DE202016000967U1 - Recording device for holding wafers - Google Patents

Abstract

A receiving device for receiving and holding a lower substrate (2), comprising: - a support surface (1o), - holding means for holding the lower substrate (2) on the support surface (1o) and - measuring means for determining a nonlinear component of a run-out Error of the lower substrate (2).

Description

The invention relates to a receiving device for receiving and holding substrates according to claim 1 and to a device according to claim 7.

Miniaturization is a perennial trend in microelectronics. Following this trend, the manufactured components or systems usually offer the same or better performance with reduced geometrical dimensions than the predecessor models. This trend can be observed in particular in consumer electronics. Year by year, faster data connections, better cameras, more precise location etc. are offered. These examples are merely the effect achieved by better electronic components, advanced software.

Although the miniaturization of the components has reached the limits of physics, the novel applications require an even higher integration of functions, transistors, etc. In order to be able to accomplish this particularly rapid development, the semiconductor industry is all members of the value creation chain from substrate production to Packaging demanded to increase the quality by reducing the error.

Years ago, for example, the cameras of the widespread mobile phones were equipped with VGA image resolution, currently the HD resolution is not uncommon. This huge increase in performance is coupled with the tighter specifications.

The increase in performance and the tighter specifications used at the same time are forcing ahead, among other things, due to the cost pressure, that more components are manufactured based on the unit area. Furthermore, the number of simultaneously processed components is increasing, which involves increasing the size of the base plates, also called substrates.

In order to promote the ever-increasing miniaturization, components must be stacked on top of each other. The stacking is still done at the substrate level. The substrates are connected to each other. The connection process is called bonding. When bonding substrates are joined together (usually cohesive, permanent), in other words bonded. A common technology for bonding is fusion bonding. Other designations include silicon direct bonding, fusion bonding, or atomic bonding. Fusion bonding is a two-stage process, in which first the substrates are held together by means of weak forces and only after this so-called pre-bonding step in the second stage a heat treatment results in a connection of the substrates whose strength can be equated to the strength of the substrate material. A special feature of the prebond is that under certain circumstances the substrate pair can be separated again from one another without destruction and the bonding is again possible. The finished processed pair of substrates (after heat treatment) can no longer be separated nondestructively.

The investments where the fusion bonds are made are called fusion bonders. In fusion bonding, among other things, for the optimal end product, purity and flatness of the contacting surface, freedom from distortion of the substrates are of elementary importance.

In particular, the following technical measures are used to achieve the freedom from distortion of the substrates to be contacted: The distortions are compensated by special recording devices or sample holders (chucks), with the recording devices existing in various designs. Decisive for the receiving devices is a flat receiving surface or mounting surface, so that the ever-smaller structures can be aligned and contacted on ever larger substrate surfaces over the entire substrate surface. This is important for the preconditioning step.

Particularly important is a high alignment accuracy of the substrates to each other as soon as an alignment accuracy or in particular distortion values of less than 2 microns to be achieved for all arranged on one or both substrates structures. In the vicinity of alignment marks this succeeds in known recording devices and devices for alignment, so-called alignment systems (English: Aligner), in particular Bondaligner, very well. As the distance from the alignment marks increases, controlled and perfect alignment with alignment accuracies or, in particular, distortion values better than 2 microns, preferably better than 1 micrometer and more preferably better than 0.25 microns, more preferably better than 0.1 microns, can not be achieved.

Problems with the prior art

The problems that exist in the current state of the art. In the prior art, there are no sample holders, with the help of which one would be able to produce a global strain or distortion by means of a targeted local thermal influence or to modify such that a run-out error between two substrates can be minimized or is avoidable. Therefore, it is not possible in the prior art to compensate for inequalities in the overlap of structures before and / or during the bonding process.

Object of the present invention is to provide a receiving device and a device with which the aforementioned technical problems are solved. It is the object of the present invention to improve generic recording devices such that a virtually distortion-free alignment and pre-bond can be achieved with them.

This object is achieved with the features of claims 1 and 7. Advantageous developments of the invention are specified in the subclaims. All combinations of at least two of the features specified in the description, the claims and / or the drawings also fall within the scope of the invention. In the case of value ranges, values lying within the limits mentioned should also be disclosed as limit values and be claimable in any combination.

The invention relates to a method and a method and a device for compensating a run-out error. The idea is based on the idea of minimizing or completely eliminating the run-out through a plurality of heating segments / heating zones of a sample holder, in particular locally. By selectively controlling the individual heating segments / heating zones of the heating zone sample holder, the local elongation or distortion of a first substrate can be adjusted so that the run-out error between the structures of the first substrate and a second substrate is minimized or eliminated altogether.

The essence of the invention is to influence the distortion of a substrate by a local, thermal influence, in particular by a Heizzoneneprobenhalter so that when bonding the substrate to a second substrate, a run-out error is minimized or even eliminated altogether.

The run-out error can have as a scaling a location-dependent component and an alignment error alignment error (translation or rotation) as a constant, location-independent component. The actual aim of the present method is to minimize the run-out error by counteracting the distortions and / or deliberately distorting the substrates relative to each other.

The knowledge of position and / or distortion and / or strain maps of the individual substrates alone are not sufficient to correct or compensate for all run-out errors, since there are also run-out errors, which only when joining at least two substrates are formed. A knowledge of the position maps thus contains no information about the run-out errors that arise during bonding.

Like all types of errors, the run-out error can be split into a random and a systematic error component. Through error analysis and statistical methods known to those skilled in the art (engineers, mathematicians, physicists), the systematic error components of the run-out error can be corrected or compensated.

In this disclosure, the correction of the run-out error is understood to be the correction of the systematic error component of the run-out error.

As is known to those skilled in the art, the statistical methods such as averaging, standard deviation, variance, scattering can be used to detect the non-random error component in order to counteract it.

In the increasingly accurate detection of the error components and extended statistics, it has surprisingly been noticed that the run-out error can be divided into linear and non-linear components. In this case, the linear component of the error is position-dependent, the position being dependent on the substrate radius, in particular proportionally.

In the case of the non-linear component of the run-out error, an inversely proportional position dependency on the substrate radius can be determined, the causes generally being in the operations of the fusion bond. The exact mathematical description needs further, extended experiments. Even if the run-out error can not be described completely mathematically closed, the compensation means for obtaining information were used experimentally.

