JP7130720B2 - Method and apparatus for aligning substrates - Google Patents

Description

The present invention relates to a method for aligning and contacting a first substrate and a second substrate, as claimed in claim 1, and a corresponding device, as claimed in claim 10.

Aligners are used in the semiconductor industry to align substrates, in particular wafers, to each other so that they can be bonded together in subsequent process steps. This joining process is called bonding.

The alignment process of providing alignment marks on the substrate surfaces to be bonded is referred to as face-to-face alignment.

In order to be able to detect and/or determine the position and orientation of the alignment marks from the outer substrate surface, opposite the substrate surface to be bonded, the substrate is opaque to the electromagnetic beam used for measurement. In some cases, alignment marks are detected between the substrates using an image detection means before the substrates are brought together.

This had various drawbacks, notably particle contamination from the camera and the large spacing between the substrates due to the placement of the camera between the two substrates for detection. This results in misalignment when brought closer together.

An improvement in face-to-face registration is shown in the registration apparatus of publication US Pat. No. 6,214,692 (US6214692B1), which can be considered the closest prior art. In this alignment device, two optical element groups with two optical elements each facing each other are used to form a system with two reference points, with respect to which the substrates alternate. is positioned at These reference points are the intersections of the optical axes of the two optical elements facing each other.

The problem with this prior art is that the disclosed open-loop control scheme can no longer meet the ever-increasing alignment accuracy requirements.

The publication US Pat. No. 6,214,692 (US6214692B1) is based on the comparison and registration correction of two images of alignment marks. The orientations of the alignment marks on the two substrates arranged face-to-face are individually detected using a camera system. From the calculated relative orientations and relative positions of the alignment marks, the positioning table (substrate holder and stage) is driven and controlled, thereby correcting the erroneous positions. The positioning table has a guide sensor and an incremental displacement sensor located near the drive.

SUMMARY OF THE INVENTION Accordingly, it is an object of the present invention to provide an apparatus and method for aligning and contacting substrates that can align and contact the substrates more accurately and more efficiently.

This task is solved by the characterizing features of claims 1 and 10 . Advantageous developments of the invention are indicated in the dependent claims. Also included within the scope of the invention are all combinations of at least two of the features mentioned in the description, claims and/or drawings. When ranges of values are specified, values within the stated boundaries are also disclosed as boundary values and may be claimed in any combination.

The idea underlying the present invention is that prior to alignment, an additional (particularly a third) alignment layer is deposited either on one of the substrates to be aligned or on the substrate holder. Additional detection of marks. In particular, this additional alignment mark is not located on the contact surface of the substrate. Preferably, this additional alignment mark is arranged on the side of the substrate or substrate holder opposite to or arranged parallel to the contact surface.

Alignment of the substrates to each other takes place indirectly, in particular by means of a plurality of alignment marks provided on the contact surfaces of the substrates. The alignment marks on the facing surfaces of the facing substrates are in particular complementary to each other. The alignment marks may be any mutually alignable objects such as crosses, circles or squares or propeller-like formations or grating structures, in particular phase gratings for the spatial frequency domain.

The alignment marks are preferably detected using an electromagnetic beam of a predetermined wavelength and/or wavelength range. These currently include infrared beams, visible light or ultraviolet beams. However, it is equally possible to use shorter wavelength beams, such as EUV or X-ray beams.

One (particularly unique) aspect of the present invention is that alignment is performed using, among other things, the detection of additional alignment marks only. "Detection only" means that the other (particularly the first and second) alignment marks are not or cannot be detected during alignment.

According to an advantageous embodiment of the invention, the first substrate and the second substrate are arranged with a spacing A between the first contact surface and the second contact surface in the Z-direction such that the first and the second substrate holder. The distance A is in particular 500 micrometers or less, particularly preferably 100 micrometers or less, in the optimal case 50 micrometers or less, in the ideal case 10 micrometers or less.

The method according to the invention is particularly suitable for any electromagnetic beam, in particular UV light, more preferably infrared light, most preferably visible light and/or the first and second alignment marks. 1. A method of face-to-face alignment of at least two substrates by sequentially opening the optical path to observe the substrate, wherein at least one substrate is positioned to accurately reproduce the substrate position and the substrate holder position. It is specified to supplement one additional optical path, which is not located or extends between the substrates when the substrates are aligned, and which can be detected externally. become.

The method according to the invention notably provides additional XY position information and/or additional XY position information detected by additionally mounted detection units and/or measurement systems and closed-loop control systems and used for open-loop control of the alignment. Positioning accuracy is improved by using directional information.

For this purpose, the device according to the invention has in particular a software-supported control unit with which the steps and components described here are controlled. In the present invention closed control loop and closed loop control shall be understood to be included in this control unit.

X-direction and Y-direction and X-position and Y-position are understood as directions extending or located in the XY coordinate system or in any Z plane of the XY coordinate system. The Z direction is arranged orthogonally to the XY directions.

The X and Y directions particularly correspond to the lateral direction.

The position features are derived/calculated from the position and/or orientation values of the alignment marks on the substrate and from the alignment marks on the substrate holder.

According to the invention, in particular, at least one additional position feature is detected by at least one additional measurement system with a new additional optical path. At least one additional alignment mark is preferably provided in the vicinity of the alignment mark on the substrate.

The method according to the invention and the device according to the invention thus have, in particular, at least one additional measuring system and/or a closed-loop control system, in which the additional measured values also determine the number of further detection units. Alignment accuracy is improved by correlation with one of a plurality of measurements of . Observing the alignment mark directly by correlating at least one of the plurality of alignment marks measured at the bonding interface between the contact surfaces with the alignment mark that is also visible when aligning the substrate. , thus allowing real-time measurement and closed-loop control during alignment. This improves the alignment accuracy of the substrate.

Additional alignment marks are arranged in particular on the back side of the substrate holder.

Between the additionally added positional features on the substrate holder and the positional features on the substrate, in particular a unique correlation is formed which preferably does not change until and during the alignment. .

The positional features additionally attached to the substrate holder that are uniquely correlated with the positional features on the substrate allow direct observation of the positional features on the substrate to be replaced by direct observation of the positional features of the substrate holder. The advantage of this is that the observable portion of the substrate holder can be placed substantially always within the field of view or measurement area of the additional measurement system.

