment with a board | substrate and a mold (contact opposing state). It is a figure which shows a mode that the lens shape | molded using the mold is arrange | positioned in the predetermined position on a board | substrate by position alignment (nanoimprint process) of a board | substrate and a mold. It is a figure which shows the operation | movement which concerns on another modification regarding a nanoimprint process.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings.

<1. First Embodiment>
<1-1. Device configuration>
FIG. 1 is a diagram showing an internal structure of an alignment apparatus 1 (also referred to as 1A) according to an embodiment of the present invention. In the following, in each figure, directions and the like are shown using an XYZ orthogonal coordinate system for convenience.

  The alignment apparatus 1 pressurizes and heats an object to be bonded (here, a substrate) 91 and an object to be bonded (here, a substrate) 92 in a chamber (vacuum chamber) 2 under reduced pressure so that both objects are bonded. An apparatus for joining the objects 91 and 92. This alignment apparatus 1 is also expressed as a bonding apparatus that bonds the substrate 92 held by the head 22 and the substrate 91 held by the stage 12. In addition, although the semiconductor substrate is illustrated here as the board | substrates 91 and 92, it is not limited to this, A glass substrate etc. may be used. Further, both the objects to be joined 91 and 92 are also expressed as objects (for alignment operation).

  Here, as shown in FIGS. 12 and 13, a situation is assumed in which a substrate (object) 91 and a substrate (object) 92 are joined. More specifically, both the substrates 91 and 92 are bonded by bonding pads (electrodes) 93 of the substrate 91 and metal bumps (electrodes) 94 of chips (components) formed on the substrate 92. As the metal bump 94, a so-called solder bump formed of an appropriate solder material can be used. Further, the metal bump 94 is not limited to this, and may be formed using various metal materials such as Au (gold), Cu (copper), Al (aluminum). In FIG. 12 and the like, the pad 93 and the metal bump 94 are exaggerated for convenience of illustration.

  In addition, the surface activation treatment is applied in advance to the bonding surfaces of both objects 91 and 92 (pad 93 and metal bump 94). Here, it is assumed that a surface activation process is performed on both objects 91 and 92 in advance in an apparatus outside the bonding apparatus 1, and then both the objects 91 and 92 are carried into the bonding apparatus 1.

  The surface activation process is a process for activating the bonding surfaces of both the objects 91 and 92, and is executed by releasing a specific substance (for example, argon) using the beam irradiation unit. The beam irradiation unit accelerates an ionized specific substance (such as argon) by an electric field and emits the specific substance toward the bonding surfaces of the objects 91 and 92, thereby joining the surfaces of the objects 91 and 92. Activate. In other words, the beam irradiation unit activates the bonding surfaces of both objects 91 and 92 by irradiating energy waves toward the bonding surfaces of both objects 91 and 92. As the beam irradiation unit, an atomic beam irradiation apparatus, an ion beam irradiation apparatus, or the like is used.

  The bonding apparatus 1 includes a vacuum chamber 2 that is a processing space for both objects 91 and 92. The vacuum chamber 2 is connected to a vacuum pump 5 via an exhaust pipe 6 and an exhaust valve 7. The vacuum chamber 2 is put into a vacuum state by reducing (reducing pressure) the pressure in the vacuum chamber 2 in accordance with the suction operation of the vacuum pump 5. Further, the exhaust valve 7 can adjust the degree of vacuum in the vacuum chamber 2 by the opening / closing operation and the exhaust flow rate adjusting operation.

  The upper object 92 is held by the head 22 (more specifically, an electrostatic chuck or a mechanical chuck provided at the tip thereof). Similarly, the lower object 91 is held by the stage 12 (more specifically, an electrostatic chuck or a mechanical chuck provided at the tip thereof).

  The head 22 is heated by a heater 22 h built in the head 22, and can adjust the temperature of the object 92 held by the head 22. Similarly, the stage 12 is heated by a heater 12h built in the stage 12, and the temperature of the object 91 on the stage 12 can be adjusted. Further, the head 22 can also cool the head 22 itself rapidly to near room temperature by an air cooling type cooling device or the like built in the head 22. The same applies to the stage 12. The heaters 12h and 22h (particularly 22h) function as heating means (melting means) for melting the metal bumps 94, and also function as cooling means (solidification means) for cooling and solidifying the metal bumps 94 again. That is, the heaters 12h and 22h (particularly 22h) function as heating / cooling means for heating or cooling the metal bumps 94.

  Both the head 22 and the stage 12 are movably installed in the vacuum chamber 2.

  The stage 12 can be moved (translated) in the Y direction by the slide moving mechanism 14. The stage 12 moves in the Y direction between a relatively far-side standby position and a relatively near-side joining position (a position just below the head 22 (see FIG. 1)). The slide moving mechanism 14 has a highly accurate position detector (linear scale), and the stage 12 is positioned with high accuracy.

  The head 22 is moved (translationally moved) in the X direction and the Y direction (two translational directions parallel to the horizontal plane) by the alignment table 23, and at the θ direction (around the axis parallel to the Z axis) by the rotation drive mechanism 25. Rotation direction). The head 22 is driven by the alignment table 23 and the rotation drive mechanism 25 based on a position detection result by a position recognition unit 28 described later, and alignment operations in the X direction, the Y direction, and the θ direction are executed. Thus, in each direction (X direction, Y direction, θ direction) (horizontal direction in short) along the plane (horizontal plane) perpendicular to the vertical direction (Z direction), the stage 12 and the head 22 are By relatively moving, the object 91 held on the stage 12 and the object 92 held on the head 22 are aligned in the horizontal direction.

  The head 22 is moved (lifted / lowered) in the Z direction by the Z-axis lifting / lowering drive mechanism 26. As the stage 12 and the head 22 move relative to each other in the Z direction, the object 91 held on the stage 12 and the object 92 held on the head 22 come into contact with each other and are pressed and joined. In addition, the Z-axis raising / lowering drive mechanism 26 can also control the applied pressure at the time of joining based on signals detected by a plurality of pressure detection sensors (load cells, etc.) 29 and 32.

  In addition, the bonding apparatus 1 further includes a controller 100.

  The controller 100 is a control unit that controls the alignment operation and the bonding operation.

  The controller 100 (FIG. 1) includes an image acquisition control unit 101, a calculation unit 102, a determination unit 103, and a position correction control unit 104.

  The image acquisition control unit 101 controls an image acquisition unit (measurement unit) configured by the imaging units 28M and 28N. The image acquisition unit acquires an overlapping image (superimposed image) GA of both alignment marks respectively disposed on the surfaces of two opposing objects.

  The calculation unit 102 calculates a positional deviation or the like between the two objects 91 and 92 in a predetermined plane based on a captured image acquired under the control of the image acquisition control unit 101.

  The determination unit 103 determines whether or not the positional deviation between the objects 91 and 92 within a predetermined plane is within an allowable error range.

  The position correction control unit 104 controls a position correction unit configured to include the drive mechanism that drives the alignment table 23 and the rotation drive mechanism 25, and relatively drives both the objects 91 and 92 to perform positional deviation or the like. Correct.

<1-2. Alignment mark>
As shown in FIGS. 2 and 3, both the objects 91 and 92 are each provided with an alignment mark (hereinafter also referred to as an alignment mark) MK. Here, two alignment marks MK1a and MK1b (see FIG. 2) are provided on one object 91, and two alignment marks MK2a and MK2b (see FIG. 3) are also provided on the other object 92.

  Two alignment marks MK1a and MK1b (also collectively referred to as MK1) provided on the object 91 are pattern marks each having a square frame shape (see also FIG. 4). On the other hand, two alignment marks MK2a and MK2b (also collectively referred to as MK2) provided on the object 92 are pattern marks having a cross shape (see also FIG. 5).

  The alignment mark MK1 is disposed on the bonding surface of the lower object 91, and the alignment mark MK2 is disposed on the bonding surface of the upper object 92. For example, the size (length) H1 of the alignment mark MK1 and the size (length) H2 of the alignment mark MK1 are each about several hundred μm (micrometers). The value H1 is slightly larger than the value H2 (for example, about 1 to 2 times).

  As shown in FIG. 2, the two alignment marks MK1a and MK1b are arranged at a large distance from each other on the substrate 91 so that the distance between the two marks MK1a and MK1b is relatively large. More specifically, these alignment marks MK1a and MK1b are arranged near both ends of the substrate 91. For example, in the substrate 91 having a diameter of 300 mm (millimeters), the alignment mark MK1a is arranged near the left end of the substrate 91 (for example, a position of 5 mm (millimeters) from the left end to the center), and the other alignment mark MK1b is disposed on the substrate 91. Near the right end (for example, a position of 5 mm (millimeters) from the right end toward the center). Note that, by increasing the distance between the two marks MK1a and MK1b, it is possible to improve the detection accuracy and alignment accuracy of the positional deviation in the rotational direction.

  As shown in FIG. 3, the two alignment marks MK2a and MK2b are similarly arranged. Specifically, the two alignment marks MK2a and MK2b are arranged on the substrate 92 so as to be separated from each other. For example, in the substrate 92 having a diameter of 300 mm (millimeter), the alignment mark MK2a is arranged near the left end of the substrate 92 (for example, a position of 5 mm (millimeter) from the left end to the center side), and the other alignment mark MK2b is disposed on the substrate 92. Near the right end (for example, a position of 5 mm (millimeters) from the right end toward the center).

  As shown in FIG. 6, in a state in which both objects 91 and 92 are vertically opposed to each other (opposed arrangement state) (see also FIG. 12), the alignment mark MK2a of the object 92 is the alignment mark MK1a of the object 91. It is arranged in the vicinity region. Similarly, in the facing arrangement state, the alignment mark MK2b of the object 92 is arranged in the vicinity of the alignment mark MK1b of the object 91.

  When both the objects 91 and 92 are correctly positioned, as shown in FIG. 6, the cross-shaped portion of the alignment mark MK2 is included inside the square frame-shaped portion of the alignment mark MK1. More preferably, the center position of the cross-shaped part and the center position of the frame-shaped part are exactly the same.

<1-3. Position recognition unit>
The joining apparatus 1 includes a position recognition unit 28 that recognizes the horizontal positions (specifically, X, Y, and θ) of the objects 91 and 92.

  As illustrated in FIG. 1, the position recognition unit 28 includes imaging units (cameras) 28 </ b> M and 28 </ b> N that acquire optical images related to the objects 91 and 92 as image data. The imaging units 28M and 28N each have a coaxial illumination system. Note that light (for example, infrared light) that passes through both the objects 91 and 92, the stage 12, and the like is used as a light source of each coaxial illumination system of the imaging units 28M and 28N.