Advantages of the invention

In the particularly preferred method and apparatus disclosed herein, the thermal deformation and / or distortion and / or interference of at least one of the substrates is employed to minimize the run-out error. An independent basic idea according to the invention is that the substrate material as a continuum reacts to local influences not only locally but also globally. Thus, a locally altered, in particular thermal size can change the global state of the substrate to the extent required.

These changes can be made in a preferred method and plant with at least one adjustable heating zone on the receiving device, more preferably with at least two heating zones, even more preferably with at least three heating zones, particularly preferably with at least five heating zones, very particularly preferably with more than 10 heating zones.

As a controlled variable, the position and / or strain maps can be used, wherein as a controlled variable both the temperature of the chuck and / or the heating current of the chuck can be applied. The strain maps and / or position and / or distortion maps of the individual substrates quantify the existing distortions for influencing.

In a particularly preferred embodiment, a receiving device may consist of a combination of a chuck which has minimal thermal deformation and high rigidity and a plurality of heating zones which can be controlled independently of one another. As a result, the disadvantageous thermal deformations of the chuck are largely eliminated and the targeted distortion of at least one of the substrates required according to the invention is achieved.

Due to the controlled structure, relatively small amounts of heat for the optimization of the substrates can be supplied to each other in terms of positions or strains. Since feedback and measurement is an important aspect of the invention, this can be used to build a knowledge and / or expert database.

To measure the amount of heat supplied, temperature sensors in at least the same number as heating zones can be used. However, it is possible to use for the detection of the temperature distribution of the receiving device with more than 10, with more than 20 or with more than 30 sensors.

Thus, substrate and / or plant-specific variables such as thermal time constants can be detected, calculated, collected, evaluated and used for optimization. With the help of such data storage a further process optimization and dynamic operation of the fusion bonder is possible, but it is necessary to check the global state of the substrates or of the pre-bonded substrate pair.

The invention is based on the Applicant's knowledge according to the European patent application EP 2 656 378 wherein detection of the entire surface, in particular the positions of the structures on the surface of each substrate as a position map of the substrate is possible. Furthermore, a device for determining by stretching and / or distortion is disclosed.

Distortion maps are generated, wherein the distortion map of a first substrate is measured against a second substrate upon joining the first substrate to local and / or global alignment errors encountered with the second substrate. The listed method steps are the determination of a first strain map of strain values along a first contact surface of the first substrate and / or the measurement of a second strain map of strain values along a second contact surface and the use of evaluation means for evaluation of the first and / or second strain maps by which the local alignment errors can be determined.

The basic idea of the present invention is to provide a receiving device consisting of a plurality of mutually independent active control elements with which the mounting surface of the receiving device, in particular in shape and / or temperature, can be influenced. In this case, the active control elements and / or control elements are used by appropriate control and / or regulation so that the known by means of the position maps and / or strain maps local alignment errors or local distortions are compensated or minimized or reduced as far as possible.

This not only eliminates local distortions, but also simultaneously minimizes or corrects, in its outer dimensions, a macroscopic, global distortion or strain of the substrate resulting from the local distortions. A device for local compensation of distortions is from the US 5094536 A known. From the WO 2009/133682 A1 is an indirect method for determining strains on the determination of positional deviations of substrate markings known.

Thus, according to the invention, it is possible, in particular in combination with the above-described inventions concerning the position maps, strain maps and / or voltage maps and the in-situ correction of alignment errors during contacting and bonding of the substrates disclosed therein, by active, in particular local, action on distortions of the substrate to achieve an even better alignment result.

According to an advantageous embodiment of the invention it is provided that the temperature of the support surface can be influenced locally by the compensation means. A local increase in temperature of the support surface leads to a local extension of the substrate held on the support surface at this position. The higher the temperature gradient, the more the substrate expands at this position. Based on the data of the position maps and / or strain maps, in particular the vector evaluation of the alignment error, in particular for each position of the position maps and / or strain maps and statistics can be acted so targeted to local distortions of the substrate or counteracted this, which exert a global effect of the substrates ,

In this context, a vector field with distortion vectors is to be understood as vector evaluation, which was determined in particular by means of one of two variants of the invention described below.

The first variant relates to use cases in which only one of the two substrates is structured. In this case, it is provided according to the invention to detect the deviation of the structures, in particular the deviation of the geometric shape from the desired shape. Of particular interest in this case is the deviation of the shape of exposure fields, in particular exposure fields, of a step-and-repeat exposure device of the nominally expected shape, which is usually rectangular. These deviations, in particular the vector field describing these deviations, can be determined on the basis of the detection of a position map of the individual alignment marks corresponding to the exposure fields EP 2656378 A1 respectively. Alternatively, this vector field may also be based on stress and / or strain maps generated by means of EP 2463892 B1 be determined. Advantageously, according to the invention, however, this vector field can also be determined and read by any other suitable measuring device. In particular, Step & Repeat lithography systems which are operated for acquiring this data with a special test mask and / or a special test routine are suitable for this measurement.

The second variant relates to applications in which both substrates are structured. In this case, it is provided according to the invention, the vector field of the alignment deviation in particular for all positions of the position maps, in particular the first and second position maps according to EP 2656378 A1 to calculate. This vector field should in particular for the, according to technological and / or economic criteria considered ideal alignment position according to the statements in EP 2656378 B1 be determined.

In particular, the control device includes software for executing / calculating corresponding routines. Control devices are conceptually included in the controller in this disclosure.

The device according to the invention advantageously comprises the above-described control device and / or control device in a central control unit and / or control unit responsible for all control processes and / or control processes. According to the invention, however, it is conceivable to provide the control device or the regulating device in the receiving device, in particular as a module of an overall device.

The method according to the invention can be further improved by providing a registration of position and / or strain maps of the first and / or second substrate after the alignment, in particular again. Thus, according to the invention, a check of the alignment success can be provided after the alignment has been carried out. Accordingly, it is conceivable to make a spin-off of a pair of substrates with too large alignment errors, for example, to align or dispose of them again according to the invention. At the same time, the acquired data can be used for self-calibration of the device, in particular by means of the control device.