Active feedback of data for positioning and position correction improves accuracy over open-loop controlled positioning in the prior art. This is because in a closed control loop the actual state of the position can be controlled. In this way, in particular, instead of traveling a pre-determined segment with a certain number of increments, the segment already traveled is measured, whereby the distance between setpoint and actual values, in particular for speed and/or acceleration, is measured. A comparison can be made. A goal of the present invention is to improve the accuracy of the alignment of two substrates, especially using a method that allows the alignment to be observed in real time from outside the bonding interface. Here, the substrates are arranged in particular at a minimum distance from each other, and there are preferably no device objects between the substrates.

The core idea of the invention is, inter alia, to introduce at least one additional measurement system and a closed-loop control system and to combine existing measured values with newly detected additional position and/or orientation values. is to connect An alignment mark measured at the bonding interface and an alignment mark, in particular directly detectable during alignment, on the substrate holder, in particular on its back surface and/or on the substrate, in particular on its back surface. By correlating , additional alignment marks can be directly observed, allowing real-time measurement and real-time closed-loop control. By such means the alignment accuracy is improved.

During detection of the first alignment mark, the second substrate and the second substrate holder are moved out of the optical axis of the first detection unit in the XY directions, thereby allowing detection. .

During detection of the second alignment mark, the first substrate and the first substrate holder are moved out of the optical axis of the second detection unit in the XY directions, thereby allowing detection. .

Apparatus The inventive apparatus for aligning at least two substrates has at least one optical system, in particular consisting of two mutually aligned optical elements or detection units, the optical paths of which preferably meet at a common focal point.

This optical system comprises, according to an advantageous embodiment, beam shaping and/or deflecting elements such as mirrors, lenses, prisms, a light source, in particular Koehler illumination, and a camera (CMOS sensor or CCD or phototransistor). image detection means such as plane or line or point detection means), movement means for focusing and evaluation means for closed-loop control of the optical system.

In one embodiment according to the invention of this apparatus, an optical and rotation system for substrate positioning is used according to the principle of turnover alignment. See Friedrich Hansen, Justierung, VEB Verlag Technik, 1964, section 6.2.4, Umschlag method. Here, at least one measurement in a given position and at least one measurement in a turned-over position rotated 180° and oriented in the opposite direction are performed. The measurement results obtained in this way are particularly free of eccentricity errors.

One development of the device according to the invention comprises two optical systems, in particular identical, of identical construction, with optical elements that are aligned with each other and can be fixed relative to each other.

One development according to the invention of this device includes more than two identical optical systems with aligned optical elements.

Further, the apparatus according to the invention includes a substrate holder for receiving the substrate to be aligned.

In another embodiment of the device according to the invention, at least one substrate holder that is at least partially transmissive, preferably more than 95% transmissive, in particular for simultaneous observation of two substrate surfaces at a given point. to use.

In a further embodiment of the device according to the invention, at least one substrate holder with a plurality of openings and/or through-holes and/or inspection windows at predetermined points, in particular for simultaneous observation of two substrate surfaces. to use.

Furthermore, the apparatus according to the invention may include a system for forming prebonds.

Furthermore, the device according to the invention preferably comprises a plurality of actuators with drive, guide, clamping and measuring systems for moving, positioning and aligning the optics and substrate holders and/or substrates with each other. mobile devices are included.

These movers can form arbitrary moves as a result of individual moves, whereby these movers preferably operate precisely with fast coarse positioning devices that do not meet accuracy requirements. and a precision positioning device for

The target value of the position to be reached is the ideal value. Mobile devices approach this ideal value. Reaching a predetermined range around the ideal value can be understood as reaching the target value.

A coarse positioning device is one in which the reachability and/or repeatability is 0.1% from the target value, based on the entire travel path or a rotation range in which 360° is one rotation for a rotatable rotary drive. Above, it is understood that the positioning device deviates preferably by 0.05% or more, particularly preferably by 0.01% or more.

Thus, for example, in a coarse positioning device with a travel path of 600 mm or more, a reaching accuracy of 600 mm×0.01%, ie 60 micrometers or more, results in a residual uncertainty.

In another embodiment of coarse positioning, the residual uncertainty in the reaching accuracy or repeatability is 100 micrometers or less, preferably 50 micrometers or less, particularly preferably 10 micrometers or less. Thermal disturbance variables should preferably also be taken into account here.

A coarse positioner performs positioning with sufficient accuracy only if there is a deviation in the processing range of the corresponding fine positioner between the actual position actually reached and the desired value for this position. Fulfill.

An alternative coarse positioning device is sufficient only if there is a deviation of half the processing range of the corresponding fine positioning device between the actual position actually reached and the target value for this position. Plays the role of positioning with precision.

A precision positioning device is one in which the residual uncertainty in reaching accuracy and/or repeatability is less than 500 ppb, preferably less than 100 ppb, ideally less than 100 ppb from the target value, based on the overall travel path or turning range. is understood to be a positioning device in which case it does not exceed 1 ppb.

Advantageously, the precise positioning device according to the invention has an ideal In typical cases, it has an absolute positioning error of less than 1 nm.

The device and corresponding method of the invention have at least two positioning devices with maximum accuracy and reproducibility. It is possible to use the concept of mutual error correction for the quality of substrate alignment. Thus, when the misalignment (twist and/or misalignment) of a substrate and its corresponding positioner is known, the correction values or vectors are used to align and correct the position of another positioner and another substrate. Thereby, positioning accuracy can be improved. How coarse and fine positioning or only coarse or only fine positioning is used for error correction by open-loop or closed-loop control in this case depends on torsion and /or a matter of magnitude and type of deviation.

In the following text, positioning devices (coarse positioning devices or fine positioning devices or combined positioning devices) and registration means are used synonymously.

Mutual alignment of multiple substrates can be performed in all six degrees of freedom of movement according to the present invention. That is, it can be performed in three translational movements according to the coordinate directions x, y and z and in three rotational movements about the coordinate axis directions. The present invention allows movement in any direction and orientation. Substrate alignment preferably includes wedge error compensation, particularly passive or active, according to the disclosure of publication EP2612109B1.