  The position recognition unit 28 can execute a position measurement operation for the alignment operation. Specifically, the position recognizing unit 28 captures images (images) relating to transmitted light and reflected light of illumination light emitted from the respective coaxial illumination systems of the imaging units 28M and 28N in a state where both the objects 91 and 92 face each other. Data) The positions of both objects 91 and 92 can also be recognized using GA. In other words, the alignment operation (alignment operation) of both objects 91 and 92 is performed by the position recognition unit (camera etc.) 28 with two sets of alignment marks (MK1a and MK2a) attached to both objects 91 and 92. ), (MK1b, MK2b).

  More specifically, as shown in FIG. 1, the light emitted from the light source (not shown) of the coaxial illumination system in the imaging unit 28M is reflected by the mirror 28e, the traveling direction thereof is changed, and the light travels upward. When the light is further reflected by the marks MK1a and MK2a of both objects 91 and 92 after passing through the window 2b (FIG. 1) and part (or all) of both objects 91 and 92, this time. Advances in the opposite direction (downward). Then, the light again passes through the window 2b and is reflected by the mirror 28e, and its traveling direction is changed to the left, and reaches the image pickup device in the image pickup unit 28M. In this way, the position recognizing unit 28 acquires the optical images (images including the marks MK1a and MK2a) related to the both objects 91 and 92 as the captured image GAa (see FIG. 6), and both the objects based on the image GAa. The position of a certain set of marks (MK1a, MK2a) attached to 91 and 92 is recognized, and the position shift between the set of marks (MK1a, MK2a) (in detail, a position shift vector) (Δxa, Δya) is obtained (see FIGS. 6 and 7). FIG. 7 is a diagram showing an example of the photographed image GAa, and shows a state in which a pair of marks MK1a and MK2a are displaced from each other.

  Similarly, the light emitted from the light source (not shown) of the coaxial illumination system in the imaging unit 28N is reflected by the mirror 28f, the traveling direction thereof is changed, and the light travels upward. If the light is further reflected by the marks MK1b and MK2b of the objects 91 and 92 after passing through part or all of the window 2b (FIG. 1) and the objects 91 and 92, this time, the opposite is obtained. Proceed in the direction (downward). Then, the light again passes through the window 2b and is reflected by the mirror 28f, and its traveling direction is changed to the right, and reaches the image pickup device in the image pickup unit 28N. In this way, the position recognizing unit 28 acquires the optical images (images including the marks MK1b and MK2b) regarding both the objects 91 and 92 as the captured image GAb (see FIG. 6), and both the objects based on the image GAb. The positions of the other set of marks (MK1b, MK2b) attached to 91 and 92 are recognized, and the position shift (position shift vector) (Δxb, Δyb) between the set of marks (MK1b, MK2b). ) Here, the photographing operations of the captured images GAa and GAb by the imaging units 28M and 28N are executed almost simultaneously.

  FIG. 8 is a top view showing positional deviations (Δxa, Δya) and (Δxb, Δyb) of two sets of marks. In FIG. 8, a position Q1 indicates the position of the alignment mark MK1a, and a position Q2 indicates the position of the alignment mark MK2a. A position Q3 indicates the position of the alignment mark MK1b, and a position Q4 indicates the position of the alignment mark MK2b. In FIG. 8, the misalignment is exaggerated.

  Thereafter, the position recognizing unit 28 and the calculating unit 102 (FIG. 1) cooperate to position misalignment (Δxa, Δya), (Δxb, Δyb) of these two sets of marks and the geometric relationship between the two sets of marks. Based on the above, a relative deviation ΔD (specifically, Δx, Δy, Δθ) of both the objects 91, 92 in the X direction, the Y direction, and the θ direction is calculated. For example, each value Δx, Δy, Δθ is calculated based on the following equations (1) to (3).

  Note that the following equation (4) may be used instead of the equation (3).

  When it is determined that the positional deviation ΔD between the two objects is not within the allowable error range (described later), for example, the relative deviation ΔD recognized by the position recognition unit 28 is reduced. As described above, the head 22 is driven in at least one of three directions of two translational directions (X direction and Y direction) and a rotational direction (θ direction). Thereby, both the objects 91 and 92 are relatively moved, and the positional deviation ΔD is corrected.

  On the other hand, when it is determined that the positional deviation between the two objects is within the allowable error range, the positional deviation correction operation is not performed and the alignment operation ends.

  In this way, the positional deviation ΔD (specifically Δx, Δy, Δθ) in the plane (horizontal plane) perpendicular to the vertical direction (Z direction) is measured, and the alignment operation for correcting the positional deviation ΔD is executed. The As will be described later, the measurement operation of the positional deviation ΔD is performed in a non-contact state of both the objects 91 and 92 and also in a contact state of both the objects 91 and 92.

<1-4. Operation in this embodiment>
Next, the operation according to the present embodiment will be described in more detail with reference to the flowchart of FIG. FIG. 11 is a flowchart showing the operation of the bonding apparatus 1.

  Here, as shown in FIG. 1 and the like, it is assumed that both the objects 91 and 92 are already arranged to face each other in a non-contact state (closely). More specifically, it is assumed that both objects 91 and 92 are opposed to each other with a minute gap (for example, about several μm (micrometer) to several tens of μm) (that is, in a non-contact state). At this time, it is assumed that both the objects 91 and 92 are roughly positioned in the horizontal direction by the rough alignment operation.

  Thereafter, in steps S11 to S13, an alignment operation in a non-contact state is performed. This alignment operation is also called fine alignment because it is more accurate than the rough alignment operation described above.

  Specifically, in step S11, first, captured images GAa and GAb (see FIG. 6) of both objects 91 and 92 (see FIG. 12) in a non-contact state are acquired. Then, based on the two photographed images GAa and GAb, the positional deviation amounts (Δx, Δy, Δθ) (“first position”) of both objects (objects) 91, 92 in the X direction, the Y direction, and the θ direction. Each of which is also referred to as “deviation”.

  Specifically, the shift amounts (Δxa, Δya) are calculated using the vector correlation method based on the image GAa obtained by simultaneously reading both marks MK1a, MK2a separated in the Z direction. Similarly, deviation amounts (Δxb, Δyb) are calculated using the vector correlation method based on the image GAb obtained by simultaneously reading both marks MK1b, MK2b separated in the Z direction. Then, based on the deviation amounts (Δxa, Δya), (Δxb, Δyb), the positional deviation amounts (Δx, Δy, Δθ) in the horizontal direction of both objects 91, 92 are measured. In particular, by analyzing the image GAa including both marks MK1a and MK2a separated in the Z direction using the vector correlation method, the position can be obtained with higher accuracy. The same applies to the image GAb.

  Here, the positional deviation amounts (Δx, Δy, Δθ) are exemplified as “first positional deviation”, but the present invention is not limited to this. For example, when the deviation in the θ direction can be ignored, the positional deviation amount (Δx, Δy) may be used as the first positional deviation. Alternatively, when performing only one-dimensional alignment, only one of the value Δx and the value Δy may be used as the first positional deviation. The same applies to the second misalignment (described later) and the third misalignment (described later).

  In addition, here, two deviation amounts (Δxa, Δya) and (Δxb, Δyb) obtained using two sets of alignment marks are used, but the present invention is not limited to this. For example, the first positional deviation may be calculated using only one deviation amount (Δxa, Δya) obtained using one set of alignment marks. Specifically, the deviation amounts (Δxa, Δya) may be used as “first positional deviation” as they are. The same applies to the second positional deviation and the third positional deviation.

  In the next step S12, it is determined whether or not the positional deviation between the objects 91 and 92 is within the allowable error range (within the first allowable error range).

  If it is determined that the positional deviation between the objects 91 and 92 is within an allowable error range (also simply referred to as an allowable range), the process proceeds to step S21. The processing after step S21 will be described later.

  On the other hand, if it is determined that the positional deviation between the objects 91 and 92 is not within the allowable error range (first allowable range), the process proceeds to step S13.

  In step S13, both objects 91 and 92 are relatively moved in order to correct the displacement (Δx, Δy, Δθ). Specifically, in a state where the stage 12 is fixed, the head 22 is moved so as to eliminate the positional deviation (Δx, Δy, Δθ), and both the objects 91, 92 are moved in the X direction, the Y direction, and the θ direction. It is moved relatively. Thereby, both the objects 91 and 92 are aligned with very high precision (for example, tolerance is within 0.2 micrometers) in the horizontal direction.

  FIG. 12 shows a state in which both objects 91 and 92 have a correct positional relationship in the horizontal direction by the positional deviation correction in step S13. In FIG. 12, both objects 91 and 92 are spaced apart from each other in the vertical direction (Z direction), and both objects 91 and 92 are not yet in contact with each other.

  After step S13, the process returns to step S11 again, and misalignment measurement is performed. Thereafter, the determination process in step S12 is performed again. Then, the branching process is performed again depending on whether or not the positional deviation between the objects 91 and 92 is within the first allowable range.

  In this way, the processes of steps S11 to S13 are repeated as necessary, and the first positional deviation is gradually reduced. In other words, both the objects 91 and 92 are aligned (also referred to as a first alignment operation) until the first positional deviation falls within the first allowable range.

  If it is determined that the positional deviation between the objects 91 and 92 is within the first allowable range, the process proceeds from step S12 to step S21. For example, when the values Δx and Δy are within the threshold value TH11 (0.2 micrometers, etc.) and the value Δθ is within the threshold value TH13 (for example, 0.00008 (deg)), both objects It is determined that the positional deviation between 91 and 92 is within the first allowable range.

  In step S21, by driving the Z-axis raising / lowering drive mechanism 26, the head 22 is lowered to bring both the objects 91 and 92 into contact (see FIG. 13). As shown in FIG. 13, in the contact state between the objects 91 and 92, for example, a pad 93 that is a protrusion on the surface of the object 91 and a metal bump 94 that is a protrusion on the surface of the object 92. Touch. FIG. 13 shows a state in which both the objects 91 and 92 are in contact with each other in a state of having a correct positional relationship in the horizontal direction.

  However, after such a contact operation, in practice, both objects 91 and 92 often deviate from the correct positional relationship (FIG. 13) as shown in FIG. Even if both the objects 91 and 92 have the correct positional relationship before the contact, a physical impact force acts when both the objects 91 and 92 are in contact with each other. It can be caused by doing so.

  In an attempt to eliminate the positional deviation after the contact as described above, the joining apparatus 1 according to this embodiment is configured so that both the objects 91 and 92 are horizontal in a state where the object 91 and the object 92 are in contact with each other. The displacement in the direction (also referred to as “second displacement”) is measured. And the said joining apparatus 1 aligns both target objects based on the said 2nd position shift. According to such an operation, it is possible to join the objects 91 and 92 in a state where the objects 91 and 92 are more accurately positioned in the horizontal direction.