The in the European patent application EP 2656378 A1 and / or the European patent application EP 2656378 B1 At the same time, disclosed inventions should be deemed to be co-disclosed embodiments of the present invention.

A problem of all embodiments of the invention so far mentioned is that all of these deal exclusively with the distortions of the individual substrates and these attempts to compensate and / or eliminate. A particular aspect has hitherto been ignored that two ideally distortion-free substrates bonded in the theoretically ideal alignment state may be subject to errors. This error is the so-called run-out, which arises only after bonding to the relative position of both substrates to each other.

For the examination of the global condition of the substrates, their characteristics, relative position to each other quality criteria must be determined and applied.

The quality criteria can be subdivided, for example, into three generic terms: input, process and output characteristics.

The input features include, for example, all substrate parameters, not changeable Machine parameters, non-influenceable environmental conditions such as the local gravitational acceleration at the site.

Process characteristics are all features that relate to the joining of the substrates: for example, substrate preparation with all Vorprozessteilen, the joining, the pre-bonding as a process itself and the aftertreatment, transport of the substrate to the next processing step.

The starting characteristics are all features that had changed the substrate until bonding, in particular fusion bonding, be it energetic, material changes or the information content of the substrate, which, for example, is encoded in the structural geometry (s).

Achieving the quality criteria presupposes an interaction of all input and process features so that the output characteristics are reproducibly achieved. This interaction is thus influenced by the bond system, the process and the materials used (substrate, plant materials) as well as, for example, substrate properties, substrate material, cleanliness, ambient conditions.

The process features are also understood as process features and used synonymously.

sample holder

The sample holder generally consists of several heating zones. A heating zone is a single controllable area, which generally consists of several heating elements. All heating elements of a heating zone can contribute to the temperature control.

A heating element is a physical element that can generate heat. Preferably, these are electrical components.

In a first embodiment of the invention, the sample holder consists of several concentric heating elements / heating zones, which are arranged around the center of the preferably circular sample holder. However, any other geometric shapes of the sample holder and / or the heating zones (square, hexagon, octagon, etc.) are possible. On the sample holder several preference concentrically operated, concentric heating zones can be arranged. There may be at least two concentric heating zones on / in the sample holder, three concentric heating zones may be preferred. With particular preference, four concentric heating zones are possible. A most preferred embodiment may include five concentric heating zones, but embodiments with six, seven, eight, and more than eight concentric heating zones, respectively, are possible.

In a second embodiment of the invention, the sample holder consists of a plurality of grid-shaped heating elements heating zones (squares, diamonds, hexagons, circles, or honeycomb structures), which can also be controlled individually. The grid consists of several fields. This number may depend on the size, shape, material, mass, volume, material properties such as heat and thermal conductivity of the sample holder. There may be two segments / heating zones on / in the sample holder. With preference, the number of fields may be formed mathematically as a combination of two series consisting of natural numbers and zero. Row A describes the number of columns, row B describes the number of rows in the grid. The variables i and / or j can describe the position of the current element as a run variable from the totality. Thus, A [1, 2, 3, 4, 5, 6, 7, 8, 9, ... n] can be formed with n.i + 1 = n.i + 1. Analogously, this applies to the variable B, B can be [1, 2, 3, 4, 5, 6, 7, 8, 9, ... n] etc ...

In particular, 4 segments / heating zone are to be designed as 2 × 2 segments / heating zones. For 9 segments / heating zones a 3 x 3 arrangement is conceivable. If the segments / heating zone differ in their size, shape, heating capacity, etc., the formation rule Element from A · Element B can be extended with additional elements:
Element of A · element of B + additional element of A + additional element of B possible. Thus, for example, a 2 × 2 arrangement with four further, irregularly formed but important for the edge uniformity elements can be formed. However, particularly preferred is a planar, uniform coverage of the sample holder with hexagonal heating elements.

In a third preferred embodiment, the sample holder may consist of a base body on which the heating elements are mounted. The heating elements can be attached to the sample holder in a material-locking, and / or positive-locking and / or non-positive manner. In the preferred embodiment, the heating elements may be attached to the sample holder by means of a heat-conducting adhesive layer. In a further embodiment, a mechanical clamping using heat conduction for the thermal contacting is possible.

In a fourth embodiment, the heating elements can be integrated in the volume of the sample holder, so that in particular thermally induced deformations of the sample holder are eliminated, which is caused by the heating elements, which are not symmetrical in volume.

In all embodiments according to the invention can with preference The heating elements are the heating elements individually, directly controlled via .... In order to minimize the control effort, a control concept composed of pulse width modulation and / or multiplexing of the individually controllable heating elements is possible. However, any control concept for operating the heating elements is conceivable.

The thermal time constant of the sample holder or substrates to be controlled is decisive for whether the multiplexing, that is to say successive operation of the individual heating elements, is acceptable and / or pulse width modulation takes place or continuous operation of the heating elements is necessary.

To achieve thermally homogeneous, isothermal zones on the sample holder, the individual heaters can be controlled by the control accordingly. For the elimination of the scattering of the heating elements, which can be measured as temperature differences on the sample holder, the sample holder can be readjusted in the controlled isothermal mode. The calculated correction values can be stored in the control as a database.

The temperature of the heating elements is detected in a preferred embodiment by means of integrated thermocouple, in particular thermistors such as Pt 100 measured via ... and / or Pt1000 thermoresistors, semiconductor temperature sensors and forwarded to the controller / controller. In preferred embodiments, pyrometric sensors and / or other optical, in particular imaging, sensors such as microbolometers can be used. The evaluation can take place in the controller or in another, responsible for the entire system computing unit.

The signals of the temperature sensors are fed to a temperature controller, which sets and / or corrects the temperature of the sample holder in a manner known to those skilled in the art in real time or with a given delay. Both analog (single) controllers such as PI or PID controllers can be used for each heating element, as well as programmed, computer-aided controllers, such as FPGA-s, where each individual controller is shown as a program part. It is also possible to use multiplexers in order to minimize the number of necessary controllers, in particular with sufficiently fast controllers, corresponding heating power, large thermal mass (high volume, high heat capacity) of the sample holder.

