JP2020123747A - Substrate holder and method for bonding two substrates - Google Patents

Abstract

To provide a substrate holder for controlled compensation of substrate runout error.SOLUTION: In a substrate holder 1 with a position fixing surface 4o for holding a substrate, a desired temperature difference ΔT is adjusted between a lower substrate and an upper substrate by configuring the substrate holder, especially an upper substrate holder such that any heat absorbed from an upper first substrate is discharged to the back side of the substrate holder, and discharged through the heat exchanger in a state of being controlled.SELECTED DRAWING: Figure 1

Description

The invention relates to a substrate holder, a device with such a substrate holder, the use of such a substrate holder, a method for bonding two substrates, and a product manufactured by such a method, in particular a substrate. It concerns a stack and the use of such a substrate holder for such a method.

In the semiconductor industry, multiple substrates, especially wafers, are aligned and bonded to each other by different methods. The process of bonding is called bonding. For best results, it is necessary to use different bonding techniques depending on the materials to be bonded.

Now, in recent years, room temperature bonding techniques have become more and more established, but metals are bonded to each other by, for example, a diffusion process at high temperature and high pressure.

Substrates with surfaces having atoms that preferably form covalent bonds are directly bonded to each other by adhesive forces. However, the adhesive force does not exhibit the maximum bond strength between the surfaces since it is merely a van der Waals bond initially. By appropriate processes, in particular heat treatment, such van der Waals bonds can be converted into covalent bonds. The bonding process in which the bonding of two surfaces by the formation of covalent bonds is carried out is called the fusion bonding process. In recent years, it has become increasingly clear that maximizing the contact area, in particular, makes a decisive contribution to improving such bonding. This opens up a whole new possibility for bonding such surfaces to one another at room temperature, in particular without heat treatment, or with only a very slight increase in temperature. Recent measurement results show that such an optimization can achieve bond strengths which approximately correspond to the theoretical strengths of the materials to be bonded to one another.

In particular in the case of fusion bonding, care must be taken not to uncontrollably distort either of the two substrates before and/or during and/or after alignment, especially by thermal loading. The distortion causes the substrate to expand or contract, which in turn causes displacement and/or misorientation of the marks to be aligned with each other, especially the chips. This displacement and/or misorientation generally increases from the center towards the edges. The error caused by this is known in the prior art, and in particular in the semiconductor industry, under the name run-out. Compensating for this error is called runout compensation. This error will be explained in more detail below.

One of the tremendous technical problems in permanent bonding of two substrates is the alignment accuracy of the functional units between the individual substrates. The substrates can be very accurately aligned with respect to each other by the alignment equipment, but distortion of the substrates can occur during the bonding process itself. Due to the distortion thus created, the functional units are not always properly aligned with respect to each other in all positions. Alignment inaccuracies at specific points on the substrate can result in distortions, scaling errors, lens defects (magnification or reduction errors), and so on. In the semiconductor industry, all thematic areas related to such problems are included in the concept of “overlay”. A corresponding overview of this subject is found, for example, in Mack, Chris, Fundamental Principles of Optical Lithography-The Science of Microfabrication (Publisher WILEY, 2007, Reprint 2012).

Each functional unit is computer designed prior to the actual manufacturing process. For example, conductor tracks, microchips, MEMS, or any other structure that can be manufactured using microsystem technology is also designed in a CAD (computer aided design) program. However, it has been found that during the manufacture of the functional unit, there is always a gap between the ideal functional unit built in the computer and the actual functional unit manufactured in the clean room. The differences may be primarily due to hardware limitations, technical engineering issues, but often also due to physical limitations. That is, the resolution accuracy of the structure manufactured by the photolithography process is limited by the size of the aperture of the photomask and the wavelength of the light beam used. Mask distortion is transferred directly to the photoresist. The linear motor of a machine can only reach a position that is reproducible within specified tolerances. Therefore, it is no wonder that the functional units of the board cannot be exactly equal to the structure built in the computer. Therefore, all substrates already have a negligible amount from their ideal state prior to the bonding process.

Comparing the positions and/or shapes of two functional units of two substrates facing each other, assuming that neither of the substrates is distorted by the coupling process, it is generally already known that It turns out that there is an incomplete match. This is because these functional units deviate from the ideal computer model due to the errors described above. The extremely frequent error is shown in Figure 8 (http://commons.wikimedia.org/wiki/File:Overlay_typical_model_terms_DE.svg, 24.05.2013 and Mack, Chris, "Fundamental Principles of Optical Lithography- The Science of Microfabrication. Chichester". (Publisher WILEY, p. 312, 2007, reprint 2012)). As shown in the drawings, overlay errors can be roughly divided into global overlay errors and local overlay errors, or symmetrical overlay errors and asymmetric overlay errors. The overall overlay error is uniform and therefore location independent. The overall overlay error causes the same deviation between two functional units located opposite each other, regardless of position. Typical global overlay errors are the errors I. and II. caused by the translation or rotation of both substrates relative to each other. The translation or rotation of the two substrates causes corresponding translational or rotational errors on the substrates for all functional units located respectively opposite one another. Local overlay errors occur in relation to location and are mainly caused by elasticity and/or plasticity problems, especially in this case elastic and/or plasticity problems caused by continuously propagating bonding waves. Caused by a problem. Of the overlay errors shown, errors III. and IV. in particular are referred to as "run-out" errors. This "runout" error is caused by distortion of at least one substrate, especially during the bonding process. Due to the distortion of the at least one substrate, the functional units of the first substrate are also distorted with respect to the functional units of the second substrate. However, the errors I. and II. can also be caused by the bonding process, but are usually significantly superimposed by the errors III. and IV., so that the errors I. and II. It becomes difficult to measure.