  Specifically, first, in step S22 (FIG. 11), a positional shift (second position) between the objects 91 and 92 in the “contact state” (contact opposing state) (see FIGS. 13 and 14). Deviation) is measured.

  For example, also in this step S22, both the mark (MK1a, MK2a) of one mark attached to the first object and one mark attached to the second object as in step S11. Is acquired at the same time, and the positional deviations (Δxa, Δya) of both marks are measured based on the captured image GAa. Similarly, a captured image GAb including both marks (MK1b, MK2b) of another mark attached to the first object and another mark attached to the second object is acquired, and the captured image is acquired. Based on GAb, the positional deviations (Δxb, Δyb) of the two marks are measured. Further, based on these two sets of positional deviations (Δxa, Δya), (Δxb, Δyb), positional deviations ΔD (specifically, Δx, Δy, Δθ) of both objects 91, 92 are calculated (measured). The

  Thereafter, in step S23, it is determined whether or not the positional deviation (second positional deviation) between the objects 91 and 92 is within the allowable error range (second allowable range). In the determination process of step S23, the same process as the determination process of step S12 is performed. Note that the second allowable range may be different from or the same as the first allowable range.

  If it is determined in step S23 that the positional deviation between the objects 91 and 92 is within the allowable error range (second allowable range), the process proceeds to step S31 (heating and pressurizing process or the like). Step S31 will be described later.

  On the other hand, if it is determined that the positional deviation between the objects 91 and 92 is not within the allowable error range (second allowable range), the process proceeds from step S23 to step S24.

  In step S24, both objects 91 and 92 are moved in a direction that is relatively separated in the Z direction, and the contact state between both objects 91 and 92 is once released (see FIG. 12). Specifically, for example, when the head 22 is raised, the contact state between the objects 91 and 92 is released.

  In the next step S25, the joining apparatus 1 uses the second positional deviation measured in step S22 for the positional deviation (also referred to as “contact-induced deviation”) that occurs when the two objects 91 and 92 contact each other. To estimate. More specifically, the contact-induced deviation that occurs in the re-contact operation (next contact operation) after the contact release in step S24 is estimated. Here, considering that there is a high possibility that the same contact-induced deviation as the contact-induced deviation that occurred in the previous step S21 contact (first contact) is likely to occur again, the previous step S21 contact (first contact). Find the contact-induced deviation that occurred in step 1.

  For example, when the positional deviation between the objects 91 and 92 is zero in step S11, the second positional deviation itself is calculated as an estimated value of “contact-induced deviation”. Note that the estimated contact-induced deviation is used in steps S26 to S28 described later.

  More preferably, the contact-induced shift is also a positional shift in a non-contact state of both objects 91 and 92 immediately before the contact of both objects 91 and 92 (contact in step S21) occurs. Estimated based on. In other words, it is preferable that the contact-induced deviation is estimated based not only on the positional deviation immediately after the contact occurs (step S22) but also on the positional deviation immediately before the contact occurs.

  For example, in the first alignment operation in steps S11 to S13, when the positional deviations of the objects 91 and 92 are not zero and there is a deviation within the first allowable range (residual deviation). The contact-induced deviation may be estimated based on the residual deviation.

  Further, the contact-induced deviation for use in the alignment operation (steps S26 to S28, etc.) after the (N + 1) th contact (where N is a natural number) is the position immediately before the (N + 1) th contact occurs. It is preferable to estimate based on both the deviation and the positional deviation immediately after the contact occurs. For example, “the relative positional relationship (adjusted between the objects 91 and 92) finally adjusted in step S26 immediately before the (N + 1) th contact operation (step S21) and finally detected (measured) in step S27. Misalignment in non-contact state ”(also referred to as“ former ”) and“ second misalignment measured in step S22 immediately after the (N + 1) th contact operation (step S21) (both objects 91). , 92 in the contact state between each other) ”(also referred to as“ the latter ”). That is, the former may be adopted as the “positional deviation immediately before the occurrence of contact” and the latter may be adopted as the “positional deviation immediately after the occurrence of contact” to estimate the contact-induced deviation.

  In steps S27, S28, and S26, a positional shift (a positional shift in the opposite direction of the contact-induced shift) is generated by reversing the contact-induced shift estimated in this way (the contact-induced shift estimated in step S25). The state to be set is set as the “target state”.

  In the next step S27, the joining device 1 determines the relative positional relationship (“third displacement” in the horizontal plane between the objects 91 and 92 in the contact release state (non-contact state) of the objects 91 and 92. Is also measured). In step S27 as well, as in step S11, the positional deviation (Δxa, Δya) is calculated based on the image GAa obtained by reading both marks MK1a and MK2a at the same time. Similarly, a positional deviation (Δxb, Δyb) is calculated based on the image GAb obtained by reading both marks MK1b and MK2b at the same time. Based on these positional deviations (Δxa, Δya), (Δxb, Δyb), positional deviations (Δx, Δy, Δθ) between the objects 91 and 92 are measured.

  Further, in step S28, it is determined whether or not the third positional deviation has achieved the target state. As described above, a state in which a positional shift (a positional shift in the opposite direction of the contact-induced shift) in which the contact-induced shift (contact-induced shift estimated in step S25) is reversed is set as the “target state”. Is done.

  Specifically, the deviation between the target state and the current state (in other words, the deviation between the positional deviation obtained by inverting the contact-induced deviation and the above-described third positional deviation) is within the allowable range (third allowable range). Whether or not). Note that the third allowable range may be different from or the same as the first allowable range.

  When the difference between the target state and the current state falls within the third allowable range, the process proceeds to step S21.

  On the other hand, when the difference between the target state and the current state exceeds the third allowable range, the process proceeds to step S26.

  In step S26, the joining apparatus 1 relatively moves both the objects 91 and 92 in the horizontal plane so that the target state relating to the relative positional positional relationship between the objects 91 and 92 is achieved. In other words, feedback control is performed to reduce the difference between the target state and the current state, and both the objects 91 and 92 are relatively moved.

  Specifically, in a non-contact state (contact release state) of both objects 91 and 92, that is, in a state where both objects 91 and 92 can freely move in the horizontal direction, both objects 91 and 92 are relatively And an alignment operation (alignment operation) is executed. Specifically, a movement operation (relative movement operation) is performed so as to realize a deviation in the reverse direction of the contact-induced deviation (step S25) after correcting the third positional deviation (step S27).

  For example, the driving amount for correcting the third positional deviation measured in step S27 and the driving amount for eliminating the “contact-induced deviation” obtained in step S25 (the reverse deviation of the “contact-induced deviation”) The combined drive amount is obtained by combining the drive amount for generating Then, both the objects 91 and 92 are relatively driven based on the combined drive amount. More specifically, when the stage 12 is fixed, the head 22 moves in the horizontal direction (X direction, Y direction, and θ direction) so as to eliminate the second misalignment measured in the contact state (step S22). The two objects 91 and 92 are relatively moved. In other words, drive control (feedback control) for reducing the deviation between the target state and the current state is performed by moving the head in the horizontal direction while the stage is fixed. For example, when it is measured in a contact state that the object 92 is displaced to the right from the object 91 (step S22), the object 92 is moved to the left with respect to the object 91 in step S26. The drive operation for realizing the state (target state) is performed.

  As described above, in step S26, the third positional deviation is adjusted so as to achieve the target state by relatively moving both objects in the contact release state of both objects.

  Thereafter, the third positional deviation is measured again in step S27.

  Further, in the next step S28, it is determined again whether or not the third positional deviation has achieved the target state. Specifically, the deviation between the target state and the current state (in other words, the deviation between the positional deviation obtained by reversing the contact-induced deviation and the latest third positional deviation) is within the allowable range (third allowable range). Whether or not). Note that the third allowable range may be different from or the same as the first allowable range.

  When the difference between the target state and the current state falls within the third allowable range, the process proceeds to step S21.

  On the other hand, when the difference between the target state and the current state exceeds the third allowable range, the process returns to step S26 again. Then, feedback control is performed to reduce the difference between the target state and the current state, and both the objects 91 and 92 are relatively moved. Then, the third positional deviation measurement is performed again at step S27, and the determination process is executed at step S28.

  In this way, the processes of steps S26 to S28 are repeated as necessary. In other words, both the objects 91 and 92 are aligned (also referred to as a third alignment operation) until the deviation from the target state related to the third positional deviation falls within the third allowable range. In particular, the target state can be realized very accurately by feedback control based on the measurement position in step S27.

  If it is determined that the deviation from the target state related to the third positional deviation is within the third allowable range, the process proceeds from step S28 to step S21.

  In step S21, the joining apparatus 1 drives the Z-axis raising / lowering drive mechanism 26, lowers the head 22, and makes the both objects 91 and 92 contact again.

  And the operation | movement of step S22, ie, the position shift measurement operation | movement (step S22) of both the target objects 91 and 92 in a contact state, etc. is performed again.

  In the next step S23, the positional deviation between the objects 91 and 92 (the second positional deviation measured again in step S22 after the second time) falls within the allowable error range (second allowable range). If it is determined that it is not, the process proceeds from step S23 to step S24, and the operations of steps S24 to S28 are executed.

  On the other hand, when it is determined in step S23 that the positional deviation between the objects 91 and 92 (second measured positional deviation) is within the allowable error range (second allowable range). Then, the process proceeds to step S31 (heating and pressurizing process, etc.).

  The above-described operation (particularly, an operation in which contact is once released to form a state (target state) in which a reverse direction of contact-induced shift has occurred in advance and then re-contacted) is repeated once or a plurality of times. By being executed, the displacement caused by the contact operation itself between the objects 91 and 92 is reduced. Therefore, both the objects 91 and 92 are aligned with very high accuracy (for example, error within 0.2 micrometers) in the horizontal direction even in the final state after contact.

  In step S31, both objects 91 and 92 are pressurized and heated (heated to a temperature TM2 (temperature higher than the melting point of the metal bump 94)), and both objects 91 and 92 are joined. Specifically, the metal bump 94 is heated and melted and bonded to the pad 93 in a state where the metal bump 94 of the object 92 is in contact with the pad 93 of the object 91. Then, after the metal bump 94 is solidified after an appropriate cooling period, the pressurized state is released. In this way, both objects 91 and 92 are well aligned and joined.

  Thereafter, a device (semiconductor device or the like) composed of a part of both the objects 91 and 92 is manufactured by appropriately cutting both the objects 91 and 92.

  As described above, according to the operation according to this embodiment, it is possible to align both objects 91 and 92 with very high accuracy. As a result, a device (semiconductor device) composed of both objects (both objects) 91, 92 and the like is manufactured with high precision.