The temperature of the heating segments / heating zone as the output is set by external energy supply. Current, Joule heat, tempered heating fluid (liquids and / or vapors and / or gases) can be used as the controlled input variable. In accordance with the input variable, the heating segments can use heating elements as resistance heaters, and / or as inductive heating elements, and / or as radiant heaters, and / or as Peltier elements and / or as heat exchangers operated in the same direction or in the opposite direction (chamber or meander structures). , and / or direct combustion chamber.

Materials of the sample holder may consist of any cleanroom suitable, form and temperature-stable material in the operating condition, which meets the requirements of thermal conductivity greater than 0.1 W / m / K, more preferably greater than 1 W / m / K, more preferably greater than 10 W / m / K , particularly preferably greater than 100 W / m / K, in the optimal case greater than or equal to 120 W / m / K, ideally greater than or equal to 150 W / m / K.

It should materials for the sample holder in particular a specific heat capacity less than 1000 J / kg / K, preferably less than 800 J / kg / K, more preferably less than 715 J / kg / K, more preferably less than or equal to 700 J / kg / K, ideally less than 500 J / kg / K.

The materials of the sample holder should in particular have the material property bending stiffness greater than 1 MPa, preferably greater than 10 MPa, more preferably greater than 30 MPa, particularly preferably greater than 100 MPa, in the optimum case greater than 1 GPa, ideally greater than 10 GPa.

In particular, the thermal expansion coefficient should be as small as possible in order to prevent distortion of the sample holder by the temperature differences. The thermal expansion coefficient is preferably less than 10 ^-4 K ^-1 , preferably less than 5 × 10 ^-5 K ^-1 , more preferably less than 10 ^-5 K ^-1 , most preferably less than 5 × 10 ^-6 K ^-1 , most preferably less than 10 ^-6 K ^-1 , most preferably less than 10 ^-7 K ^-1 .

The materials of the sample holder may be metals and / or alloys and / or non-stoichiometrically defined materials and / or metal foams. These materials may include iron, nickel, chromium, manganese, aluminum, tungsten, molybdenum, copper, silver, gold, cobalt, platinum, cerium, lead, titanium, tantalum, zinc and / or tin.

The materials of the sample holder may include ceramic and / or glass-ceramic materials and / or composite materials containing lithium, aluminum, silicon, carbon, nitrogen. Furthermore, glass-ceramic materials can be used which are free of alkali oxide.

The flatness and surface roughness of the sample holder are essential design features.

In the further course of the flatness is used as a measure of the perfection of a planar surface, in particular a surface. Deviations from a planar surface result from ripples and roughness. The waviness of a surface is characterized by a certain periodic increase and decrease in the surface area, in particular in the millimeter range, more rarely in the micrometer range. Roughness, on the other hand, is a rather aperiodic phenomenon in the micro or nanometer range. The exact definition of such surface properties is known to any person skilled in surface physics, tribology, mechanical engineering or materials science. To treat the different deviations from the ideal surface, the term roughness is used interchangeably in the further course of the patent for the superposition of all such effects.

Roughness is reported as either average roughness, square roughness or average roughness. The values determined for the average roughness, the squared roughness and the average roughness depth generally differ for the same measuring section or measuring surface, but are in the same order of magnitude. Therefore, the following numerical value ranges for roughness are to be understood as either average roughness, squared roughness, or averaged roughness. The roughness is less than 100 microns, preferably less than 10 microns, more preferably less than 1 micron, most preferably less than 100 nm, most preferably less than 10 nm.

The adhesive properties of the sample holder should cover two states. For the fixation of the respective substrate a switchable, high adhesive force should be used. For handling the adhesive property should not be present. The adhesive property is to be expressed as sticking energy. The energy is given in J / m ² . The energy per unit area is in the deactivated state of the switchable surface, therefore, in the state in which the adhesion is to take a minimum, less than 2.5 J / m ² , preferably less than 0.1 J / m ² , more preferably less than 0.01 J / m ^2, with the greatest preference less than 0.001 J / m ^2, with preference allergrößtem less than 0.0001 J / ^m2, at bevorzugtesten less than 0.00001 J / ^m2. The energy per unit area is in the activated state of the switchable surface, therefore, in the state in which the adhesion is to take a maximum, greater than 0.00001 J / m ² , with preference greater than 0.0001 J / m ² , more preferably greater than 0.001 J. / m ² , most preferably greater than 0.01 J / m ² , most preferably greater than 0.1 J / m ² , most preferably greater than 1 J / m ² .

The sample holder should physically and / or chemically under normal laboratory conditions according to the invention below 70 ° C, more preferably below 100 ° C, more preferably below 200 ° C, more preferably below 250 ° C, in the optimal case below 300 ° C, ideally below 650 ° C. be stable. This means inert properties, ie mass loss by outgassing or oxidation less than 0.1%, more preferably less than 0.01%, particularly preferably less than 0.001%, based on an initial mass, measured at 20 ° C. under a protective gas atmosphere free from water or water vapor.

In an embodiment of the sample holder disclosed in accordance with the invention, a substrate can be attached thereto by vacuum in accordance with the switchable adhesive property. If the vacuum is generated by appropriate cavities and lines on the surface of the sample holder, this sucks the substrate. The substrate is detachable from the surface after the vacuum is interrupted by respective actuators. Pressurization of the substrate is also possible to achieve accelerated separation of the substrate from the sample holder.

In a further embodiment of the sample holder according to the invention, a substrate with electrostatically operated means can be attached to the sample holder. If the attachment is active, the substrate is fixed on the sample holder. This may correlate with the switchable adhesive property of the sample holder.

In a further embodiment according to the invention, the sample holder can be fastened by means of a mechanical clamping device (or a plurality of mechanical clamping devices), in particular without compulsion. As compulsory freedom is understood in a substrate to prevent the movements, dislocations in the exact six degrees of freedom (3 rotation and 3 translation).

In a further embodiment according to the invention, the sample holder may have local, adjustable, highly adhesive zones which fasten the substrate, in particular, on the outer diameter. The dissolution of the substrate can be done by irradiation by means of high-energy radiation such as laser.

In a further embodiment of the invention, the sample holder has at least at the edge over a highly adhesive layer. The sample holder can, however, in any pattern or unstructured full surface a highly adhesive layer own, on which the substrate is attached. The dissolution of the substrate can be done by means of targeted applied overpressure and / or lifting devices, in particular pins.