The biggest problem with bringing two substrates into close proximity is that the surrounding environment is generally not in thermodynamic equilibrium with the substrates. There is always a thermodynamic equilibrium if all thermodynamic variables, especially in the special case temperature, are equal for all subsystems to be considered. In many cases, one of the substrates, especially the one fixed to the lower substrate holder, has a relatively high temperature.

In many cases, it is desirable, or even desirable, to set another temperature for the lower substrate that is higher than the temperature of the upper substrate in order to compensate for the previously described runout error of the substrate in a controlled manner. It is contemplated to do so. In this case, it may be necessary to adjust the temperature of the lower substrate accordingly, especially heating or cooling.

Here, when the first upper substrate fixed on the substrate holder according to the present invention is brought closer to the second lower substrate, the second lower substrate, and especially the lower substrate holder as a whole And heat the upper first substrate, causing it to expand thermally and to be subjected to a particularly complex heating distribution. This heating distribution is obtained by a temperature/time curve. In this case, even a slight temperature difference between the first substrate and the second substrate may cause a significant strain of the upper first substrate, or the upper first substrate may follow a complicated temperature transition. Can heat the substrate. The temperature of the upper substrate rises as the spacing between both substrates increases and remains constant in the saturation region for a short time, after which it is reduced exponentially by further processes, after which the boundary conditions Remains constant if does not change. The prior art is particularly problematic in that the substrates are bonded together in a temperature range where the temperature varies as a function of time. That is, the bonding waves are exposed to different temperatures at different times, or in other words different positions, thereby producing the runout error described above.

It is therefore the object of the present invention to provide an improved substrate holder and an improved method, which are able to overcome the drawbacks of the prior art and in particular to compensate for runout errors and in particular to avoid them altogether. Is.

The above problem is solved by a substrate holder according to the invention, a device according to the invention, a use according to the invention, a method according to the invention and a product according to the invention, as well as the use according to the invention, which are described in the independent claims. ..

Advantageous developments of the invention are described in the dependent claims. All combinations of at least two features mentioned in the description, the claims and/or the drawings are also within the scope of the invention. Within the stated range of values, the values themselves, which are within the stated range, are to be considered as disclosed as limits and should be claimable in any combination.

The core of the invention is that any heat absorbed is controlled in order to be able to adjust the position-fixed, in particular upper substrate (also referred to below as the first substrate) in a targeted manner. The substrate holder according to the invention, in particular the upper substrate holder (hereinafter also referred to as the first substrate holder), in such a way that it is discharged, in particular, to the back side of the substrate holder and is discharged there via a heat exchanger. To do.

A further important feature of the invention is that the thermal resistance of the device is intended to be optimized in order to be able to adjust the temperature difference ΔT according to the invention. is there.

An important feature of the present invention is that the desired temperature difference ΔT between the lower substrate and the upper substrate is adjusted by proper selection of the thermal resistance. This temperature difference ΔT is generally a function of time or a function of the spacing between both substrates. However, according to the invention, the temperature difference ΔT in the temperature range of the temperature saturation of the lower substrate is of primary importance, and in the following description of this specification, this temperature range is marked d. In this temperature region d, it is required to keep the temperature difference ΔT constant. By adjusting and maintaining the temperature difference ΔT in a targeted manner, the adverse “runout” error can be reduced or even eliminated altogether.

The temperature difference ΔT, especially in the temperature saturation region d, is generally (i) due to the thermal resistance and/or (ii) especially due to the heating element of the heating device in the lower substrate holder and/or (iii) cooling It is adjusted as desired by the elements, in particular the cooling fluid.

According to the invention, the substrate holder has a position fixing surface for holding a substrate, the substrate holder preferably and/or the position fixing surface for discharging heat from the position fixing surface. It has a heat conductor for supplying heat to.

A further subject of the invention is a device for bonding a first substrate to a second substrate, said device according to any one of the preceding embodiments for holding at least one of the two substrates. According to claim 1, comprising an apparatus. In this regard, reference is made in particular to the description relating to the substrate holder.

A further subject of the invention relates to the use of a substrate holder according to the invention as an upper substrate holder.