  In particular, after the third positional deviation is measured in step S27, in step S26 which is executed again via step S28, both objects 91 and 92 are relatively moved in the horizontal direction, and the target state is achieved. Thus, the third positional deviation is corrected. According to this, it is possible to realize the target state related to the third positional deviation very accurately in the non-contact state in which the previous contact is released and the next contact (re-contact) preparation stage. As a result, it is possible to reduce the deviation of both objects 91 and 92 to a very small value (for example, within the second allowable range) after the re-contact immediately after the preparation stage.

<1-5. First operation example>
15 to 19 are diagrams for explaining an outline of an example of operation according to the present embodiment (first operation example). In these figures, both the objects 91 and 92 are simplified by a single thick line. The representative positions of the objects 91 and 92 are indicated by black dots. Here, for the sake of explanation, the right end position of each of the objects 91 and 92 is set as the representative position, but the present invention is not limited to this. Further, here, for simplification of description, only the positional deviation in one direction (left and right direction in the drawing) parallel to the paper surface is considered. In addition, both the upper and lower objects 91 and 92 may move, but here, the lower object 91 does not move (relative to the reference position P0) and only the upper object 92 moves. As shown.

  In FIG. 15, a situation is assumed in which the first misalignment (the misalignment in the non-contact state) is zero by the alignment operation in steps S11 to S13.

  In this situation, when the contact operation in step S21 is performed, a shift due to the contact (contact-induced shift) occurs (see FIG. 16). For example, a contact-induced shift of about 5 μm (micrometer) occurs.

  In step S22, the positional deviation (second positional deviation) in the contact state between the objects 91 and 92 is measured. In FIG. 16, the deviation d2 (rightward) is measured as the second positional deviation.

  If it is determined in step S23 that the deviation d2 (for example, 5 micrometers) is not within the second allowable range, the contact between the objects 91 and 92 is once released in step S24.

  In response to the contact release (step S24), the positional deviation that has occurred due to the contact is naturally resolved once. FIG. 17 shows a state in which the state has returned to the state immediately after the first alignment (the state in which the deviation amount is zero here).

  In step S25, (contact-related deviation) in the next contact operation (re-contact operation) (step S21) is estimated. Here, the deviation d2 itself measured in the previous contact operation is estimated as a contact-induced deviation.

  Here, it is assumed that after step S26 is performed without performing steps S27 and S28, the process returns to step S21 again.

  Specifically, in step S26, as shown in FIG. 18, a relative movement operation is performed in advance so as to eliminate the contact-induced deviation (here, equal to the second positional deviation). Specifically, for example, as shown in FIG. 16, when a rightward shift d2 occurs in the contact state of both objects 91 and 92, the movement corresponding to the shift d2 is performed in the reverse direction (leftward). Is called. As a result, ideally, a state in which a positional deviation in the opposite direction of the contact-induced deviation occurs is reached. This state is a “target state”. In other words, both alignment marks do not match, but rather a displacement caused by the contact (here, (rightward) displacement d2 (+5 μm)) (for example, leftward displacement (−d2) (−5 μm)) ) Occurs (FIG. 18).

  In the case where the target state as shown in FIG. 18 has been accurately realized, the process returns to step S21 and then the objects to be joined again come into contact with each other. (For example, rightward displacement d2) occurs again (see FIG. 19).

  In a state in which a deviation (for example, a leftward shift) opposite to the estimated contact-induced deviation occurs (FIG. 18), a deviation (for example, a rightward deviation d2) similar to the estimated contact-induced deviation actually occurs. As a result, in the re-contact state, it is possible to bring the positional deviation of both objects to be close to zero (e.g., (ideally) exactly match both alignment marks). In FIG. 19, in the second step S21, the same deviation as the deviation d2 in the first step S21 occurs, and the positional deviation between the objects 91 and 92 is zero (ideal state). It is shown.

  According to such an operation, both the objects 91 and 92 can be positioned with high accuracy.

  However, as described above, the inventor of the present application has clarified that “the misalignment that occurs when the two objects to be joined contact each other cannot be sufficiently reduced due to various circumstances”. That is, the inventor of the present application has clarified that the positioning cannot be performed with sufficient accuracy only by the above-described operation.

  Examples of the circumstances include the following.

  For example, as described above, when the contact between the objects 91 and 92 is once released (in other words, when the objects 91 and 92 are peeled off), the object (for example, 92) and the target object may shift from each other, and the positional relationship between the holding member and the target object may shift (see FIG. 21). As can be seen from comparison with the ideal state of FIG. 17, FIG. 21 shows a situation in which a deviation d3 (for example, about 2 μm (micrometer)) occurs immediately after the contact is released.

  Alternatively, when there is an inclination between the two objects to be joined, the head (holding member holding one object) is horizontal along the inclination of the stage (holding member holding the other object). While moving in the direction (for example, the right direction), the objects to be joined come into contact with each other. In such a situation, when the contact between the objects 91 and 92 is released (for example, when the head is lifted), if the rigidity of the holding member (for example, the head) that holds one object is weak, the holding is performed. The member (and the one target object) does not return to the original position (position before contact) (see FIG. 17) in the horizontal direction, and a deviation (horizontal deviation) from the original position may occur. is there. Also in this case, the deviation d3 (see FIG. 21) may occur.

  Furthermore, as shown in FIG. 18, when a relative movement is performed to eliminate the deviation d2, the movement to the position for eliminating the deviation d2 is not always sufficient due to the head drive error. It may not be accurate (the exact target state has not been reached). More specifically, even if the head 22 is driven based on a command value for driving leftward by 5 μm (micrometer), only 4.7 μm (micrometer) may actually be driven.

  Therefore, in this embodiment, operations of steps S26 to S28 (particularly, steps S27 and S28) are executed after steps S24 and S25. Specifically, in the contact release state (see FIG. 18) of both objects 91 and 92 (after the contact between both objects 91 and 92 is released), both the objects are relative to each other in a plane perpendicular to a predetermined direction. And the third misalignment is adjusted to achieve the target state. Specifically, after the third positional deviation is measured in step S27, a reverse deviation (a deviation obtained by inverting the contact-induced deviation) of the contact-induced deviation (contact-induced deviation estimated in step S25) is generated. Such a relative movement operation (step S26) is performed.

  More specifically, immediately after steps S24 and S25, the measurement operation of step S27 is performed, and a deviation d3 (for example, 2 μm rightward (see FIG. 21)) is measured. Then, it is determined in step S28 that the deviation d3 is not within the third allowable range, and the process proceeds to step S26.

  Next, in step S26, a shift (shift) obtained by combining (summing) a shift in the opposite direction of the shift d3 (for example, a shift of 2 μm to the left) and a shift in the reverse direction of the shift d2 (for example, a shift of 5 μm in the left direction). A drive operation (a drive operation for driving the head 22 to the left by 7 μm) is performed (see FIG. 22). The driving operation that generates the relative movement in the reverse direction of the shift d3 is a driving operation that aims to return the relative positional relationship between the objects 91 and 92 to an ideal state (state of zero shift) (see FIG. 17). . Further, the driving operation for generating the reverse shift of the shift d2 intentionally generates the reverse shift of the contact-induced shift in the relative positional relationship between the objects 91 and 92 (see FIG. 18 and FIG. 22). ) Is a driving motion aiming at.

  Then, after the measurement operation in step S27 is performed again, the determination operation in step S28 is performed again. If the error from the target state is not within the third allowable range due to a drive error or the like, the drive operation (step S26) is performed again to eliminate the error. By repeating such an operation, an error from the target state falls within the third allowable range. For example, a state (target state) (see FIG. 18) in which a deviation d2 in the left direction (for example, 5 μm in the left direction) has occurred is accurately (ideally) realized. The driving error in the driving operation in step S26 is also gradually reduced by feedback control (visual feedback control) reflecting the detection result in step S27.

  Thereafter, when the re-contact operation in step S21 is performed, it is possible to position both objects 91 and 92 with high accuracy as described above (see FIG. 19).

  In the second step S21, if a contact-induced deviation (d2 + α) that is different from the deviation d2 in the first step S21 occurs (even if the ideal target state is realized in step S28). ) A shift α may occur as a positional shift between the objects 91 and 92 (see FIG. 20).

  When the shift α is within the second allowable range, the process proceeds from step S23 to step S31, and the alignment of both objects 91 and 92 is completed.

  On the other hand, when the deviation α is not within the second allowable range, the process proceeds from step S23 to step S24, and the processes of steps S24 to S28 and S21 to S23 are executed again. At this time, the contact-induced deviation is estimated based on the deviation (−d2) immediately before the contact in FIG. 20 (FIG. 18) and the deviation α measured immediately after the contact in FIG. Specifically, the difference between the two (+ α − (− d2)) (the rightward shift (d2 + α)) is estimated as the contact-induced shift, and this time the shift (d2 + α) is reversed (reverse direction). ) (− (D2 + α)) is set as the target state, and the processing of steps S26 to S28 is executed. Thereafter, the contact operation (step S21) is performed again, and the second misalignment measurement is performed again in step S22. By repeating such an operation, the deviation between the objects 91 and 92 falls within the second allowable range, and the alignment is completed. In particular, as described above, the target state is realized relatively well by being accompanied by the alignment operations in steps S26 to S28. Therefore, a relatively small number of retries (the number of recontact operations after once releasing contact) Thus, it is possible to keep the deviation between the objects 91 and 92 within the second allowable range.

  Thus, in step S25 after the second time, the contact-induced deviation for use in the alignment operation (steps S26 to S28, etc.) after the (N + 1) th (where N is a natural number) contact release is ( N + 1) It is estimated on the basis of both the positional deviation immediately after the occurrence of the contact (step S21) (step S22) and the positional deviation immediately before the occurrence of the contact (step S27). Specifically, “the relative positional relationship adjusted in steps S26 and S27 immediately before the (N + 1) -th contact operation (step S21) (the positional deviation between the objects 91 and 92)” and “the relevant (N + 1 This is estimated as the difference between the difference between “the second position shift measured in step S22 immediately after the second contact operation”.

  According to the operation as described above, the third position in the non-contact state at the stage where the previous contact is released and the preparation for the next contact (re-contact) is completed especially by the processing of steps S26, S27, and S28. It is possible to realize the target state relating to the deviation (see FIGS. 18 and 22) very accurately. As a result, it is possible to reduce the deviation of both objects 91 and 92 to a very small value (for example, within the second allowable range) after the re-contact immediately after the preparation stage.

<1-6. Second example of operation>
23 to 27 are diagrams for explaining an outline of another operation example (second operation example) according to the present embodiment. These drawings are similar to FIGS. 15 to 19.

  However, in this operation example, for example, the residual deviation d1 in the alignment operation in steps S11 to S13 is acquired as the positional deviation immediately before the contact operation (step S21). Then, the second positional deviation d2 acquired in step S22 immediately after the contact operation (step S21) is acquired as the positional deviation immediately after the contact operation (step S21). Further, the contact-induced deviation is estimated based on the positional deviation d1 immediately before the contact operation (step S21) and the positional deviation d2 immediately after.