In a further embodiment of the invention, the substrate may be attached to the sample holder via switchable magnetic forces.

• – Mechanische Fixierungen, insbesondere Klemmen,
• – Vakuumfixierungen, insbesondere mit einzeln ansteuerbare Vakuumbahnen und/oder miteinander verbundenen Vakuumbahnen,
• – Elektrische Fixierungen, insbesondere elektrostatische Fixierungen,
• – Magnetische Fixierungen
• – Adhäsive Fixierungen, insbesondere Gel-Pak Fixierungen und/oder
• – Fixierungen mit adhäsiven, insbesondere ansteuerbaren, Oberflächen.

In summary, the following fixations for a sample holder come into question:

• - Mechanical fixations, in particular clamps,
• Vacuum fixings, in particular with individually controllable vacuum webs and / or interconnected vacuum webs,
• - Electrical fixations, in particular electrostatic fixations,
• - Magnetic fixations
• - Adhesive fixations, in particular gel-Pak fixations and / or
• - Fixations with adhesive, especially controllable, surfaces.

The fixings are in particular electrically controlled. Vacuum fixation is the preferred type of fixation. The vacuum fixation preferably consists of a plurality of vacuum webs, which emerge on the surface of the sample holder. The vacuum paths are preferably individually controllable. In a technically more feasible application, some vacuum webs are combined to form vacuum web segments, which can be individually controlled, therefore evacuated or flooded. Each vacuum segment, however, is independent of the other vacuum segments. This gives the possibility of building individually controllable vacuum segments. The vacuum segments are preferably designed annular. As a result, a targeted, radially symmetrical, in particular from inside to outside carried out fixing and / or detachment of a substrate from the sample holder allows.

method

The general method, which achieves the fusion bonding of substrates and can be carried out by means of fusion bonding systems, consists essentially of the following substeps, wherein known subdivisions of the substeps are assumed as general knowledge of the person skilled in the art.

In a recipe, the process characteristics are summarized purposefully, the sub-steps created in a meaningful combination. The recipes can be provided both as instruction sets in machine-understandable form or for the operator. Execution of the recipe to implement the process features performed on the fusion bonding equipment is described here. Preparatory process steps will be omitted here.

A first possible method according to the invention has the following steps.

In a first method step according to the invention, a first substrate is charged to a first chuck.

In a second method step according to the invention, the position and / or expansion card can be created for this substrate.

After loading a second substrate on a second chuck in the third method step according to the invention, the position and / or expansion card in the fourth method step according to the invention can likewise be generated for the second substrate.

In a fifth method step according to the invention, the coarse positioning errors, in particular from the position and / or strain maps, are detected, calculated and corrected by means of corresponding positioning means.

The sixth method step according to the invention, in particular iteratively, relates to the measurement and / or determination of the residual error between a first and second substrate with particular regard to the nonlinearities of the scaling error.

The seventh method step according to the invention, in particular iteratively executed, involves the targeted application of a recording device for local and / or global, regulated supply of heat quantities for influencing a first or second substrate.

In an eighth, in particular iterative, method step according to the invention, the compensation performed in the seventh method step can be checked, wherein after reaching a threshold value, the iteration can be terminated.

In a ninth method step according to the invention, the bonding wave is initiated and the pre-bonded substrate pair is generated.

In a tenth method step according to the invention, the substrate stack can be generated again with a modified position and / or expansion card for controlling the run-out error.

After successful, necessary calculation steps, the substrate pair can be released for further processing in the eleventh method step according to the invention or separated and cleaned for correction. At the same time, the statistically determined new results for the further error correction can be fed into the data memory so that the correction parameters can be calculated.

In a twelfth possible method step according to the invention, the correction parameters can be used for a further substrate pair to be processed.

Further advantages, features and details of the invention will become apparent from the following description of preferred embodiments and from the drawings. The drawings contain rough schematic representations without proportional accuracy. For the sake of understanding, effects or details may be exaggerated. The show in:

1 FIG. 2 is a schematic representation of a fusion bonding device with a receptacle / sample holder according to the invention, FIG.

2a shows a first embodiment of a sample holder / receiver,

2 B FIG. 2 illustrates a second embodiment of a receptacle / sample holder according to the invention, FIG.

2c FIG. 3 illustrates a third embodiment of a receptacle / sample holder according to the invention, FIG.

2d FIG. 4 illustrates a fourth embodiment of a sample holder according to the invention, FIG.

2e illustrates a fifth embodiment of a sample holder according to the invention,

2f FIG. 5 illustrates a sixth embodiment of a sample holder according to the invention, FIG.

2g FIG. 7 illustrates a seventh embodiment of a sample holder according to the invention, FIG.

2h illustrates an eighth embodiment of a sample holder according to the invention,

2i FIG. 9 shows a ninth embodiment of a sample holder according to the invention, FIG.

2y illustrates a tenth embodiment of a sample holder according to the invention,

3a illustrates a first embodiment of a sample holder with sensor.

3b shows a second embodiment of a sample holder with sensor according to the invention,

3c shows a third embodiment of a sample holder with sensor according to the invention,

4a shows a first embodiment of a sample holder with heating zones,

4b shows a second embodiment of the invention a sample holder with heating zones,

5a represents a first variant of a method according to the invention,

5b represents a second variant of a method according to the invention,

5c represents a third variant of a method according to the invention,

5d represents a fourth variant of a method according to the invention,

6 shows the shares of the run-out error with or without compensation as a diagram,

7a shows a first schematic embodiment of a sample holder with heating zones in plan view,

7b shows a second embodiment according to the invention of a sample holder with heating zones in plan view,

7c shows a third embodiment of a sample holder according to the invention with heating zones in plan view, and

7d shows a fourth embodiment of a sample holder according to the invention with heating zones in plan view.