A further object of the invention, inter alia independent, is a method for bonding a first substrate to a second substrate, wherein in a first step the substrates are brought close to each other, whereby the first substrate The temperature of the substrates rises, and in the second step, the substrates are stopped from approaching each other, and the distance between the substrates is kept constant, and when the distance is constant, at least a predetermined period of time is maintained. A constant temperature of one substrate is generated, and in a third step, the two substrates are at least temporarily bonded to each other within the time period in which the first substrate is at a constant temperature.

This situation can also be explained as that the temperature difference ΔT between both substrates is constant in the well-defined temperature region d. Further, the magnitude of the temperature difference ΔT can be adjusted by properly selecting the thermal resistance.

A further subject of the invention relates to a product having a first substrate and a second substrate, in particular a substrate stack, wherein said substrates are bonded to one another by the method according to the invention.

A further subject of the invention relates to the use of such a substrate holder for holding a substrate during the performance of such a method.

In general, the substrate holder, especially the upper substrate holder, should be thermally coupled to ambient temperature as best as possible. This can result in heat supply and/or heat removal. By bringing the two substrates close together, in particular the upper substrate is heated by the lower substrate or the lower substrate holder. Due to the particularly large thermal mass of the upper substrate holder and the highest possible thermal conductivity of this substrate holder, heat is dissipated from this upper substrate. Therefore, in the substrate holder according to the present invention, the temperature distribution of the substrate holder when approaching the lower substrate or the lower substrate holder, and particularly also the temperature distribution of the upper substrate, can be adjusted as desired. Is configured.

In this case, the thermal resistance of the substrate holder according to the invention is designed in such a way that the temperature adaptation of the heat conductor, and thus of the upper substrate, to the cooling fluid takes place as quickly and efficiently as possible. Therefore, thermal resistance is preferably minimized. The cooling fluid is preferably ambient. Therefore, the temperature of the cooling fluid is preferably room temperature.

By knowing the temperature-time curve in the upper substrate holder in particular or in the upper substrate in particular, it is possible to identify a point in time at which the throughput is increased at the same time which is optimal for the bonding in particular. The corresponding process or method is also an important, especially independent, and inventive feature of the present invention.

All temperature distributions disclosed herein can be considered as the temperature distribution of the substrate on the substrate holder or the temperature distribution of the substrate holder. The thermal coupling between the substrate and the substrate holder is preferably efficient so that the temperature shift can be ignored. In reality, the temperature of the lower substrate during heating of the lower substrate holder may be slightly lower than the temperature of the lower substrate holder. The temperature of the upper substrate is generally slightly higher than the temperature of the upper substrate holder according to the invention. This slight temperature difference is associated with a non-zero thermal resistance between the substrate holder and the substrate.

According to the substrate holder according to the invention, which is also referred to below as the sample holder, any heat generated as already explained above is radiated to the back side in a controlled manner, where it is converted via a heat exchanger and according to the invention It is ejected from the substrate holder. Furthermore, the large thermal mass of the upper substrate holder, among other things, results in temperature stabilization of the upper substrate, among others, which minimizes thermal fluctuations of the surrounding ambient environment. A further important feature of the present invention is that the relatively large thermal mass stabilizes the temperature of the upper substrate during the bonding process or the temperature difference ΔT between the lower and upper substrates. is there.

Furthermore, by knowing the heat transport through the substrate holder according to the invention, it is possible to specify, inter alia, the temperature-time diagram for the upper substrate holder or the upper substrate, and also for the substrate holder according to the invention. This diagram can be changed by changing the parameters.

The substrate holder according to the invention can be used as an upper and/or lower substrate holder. The substrate holder according to the invention is designed in particular as an upper substrate holder, so that the upper first substrate, which is fixed on it, is deformed in the direction of gravity, unless it is fixed on the whole surface. It

In the following, reference will be made many times to the roughness of the surface. Roughness is expressed herein as either arithmetic mean roughness, squared roughness, or mean roughness depth. The calculated value of arithmetic mean roughness, the calculated value of squared roughness, and the calculated value of mean roughness depth are generally different even if the measurement path or measurement surface is the same, but within the same number of digits range. is there. Therefore, the following numerical ranges for roughness should be understood as being values of either arithmetic mean roughness, square roughness or mean roughness depth.

The substrate holder according to the invention is capable of heating and/or cooling the upper first substrate, among others. The heat is dissipated via the heat conductor, in particular from the upper first substrate, and is preferably conducted to the cooling fluid. In this case, the heat conductor is a cooling body. However, it is also conceivable that the fluid is a heating fluid that radiates heat to the heat conductor to heat the upper first substrate. In this case, the heat conductor is a heating body.

The cooling fluid is preferably ambient. The temperature of the cooling fluid is preferably room temperature.

In a preferred embodiment, the heat conductor has ribs for discharging and/or supplying the heat, in particular on the side of the heat conductor opposite the fixing surface (also referred to below as backside). .. These ribs can be arranged above all over the back side of the heat conductor, which makes it possible to improve the heat exchange.