A robot for substrate transport is included in the moving device. The fixed part may be integrally or functionally integrated with the movement device.

Furthermore, the device according to the invention is preferably a closed-loop control system and/or for carrying out the described steps, in particular the movement flow, for making corrections, for analyzing and storing the operating state of the device according to the invention. Or includes a rating system, especially a calculator.

The process is preferably written as a protocol and executed in a machine-readable form. A protocol is a set of optimal values for functionally or process-technically relevant parameters. Reproducibility of the manufacturing flow can be guaranteed by using the procedure manual.

Furthermore, the device according to the invention, according to an advantageous embodiment, comprises a power supply system and an auxiliary and/or auxiliary system (compressed air, vacuum, electrical energy, liquids such as hydraulic pressure, refrigerants, heat transfer medium, temperature means and/or devices for stabilization, electromagnetic shielding).

In addition, the device according to the invention includes a frame, a sheath, active or passive subsystems that dampen or dampen or extinguish vibrations.

Measuring Furthermore, the device according to the invention comprises at least one measuring system, which preferably comprises a measuring unit for each axis of motion, which can be implemented in particular as a displacement measuring system and/or an angle measuring system.

It is also possible to use tactile or non-tactile measurement methods. A measurement standard scale, a unit of measurement, may be a physical object, in particular a reference, or implicit in the measurement method, such as the wavelength of the beam used.

At least one measurement system can be selected and used to reach the alignment accuracy. A measuring system implements a measuring method. Here, in particular
inductive method, and/or capacitive method, and/or resistive method, and/or comparison method, in particular optical image recognition method, and/or (especially glass reference scale or interferometer as reference device, in particular Incremental or absolute value methods (using laser interferometers or electromagnetic reference scales), and/or transit time measurements (Doppler method, time-of-flight method) or other time detection methods, and/or trigonometry, especially laser triangulation,
• Autofocus methods, and/or • Intensity measurement methods, such as fiber optic rangefinders, can be used.

Additional Measurement System with Modifications, Substrate Holder Furthermore, one particularly preferred embodiment of the device according to the invention provides an XY measurement of one of the substrates and/or one of the substrate holders. It comprises at least one additional measuring system for detecting the position and/or alignment direction and/or angular position with respect to a predetermined reference, in particular with respect to the frame. As a frame, in particular a part made of hard natural ore or mineral cast or nodular graphite cast or hydraulic concrete can be provided, which frame in particular is vibration-damping and/or vibration-insulating. and/or configured to dampen vibrations.

In the present invention, the measurements can be combined and/or cross-referenced and/or correlated with one another, whereby measuring one alignment mark results in another position in relation to it. The position of the alignment marks can be estimated.

In one embodiment according to the invention, the position of the substrate holder with respect to the fiducial, in particular with respect to the first alignment mark of the first substrate and/or the second alignment mark of the second substrate, is one Detected at points (or locations or measurement spots or fields of view).

In another embodiment according to the invention the position of the substrate holder is detected at exactly two points relative to the reference.

In another embodiment according to the invention, the position of the substrate holder is detected at exactly three points relative to the reference, thereby identifying the position and orientation of the substrate holder.

For localization at one point or two points or three points or any number of points, a first embodiment of the invention provides optical pattern recognition using a camera system and Patterns can be used. The pattern is detected in a real-time system, especially continuously during registration.

For localization, the cited measurement methods can likewise be used.

The invention contemplates the reverse, in particular by attaching the detection unit to the substrate holder and the alignment mark to the frame.

In order to be able to perform detection, evaluation and open-loop control at any point in time, in particular permanently, in an advantageous embodiment the closed-loop control unit (and/or the open-loop control unit) has, in particular continuously For the purpose of providing measurements, the pattern is distributed over an area larger than the field of view of the image detection unit of the detection unit.

In another embodiment according to the invention of the apparatus, at least one of the substrate holders has a through opening for detecting the alignment mark from the mounting surface of the substrate.

Advantageously according to the invention, at least one of the substrates has at least one alignment mark on the surface to be bonded (contact surface) and at least one alignment mark on the opposite surface (mounting surface). This is the case with one alignment mark.

These alignment marks are preferably a plurality of preferably evenly distributed alignment marks that can be associated particularly uniquely on the mounting surface of at least one of the substrates. An alignment mark on the mounting surface can be correlated, at least in its XY position and/or alignment direction, to the XY position and/or alignment direction of an alignment mark on the same substrate. This allows for continuous positioning during substrate loading and alignment until the substrates are in contact.

In one development of the invention, the alignment marks on the bearing surface are distributed particularly evenly, except for the edge zone (edge exclusion zone) of the substrate.

In another embodiment of at least one of the plurality of substrates, the plurality of alignment marks of the mounting surface are uniquely distributed with respect to the contact surface and alignment marks arranged on this contact surface of the same substrate. be done.

For determining the XY position in at least one point, in another embodiment according to the invention, to determine the XY position and/or to determine the orientation of the substrate holder, correspondingly: In particular, at least one interferometer with a monolithically constructed reflector can be used. The number of interferometers is in particular the same as the number of reflecting surfaces of the reflector.

A substrate holder, especially formed from a monolithic block, preferably has at least two of the following functions. i.e.
・Fixation of substrates using vacuum (vacuum tracks, joints)
- preferably mechanical and/or hydraulic and/or piezoelectric and/or pyroelectric and as disclosed in particular in EP2656378B1, WO2014191033A1 / or shape compensation for deforming the substrate using electrothermal control elements (measuring standards, reflecting surfaces and/or prisms, in particular reflectors for interferometers, dragonfly and/or dragonfly fields, two-dimensionally formed datums, three-dimensional datums, in particular stages) for position determination and/or orientation determination and movement (guide tracks)
have

The displacement device according to the invention, which is not used for fine adjustment, is configured in particular as a robot system, preferably with incremental displacement sensors. The accuracy of this movement device for the auxiliary movement is decoupled from the accuracy of alignment of the substrate stack, whereby this auxiliary movement is less than 1 mm, preferably less than 500 micrometers, in particular It is preferably performed with a low repeatability of 150 micrometers or less.