  More specifically, in the second operation example, the first misalignment (the misalignment in the non-contact state) is not zero due to the alignment operation in steps S11 to S13 (the misalignment within the first allowable range ( Assume a situation (see FIG. 23) in which a residual deviation) d1 (for example, 0.1 micrometer) exists.

  In this situation, when the contact operation in step S21 is performed, a shift due to the contact (contact-induced shift) occurs (see FIG. 24).

  In step S22, the positional deviation (second positional deviation) in the contact state between the objects 91 and 92 is measured. In FIG. 24, the deviation d2 is measured as the second positional deviation.

  If it is determined in step S23 that the deviation d2 (for example, 5 micrometers) is not within the second allowable range, the contact between the objects 91 and 92 is once released in step S24.

  In response to this contact release, the positional shift that has occurred due to the contact is once resolved naturally. FIG. 25 shows a state in which the state has returned to the state immediately after the first alignment (a state where a slight deviation d1 exists here).

  In step S25, (contact-related deviation) in the next contact operation (re-contact operation) (step S21) is estimated. Here, as shown in FIG. 24, assuming that a rightward shift d2 occurs in the contact state of both objects 91 and 92 and the shift d1 remains in the first alignment operation, the shift d2 is assumed. The difference (d2−d1) between the deviation d1 and the deviation d1 is estimated as a contact-induced deviation.

  Here, if step S26 is performed without performing steps S27 and S28 and then the process returns to step S21 again, the following operation is performed.

  Specifically, in step S26, as shown in FIG. 26, a relative movement operation that eliminates the contact-induced deviation (here, (d2-d1)) is performed. As a result, ideally, a state in which a positional shift in the direction opposite to the contact-induced shift (d2-d1) occurs is reached. This state is the “target state”. In other words, the alignment marks do not match, but rather shift to a state in which there is a shift in the opposite direction (for example, a shift in the left direction) from the contact-induced shift (here, “shift (d2-d1)”) ( FIG. 26).

  Thereafter, returning to step S21, when both the objects to be joined come into contact again, a displacement (contact-induced displacement) (for example, a rightward displacement (d2-d1)) due to the recontact occurs again (see FIG. 27).

  In a state (FIG. 26) in which a shift in a direction opposite to the estimated contact-induced shift (for example, a shift in the left direction (-(d2-d1))) occurs (for example, a shift in the right direction). As a result of (d2-d1)), it is possible to make the positional deviation of both objects to be close to zero in the re-contact state (e.g., (ideally) just match both alignment marks). is there. In FIG. 27, in the second step S21, the same deviation as the deviation (d2-d1) in the first step S21 occurs, and the positional deviation between the objects 91 and 92 is zero (ideal). State).

  However, as described above, due to various circumstances, the deviation that occurs when the objects to be bonded come into contact with each other may not necessarily be sufficiently reduced.

  On the other hand, also in the second operation example, the operations of steps S26 to S28 (particularly, steps S27 and S28) are executed after steps S24 and S25.

  Specifically, immediately after steps S24 and S25, the measurement operation of step S27 is performed, and a deviation d3 (for example, 2 μm to the right (see FIG. 21)) is measured. Then, it is determined in step S28 that the deviation d3 is not within the third allowable range, and the process proceeds to step S26.

  Next, in step S26, a shift in the opposite direction of the shift d3 (for example, a shift of 2 μm to the left) and a shift in the reverse direction of the shift (d2-d1) (for example, a shift of 4.9 μm in the left direction) are combined ( A driving operation (driving operation for driving the head 22 to the left by 6.9 μm) is performed so as to realize the shifted (summed vector). The driving operation that generates the relative movement in the reverse direction of the shift d3 is a driving operation that aims to return the relative positional relationship between the objects 91 and 92 to an ideal state (state of zero shift) (see FIG. 17). . Further, the driving operation for generating a shift in the opposite direction of the shift (d2-d1) intentionally generates a shift in the reverse direction of the contact-induced shift in the relative positional relationship between the objects 91 and 92 (see FIG. 18). ) Is a driving motion aiming at.

  Then, after the measurement operation in step S27 is performed again, the determination operation in step S28 is performed again. If the error from the target state is not within the third allowable range due to a drive error or the like, the drive operation (step S26) is performed again to eliminate the error. By repeating such an operation, an error from the target state falls within the third allowable range. For example, a state (target state) (see FIG. 26) in which a shift (d2-d1) in the left direction (for example, 4.9 μm in the left direction) has occurred is realized accurately (ideally). The driving error in the driving operation in step S26 is also gradually reduced by feedback control (visual feedback control) reflecting the detection result in step S27.

  According to this, in the non-contact state at the stage where the previous contact is released and the preparation for the next contact (re-contact) is completed, the target state (see FIG. 26) regarding the third misalignment is realized very accurately. Is possible. As a result, it is possible to reduce the deviation of both objects 91 and 92 to a very small value (for example, within the second allowable range) after the re-contact immediately after the preparation stage.

  Further, in step S23 immediately after one re-contact operation (second step S21), if the second positional deviation does not fall within the second allowable range, the operations in steps S24 to S28 are performed again. Repeated.

  In step S25 after the second time, the contact-induced deviation for use in the alignment operation (steps S26 to S28, etc.) after the (N + 1) th (where N is a natural number) contact release is the (N + 1) th time. It is estimated based on both the positional deviation immediately after the contact occurs (step S21) (step S22) and the positional deviation immediately before the contact occurs (step S27).

  For example, in the second step S21, when a contact-induced shift (d2-d1 + α) different from the contact-induced shift (d2-d1) in the first step S21 occurs (see FIG. 28), both targets A shift α may occur as a positional shift between the objects 91 and 92 (see FIG. 28). Also in this case, the same operation as in the first operation example may be performed. However, in this case, the deviation (-(d2-d1)) (FIG. 26) measured in step S27 immediately before the contact in FIG. 28 and the deviation measured in step S22 immediately after the contact in FIG. Based on α, the contact-induced deviation is estimated. Specifically, the difference between the two (α − (− (d2−d1))) (rightward displacement (d2−d1 + α)) is estimated as the contact-induced displacement, and this time (d2−d1 + α). A state in which a deviation (reverse deviation) (− (d2−d1 + α)) is generated as a target state is executed, and the processes of steps S26 to S28 are executed. Thereafter, the contact operation (step S21) is performed again, and the second misalignment measurement is performed again in step S22.

  By repeating such an operation, the deviation between the objects 91 and 92 falls within the second allowable range, and the alignment is completed. As described above, in particular, the target state is realized relatively well by being accompanied by the alignment operations in steps S26 to S28. Therefore, the deviation between the objects 91 and 92 can be reduced to the second allowable range with a relatively small number of times. It is possible to fit in.

<2. Second Embodiment>
The idea of the first embodiment may be applied in alignment when one substrate is bent from the center thereof toward the outer peripheral portion and the one substrate and the other substrate are joined. . The second embodiment is a modification of the first embodiment. Hereinafter, the difference from the first embodiment will be mainly described.

  First, an outline of the operation of bending and bonding the substrates will be described.

  FIGS. 29-32 is a figure which shows the outline of joining operation | movement (joining operation | movement accompanied by a deformation | transformation (deflection) of a board | substrate) which concerns on 2nd Embodiment.

  In FIG. 29, the substrate 91 is held on the stage 41 and the substrate 92 is held on the head 42.

  In the second embodiment, a stage 41 is provided instead of the stage 12 (see FIG. 1), and a head 42 is provided instead of the head 22 (see FIG. 29, etc.). In the second embodiment, a vacuum suction type holding mechanism 45 is employed as a holding mechanism for the substrates 91 and 92. As shown in FIG. 33, the holding mechanism 45 includes an outer holding portion 45a and an inner holding portion 45b. The outer holding portion 45a and the inner holding portion 45b can function their adsorption actions individually. The projection mechanism 43 pushes down the central portion of the substrate 92 downward while the substrate 92 is held only by the adsorption groove of the outer holding portion 45a out of the adsorption grooves of the outer holding portion 45a and the inner holding portion 45b. By (center push), the upper substrate 92 can be bent (elastically deformed) so as to protrude downward from the center toward the outer peripheral portion. Conversely, the pressing force of the protruding mechanism 43 is eliminated, and the substrate 92 is held in a flat state by adsorbing the substrate 92 by both the adsorption groove of the outer holding portion 45a and the adsorption groove of the inner holding portion 45b. Is possible.

  In FIG. 29, the substrate 92 is held by the head 42 in a flat state. Moreover, the board | substrate 92 and the board | substrate 91 are opposingly arranged by the non-contact state.

  From the state shown in FIG. 29, when the center push (pressing against the central portion of the substrate) is performed by the protruding mechanism 43, the upper substrate 92 is in a state where the central portion protrudes downward from the outer peripheral portion (convex downward). To be bent (elastically deformed) (see FIG. 30).

  Next, the head 42 is lowered to bring the central portion of the upper substrate 92 into contact with the central portion of the lower substrate (FIG. 31). Then, the pressing force (projection amount) of the protrusion mechanism 43 is gradually reduced to gradually eliminate the bending of the upper substrate 92, and the contact portion with the lower substrate 91 is gradually moved from the central portion toward the outer peripheral portion. By increasing the number, the upper substrate 92 and the lower substrate 91 are joined (FIG. 32). In this way, contact between the two objects is started from a state in which one of the objects (substrate) 92 is bent and the objects 91 and 92 partially contact each other (see FIG. 31). Finally, they are brought into contact with each other (see FIG. 32).

  Such bonding is used, for example, when the surfaces are made hydrophilic and the two substrates are bonded together. According to this, it is possible to avoid or suppress the generation of voids due to air entrainment.

  Now, in such bonding, the two substrates face each other (see FIGS. 30 and 31) (particularly, one substrate is bent and the central portion of the one substrate and the central portion of the other substrate are In the state in which the two substrates are in contact (see FIG. 31), the above-described idea may be applied when the two substrates are aligned.

  In particular, in such joining, particularly when the contact between both objects 91 and 92 is once released (in other words, when both objects 91 and 92 are peeled off), the object is affected by an impact associated with the contact release. The holding member (head 42) that holds the object (for example, the object 92) and the object are shifted from each other, and the positional relationship between the holding member and the object is shifted (see FIG. 21). Is likely to occur.

  In the second embodiment, in consideration of such circumstances, the same operation as that of the first embodiment is executed.

  In steps S11 to S13 (first alignment operation in a non-contact state), the object (substrate) 92 is not bent yet and is held by the head 42 in a flat state.