The heat can be distributed via the ribs along a relatively large surface, the so-called rib surface. The ribs can be arranged, inter alia, perpendicular to the fixing surface. The ribs are preferably arranged parallel to each other. If a heat conductor that functions as a cooling body is used, it is a cooling rib. If heat conductors functioning as heating bodies are used, the ribs can also be referred to as heating ribs, these heating ribs optimally conducting heat from the fluid to the heat conductor. The ribs will now be described in the following description. Unless explicitly stated, in the following description, heat conductors are mainly regarded as cooling bodies, ribs are regarded as cooling ribs, and fluids are regarded as cooling fluids.

Embodiments of the substrate holder according to the invention are preferably designed such that the ribs are arranged in the encapsulation, for example in the housing. The encapsulation part preferably has at least two access parts. One of the access parts is used for supply of fluid and the other for discharge. This allows the fluid to flow continuously through the ribs of the heat conductor, in particular spatially isolated from the surrounding environment. Such a compact structure also makes it possible to isolate the embodiment according to the invention from surrounding components. If the cooling is gas cooling, in particular air cooling, then it may be sufficient to let the fan flow a gas stream, in particular an air stream, to the ribs in order to ensure efficient cooling. In a very particularly preferred embodiment, the cooling ribs are cooled only by the ambient atmosphere.

Preferably, the flow rate of the fluid can be controlled. In this case, the flow velocity is faster than 1 mm/sec, preferably faster than 1 cm/sec, even more preferably faster than 10 cm/sec, most preferably faster than 1 m/sec. The compact encapsulation also allows the fluid to be placed under pressure. In this case, the pressure of the fluid preferably corresponds to the ambient pressure. However, the fluid can also be used under superatmospheric pressure. In this case, the pressure is more than 1 bar, preferably more than 2 bar, even more preferably more than 5 bar, most preferably more than 10 bar, most preferably more than 20 bar. The fluid supply to the encapsulation and thus to the ribs is preferably carried out via a hose system connected to the access.

Optional Cooling and Heating Elements In addition to the heat conductors described below and the heat exchangers behind them, the substrate holder according to the invention provides additional actively controllable cooling and/or heating elements. Can have These additional cooling and/or heating elements are preferably incorporated in the substrate holder according to the invention, in particular in the heat conductor. It is also conceivable to mount cooling and/or heating elements around the heat conductor in order to keep it as homogeneous as possible and to prevent temperature jumps due to the additionally incorporated components.

The heating element is preferably an induction heater. However, since it is only necessary to perform temperature compensation for a relatively small temperature difference, it is possible to control more accurately, more quickly, and more efficiently, and the temperature of the heat conductor can be controlled by radiant heat. It is also conceivable to place an infrared source, which can be raised in the range of °C, on the side surface of the heat conductor.

It is also possible for the cooling element to be an additionally installed Peltier element, which is independent of the original heat conductor according to the invention and which is an additional substrate holder according to the invention, in particular a heat conductor. Allows for efficient cooling. The Peltier element is preferably mounted outside the heat conductor so as not to destroy the material homogeneity of the heat conductor.

The original inventive feature of the present invention is the heat conductor.

Thermal conductor A thermal conductor is a component that has the greatest possible thermal mass. Thermal mass is the product of the specific heat capacity and the mass of the body. If the density distribution is constant, the product of density and volume may be used instead of mass.
C _th =m· _cm =ρ·V· _cm

The concept of thermal mass is mainly used in the engineering field. In the scientific field heat capacity is used, which is a more commonly used concept. The unit of heat capacity is J/K. Heat capacity is a measure of an object's ability to store heat at a particular temperature. Objects with a large heat capacity are regenerators that can be used as cushioning elements.

In general, if the temperature Tk of the cooling fluid used is different from the temperature of the upper substrate, the temperature gradient will decrease via the heat conductor. It is also possible to consider the average temperature instead of the temperature gradient. In the following description of the present specification, Tw is attached to the temperature gradient or the average temperature. The temperature of the cooling fluid is preferably kept constant during the performance of the process according to the invention, while the temperature gradient or the average temperature Tw generally changes. The temperature Tw preferably always coincides with the temperature of the upper substrate and deviates only slightly from this temperature.

According to an important recognition according to the invention, the temperature of the upper substrate and the temperature of the heat conductor according to the invention or of the upper substrate holder according to the invention, especially if the thermal resistance Rth4 between the two substrates is infinite. Should correspond to the temperature of the cooling fluid, namely to the ambient temperature, among others. However, due to the finite value of the thermal resistance Rth4, a heat flow rate from the lower substrate to the upper substrate may occur.

According to the invention, in order to reduce or preferably completely eliminate the "runout" error, in particular the temperature difference ΔT between the distances d between the substrates is known and adjusted as desired. What is possible is especially important.