The open-loop control and/or closed-loop control of the displacement device according to the invention for (lateral) alignment (fine adjustment) is particularly useful in the XY position and/or alignment direction detected by further measuring means. done based on The accuracy of this displacement device is preferably ≤200 nm, preferably ≤100 nm, particularly preferably ≤50 nm, very particularly preferably ≤20 nm, optimal case ≤10 nm, ideal case ≤1 nm.

Method A first embodiment of the alignment method according to the invention comprises the following, in particular at least partially sequential and/or simultaneous steps, in particular in the following order.

First processing step: loading the first/lower substrate face down into the first/lower substrate holder. A first alignment mark is provided (exists) on the facing surface (contact surface).

2nd process step: using a movement device for a particularly coarse adjustment, the first/lower substrate is brought into the field of view of the detection position of the first/upper detection unit of the optical system by the substrate holder. move with.

Third processing step: If the optics have not already been processed by the second processing step, move them to the detection position. Optionally self-calibration of the optical system can be done already at this stage or before starting the movement to the detection position.

Fourth processing step: Focusing the first/upper detection unit on the discriminative pattern of first alignment marks located on the contact surface to be bonded of the substrate.

Optionally, the optics may align the second/lower detection unit to the focal plane of the first/upper detection unit.

Fifth processing step: Detecting the first alignment mark, especially using pattern recognition. At the same time, in particular by synchronizing with the first detection unit, the additional measuring system according to the invention (with a third detection unit) allows the detection of the first substrate, preferably in a plane other than the contact plane. Detecting the XY position and/or alignment orientation of the holder and/or the first substrate.

Sixth processing step: Clamping the position of the optical system, especially mechanically and/or electronically and/or magnetically, especially by zeroing out all degrees of freedom.

Seventh processing step: alignment of the second/lower detection unit to the focal plane of the first/upper detection unit by means of the optical system.

Optionally, this alignment is omitted, performed subsequent to the fourth processing step.

Eighth processing step: Abstraction and/or calculation by a closed-loop control calculator (especially in the open-loop control unit) and an evaluation calculator to obtain measurement results for open-loop control of the alignment. That is, the closed-loop control calculator and the evaluation calculator inter alia form a correlation of the measurement results of the optical system and the (third) additional measurement system, in particular the first/lower substrate holder and the first/ Store this result as a target value for the XY position and/or alignment direction of the underlying substrate.

Ninth process step: moving the first/lower substrate holder out of the field of view of the optics (beam path for detection). This prevents the light path to the first detection unit from being blocked. The optics preferably subsequently remain stationary.

Tenth processing step: The second/upper substrate is loaded into the second/upper substrate holder. This processing step can precede one of the preceding processing steps.

Eleventh process step: moving the second/upper substrate holder together with the second/upper substrate into the field of view of the optical system.

Twelfth processing step: seek and detect the alignment marks on the second/upper substrate by the second/lower detection unit of the optical system, similar to the first/lower detection unit. The optical system does not move mechanically in this case, but a focus correction is conceivable. Preferably, however, no focusing movement is performed either.

13th processing step: blocking the optical path of the first/upper detection unit by the second/upper substrate and the second/upper substrate holder, so that in the aligned position, the first/upper The contact surfaces to be bonded of the underlying substrate are no longer directly observable/detectable.

Fourteenth processing step: Determining the misregistration by means of the closed-loop control calculator and the evaluation calculator. See the disclosure of publication US Pat. Specifically, a registration error vector is created from the registration errors. Following this, in particular at least one correction vector is calculated. The correction vector may be a vector parallel and opposite to the misregistration vector, such that the sum of the misregistration vector and the correction vector is zero. In special cases, another parameter can be considered in the calculation of the correction vector, which makes the result different from zero.

Fifteenth processing step: The XY position and/or alignment direction of the second/upper substrate (including the substrate holder) is set according to the correction vector and subsequently clamped, whereby the calculated position Misregistration is at least minimized and preferably eliminated.

16th processing step: After clamping/fixing the second/upper substrate, detect/check the XY position of the upper second substrate again by the second/lower detection unit . This allows any shear and/or twist caused by the clamping to be identified and minimized, especially eliminated, by repeating the 12th to 15th processing steps.

Alternatively, detected deviations and/or torsions can be taken into account to create correction values and/or correction vectors, which can be taken into account in subsequent processing steps to remove them. . The correction value for deviations is preferably 5 μm or less, preferably 1 μm or less, particularly preferably 100 nm or less, very particularly preferably 10 nm or less, in the optimum case 5 nm or less, ideally In the case it is less than 1 nanometer. The correction value for torsion is in particular less than or equal to 50 microradians, preferably less than or equal to 10 microradians, particularly preferably less than or equal to 5 microradians, very particularly preferably less than or equal to 1 microradians, in the optimum case less than or equal to 0.1 microradians, ideal less than 0.05 microradians in the case.

In the first five processing steps, as in the first two processing steps, at least one measurement of the position before fixing of the upper substrate and at least one measurement of the position after fixing of the upper substrate. can be done. According to the invention, it is possible to consider the correction of repeated errors, whereby the next processing step is executed only when a given accuracy requirement (threshold) is reached.

First 6 process steps: return the first/lower substrate holder to the already detected position and orientation of the target value (see eighth process step). In this case, the XY position and alignment direction of the lower substrate holder (together with the substrate) are controlled, especially in real time, by an additional measurement system. The first alignment mark of the first substrate is not visible because the optical path is blocked by the clamped upper substrate holder.

First 7 processing steps: The actual XY position and alignment direction of the first/lower substrate holder is modified until the difference with the target value is zero, but at least above a predetermined threshold. It will be fixed until it no longer exists.

Optional first eight processing steps: bonding the substrates. See the disclosure of publication WO 2014/191033 (WO2014191033A1).

First 9 processing steps: Unloading the substrate stack from the apparatus.

Included in an alternative second embodiment of the alignment method according to the invention are the following changes to the first embodiment of the loading order for the upper and lower substrates.

As a first processing step according to the invention, a first/lower substrate is loaded into the first/lower substrate holder.

As a second processing step according to the invention, the second/upper substrate is loaded into the second/upper substrate holder.

Subsequently, the processing steps of the first embodiment are similarly applied.