  Thereafter, immediately before the transition to step S21, one of the objects 91, 92 (here, the object 92) is bent, and the contact operation of the objects 91, 92 is started (see FIG. 31). . In the contact operation in step S21, finally, the object 92 transitions to a flat state (see FIG. 32). That is, after the contact between the objects 91 and 92 is started (see FIG. 31), the objects are brought into contact with each other in a state where the one object that has been bent returns to a flat state (see FIG. 32). reference). In step S21, as described above, a deviation (contact-induced deviation) occurs due to the contact operation.

  Next, the second positional deviation is measured in step S22. In step S <b> 22, the horizontal displacement between the object 92 having a flat state and the object 91 facing the object 92 is measured as a “second displacement”.

  Further, after step S23, the process proceeds to step S24.

  In step S24, for example, the reverse operation of the contact operation in step S21 (contact state → non-contact state) is performed. Specifically, one object (for example, 92) returns from a flat state (contact state) to a downward convex state (a state where the substrate 92 is bent and only the central portion of the substrate 92 is in contact). After the contact, the contact between the objects 91 and 92 is released (non-contact state).

  However, in the contact release operation (in other words, the peeling operation of both objects) in step S24, as described above, the substrate 42 and the head 42 that holds the substrate 92 due to, for example, an impact associated with the contact release. And the positional relationship between the head 42 and the substrate 92 may be shifted (see FIG. 21).

  Therefore, also in the second embodiment, the positional relationship between the objects 91 and 92 is adjusted by the operations of steps S25 to S28. Specifically, the positional relationship between the objects 91 and 92 is adjusted so as to reach the target state.

  In the second embodiment, in step S27, the third displacement is measured in a state where the deflection of the object 92 is returned and the object 92 is changed to a flat state (see FIG. 29). More specifically, the pressing force of the projecting mechanism 43 is eliminated, and the object 92 is flattened by re-adsorbing the object 92 by both the adsorption groove of the outer holding part 45a and the adsorption groove of the inner holding part 45b. Hold in state. Then, the relative positional relationship between the object 92 re-adsorbed by the holding mechanism 45 and held in a flat state and the opposite object 91 is measured. For example, the same measurement operation as in the first embodiment is performed. Further, the operations in steps S26 to S28 are repeated until the deviation from the target state falls within the third allowable range.

  Further, the operations of steps S24 to S28 and S21 to S23 are repeatedly executed until the second positional deviation amount falls within the second allowable range.

  According to the operation as described above, it is possible to obtain the same effect as in the first embodiment.

  In particular, when the object 92 is once returned to the flat state in step S27, it is possible to avoid the occurrence of a new position error due to the bending of the object 92. If the object 92 is bent, the mark 92 of the object 92 is caused by the skew because the object 92 is skewed with respect to the horizontal direction, particularly on the outer peripheral side of the object 92. Will slip out. On the other hand, in step S27 of the second embodiment, since the third positional deviation is measured after the object 92 is once returned to the flat state, the deviation due to the above-mentioned skew (target It is possible to avoid the occurrence of a new position error due to the bending of the object 92. Thus, since the measurement in step S27 (and / or step S11) and the measurement in step S22 are performed in a state where both the objects 91 and 92 are flat, a new position resulting from the deflection of the object 92 is obtained. It is possible to avoid the occurrence of errors. As a result, it is possible to more accurately estimate the contact-induced deviation.

  Note that it is not always necessary that the entire range of at least one of the objects 91 and 92 has returned to a flat state after the completion of the contact operation in step S21. For example, the range (partial region) extending from the central part of both objects to a part slightly inside the outer peripheral edge of both objects (for example, a part near the mark arrangement part) returns to a flat state and is in contact with each other. However, the state etc. which the outer peripheral edge part of both the objects is not contacting each other may be sufficient. More specifically, in the entire period of the contact operation in step S21 (the entire period from the start to the completion), the bonding apparatus 1 (head 42) maintains the pressing force of the protrusion mechanism 43 (the protrusion mechanism 43 protrudes). While the object 92 is pressed against the object 91 with a relatively strong force, the adsorption groove of the outer holding part 45a and the adsorption groove of the inner holding part 45b (adsorption by the adsorption groove of the inner holding part 45b) The substrate 92 is sucked only by the suction groove of the outer holding portion 45a (with the force released). In particular, as the contact operation proceeds from the start time to the completion time (the distance between the two objects is reduced), the pressure between the two objects (the downward pressing force of the head 42) is gradually increased. Go. According to this, although the contact part of both objects gradually expands from the center part of each object toward the outer peripheral part with an increase in the applied pressure, both objects (even when the contact operation is completed). A range (partial region) from the center part of both objects to the mark arrangement part (part slightly inside the outer edge) of both objects is brought into contact with each other in a flat state without bringing the outer periphery ends into contact with each other. It is possible. Such a contact state is a state similar to FIG. 31 and is a state in which both objects are in contact with each other over a wider range than the contact range shown in FIG. 31 (a range only in the vicinity of the central portion). . By performing temporary joining (step S21) in such a state, both the objects 91 and 92 are not in contact (joining) with each other in the vicinity of the outer peripheral portion (outer edge portion). In the contact release process (substrate peeling process), the peeling from the outer peripheral part is started relatively easily, and the peeling of the objects 91 and 92 gradually proceeds from the outer peripheral part to the central part. Therefore, compared with the case where both the objects 91 and 92 contact each other also in the outer peripheral part in step S21, it is possible to peel both objects 91 and 92 from each other comparatively easily.

  Further, when the pressing force of the protruding mechanism 43 is maintained even immediately after the completion of step S21, in the subsequent step S24, the contact between the objects 91 and 92 is released by eliminating the pressing force of the protruding mechanism 43. It may be. More specifically, for example, the protrusion amount of the protrusion mechanism 43 is reduced to eliminate the pressing force of the protrusion mechanism 43, and the object 92 is moved by both the suction groove of the outer holding portion 45a and the suction groove of the inner holding portion 45b. You may make it the contact of both the objects 91 and 92 cancel | release by adsorb | sucking.

  Moreover, in the said 2nd Embodiment, although the vacuum suction type holding | maintenance mechanism is illustrated, it is not limited to this. As a substrate holding mechanism, for example, various methods such as an electrostatic chuck method and a mechanical method may be employed. Further, as shown in FIG. 34, a substrate (for example, 92) may be bent using a pressing plate 44 that is thin and flexible. More specifically, the pressing plate 44 may be elastically deformed according to the protrusion of the protruding mechanism 43, and the substrate 92 may be bent along the elastic deformation of the pressing plate 44. In FIG. 34, the substrate 92 is bent with respect to the outer peripheral portion thereof so that the inner peripheral portion (substrate central portion) protrudes downward (in a convex state downward).

  Here, only the upper substrate 92 is bent, but the present invention is not limited to this, and conversely, only the lower substrate 91 may be bent. Alternatively, both the upper substrate 92 and the lower substrate 91 may be bent. In order to bend the lower substrate 91, various mechanisms (projection mechanism and holding mechanism 45) provided on the head 42 side may be provided on the stage 41 side. In these cases, the same operation as described above may be performed.

<3. Modified example>
Although the embodiments of the present invention have been described above, the present invention is not limited to the contents described above.

  For example, in each of the above-described embodiments, the process of step S25 is performed immediately after the process of step S24. However, the present invention is not limited to this, and conversely, immediately before the process of step S24 (steps S23 and S24). It may be performed during (between). Alternatively, the process of step S25 may be performed immediately before step S23 (between step S22 and step S23).

  In the first embodiment (first operation example, second operation example, and the like), the position shift (first position shift (step S11) or both of the objects 91 and 92 immediately before the previous contact operation or The contact-induced deviation is calculated based only on the difference between the adjusted third positional deviation (step S27) and the positional deviation (second positional deviation) of the objects 91 and 92 immediately after the previous contact operation. (Step S25), but is not limited to this.

  For example, in step S25, the latest contact-caused deviation may be calculated based on the calculated value of the contact-caused deviation at the time of a plurality of past contacts. Specifically, an average value (for example, 5.6 μm) of past and present estimated values may be calculated as the latest contact-induced deviation. In other words, statistical processing (average processing or the like) regarding a plurality of past estimation results may be performed, and the latest contact-induced deviation may be estimated based on the results of the statistical processing.

  Here, in the first embodiment, the same contact cause deviation as the contact cause deviation that occurred in the previous contact operation (for example, the first contact operation) in step S21 is the next contact operation (for example, the second contact operation). In view of the above, the contact-induced deviation is estimated as described above.

  However, in actuality, in the (N + 1) th (for example, second) contact operation, a contact-induced shift different from the contact-induced shift that occurred in the immediately preceding Nth (for example, the first contact) can occur. That is, the displacement caused by contact (contact caused displacement) is not the same every time, and there is “variation” in the contact caused displacement. More specifically, the contact-induced deviation is “6.0 μm” in the first contact operation, “5.2 μm” in the second contact operation, “5.6 μm” in the third contact operation, and the fourth contact. There are cases where the contact operation has different values (“variation”) such as “5.9 μm” in the operation and “5.3 μm” in the fifth contact operation. For this reason, the next contact-induced deviation may be different from the previous contact-induced deviation.

  On the other hand, according to the statistical processing as described above, it is possible to estimate the “contact-induced deviation” well in consideration of the variation (estimate the amount that is most likely to occur). Is possible. Therefore, it is possible to increase the possibility that the second positional deviation is within the second allowable range in step S23.

  Note that the past estimation result regarding the contact-induced deviation is not limited to the past estimation result for the set of objects 91 and 92 currently being joined, but is the past estimation result for another set of objects 91 and 92. There may be.

  In each of the above-described embodiments, the alignment marks MK1a and MK2a of the mark pair are set using reflected light (and transmitted light) from each mark pair (for example, MK1a and MK2a) (see FIG. 12 and the like). The captured image GA captured at the same time is acquired (see FIG. 1 and the like), but is not limited thereto. For example, a photographed image GA obtained by simultaneously photographing both alignment marks MK1a and MK2a of the mark pair may be acquired using only transmitted light that has passed through each mark pair (for example, MK1a and MK2a). . More specifically, illumination (light source) is provided on the upper side (one side) of both objects 91 and 92, and the light emitted from the light source is used to mark portions of both objects 91 and 92 (alignment marks MK1a and MK2a). Make it transparent. And what is necessary is just to image the transmitted light using the imaging part provided in the downward side (other side) of both the objects 91 and 92. FIG.

  In each of the above-described embodiments and the like, between the objects 91 and 92 based on the image GAa obtained by simultaneously reading the upper and lower marks MK1a and MK2a and the image GAb obtained by simultaneously reading the upper and lower marks MK1b and MK2b. A misalignment (third misalignment) is measured.