The object according to the invention of the embodiment according to the invention is to discharge the temperature of the substrate in a controlled manner as much as possible and to strongly stabilize it accordingly, so that the heat conductor has a heat capacity as high as possible. The heat capacity of the heat conductor is as large as possible in order to enable efficient heat storage or to compensate for thermal fluctuations as efficiently as possible. The temperature stability is also reflected in the stability of the temperature difference ΔT. For the majority of solids, the heat capacity at constant volume differs only slightly from the heat capacity at constant pressure, given moderate temperature and pressure. Therefore, the following description does not distinguish between these two heat capacities. Furthermore, the specific heat capacity is indicated. The specific heat capacity of the heat conductor is especially greater than 0.1 kJ/(kg·K), preferably greater than 0.5 kJ/(kg·K), and even more preferably greater than 1 kJ/(kg·K), and most It is preferably greater than 10 kJ/(kg·K), and most preferably greater than 20 kJ/(kg·K). If the density and geometry of the heat conductor is known, the specific heat capacity can be converted to an absolute heat capacity by the above equation.

The material of the thermal conductor should have the highest possible thermal conductivity, as it must transport heat as quickly as possible. The thermal conductivity is between 0.1 W/(m·K) and 5000 W/(m·K), preferably between 1 W/(m·K) and 2500 W/(m·K), and even more preferably 10 W. /(M·K) to 1000 W/(m·K), most preferably 100 W/(m·K) to 450 W/(m·K). Copper, the most commonly used constituent material for dissipating heat, has a thermal conductivity of, for example, about 400 W/(m·K). Thermal conductivity defines how much energy is transferred per unit time over a given distance for a given temperature difference. The energy or heat quantity transported per unit time is called the heat flow rate. The heat flow rate is greater than 1 J/sec, preferably greater than 10 J/sec, even more preferably greater than 100 J/sec, most preferably greater than 200 J/sec, and most preferably greater than 500 J/sec.

The heat conductor is cooled, preferably actively or passively, on the back side of the heat conductor. Passive cooling is carried out, inter alia, by radiating heat through as large a surface area as possible. Active cooling is performed by a cooling fluid. The cooling fluid can be a gas or a liquid. For example, the following can be considered:
●Liquid, especially ○Water ○Oil ●Gas, especially ○Rare gas ・Helium ・Argon ○Molecular gas ・HFCKW
・HFKW
・FCKW
・PFKW
・CO2
・N2
・O2
●Mixed gas, especially ○Air, especially ・Ambient air

The cooling fluid absorbs heat via the heat conductor and is thereby heated thereby simultaneously cooling the heat conductor. The heated cooling fluid is preferably circulated in the cooling circuit, discharging heat at different points in the circuit system, at which time it is cooled again and fed back to the cooling circuit. A cooling gas is preferably used because it is easy to handle. If the cooling fluid is ambient air, the cooling is accomplished by discharging heat from the heat conductor into the ambient air. The locally heated ambient air then propagates in the ambient atmosphere, resulting in temperature adaptation and cooling.

Dissipating the heat through the relatively large surface improves the efficiency of heat dissipation or heat transfer to the cooling fluid. By consciously increasing the surface roughness, the surface area can be increased even further. In this case, the roughness is greater than 10 nm, preferably greater than 100 nm, even more preferably greater than 1 μm, most preferably greater than 10 μm, and most preferably greater than 100 μm.

It is also conceivable to use a heat conductor without ribs, which can simplify the production of the heat conductor.

In a further embodiment according to the invention it may be conceivable to provide open pores on at least the front side of the heat conductor. In this case, the pore size is larger than 100 nm, preferably larger than 1 μm, even more preferably larger than 10 μm, most preferably larger than 100 μm, and most preferably about 1 mm. The cooling fluid flows through the open pores, where it more efficiently absorbs heat due to its large surface area. In order to further increase the surface area of the ribs, it is also conceivable to provide the ribs with open pores only.

The main problem of the substrate holder according to the invention, in particular of the heat conductor, is the regulation and stabilization of the temperature of the substrate, or the regulation and stabilization of the temperature difference between the lower substrate and the upper substrate. Is. In addition to this, the substrate holder according to the invention supplies heat to and/or removes heat from the substrate as the substrate needs to be cooled and/or heated. The substrate holder according to the invention makes it possible, in particular, to adjust the maximum temperature or the temperature difference ΔT between the upper and the lower substrate in a targeted manner, and more particularly to both substrates. The temperature stability of the maximum temperature or the temperature difference ΔT is guaranteed for the same period of time required for bonding, and even more preferably for a longer period.

In the further description, reference is made to a plurality of embodiments according to the invention, each differing in at least one characteristic. All the embodiments according to the invention described above can be arbitrarily combined with one another so as to make it possible to create further corresponding embodiments according to the invention combining the features described above.