In a third embodiment of the method according to the invention, the first method and/or the second method are modified so that both in the device and in the method according to the invention the up and down directions are interchanged. . Therefore, the upper substrate holder in particular is observed by at least one additional measuring system.

A fourth embodiment of the method according to the invention modifies the described process, thereby changing the order of substrate loading while the process result remains the same.

In a fifth embodiment of the method according to the invention, the speed is increased by parallelizing the processing steps, in particular the loading of the second substrate is already performed during the pattern recognition step on the first substrate.

In a sixth embodiment of the method according to the invention and a corresponding apparatus, the position and/or orientation of both the upper substrate holder and the lower substrate holder and/or the upper substrate are determined by an additional measuring system. and/or the underlying substrate can be detected.

In a seventh embodiment of the method according to the invention and a corresponding device, an additional measuring system makes it possible to detect the position and/or orientation of the upper substrate holder and/or the upper substrate.

All technically conceivable combinations and/or permutations and multiplexing of functional and/or material parts of the apparatus and consequent modifications may be incorporated into at least one of the multiple processing steps or methods. deemed to be disclosed.

Where an apparatus feature is disclosed herein and/or in the description of the figures that follow, this feature is also disclosed and valid as a method feature, and vice versa.

Further advantages, features and details of the invention result from the following description of preferred embodiments and can be obtained from the drawing.

1 is a schematic cross-sectional view of one embodiment of a device according to the invention; FIG. 2 is a schematic enlarged cross-sectional view of the embodiment shown in FIG. 1 at a first processing step; FIG. 2 is a schematic enlarged cross-sectional view of the embodiment shown in FIG. 1 in a second processing step; FIG. 2 is a schematic enlarged cross-sectional view of the embodiment shown in FIG. 1 in a third processing step; FIG.

In the drawings, according to embodiments of the invention, advantages and features of the invention are indicated by reference numerals that respectively identify them, and components or features of the same function or function having the same action are indicated by the same reference numerals. is indicated.

FIG. 1 depicts a schematic and not-to-scale functional view of the main components of a first embodiment of an alignment device 1 . Alignment device 1 includes substrates 16 (referred to as first substrate and/or lower substrate), substrates 17 (referred to as second substrate and/or upper substrate) which are not inscribed in FIG. ) and/or the substrate stacks can be aligned with each other and at least partially bonded to each other and/or temporarily bonded (so-called pre-bonding). The possible movements/degrees of freedom of the functional components described below are symbolically depicted with arrows.

A first substrate 16 can be loaded onto the first substrate holder 10 and secured to the first substrate holder 10 without leaving any degrees of freedom of the substrate 16 relative to the first substrate holder 10 . A second substrate 17 can be loaded onto the second substrate holder 13 and secured to the second substrate holder 13 without leaving any degrees of freedom of the substrate 16 relative to the second substrate holder 13 .

In particular, the lower first substrate holder 10 is arranged on a first displacement device 11 for holding the first substrate holder 10 and for performing feeding and alignment movements (alignment).

In particular, the upper second substrate holder 13 is arranged on a second displacement device 14 for holding the second substrate holder 13 and for performing feed and alignment movements (alignment). The displacement devices 11, 14 are fixed to a particularly common rigid base or frame 9 in order to reduce/minimize vibrations of all functional members.

In order to observe (detect) the first alignment mark 20 and the second alignment mark 21, the optical system 2 consists of the following. i.e.
- a first detection unit 3, in particular an image detection means, for detecting the first alignment marks 20 of the first substrate 16;
a second detection unit 4, in particular an image detection means, for detecting the second alignment marks of the second substrate 17;

When the first substrate 16 and the second substrate 17 are arranged for alignment, the optical system 2 is arranged between the first substrate 16 and the second substrate 17, preferably Focusable to a common focal plane 12 . Movement of the optical system 2, in particular in the X, Y and Z directions, takes place using a positioning device 5 which positions the optical system 2. FIG. The positioning device 5 is fixed to a particularly sturdy base or frame.

In one particularly preferred embodiment, not shown, the overall optical system 2 (comprising positioning means 5, detection units 3 and 4, etc.) is used in a double mirror symmetrical embodiment.

The at least one additional, in particular optical, measuring system 6 with at least one third detection unit 7 detects the third alignment marks 22 to improve the alignment accuracy according to the invention. used to let Additional measuring system movements are performed by the positioning device 8 .

If the measuring device is an optical measuring system 6, the positioning device 8 can focus on the third alignment mark 22 by moving the third detection unit 7 in the Z direction. Positioning in the XY direction is also conceivable, wherein the measuring system 6 is preferably fixed to a table/frame during positioning. Alternatively, the exact XY position of measurement system 6 must be known.

In the illustrated embodiment according to the invention of the alignment device 1, the additional measuring system 6 allows the XY position and/or orientation (especially also the rotational orientation) of the lower substrate holder 10 to be determined with particularly high accuracy. detected by

In a further embodiment according to the invention, not shown, of the alignment device 1, the position and/or orientation of the upper substrate holder, in particular the position and/or the orientation, of the upper substrate holder is detected with particularly high accuracy by means of an additional measuring system.

In a further embodiment according to the invention, not shown, of the alignment device 1, an additional measuring system detects particularly the position and/or orientation of the upper and lower substrate holders with particularly high accuracy. .

In a further embodiment according to the invention, not shown, of the alignment device, the at least one measuring system detects in particular the position and/or orientation of at least one of the two substrates with particularly high accuracy. For this purpose, the alignment marks on the contact surface and the marks on the surface facing away from the contact surface are detected.

FIG. 2a shows a schematic representation of the processing steps according to the invention for correlating the measurements of at least two detection units, in particular the different side surfaces of the first substrate 16 and/or the first substrate holder 10. There is shown a schematic diagram of the processing steps according to the invention for correlating the measurements of two detection units respectively detecting at least two XY positions of alignment marks 20, 22 arranged in .