In the figures, the same components or components with the same function with the same reference numerals. 1 represents a schematic, not to scale, schematic diagram of a fusion bonding plant ( 6 ) with a heating zone sample holder ( 1 ). In the fusion bonding plant ( 6 ) is the heating zone sample holder ( 1 ) on active and / or passive position and position control elements ( 8th . 8th' ), which can be used as vibration damping. The lower substrate ( 2 ) with the outer functional surface ( 2a ) on the functional surface ( 1o ) of the heating zone sample holder and placed without fixing or by means of a fixation, in particular with vacuum webs ( 5 ' ) attached to it. The lower substrate ( 2 ) has the diameter D1, which can significantly influence the proportional run-out error. The functional surface to be bonded ( 2i ) pointing against the gravitational direction, in the method with the functional surface ( 3i ) of the upper substrate ( 3 ) with a diameter (D2). The upper substrate has an outer functional surface ( 3a ), which is by means of vacuum webs ( 5 ) on the functional surface ( 4o ) of the upper sample holder ( 4 ) is attached. The Prebond is preferred with a pin ( 7 ). For the creation of the position maps and / or for the observation of the prebond and / or for the determination of the run-out error and / or acquisition of all necessary measurable parameters, the sensor ( 10 ) and / or sensors not shown. The sensor data and / or control and / or regulating signals and / or supply of compressed air and / or vacuum and / or electrical energy and / or optical information is transmitted by means of signal and / or supply lines ( 11 . 11i . 11II . 11III . 11iv ) connected to one and / or more computing and / or control and / or regulating and / or supply unit, which provides the required energy, information or substance and the respective controlled variable.

According to the invention, in particular the control of the heating zone sample holder and their observation are of great importance.

2a FIG. 2 illustrates a first schematic, not to scale, embodiment of a heating zone sample holder (FIG. 1 ) on the bottom are different, separate and separately controllable heating zones ( 12 . 12 ' . 12 '' ) appropriate. The following features are not shown in the picture. The heating zones may preferably be galvanically connected to the control and / or regulation. According to the invention, the energy supply by means of local radiation (heat and / or microwaves) is also possible. The connection of the heating zones to the base body of the heating zone sample holder can be carried out by means of adhesive. It is possible according to the invention to attach the heating zones by lamination and / or vapor deposition and / or non-positively pressed in particular with springs. In a further embodiment, the formation of the heating zones by means of implanted, integrated heating elements is possible.

2 B represents a second embodiment of a sample holder according to the invention.

As an extension to 2a be on a functional surface of the heating zone sample holder ( 1 ) Pimples ( 14 ), which can have better bonding properties by reducing the surface area. On the bottom are different, separate and separately controllable heating zones ( 12 . 12 ' . 12 '' ) appropriate.

2c represents a third embodiment of a sample holder according to the invention. In extension to 2a separates the heating zone sample holder ( 1 ) by thermal insulators ( 13 . 13 ' ) preferably, on the underside of different, separate and separately controllable heating zones ( 12 . 12 ' . 12 '' ). Air is particularly preferably used as a thermal insulator.

2d represents a fourth embodiment of a sample holder according to the invention. As an extension to 2a and 2c be on the heating zone sample holder ( 1 ) for reducing the undesirable adhesion of the surface nubs ( 14 ) arranged. The heating zones ( 12 . 12 ' . 12 '' ) are preferred with thermal insulators ( 13 . 13 ' ) separated from each other.

2e illustrates a fifth embodiment of a sample holder according to the invention. In extension to 2a separates the heating zone sample holder ( 1 ) by thermal insulators ( 13 . 13 ' ) preferably, on the underside of different, separate and separately controllable heating zones ( 12 . 12 ' . 12 '' ). Dielectrics are particularly preferably used as thermal insulators. The insulator material should have a similar, preferably identical thermal thermal expansion as the sample holder, so that thermally induced deformations are excluded.

2f represents a sixth embodiment of a sample holder according to the invention. As an extension to 2a and 2c and 2e be on the heating zone sample holder ( 1 ) for reducing the undesirable adhesion of the surface nubs ( 14 ) arranged. The heating zones ( 12 . 12 ' . 12 '' ) are preferred with thermal insulators ( 13 . 13 ' ) separated from each other.

2g shows a seventh embodiment of a sample holder according to the invention. The heating zone sample holder ( 1 ) consists exclusively of the heating zones ( 12 . 12 ' . 12 '' ), which can be controlled independently.

2h shows an eighth embodiment of a heating zone sample holder according to the invention ( 1 The embodiment represents an extension of 2g wherein the functional surface of the sample holder with nubs ( 14 ) is equipped to reduce the adhesion properties.

2i shows a ninth embodiment of a heating zone sample holder according to the invention ( 1 This embodiment may be a combination of 2g and 2c be executed. The heating zones ( 12 . 12 ' . 12 '' ) as a solid body the carrying functionality of the sample holder and the thermal insulators ( 13 . 13 ' ) separate the heating zones thermally from each other.

2y FIG. 10 illustrates a tenth embodiment of a heating zone sample holder according to the invention (FIG. 1 ). 2y represents an extension of the 2i with pimples ( 14 ). The heating zones ( 12 . 12 ' . 12 '' ) can thermally isolated from each other with a dielectric (not shown) and / or air ( 13 . 13 ' ) become.

3a shows a first embodiment of a heating zone sample holder ( 1 ) with sensor ( 15 The sensor can be mounted on the heating zone sample holder in close proximity to the heating zones ( 12 . 12 ' . 12 '' ) lie. Advantage of this embodiment may be the information of the example. Temperature of the heating zone. It can be mounted in a non-illustrated embodiment corresponding to each heating zone sensors.

3b shows a second embodiment according to the invention of a heating zone sample holder with sensor ( 15 ). The difference to 3a is the integration of the sensor in the solid, so that, for example, the temperature fluctuations from the environment can be better filtered and the temperature of the heating zone sample holder can be measured more accurately. So the effect of the heating zones ( 12 . 12 ' . 12 '' ) are measured better. The attachment of the sensor can be done in a designated cavity and / or be integrated and / or embedded in the carrier material of the heating zone sample holder.

3c FIG. 3 illustrates a third embodiment of a heating zone sample holder according to the invention (FIG. 1 ) with sensor ( 15 ). As difference to 3a and 3b Here, the sensor is mounted in close proximity to the substrate. The interaction of heating zones ( 12 . 12 ' . 12 '' ) on the substrate can be measured better.