In an exemplary embodiment according to the invention, the substrate holder according to the invention has a separate position fixing part on which the heat conductor is arranged. That is, the heat conductor and the position fixing part are two different components, which are different from each other but are coupled to each other. The most efficient thermal coupling of the two components is carried out via a surface which is as flat as possible. In this case, the roughness of the contacting surfaces of the position fixing part or the heat conductor is less than 100 μm, preferably less than 10 μm, even more preferably less than 1 μm, most preferably less than 100 nm, most preferably less than 10 nm. is there. Heat transfer can be further improved by using a thermally conductive paste.

In a further preferred embodiment, the fixing surface is integrally formed with the heat conductor. In other words, the heat conductor itself is configured as the position fixing portion. The heat conductor and the position fixing portion or the position fixing surface are integrally formed. However, it does not critically affect the functionality of the present invention and will not be further handled, illustrated, or described. According to the embodiment of the present invention, since there is no boundary surface between the position fixing portion and the heat conductor, it is possible to improve heat conduction.

Since the embodiment according to the invention in which the heat conductor is integral or consists of one part is the optimum embodiment according to the invention, all variants below are referred to as being related to this prototype. To do. Therefore, the position fixing part and the heat conductor are used as synonyms in the following description.

In a further particularly preferred embodiment according to the invention, the substrate holder comprises at least one, in particular movable, preferably drivable deformation element for deforming the substrate, the at least one deformation element comprising: It is preferably arranged in the center of the substrate holder. The at least one deformable element may be movable, in particular perpendicular to the fixed surface or the substrate to be fixed, and in particular drivable. The at least one deformation element is preferably configured to allow the substrate to deform in a direction away from the position fixing surface. The substrate holder or the heat conductor can have, inter alia, a centrally formed and/or a coherently extending hole, in which the at least one deformation element is particularly movable, preferably Is drivably arranged or this hole allows access to at least one deformation element which is capable of deforming the fixed substrate. The at least one deformation element is, for example:
● Pin ● Spike ● Ball ● Nozzle, especially ○ Gas nozzle

The deformation element is manipulated or controlled in such a way that it can be deformed in a targeted manner to deform the substrate at least locally, preferably in the center. In this case, the deformation is preferably concave when viewed from the side of the deformation element. Deformation is used, inter alia, for the process of separating the substrate from the fixed part or fixed surface.

In a further embodiment according to the invention, the heat conductor has at least one contact with the fixing surface in order to ensure that the contact between the substrate and the fixing surface or the contact between the substrate and the material of the heat conductor is as low as possible. It has three recesses and/or recesses. This reduces the so-called effective position fixing surface area. The effective position fixing surface area is the area of the position fixing surface that is actually in contact with the substrate. Preferably, at least one recess is arranged on the fixing surface so that the substrate can be held away from the fixing surface. The advantage of this embodiment according to the invention is that the contamination of the substrate by the surface of the heat conductor is reduced. In order to carry out the heat transfer efficiently, a gas having a correspondingly high thermal conductivity and a correspondingly high heat capacity is introduced, in particular into at least one recess and/or recess. You can In this case, the substrate is fixed only in a few fixing elements, especially in the peripheral and/or central fixing elements. Such an embodiment is disclosed in International Publication No. WO 2013/023708 (WO2013/023708A1), and the disclosure content of the document regarding this embodiment is explicitly included in the disclosure content of the present application.

In a further embodiment according to the invention, protrusion-shaped (protrusions) and/or needle-shaped and/or column-shaped elements are arranged in the at least one recess, by means of which these elements move the substrate from the fixing surface. They can be held apart and these elements can extend sharply in the direction of the substrate, among others. These elements reach the surface of the heat conductor and support the fixed substrate. In order to ensure a thermal coupling between the fixed substrate and the heat conductor, in this embodiment according to the invention a fluid with a large heat capacity is provided in the intermediate space of the protrusions and/or the needles and/or columns. It is also possible to let it flow.

Position Fixing Element All embodiments according to the invention of the present disclosure are able to fix a substrate, in particular a wafer, and even more preferably a semiconductor wafer. In this case, the position fixing can be performed by any arbitrary position fixing element. Preferably, a position fixing element for fixing the substrate is arranged inside, in particular, and/or above the position fixing surface, in particular entirely. The following can be considered:
●Vacuum position fixing ●Electrostatic position fixing ●Magnetic position fixing ●Mechanical position fixing, especially ○ Clamp ●Adhesive position fixing, especially ○Adhesive film position fixing

Particularly preferred are vacuum position fixing or vacuum passages (hereinafter also referred to as vacuum channels), which are arranged entirely distributed over the position fixing surface. Vacuum fixturing consists of multiple vacuum channels terminating in vacuum openings provided in the fixturing surface of the substrate holder.

In another embodiment according to the invention, the vacuum channels are coupled to one another so that the vacuum channels can be evacuated and/or inflated simultaneously.

In another embodiment according to the invention, at least the individual vacuum channels are coupled to one another and form a corresponding group of vacuum channels. In this case, it is possible to control each vacuum channel group individually, whereby position fixing and/or dissociation of each section of the substrate can be realized. In a particular embodiment according to the invention, the plurality of vacuum openings are arranged in a plurality of centered circles each having a different radius to form a vacuum channel group. Advantageously, all the vacuum channels contained in the same circle are controlled simultaneously, so that the fixing and/or dissociation of the substrate starts at the center and progresses radially outwards. You can This provides a particularly efficient means for controlled fixation and/or dissociation of the substrate.