  However, the present invention is not limited to this. For example, you may make it read both marks (MK1a, MK2a) of one mark attached | subjected to the lower side object, and the other mark attached | subjected to the upper side object at different timings. In other words, an image including the one mark and an image including the other mark may be taken separately (at different times). Similarly, both marks (MK1b, MK2b) of other marks attached to the lower object and other marks attached to the upper object may be read at different timings. Good. However, according to the simultaneous reading of the marks attached to the upper and lower objects as in the above embodiments, the vibration of the apparatus between the upper mark reading time and the lower mark reading time is reduced. It is possible to suppress the influence of the measurement error or the drive error of the image pickup system (focus drive system or the like) due to this.

  Further, the two alignment marks MK1a and MK2a may be taken by different cameras. The same applies to the other two alignment marks MK1b and MK2b.

  When measuring the third misalignment (step S27) or the like, if one holding member (for example, the lower holding member) has very high rigidity and cannot be considered to move, Only the mark of the object held by the holding member (for example, the upper holding member) may be newly read.

  For example, in step S27, only one mark MK2a is read out of both marks (MK1a, MK2a) of the mark MK2a attached to the upper object 92 and the mark MK1a attached to the lower object 91. You may do it. In addition, for the other mark MK1a of the two marks (MK1a, MK2a), the position measured before the contact release operation (step S24) (for example, the position measured in the immediately preceding step S22 or the like). You may make it consider that it is effective as it is. Then, based on the difference between the position of the lower mark MK1a measured in the immediately preceding step S22 and the position of the upper mark MK2a measured in step S27, the relative positional relationship between both marks (MK1a, MK2a) in step S27. May be acquired. The same applies to the other mark pairs (MK1b, MK2b). As described above, the third positional deviation may be measured substantially based only on the position of the upper mark MK2a (, MK2b) in step S27.

  In each of the above embodiments, the two captured images GAa and GAb are captured in parallel (substantially simultaneously) using the two cameras 28M and 28N. However, the present invention is not limited to this. For example, the captured images GAa and GAb may be sequentially captured by moving one camera 28M in the X direction and / or the Y direction. In other words, the two alignment marks MK1a and MK1b that are spaced apart in the horizontal direction may be photographed by different cameras. The same applies to the other two alignment marks MK2a and MK2b.

  Moreover, in each said embodiment etc., what emits infrared light is used as a light source of illumination, However, It is not limited to this, For example, visible light etc. may be radiate | emitted.

  Further, in each of the above-described embodiments and the like, a mode in which the present invention is applied to alignment between a substrate and a substrate (wafer-on-wafer (WOW) bonding technique) is illustrated. It is not limited. In other words, the alignment processing target is not limited to the substrates.

  For example, the present invention may be applied to alignment (chip-on-wafer (COW) bonding technology) between one substrate and each of a plurality of chips. More specifically, two alignment marks for each of a plurality of chips are provided on the substrate (an alignment mark twice the number of chips is provided on the substrate). Then, the alignment between each chip and the substrate is performed by aligning the two alignment marks provided on the substrate for each chip with the two alignment marks provided on each chip by the above-described method. That's fine.

  Alternatively, the present invention may be applied to alignment between chips (chip-on-chip (COC) bonding technology).

  In addition, the present invention may be applied to alignment between a substrate and a mold (such as a transparent mold) in nanoimprint processing. More specifically, the above concept is applied when both the substrate and the mold are targets (alignment) and the two are aligned in a plane perpendicular to the opposing direction of the substrate and the mold. May be. Note that such treatment is performed by bringing both the substrate and the mold facing each other into contact with each other in a state where a resin (such as a photocurable resin or a thermoplastic resin) is interposed between the substrate and the mold. What is necessary is just to perform in the apparatus (it also calls a pressurization apparatus) etc. which press. The both are also referred to as objects to be pressed. Furthermore, the said pressurization apparatus is expressed also as a resin device manufacturing apparatus which manufactures the various resin devices using resin.

  For example, as shown in FIGS. 35 and 36, a substrate 82 coated with a resin 88 on its surface and a mold 81 to be pressed against the resin are arranged in a horizontal direction (a direction parallel to the main surface of the substrate 82) (FIG. The above-mentioned idea may be applied when the position is accurately aligned in the horizontal direction or the like at 35 or the like. 35 is a diagram illustrating a state where the substrate 82 and the mold 81 are separated from each other, and FIG. 36 is a diagram illustrating a state where the substrate 82 and the mold 81 are in contact with each other.

  In addition, as shown in FIG. 37, a resin layer 73 is molded and cured on one surface of a substrate (glass substrate or the like) 72 using a mold 71 having a plurality of concave portions each formed in accordance with a lens shape. Thus, when the plurality of lenses (73) are formed on one surface of the substrate 72, the above idea can be applied. In particular, by using the alignment technique described above, it is possible to accurately arrange each lens at a predetermined position in the substrate.

  FIG. 37 shows a state where the mold 71 is accurately positioned with respect to the substrate 72 after the resin is applied to the mold 71 to form a plurality of lenses 73 (or in the middle of the formation). Has been. In FIG. 37, a relatively small number of lenses 73 are shown for convenience of illustration, but in many cases, a very large number of lenses 73 are actually formed.

  In particular, in the nanoimprint process or the like (positioning operation accompanied by an operation of contacting through an uncured resin, etc.), the contact may be temporarily released in step S24 as in the above embodiments. However, the present invention is not limited to this. For example, instead of releasing the contact in step S24, the pressure applied between the two objects is reduced to a value smaller than a certain value PR, and both the objects are moved to a state where both the objects are easily moved. The state (contact state) may be changed. Hereinafter, such a modification will be described in detail with reference to FIG. In FIG. 38, basically the same operation as in the first embodiment (see FIG. 11) is performed. Hereinafter, the difference from the first embodiment (see FIG. 11) will be mainly described.

  First, steps S11 to S13 are performed in the same manner as in the first embodiment.

  Thereafter, the first steps S21 (S21b) to S23 (S23b) are performed in the same manner as in the first embodiment. In step S21 (S21b) immediately after steps S11 to S13, as a result of the relative movement of both objects in the Z direction, the objects are in contact with each other and a constant value is set between the objects. It is also expressed that a state in which a pressure higher than PR is applied (also referred to as a “high pressure contact state”) is formed. In short, the high-pressure contact state is a state in which both objects are pressed and contacted with a relatively strong force (a force stronger than a low-pressure contact state described later). In step S23, the second displacement is measured in the high-pressure contact state.

  In the next step S24 (S24b), both the objects are moved in accordance with the relative movement in the Z direction (moving in a direction away from each other) to reduce the pressure between the objects. Transition from the high pressure contact state to the low pressure contact state. Here, the “low pressure contact state” is a state in which the two objects are in contact with each other in a state where a pressure smaller than a predetermined value PR is applied between the two objects (for example, light touching). Contact state). In this step S24b, the contact state is continued, but the pressure acting between the two objects is reduced. In the “high pressure contact state”, both objects are difficult to move freely in the horizontal plane, whereas in the “low pressure contact state”, both objects move freely in the horizontal plane. Easy.

  In step S25 (S25b), a positional shift (also referred to as “high-pressure contact-related shift”) that occurs with the formation of the high-pressure contact state for both objects is estimated based on the second positional shift. “High-pressure contact-induced shift” is a shift that occurs with the transition from the low-pressure contact state to the high-pressure contact state, and is caused by the above-described various situations (for example, the inclination between the head and the stage). Can occur. The “high-pressure contact-induced deviation” is estimated in the same manner as the “contact-induced deviation”. For example, in step S25b after the next time, the shift ("high-pressure contact-induced shift") that occurs with the state transition (transition from the low-pressure contact state to the high-pressure contact state) in the next step S21b is "contact-induced shift". It is estimated in the same manner as the above-described various methods. In the first step S25b, it is highly likely that the “high-pressure contact-induced shift” is a shift similar to the shift that occurs with the transition from the non-contact state to the high-pressure contact state. The “high-pressure contact-induced shift” is estimated using the contact-induced shift that has occurred in the contact (first contact).

  In step S27 (S27b), the third positional deviation is measured in the low-pressure contact state of both objects.

  In step S28 (S28b), it is determined whether or not the third positional deviation has achieved the target state. Here, a state in which a positional shift in the direction opposite to the “high-pressure contact-induced shift” occurs is set as the “target state”.

  In step S26 (S26b) subsequent to step S28 (S28b), in a low-pressure contact state (movable state) of both objects, the both objects are relatively moved in the horizontal plane to achieve the “target state”. Thus, the third positional deviation is adjusted.

  After steps S26, S27, S28 are executed as appropriate (repeated once or a plurality of times), if it is determined that the deviation between the target state and the current state falls within the third allowable range, step S21 (S21b )

  In step S21b after the second time (step S21b after the third positional deviation is adjusted), the two objects are moved (approached) relative to each other in the Z direction. And the state of both objects transitions from the low pressure contact state to the high pressure contact state. In other words, an operation (repressurization operation) is performed in which both objects are again changed to the high pressure contact state.

  And operation | movement of step S22 (S22b), S23 (S23b), etc. is performed again. Thereafter, similarly to the first embodiment, the operations of steps S24 (S24b) to S28 (S23b), S21 (S21b), and S22 (S22b) are repeatedly executed as necessary. In step S31 (S31b), a resin curing process is performed.

  The operation according to the modified example as described above may be performed.