A first/lower substrate 16 is fixed to the first fixing surface 10 a of the first substrate holder 10 . The fixation may be due in particular to mechanical and/or electrostatic clamping or to pressure differences between the normal atmospheric ambient environment and the negative pressure on the first substrate holder 10, also referred to as vacuum clamping. is used. This fixation is made in particular such that the first substrate 16 does not undergo parasitic or undesired movement relative to the first substrate holder 10 during the overall method according to the invention. , where the first substrate holder 10 and the first substrate 16 each have corresponding coefficients of thermal expansion, preferably linearly corresponding progressions, the difference in coefficients of thermal expansion and/or The thermal expansion can be prevented or reduced in particular if the linear progression of the coefficients of thermal expansion preferably differs by 5% or less, preferably 3% or less, particularly preferably 1% or less.

The apparatus described above preferably has a temperature variation during the alignment cycle of 0.5 Kelvin or less, preferably 0.1 Kelvin or less, particularly preferably 0.05 Kelvin or less, in the optimal case 0.01 Kelvin or less. It is operated in a sustainable temperature-stabilized ambient environment, especially in clean rooms.

The first substrate 16 and the first substrate holder 10 , which are immovably fixed relative to each other, can be understood as a quasi-monolithic body for carrying out the movement of the first substrate 16 .

This fixing of the substrate can be effected by form-fitting and/or preferably by force-fitting. Due to the quasi-monolithic bonding, a plurality of influences that can lead to misalignment and/or torsion and/or deformation between the substrate holder and the substrate are at least reduced, preferably by at least an order of magnitude, particularly preferably removed.

The above multiple effects can be thermal and/or mechanical and/or flow induced and/or material (particles).

The substrate is fixed to the substrate holder by means of form-fitting or force-fitting, which makes it possible in particular to suppress differential thermal expansion. Further, the substrate holder can reduce, eliminate and/or compensate for deformation of the substrate itself. See EP 2656378 B1 for this.

During detection of the first alignment mark 20 the first substrate holder 10 is in the optical path of the (upper) first detection unit 3 . A first alignment mark 20 is arranged on the contact surface 16i of the first substrate 16 to be bonded, in the field of view of the (upper) first detection unit 3, in particular in the optical path. The first detection unit 3 forms an image, in particular a digital image, which is schematically shown here as an alignment mark in the form of a cross. The first alignment mark 20 can also consist of a plurality of alignment marks. From the images of the alignment marks, measurements are formed/calculated that characterize, inter alia, the XY position and/or the alignment direction (especially in the direction of rotation about the Z direction) of the first substrate 16, ie the alignment state. be.

During detection of the first alignment mark 20 the second/lower detection unit 4 preferably supplies no measurements. This is because in particular the second substrate 17 is/was arranged outside the optical path of the first detection unit 3 .

The first/lower substrate holder 10 and/or the first/lower substrate 16 have a third alignment mark 22 and based on the third alignment mark 22 the substrate holder 10 and/or the XY position and/or the alignment direction (especially in the direction of rotation about the Z direction), ie the alignment state, of the first substrate (16) is detected, in particular from another direction, preferably The first detection is detected diametrically from the opposite direction in the Z direction.

Preferably the relative movement of the first detection unit 3 with respect to the third detection unit 7 is measurable, more preferably from the detection of the first and third positioning marks 20 and 22: No relative movement takes place between the first detection unit 3 and the third detection unit 7 until the first substrate 16 and the second substrate 17 come into contact.

An (additional) third detection unit 7 of the additional measurement system 6 determines from the measurement of the third alignment marks 22 of the first/lower substrate holder 10 the first/lower substrate XY position and/or orientation measurements of the holder 10 are provided. This measured value is formed in particular from a preferably digital image which is symbolically drawn as a cross. The third alignment mark 22 can also consist of multiple alignment marks.

Two measurements (the XY position and/or alignment direction of the first substrate 16 and the XY position and/or alignment mark 22 of the first substrate holder 10 or the first substrate 16) or alignment direction) are associated and/or correlated with each other such that the XY position and/or alignment direction of the first alignment mark 20 is always this XY position and/or alignment direction. The third alignment mark can be uniquely associated with/or based on the alignment direction, in particular by detecting the XY position and/or the alignment direction.

This processing step provides XY alignment of first alignment mark 20 and/or second alignment mark 21 while aligning and/or contacting first substrate 16 and second substrate 17 . Alignment is achieved without direct detection of position and/or alignment direction. Additionally, the spacing between substrates can be minimized during alignment. This spacing can preferably correspond to the spacing of the substrates already during detection of the first and second alignment marks.

In other words, in particular between the substrate holder 10 and the additional measurement system 6, an unobstructed optical path is obtained, with which alignment of the substrate can be performed in an open control loop, or Substrate alignment is performed. This processing step adjusts the XY position and/or alignment direction of the first substrate holder 10 and, in turn, the XY position and/or alignment direction of the first substrate secured to the first substrate holder 10. can be accurately identified and reproducibly reconstructed.

In particular, restoration of the XY position and/or alignment direction of the substrate holder 10, and in particular restoration of the XY position and/or alignment direction of the substrate monolithically bonded thereto, is a particularly unique core one. It is one aspect. The processing steps for this have already been discussed elsewhere.

In particular, the positioning repeatability (measured as the relative misalignment between the two substrates), also known as flip play, is less than or equal to 1 micrometer, preferably less than or equal to 100 nm, particularly preferably less than or equal to 30 nm. , very particularly preferably amounts to 10 nm or less, in the optimum case to 5 nm or less and in the ideal case to 1 nm or less. This inversion play may be repeated reaching of a predetermined position with the movement devices 11,14 and/or 5,8. This reversal play results from movement of the moving device that has changed only the sensed position, thereby causing the measured quantity to be in relative misalignment. Positioning accuracy of the mover that does not affect the substrate can be considered as irrelevant flip play.

In order to further improve the positioning accuracy according to the invention, it is preferred to operate the first detection unit 3 and the third detection unit 7 synchronously in time, in particular 10 min. 1 second or less, preferably 1 millisecond or less, particularly preferably 10 microseconds or less, very particularly preferably 1 microsecond or less, optimal case 1 ns or less, ideal case 0.0 ns or less. It is to operate with a time difference. This is particularly advantageous. This is because disturbing influences such as mechanical vibrations can be eliminated. Mechanical vibrations propagate in materials, especially solid-borne sound at thousands of m/s. If the closed-loop control and detection means operate faster than the propagation velocity of solid-borne sound, the disturbance is reduced or eliminated.