Thermal Resistance: Equivalent Circuit Diagram A further important inventive feature of the invention is, inter alia, the optimization of the heat flux through the substrate holder according to the invention. The heat flux between the heat source and the heat sink is critically influenced by the thermal resistance. Each static multiparticulate system, and thus fluids such as gases and liquids, and solids have thermal resistance. The definition of thermal resistance is well known to those skilled in the art. Thermal resistance is not a pure material parameter. Thermal resistance depends on thermal conductivity, thickness, and cross-sectional area.
R=d/(A·λ)

In the rest of the description, it is assumed that the heat flow always flows through the same cross-sectional area, so the thermal resistance at a constant cross-sectional area is determined by the thermal conductivity and thickness of each material under consideration. Should be regarded as a function of. Thermal resistance is abbreviated in the drawings using Rth and subscripts. According to the invention, there are, inter alia, eight associated thermal resistances. Rth1 to Rth8 are (i) thermal resistance of the lower substrate holder, (ii) thermal resistance of fluid or vacuum between the lower substrate holder and the lower substrate, and (iii) heat of the lower substrate. Resistance, (iv) thermal resistance of fluid or vacuum between both substrates, (v) thermal resistance of upper substrate, (vi) thermal resistance of fluid or vacuum between upper substrate and upper substrate holder, ( vii) the thermal resistance of the heat conductor, and (viii) especially the thermal resistance of the fluid flowing between the cooling ribs.

The heat flow rate is directly proportional to the temperature difference applied between the heat source and the heat sink. Thermal resistance is a proportional constant. Therefore, the following applies.
R=(1/ΔT)・(dQ/dt)

Among other important inventive features of the present invention is to minimize the thermal resistance above and/or below the substrates and maximize the thermal resistance between the substrates. Therefore, according to the invention, the thermal resistance should be designed, inter alia, as follows:
Rth1 is minimized by choosing a material with a particularly high thermal conductivity,
Rth2 is minimized by choosing a fluid with a particularly high thermal conductivity,
Rth3 should be minimized by choosing a substrate with high thermal conductivity,
Rth4 is maximized by passing a gas with a particularly low thermal conductivity, and/or by a vacuum, and/or by optimized process control, inter alia by carefully selecting the spacing,
Rth5 should be minimized by choosing a substrate with high thermal conductivity,
Rth6 is minimized by choosing a fluid with a particularly high thermal conductivity,
Rth7 is minimized by choosing a material with a particularly high thermal conductivity, and/or Rth8 is minimized by choosing a fluid with a particularly high thermal conductivity.

Among other important features of the embodiments according to the present invention are that the temperature of the upper substrate or the temperature difference ΔT between the lower substrate and the upper substrate can be adjusted as desired and the bonding process It is to keep it as constant as possible during. This is achieved by the correct choice of thermal resistance according to the invention. By maximizing the thermal resistance Rth4, the heat flow from the lower substrate to the upper substrate is minimized and, on the contrary, preferably completely blocked. However, since the complete cutoff of the heat flow rate cannot be actually realized, the temperature change of the upper substrate always occurs in practice. The temperature difference ΔT is especially below 20° C., preferably below 10° C., even more preferably below 5° C., most preferably below 1° C., most preferably below 0.1° C.

On the other hand, above all, it should be possible to precisely adjust the temperature of the lower substrate by means of a heating device in the lower substrate holder. In particular, the temperature of the lower substrate should match the temperature of the lower substrate holder. The lower substrate holder is thermostated above all above 100° C., preferably below 75° C., even more preferably below 50° C. and most preferably below 30° C.

Furthermore, above all, the temperature of the upper substrate should be matched to the temperature of the cooling fluid and/or the heat conductor. In a very special embodiment according to the invention, the temperature of the cooling fluid particularly corresponds to the ambient temperature. This is especially true when the atmosphere itself is used as a cooling fluid. The cooling fluid is thermostated to a temperature above 100°C, preferably below 75°C, even more preferably below 50°C, most preferably below 30°C. In a very special embodiment according to the invention, the ambient atmosphere is used as cooling fluid and thus has room or ambient temperature.

The diameter of the substrate cannot be changed. The thermal conductivity and thickness of the substrates used are also often pre-determined by manufacturing conditions and are therefore often unavailable for optimization according to the invention. By properly selecting the thermal resistance according to the invention, among others, the heat flux from the lower substrate to the upper substrate is preferably minimized and the heat flux from the upper substrate to the cooling fluid is maximized. .. Therefore, the temperature difference ΔT remains constant according to the invention.