1 Joining device (alignment device)
12, 41 Stage 22, 42 Head 23 Alignment table 28M, 28N Camera (imaging part)
28e, 28f Mirror 25 Rotation drive mechanism 26 Z-axis raising / lowering drive mechanism 28 Position recognition part 43 Protrusion mechanism 44 Pressing plate 45 Holding mechanism 45a Outer holding part 45b Inner holding part 91, 92 Substrate (object)
GAa, GAb Photographed image MK1a, MK1b, MK2a, MK2b Alignment mark

Claims (26)

An alignment method for aligning both objects of a first object and a second object,
a) In a state where the first object and the second object face each other in a non-contact manner, the first object is a positional deviation between the two objects and is a positional deviation in a plane perpendicular to a predetermined direction. Measuring the displacement, correcting the first displacement and aligning the objects; and
b) After the step a), the first object and the second object are relatively moved in the predetermined direction to bring the first object and the second object into contact with each other. Steps,
c) In a contact state between the first object and the second object, a second position shift that is a position shift between the two objects and is a position shift in a plane perpendicular to the predetermined direction. Measuring step;
d) after the second misalignment measuring operation, moving both the objects relative to each other in the predetermined direction to temporarily release the contact state between the objects;
e) estimating a contact-induced shift, which is a positional shift that occurs with the contact between the two objects, based on the second positional shift;
f) measuring a third positional shift that is a relative positional relationship between the two objects in a contact release state of the two objects and is a relative positional relationship in a plane perpendicular to the predetermined direction;
g) A state in which a positional deviation opposite to the contact-induced deviation occurs is set as a target state, and the two objects are relatively moved within a plane perpendicular to the predetermined direction in the contact release state of the two objects. And adjusting the third misalignment to achieve the target state;
h) after the step g), performing a re-contact operation for bringing the two objects into contact again;
An alignment method comprising: The alignment method according to claim 1,
Said step f)
f-1) reading a first mark attached to the first object and a second mark attached to the second object;
f-2) measuring the third misalignment based on the misalignment between the marks measured using both the first mark and the second mark;
An alignment method comprising: The alignment method according to claim 2,
In the step f-1), a photographed image including both the first mark attached to the first object and the second mark attached to the second object is obtained. Acquired,
In the step f-2), the positional deviation between the marks is measured based on the photographed image. The alignment method according to claim 2,
The step f-1)
f-1-1) reading the first mark attached to the first object;
f-1-2) reading the second mark attached to the second object at a timing different from the step f-1-1);
An alignment method comprising: The alignment method according to claim 1,
Said step f)
f-1) reading only one of the two marks, the first mark attached to the first object and the second mark attached to the second object;
f-2) a position measured before step f) with respect to the other of the two marks, and a position measured using the one mark read in step f-1) Obtaining a relative positional relationship between the two marks based on:
An alignment method comprising: In the alignment method in any one of Claims 1-5,
In step a), the two objects are aligned until the first misalignment falls within the first tolerance, and the residual misalignment within the first tolerance is determined,
In the step e), the contact-induced deviation is estimated based on the residual deviation that is the deviation remaining in the step b). The alignment method according to any one of claims 1 to 6,
Step c) is performed again after step h),
Steps d) to h) and step c) are repeatedly executed in this order until it is determined in step c) that the second misalignment falls within a second allowable range. Alignment method. The alignment method according to claim 7,
In the step e), the positional deviation in the non-contact state of the two objects finally detected in the step g) immediately before the step e) and the step c just before the step e). The alignment method is characterized in that the contact-induced deviation is estimated based on the second positional deviation obtained in (1). The alignment method according to any one of claims 1 to 8,
In the step e), the contact-induced deviation is estimated based on a plurality of past estimation results relating to the contact-induced deviation. The alignment method according to any one of claims 1 to 9,
Until the deviation from the target state falls within the third allowable range, the measurement operation of the third positional deviation in the step f) and the relative movement operation of the two objects in the step g) are repeated. An alignment method characterized by being performed. In the alignment method in any one of Claims 1-10,
Each of the first and second objects is a substrate;
In the step b), the alignment method is characterized in that at least one of the objects is bent and contact between the objects is started in a state where the objects are partially in contact with each other. The alignment method according to claim 11,
In the step f), the third position is in a state where the contact between the two objects is released, and in a state where the bending of the at least one object is returned and the at least one object is flat. An alignment method characterized in that a deviation is measured. An alignment apparatus for aligning both objects of a first object and a second object,
In a state where the first object and the second object are opposed to each other in a non-contact manner, the first object is a position error between the two objects and is a position error in a plane perpendicular to a predetermined direction. Measuring means for measuring,
Alignment means for correcting the first positional deviation and aligning the two objects;
After the correction of the first displacement, the first object and the second object are moved relative to each other in the predetermined direction by moving the first object and the second object. Relative movement means for contacting
With
In the state where the first object and the second object are in contact with each other by the relative movement operation by the relative movement means, the measuring means is a positional deviation between the two objects and is perpendicular to the predetermined direction. Measuring a second misalignment that is a misalignment in a plane;
The relative movement means, after the second misalignment measurement operation, moves both the objects relative to each other in the predetermined direction to temporarily release the contact state between the objects.
The alignment means estimates a contact-induced displacement that is a displacement that occurs with the contact of the two objects based on the second displacement;
The measuring means measures a third positional shift, which is a relative positional relationship between the two objects and is a relative positional relationship in a plane perpendicular to the predetermined direction in the contact release state of the two objects. ,
The alignment means sets, as a target state, a state in which a positional deviation in the direction opposite to the contact-induced deviation occurs, and both the objects in a plane perpendicular to the predetermined direction in the contact release state of the both objects. To adjust the third misalignment so as to achieve the target state,
The relative movement means performs a re-contact operation for bringing the objects into contact again after the contact state of the objects is once released and the objects are relatively moved. Alignment device. The alignment apparatus according to claim 13,
When measuring the third misalignment, the measuring means uses both the first mark attached to the first object and the second mark attached to the second object. Alignment that reads and measures the third misalignment based on the misalignment between the marks measured using both the first mark and the second mark. apparatus. The alignment apparatus according to claim 14, wherein
In the measurement of the third misalignment, the measuring means includes the first mark attached to the first object and the second mark attached to the second object. An alignment apparatus characterized in that a captured image including both marks is acquired and a positional deviation between the marks is measured based on the captured image. The alignment apparatus according to claim 14, wherein
The measuring means reads the first mark attached to the first object when measuring the third positional deviation, and at a timing different from the reading timing of the first mark, An alignment apparatus that reads the second mark attached to a second object. The alignment apparatus according to claim 13,
In the measurement of the third positional deviation, the measuring means
While reading only one of the two marks, the first mark attached to the first object and the second mark attached to the second object,
The relative position of both marks based on the position measured using the one mark read and the position measured before the third misalignment measurement time for the other of the marks. An alignment apparatus characterized by acquiring a positional relationship. The alignment apparatus according to any one of claims 13 to 17,
The alignment means aligns the objects until the first displacement is within a first allowable range;
The measuring means obtains a residual deviation within the first allowable range still remaining after the alignment;
The alignment device, wherein the alignment means estimates the contact-induced deviation based on the residual deviation. The alignment apparatus according to any one of claims 13 to 18,
Until the second misalignment falls within the second tolerance range,
A contact release operation for once releasing the contact state of both objects;
An operation of estimating the contact-induced deviation;
A measurement operation of the third misalignment;
An adjustment operation for adjusting the third misalignment so as to achieve the target state;
The re-contact operation for bringing the two objects into contact again;
A measurement operation of the second misalignment;
Is repeatedly executed. The alignment apparatus according to claim 19,
The alignment means includes a positional deviation in a non-contact state of the two objects finally detected in the third positional deviation adjustment operation immediately before the recontact operation, and a position immediately after the recontact operation. The alignment apparatus, wherein the contact-induced deviation is estimated based on the second positional deviation obtained in the second positional deviation measuring operation. The alignment apparatus according to any one of claims 13 to 20,
The alignment device estimates the contact-induced deviation based on a plurality of past estimation results related to the contact-induced deviation. The alignment apparatus according to any one of claims 13 to 21,
Until the deviation from the target state falls within a third allowable range, the measurement operation for the third positional deviation and the adjustment operation for adjusting the third positional deviation to achieve the target state are repeated. An alignment apparatus that is executed. The alignment apparatus according to any one of claims 13 to 22,
Each of the first and second objects is a substrate;
The relative movement means deflects at least one of the objects and starts the contact of the objects from a state where the objects are partially in contact with each other. The alignment apparatus according to claim 23.
In the state in which the contact between the two objects is released and the bending of the at least one object is returned and the at least one object is flat, the measuring means detects the third positional deviation. An alignment apparatus characterized by measuring. An alignment method for aligning both objects of a first object and a second object,
a) In a state where the first object and the second object face each other in a non-contact manner, the first object is a positional deviation between the two objects and is a positional deviation in a plane perpendicular to a predetermined direction. Measuring the displacement, correcting the first displacement and aligning the objects; and
b) After the step a), the first object and the second object are moved relative to each other in the predetermined direction so that the objects are in contact with each other and between the objects. Forming a high-pressure contact state in which a pressure equal to or greater than a certain value is applied to
c) a second positional shift that is a positional shift between the two objects in a high-pressure contact state between the first target object and the second target object and that is a positional shift in a plane perpendicular to the predetermined direction. Measuring steps,
d) After the second misalignment measurement operation, the two objects are moved relative to each other in the predetermined direction so that the objects are moved from the high-pressure contact state to each other between the objects. Transition to a low pressure contact state where a pressure smaller than a certain value is applied and the two objects are in contact with each other;
e) estimating a high-pressure contact-induced shift that is a positional shift that occurs in association with the formation of a high-pressure contact state for both objects based on the second positional shift;
f) measuring a third positional shift which is a relative positional relationship between the two objects in a low-pressure contact state between the two objects and which is a relative positional relationship in a plane perpendicular to the predetermined direction;
g) A state in which a positional shift in the opposite direction to the high-pressure contact-induced shift occurs is set as a target state, and the two objects are relative to each other in a plane perpendicular to the predetermined direction in the low-pressure contact state of the both objects. Adjusting the third displacement so as to achieve the target state,
h) after the step g), performing a re-pressurization operation for causing both the objects to transition to the high-pressure contact state again;
An alignment method comprising: An alignment apparatus for aligning both objects of a first object and a second object,
In a state where the first object and the second object are opposed to each other in a non-contact manner, the first object is a position error between the two objects and is a position error in a plane perpendicular to a predetermined direction. Measuring means for measuring,
Alignment means for correcting the first positional deviation and aligning the two objects;
After the correction of the first misalignment, the first object and the second object are relatively moved in the predetermined direction, the objects are in contact with each other, and the objects Relative movement means for forming a high-pressure contact state in which a pressure of a certain value or more is applied between them;
With
The measuring means is a displacement of both objects in a state where the high pressure contact state between the first object and the second object is formed by a relative movement operation by the relative movement means. Measuring a second positional deviation which is a positional deviation in a plane perpendicular to the predetermined direction,
The relative movement means moves the both objects relative to each other in the predetermined direction after the second misalignment measuring operation, so that the both objects are moved from the high-pressure contact state to the both objects. A pressure smaller than the predetermined value is applied between them to make a transition to a low-pressure contact state where the two objects are in contact with each other;
The alignment means estimates a high-pressure contact-induced shift, which is a positional shift that occurs with the formation of a high-pressure contact state regarding the two objects, based on the second positional shift,
The measuring means measures a third positional shift, which is a relative positional relationship between the two objects in a low-pressure contact state between the two objects and is a relative positional relationship in a plane perpendicular to the predetermined direction. ,
The alignment means sets, as a target state, a state in which a positional shift in a direction opposite to the high-pressure contact-induced shift occurs, and the two objects in a plane perpendicular to the predetermined direction in the low-pressure contact state of the two objects. Relatively moving an object to adjust the third misalignment to achieve the target state;
The relative movement unit performs a re-pressurization operation for causing both the objects to transition to the high-pressure contact state again after the third positional deviation is adjusted.

Images