A disturbance causes the orientation of the first substrate 16 on the first substrate holder 10 to change, which causes the first detection unit 3, which has already recorded the measured values, to be additionally provided with the detection unit 7. If the correct measurement system 6 has not yet recorded measurements, this disturbance can be part of the reduced alignment accuracy. This is because, in the time between the recording of the measurements of the detection means 3 and 7, fast position changes, especially due to vibrations, can occur on the order of nanometers or micrometers. If the recording of the measurements is delayed in time (on the order of seconds or minutes), other interfering influences such as temperature-induced shape or length changes can likewise affect the alignment accuracy. can be lowered.

If the first detection unit 3 and the third detection unit 7 are synchronized with each other (especially by simultaneous triggering of detection and adjustment of the detection time and/or the same integration time for the camera system), some Disturbing effects can be reduced and, in optimal cases, eliminated. This is because the detection is performed at a time when the disturbing influence has the least possible impact on the detection accuracy.

In one possible embodiment of the device, it is possible to perform the detection especially synchronously to the extremes of the vibration, especially when the periodic disturbing influences are known. For this purpose, the interfering influences are preferably recorded in advance by means of vibration sensors (accelerometers, interferometers, vibrometers) at locations of the device that are relevant to the accuracy and are processed in order to eliminate them, in particular in a computing unit, a computer. be able to. In another embodiment, multiple vibration sensors can be fixedly incorporated into the device at distinctive locations.

In another possible embodiment, the device is in particular an active and/or passive vibration damper, an active and/or passive vibration quencher and/or an active and/or passive vibration Combinations of insulators can be implemented and used, and these can also be implemented and used in cascade. Additionally, it is possible to superimpose the vibration as a disturbing effect with the forced vibration, so that the detection of the alignment mark can be performed in a so-called lock-in manner. Modal analysis and/or FEM can be used to characterize the device and to check the vibration conditions. This as well as the design of such devices are known to those skilled in the art.

Clamping the first detection unit 3 and the third detection unit 7 before, during or after detecting the first alignment mark 20 and the third alignment mark 22, at least absolutely and /or reduce the degrees of freedom in the relative X and Y directions to zero. Relative refers to the movement of the first detection unit 3 with respect to the third detection unit 7 .

FIG. 2b is a schematic representation of the processing steps according to the invention for measuring the second/upper substrate 17. FIG. The optical system 2 is in particular already in the clamped state and the image and/or the XY position of at least one alignment mark of the first/lower substrate 16 is stored (not shown). ). In the clamped state, at least the degrees of freedom in the X and Y directions, at least absolute and/or relative to the first detection unit, of the second detection unit are reduced to zero.

In the clamped state, the XY positions and/or alignment directions of the first, second and third alignment marks can be related to the same XY coordinate system. Alternatively or additionally, the first, second and third detection units are calibrated in the same XY coordinate system.

The second/upper substrate 17 fixed to the second/upper substrate holder 13 is moved to the detection position by the second moving device 14 that moves the second substrate holder 13, and the second/lower substrate holder 13 is moved to the detection position. By the side detection unit 4 the XY position and/or the alignment direction of the second alignment mark 21 of the second substrate 17 is detected/captured.

The goal is as perfect as possible alignment of the second substrate 17 with respect to the XY position and/or orientation of the first/lower substrate 16 . The lower substrate 16 can be an obstruction in the optical path for observing the second alignment marks 21 of the second/upper substrate 17 by the second detection unit 4 and thus particularly in the X and/or Y direction. Moving in the direction moves the first/lower substrate 16 out of this optical path, preferably without moving in the Z direction.

Correction of the XY positions and relative orientations of the two substrates is performed by comparing the XY positions and/or alignment directions of the first and second alignment marks in a third alignment. It is done correlated to the mark.

After the second/upper substrate alignment state has been achieved, in particular with minimal misalignment, preferably with misalignment eliminated, the upper substrate holder is clamped in this processing step, i.e. at least It takes away the degrees of freedom in the X and Y directions.

FIG. 2c schematically illustrates the processing steps according to the invention, in particular for aligning the first substrate 16 with respect to the clamped second substrate 17. FIG.

The optical path of the optical system 2 is interrupted during alignment by the first substrate holder 10 and the second substrate holder 13 between the first detection unit 3 and the second detection unit 4, which Therefore, the detection units 3, 4 cannot be used during alignment.

An additional measuring system 6 with a third detection unit 7, which in a particularly preferred embodiment is a camera system with a microscope, preferably in real time, in particular continuously, provides XY position closed-loop control and/or or form image data that can be used as raw data for directional closed-loop control.

The (theoretical or average) spacing of the contact surfaces 16i, 17i to be bonded (especially without taking into account possible preloads or deflections) is in particular below 1 mm, preferably below 500 micrometers, particularly preferably below 100 micrometers. Hereafter, it is 50 micrometers or less in the optimal case, and 10 micrometers or less in the ideal case. In particular, this interval can be set by mobile device 11 and/or mobile device 14 .

For the alignment of substrate 16 and substrate 17, in particular the target values determined/established according to FIG. 2a are used. This target value includes inter alia image data of the third alignment mark 22 of the first substrate holder 10 and/or determined XY position data and/or of the first substrate holder 10 relative to the movement device 11 or alignment direction data, and/or closed-loop control parameters such as trajectory for optimally reaching the XY position, and/or machine-readable values for the drive, among others.

In another embodiment, residual errors not removed in positioning the upper and/or lower substrates can be considered as corrections to the positioning of another (lower or upper) substrate. be.

The first substrate holder 10 is moved until the alignment error calculated from the additional measurement system target and the actual position and/or orientation of the substrate holder is minimized and in the ideal case eliminated, or Until the decision to interrupt is reached, the first movement device 11 moves in closed-loop controlled positioning and in particular in closed-loop controlled direction. In other words, the lower substrate holder 10 is returned to the already known measured XY alignment position in a controlled, closed-loop controlled manner.

In another embodiment, residual errors not removed in positioning the upper substrate and/or the lower substrate are similarly considered as correction values for the positioning of another (lower or upper) substrate. It is possible.