4a shows a first embodiment of a heating zone sample holder ( 1 ) with heating zones ( 12 . 12 ' . 12 '' The heating zones are integrated in the mechanically neutral zone of the heating zone sample holder, so that deformation of the sample holder as a result of heat can be reduced. It is also possible for the heating zones to be separated from each other by thermal isolators ( 13 . 13 ' ) to separate from each other. 4a thus represents an alternative embodiment of the sample holder shown, for example. In 2a represents.

4b FIG. 2 illustrates a second embodiment of a heating zone sample holder according to the invention (FIG. 1 ) with heating zones ( 12 . 12 ' . 12 '' ). As difference to 2a or too 4a Here are the heating zones and the thermal insulators ( 13 . 13 ' ) applied directly to the functional surface of the sample holder and the base body serves for mechanical stability and freedom from deformation.

5a represents a first variant of a method according to the invention is possible.

In the process step 100 a first substrate is loaded on one of the sample holders.

In the process step 101 a position and / or distortion and / or expansion card is created from this first substrate, processed as a data set and stored.

In the process step 102 a second substrate is loaded on one of the sample holders.

In the process step 103 a position and / or expansion card is created by this second substrate, processed as a data set and stored.

In the process step 104 From the data of the two position and / or strain maps, a first coarse correction is performed. This can be done with a mechanical adjustment of the position of the sample holder to each other. For this purpose, the sample holders can be attached to solid-body joints or traversing tables (xy-cross tables), which are moved into position with translated stepper and / or direct drives and held in position. In a further embodiment, the positions of the coarse correction can be determined by means of imaging methods, which are carried out in particular by means of shock adjustment, piezo and / or voice coil actuators in a control loop. A particular advantage of the shock adjustment is that the actuators do not work in continuous operation, which reduces the heat generated by the waste heat from the drives. This eliminates external sources of interference.

In the process step according to the invention 105 a fine measurement of the substrates takes place. It must measure the relative positions of the substrates with better 10 microns, preferably better 1 micrometer, more preferably better 100 nanometers, most preferably better 50 nanometers, ideally better 10 nanometers, ideally better 1 nanometer. During the measurement, the data of the position and / or strain and / or distortion maps should be taken into account and the relative positions created with error correction. The relative positions can be used to determine the detectable residual errors.

In the process step according to the invention 106 a correction is made with the heating zone sample holder. From the data of the strain and / or distortion and / or position maps of the two substrates, from the position data of the coarse correction, from the data of the fine measurement a residual error of the substrates was determined. The listed errors cause errors in the Prebond in most cases. In order for these to be corrected, the heating zone sample holder preferably preferably locally heats the lower substrate mounted on the heating zone sample holder. Due to the local heating, the position and / or strain and / or distortion state is changed in a targeted manner both locally and globally on the substrate. By individually controllable, controlled heating zones temperature differences of less than 30 K, preferably less than 10 K, more preferably less than 1 K, most preferably less than 0.1 K, in the optimum case less than 0.01 K are generated. The temperature differences produce the desired distortion, which is opposite to the distortion vector and thereby reduces, compensates, optimally eliminates. Both transient processes and quasi-stationary states of the heating zone sample holder can be used for this purpose. For the operating mode transient processes, the state of the heating zone sample holder should be present as a multi-parameter model in order to be able to set the effect of the amount of heat supplied, the time constants, etc. as distortion of the respective substrates, possibly exactly (in ppm, more preferably in ppb order of magnitude). By using a data store, this can be used as an expert database for the other substrates.

In the quasi-stationary state, the correction by the heating zone sample holder is performed so that there is no change between two observable states of the substrates. However, the individual heating zones can go through different temperature cycles, it is not necessary that the heating zones themselves are operated isothermally. There should only be no temporally changeable processes in the correction with the heating zone sample holder.

In the process step according to the invention 107 are known in the art, the substrates connected to each other.

5b represents a second variant of a method according to the invention is possible. In contrast to 5a become the introductory procedural steps 100 to 103 , Charging both substrates and creation of position and / or stress and / or strain maps carried out parallel in time. This makes it possible to shorten the time required for fusion bonding.

5c represents a third variant of a method according to the invention. This method represents a modification of the in 5a and 5b illustrated method. The introductory procedural steps 100 to 105 are with 5b identical. After the rough correction 104 and fine measurement 105 an iteration loop is started with the decision Q1. It is checked whether the positions and / or states of the two substrates are sufficiently accurate for the pre-bonding 107 are. This can be achieved with a preset threshold and defined tolerance zones. If the calculated run-out error is below the threshold, the positioning and / or state is sufficiently accurate, the loop is left and pre-bonding is performed. If the positioning and / or condition is insufficient, the correction will be made by heating zone sample holder 106 performed, how to 5a has been described in detail. After correction, a new measurement takes place 105 , the result of the measurement is re-evaluated by Q1 and the iteration can be continued or aborted.

It is possible according to the invention to add further states and decisions to Q1 so that no endless loops are formed.

5d represents a fourth variant of a method according to the invention, which is an addition to the method in 5c is. The process steps 100 to 107 , Creation of the Prebondes are identical. This is followed by a new measurement of the bonded substrate stack 105 , In this case, not only the position and / or strain and / or distortion maps of the individual substrates can be used for the correction, but it is the actual run-out error measured.

In the further method step Q2, a second iteration loop is introduced which rates the quality of the prebond. Not only alignment status but also errors in the bond interface and other parameters can be considered. If the condition is sufficient, the processing with the production of the bond in an external system will leave the loop and this in the process step 109 carried out.

If in Q2 the quality of the pre-bond is rated as insufficient, for example, by exceeding the previously defined tolerances, the substrate stack becomes in the process step 108 separated from each other. Then the substrates are cleaned, processed and again the bonding process as substrates for the process steps 100 respectively. 102 provided. This iteration loop can also be supplemented with further abort criteria, not listed, which in particular forbids infinite loops. It can be a knowledge base from the results of fine surveying 105 the substrate stack on the one hand, the better combinability of the substrate pairs as well as a vote and statistical assurance of fine correction 106 be used by heating zone sample holder.

6 shows the shares of the run-out error with or without compensation as a diagram. The abscissa represents the radial position with respect to the center of the substrate, and the ordinate plots the run-out error.

The data comes from distortion maps, here the individual lengths of the distortion vectors are represented.