Further correction factors can be obtained from the vibrating conditions of the apparatus or parts of the apparatus, among others, as previously explained, where these correction factors are used in positioning the substrate to account for residual positioning uncertainties. and improve alignment accuracy. To verify the final position, again all known disturbing factors and influences can be taken into account and corrected accordingly.

Finally, in this process step according to the invention, the clamping of all drives prevents movement of the substrate holders 10, 13 in the XY direction.

After aligning the substrates 16, 17 according to one of the methods according to the invention, at least one of the substrates is subjected to a substrate deformation device 15 in order to join the substrates together by pre-bonding. can be deformed in the direction of another substrate.

1 alignment device 2 optical system 3 first/upper detection unit 4 second/lower detection unit 5 positioning device 6 additional measurement system 7 third/additional detection unit 8 additional measurement Positioning Device of the System 9 Frame 10 First Substrate Holder 10a First Fixed Surface 11 First Moving Device 12 Common Theoretical Focal Plane of the Optical System 13 Second/Upper Substrate Holder 13a Second Fixed Surface 14 second moving device 15 substrate deformation device 16 lower substrate 16i first contact surface 16o first mounting surface 17 upper substrate 17i second contact surface 17o second mounting surface 20 first position alignment mark 21 second alignment mark 22 third alignment mark

Claims (18)

A method for aligning and contacting a first contact surface (16i) of a first substrate (16) and a second contact surface (17i) of a second substrate (17), said method teeth,
- fixing a first mounting surface (16o) of said first substrate (16) to a first substrate holder (10) and a second mounting surface (17o) of said second substrate (17); to a second substrate holder (13) positionable opposite said first substrate holder (10);
- detecting the first XY position and/or the first alignment direction of the first alignment mark (20);
- detecting a second XY position and/or a second alignment direction of the second alignment mark (21);
- aligning said first substrate (16) with respect to said second substrate (17);
- bringing into contact said first substrate (16) and said second substrate (17) aligned with respect to said second substrate (17);
in a method comprising
Prior to said alignment, a third XY position of a third alignment mark (22) and/or a third alignment mark (22) of said first substrate holder (10) and/or said first substrate (16) additionally detecting three alignment directions by a third detection unit (7) and using said third detection unit (7) to open-loop control said alignment;
for alignment of said first substrate (16) and said second substrate (17), said third XY position and/or said third alignment direction; XY position and/or with at least one of said first alignment direction and b) said second XY position and/or said second alignment direction are correlated to provide real-time measurements and closed-loop control is performed during said alignment;
A method characterized by: Said first alignment mark (20) and said third alignment mark (22) are positioned such that said first substrate (16) or said first substrate holder (10), viewed in the Z direction located on opposite sides of each other,
The method of claim 1. said detection of said first alignment mark (20) and said third alignment mark (22) in associated processing steps by a first detection unit (3) and said third detection unit (7); by synchronizing with the
3. A method according to claim 1 or 2. said detection of said first alignment mark (20) and said third alignment mark (22) being performed simultaneously in associated processing steps;
4. The method of claim 3. fixing the second substrate holder (13) in at least the XY directions during the alignment;
A method according to any one of claims 1 to 4 . The first detection unit (3) and the second detection unit (4) are mounted on a common XY positioning device and/or the optical axis of said first detection unit (3) and the optical axes of said second detection unit (4) are aligned or associated with each other,
A method according to any one of claims 1 to 5 . said third detection unit (7) with respect to an optical system (2) consisting of a first detection unit (3) and a second detection unit (4) during said detection at least until aligned; , arranged in a fixed position in the X and Y directions and having no degrees of freedom ,
A method according to any one of claims 1 to 6 . performing said open-loop control of said alignment only on said detection of said third XY position and/or said third alignment direction;
A method according to any one of claims 1 to 7 . providing the first substrate and the second substrate with a spacing A between the first contact surface and the second contact surface in the Z-direction during the detection and until the alignment; positioned between the first substrate holder and the second substrate holder;
A method according to any one of claims 1 to 8 . The spacing A is 500 micrometers or less ,
10. The method of claim 9 . The spacing A is 100 micrometers or less,
10. The method of claim 9. The spacing A is 50 micrometers or less,
10. The method of claim 9. The spacing A is 10 micrometers or less,
10. The method of claim 9. A device for aligning and contacting a first contact surface (16i) of a first substrate (16) and a second contact surface (17i) of a second substrate (17), comprising:
- a first substrate holder (10) for fixing a first mounting surface (16o) of said first substrate (16);
- a second substrate holder (13) positionable opposite said first substrate holder (10), fixing a second bearing surface (17o) of said second substrate (17);
- a first detection unit (3) for detecting a first XY position and/or a first alignment direction of the first alignment mark (20);
- a second detection unit (4) for detecting a second XY position and/or a second alignment direction of the second alignment mark (21);
- alignment means for aligning said first substrate (16) with respect to said second substrate (17);
- contacting means (15) for contacting said first substrate (16) and said second substrate (17) aligned with respect to said second substrate (17);
in a device having
The apparatus comprises a third detection unit ( 7), wherein the alignment is open-loop controllable using the third detection unit (7);
for alignment of said first substrate (16) and said second substrate (17), said third XY position and/or said third alignment direction; XY position and/or with at least one of said first alignment direction and b) said second XY position and/or said second alignment direction are correlated to provide real-time measurements and closed-loop control is possible during said alignment;
A device characterized by: said first detection unit (3) and said third detection unit (7) are synchronizable or synchronized;
15. Apparatus according to claim 14 . during said alignment said second substrate holder (13) is configured to be fixable in at least the XY direction;
15. Apparatus according to claim 14 . Said first detection unit (3) and said second detection unit (4) are mounted on a common XY positioning device and/or the optical axis of said first detection unit (3) and the optical axes of said second detection unit (4) are aligned or associated with each other,
15. Apparatus according to claim 14 . Said third detection unit (7) is aligned at least during said detection with respect to an optical system (2) consisting of said first detection unit (3) and said second detection unit (4) are fixedly positioned in the X and Y directions and have no degrees of freedom ,
15. Apparatus according to claim 14 .

Images