K1 represents the radius-dependent, linear component of the run-out error. This fraction has been adequately described in the literature and is known to the person skilled in the art.

However, non-linear error components were found, which are due to several, not clearly identifiable processes. There are, for example, nonlinearities as the effect of temperature gradients and / or thermal gradients during bonding. The effect of the pin for bond initiation can lead to non-linearity. Turbulence in the bonding process can also lead to run-out errors. These nonlinearities can be approximated, for example, with a quadratic function.

If the heating zone sample holder is operated such that the resulting distortion and / or strain and / or position represents a quadratic function of opposite sign, as curve K2 illustrates, the resulting run-out K3 can be understood as K1 + K2.

It is an aspect of the invention according to the invention and the processes according to the invention to make the resulting run-out error K3 disappear at every position r, therefore to produce a curve which is as congruent as possible to the abscissa, ie the property K4 (r). = 0 for all r owns. Since it will be technically very difficult to generate a curve with the property K4 (r) = 0 for all r, in the 6 not a curve, but a set of curves K4 (r) specified.

7a shows a first schematic embodiment of a sample holder with heating zones ( 1 ) in plan view. The individual heating zones 12 . 12 ' . 12 '' . 12n are executed as concentric circles around a common center. A total of n heating zones can be formed, whose number and size can be adapted to the needs. n can be from the set { 1 . 2 . 3 . 4 . 5 . 6 . 7 . 8th . 9 . 10 } or more. With this embodiment, the errors that vary proportionally with the substrate radius can be well compensated. As a disadvantage, the different surface area and, as a result, the different sized heating power can be regarded as a homogeneous temperature of the heating zone sample holder. In a further embodiment, the individual circular rings of the heating zones can be further divided into circular ring sections.

7b FIG. 2 illustrates a second embodiment of a heating zone sample holder according to the invention (FIG. 1 ) with heating zones in plan view. The individual heating zones ( 12 . 12 ' . 12 '' . 12n ) are executed as circular sections around a common center. A total of n heating zones can be formed, whose number and size can be adapted to the needs. n can be from the set { 1 . 2 . 3 . 4 . 5 . 6 . 7 . 8th . 9 . 10 } or more. Advantage of the embodiment may be a uniform area distribution. The disadvantage is that in this embodiment, no influence of the substrate along its radius is possible.

7c FIG. 3 shows a third preferred embodiment of a heating zone sample holder according to the invention (FIG. 1 ) with heating zones in plan view. The individual heating zones ( 12 . 12 ' . 12 '' . 12n ) are designed as a regular n-corner, here as hexagons nationwide. A total of n heating zones can be formed, whose number and size can be adapted to the needs. n can be from the set { 1 . 2 . 3 . 4 . 5 . 6 . 7 . 8th . 9 . 10 . 15 . 20 . 25 . 30 } or more. Advantage of the embodiment may be a uniform area distribution. By independently controllable heating zones can thus be the advantages of the embodiments of 7a and 7b combine with each other.

7d shows a fourth embodiment according to the invention of a heating zone sample holder ( 1 ) with heating zones in plan view. The individual heating zones ( 12 . 12 ' . 12 '' . 12n ) are designed as irregular zones. This embodiment may be advantageous in special substrates and / or substrates with special sizes and features.

LIST OF REFERENCE NUMBERS

1

Heating zones grips

1o

Functional surface of the heating zone sample holder

2i

lower substrate

2i

bonding surface of the lower substrate to be bonded

2a

outer bearing surface of the lower substrate

3

upper substrate

3i

bonding surface of the upper substrate to be bonded

3a

outer bearing surface of the upper substrate

4

upper sample holder

4o

Functional surface of the upper sample holder

5, 5 '

vacuum tracks

6

Fusion Bond Investment

7

Bond initiation spin

8, 8 '

active and / or passive position control and / or vibration damping

9

Auxiliary units: electrical system, electronics, control unit, computing unit, vacuum, compressed air

10

Monitoring means for the run-out error

11, 11 ', 11' ', 11' '', 11iv

Signal, supply and data lines

12, 12 ', 12' ', 12n

heating zones

13, 13 ', 13' ', 13' ''

Thermal insulator

14

Pins, pimples

15

Sensors, in particular temperature sensors

QUOTES INCLUDE IN THE DESCRIPTION

This list of the documents listed by the applicant has been generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Cited patent literature

• EP 2656378 [0029]
• US 5094536 A [0032]
• WO 2009/133682 A1 [0032]
• EP 2656378 A1 [0036, 0037, 0041]
• EP 2463892 B1 [0036]
• EP 26563781 B [0037, 0041]

Claims (7)

Receiving device for receiving and holding a lower substrate ( 2 ) with the following features: - a mounting surface ( 1o ), Holding means for holding the lower substrate ( 2 ) on the mounting surface ( 1o ) and - measuring means for determining a non-linear component of a run-out error of the lower substrate ( 2 ). Recording device according to Claim 1, in which compensating means for compensating for a run-out error, in particular before contacting the lower substrate, are detected by the values determined by the measuring means ( 2 ) with an upper substrate ( 3 ), are controllable. Receiving device according to one of the preceding claims, in which by the compensation means the elongation of the support surface ( 1o ), in particular by arrangement of preferably individually controllable, piezoelectric elements on a rear side of the support surface, is locally influenced. Receiving device according to one of the preceding claims, characterized in that by the compensation means the shape of the support surface ( 1o ), in particular by, preferably mechanical, action in a Z-direction, is locally influenced. Receiving device according to one of the preceding claims, in which the mounting surface ( 1o ) from a back side ( 1r ) of the support surface ( 1o ) is acted upon by the compensation means locally, in particular hydraulically and / or pneumatically, with pressure. Recording device according to claim 5, wherein each control element for controlling the compensation means or groups of control elements is separately controllable / are. Device for aligning a first substrate ( 2 ) with a second substrate ( 3 ) having the following features: - a device for determining strains and / or distortions (run-out errors) of the first substrate ( 2 ) with respect to the second substrate ( 3 ), - at least one receiving device according to one of the preceding claims for receiving at least one of the substrates ( 2 . 3 ) and alignment means for aligning the wafers, taking into account the run-out errors and simultaneous compensation by the compensation means.

Images