A further goal of the choice of thermal resistance according to the invention is, inter alia, to keep the temperature of the upper substrate constant, in particular at ambient temperature, so that the influence of other heat sources, in particular of the lower substrate, is minimized. It is to turn. If the temperature of the lower substrate holder, and thus of the lower substrate, is kept constant, this is especially true during the bonding process in the temperature region d between the upper substrate and the lower substrate. This is synonymous with maintaining the temperature difference ΔT of. This is achieved in particular by maximizing the thermal resistance Rth4 between the substrates. On the other hand, the temperature T1u of the lower substrate should be controllable as efficiently as possible by the heating device. In this case, Tp is attached to the temperature of the lower substrate holder. Preferably, the temperature Tp of the lower substrate holder is the same as the temperature T1u of the lower substrate at any time point. The conduction of heat from the heater to the underlying substrate is achieved especially by minimizing the thermal resistances Rth1 and Rth2.

Process The method according to the invention or the process according to the invention can be explained on the basis of a so-called temperature-time diagram. In this temperature-time diagram, the temperature, in particular the temperature T of the substrate fixed in the substrate holder according to the invention, is shown as a function of time t (temperature graph). In this case, the temperature is shown on the leftmost vertical axis of the temperature-time diagram. An interval/time curve (an interval graph) can be described in the temperature/time diagram, and from this curve, it is possible to read at what point and how large the interval between both substrates is. In this case, the vertical axis of the interval/time curve is shown at the right end of the temperature/time diagram. The space-time curve is preferably logarithmically scaled (scaled) as it represents the space from mm to nm range. However, for the sake of clarity, the interval-time curves in the drawing are shown by linear scaling. In the following description, for simplification, reference is made only to the temperature/time diagram, or the Tt diagram for short. Besides the Tt diagram for a position-fixed substrate, it is also possible to describe the Tt diagram for a substrate holder according to the invention. However, these two T-t diagrams differ only slightly from each other, especially in view of the minimal deviation along the temperature axis. Therefore, in the following description of the present specification, the T-t diagram is used synonymously for the temperature-time diagram of a positionally fixed substrate and/or a substrate holder according to the invention. This assumption is justified especially when the thermal resistances Rth2 and Rth6 are minimal. In this case, the thermal coupling between the substrate holder and the substrate is so good that it can be assumed that the temperature of the substrate holder and the temperature of the substrate are more or less the same.

Each diagram can be divided into six sections in general, and time sections in particular.

In the first first section a, the substrates are brought close to each other in a state where the distance is relatively large. The spacing between both substrates in section a is greater than 1 mm, preferably greater than 2 mm, even more preferably greater than 3 mm, most preferably greater than 10 mm, and most preferably greater than 20 mm. The movement of the substrate in section a is such that the temperature rise by the other substrate, in particular the lower second substrate, or the other substrate holder, which in particular can be heated above room temperature, in particular the lower second substrate holder. Does not cause a temperature rise. To the extent that the effect of thermal radiation of the second lower substrate or the second lower substrate holder and/or the effect of thermal convection of the gas around the upper first substrate occurs. As the spacing decreases, a moderate temperature rise occurs on the upper first substrate.

The region b where the temperature rises gently is called a coarse approach region. In this case, the distance between both substrates is between 10 mm and 0 mm, preferably between 5 mm and 0 mm, even more preferably between 1 mm and 0 μm, most preferably between 100 μm and 0 μm.

When the substrates are brought closer to each other, the temperature of the upper first substrate is rapidly increased at the end of the coarse approach region b. There is a kind of thermal bond between the two substrates. Since the interval/diameter ratio of the substrate distance to the substrate diameter is small, the upper first substrate is heated by heat. The ambient gas heated by the thermal radiation can no longer diffuse at a sufficient rate from the intermediate space between the two substrates, so that heat is preferably transferred directly from the lower second substrate to the upper first substrate. To communicate. Similar considerations apply to thermal radiation, which only substantially has the possibility to reach the surface of the upper first substrate. This area where the substrate is heated strongly is called the proximity approach area c. In this case, the distance between both substrates is between 1 mm and 0 mm, preferably between 100 μm and 0 μm, even more preferably between 10 μm and 0 μm, and most preferably between 1 μm and 0 μm.

The transition of the temperature distribution from the close proximity region c to the so-called temperature saturation region d is preferably carried out by a transition that is mathematically as stationary as possible but indistinguishable. It is also conceivable to carry out the transition continuously, so that the separation of the regions c and d cannot be carried out uniquely. The shape of the temperature-time diagram looks like a "shark fin". However, other shapes are also conceivable.

The bonding process according to the invention is preferably carried out in the temperature saturation region d. The translational approach between the substrates is stopped, ie the spacing between the substrates remains constant. At this point, the upper first substrate has a constant temperature T4o for a well-defined period t1 corresponding to the length of the temperature saturation region d. A constant temperature T4o means that the maximum temperature fluctuation is at most 4K, preferably at most 3K, more preferably at most 2K, most preferably at most 1K, most preferably at most 0.1K. .. The spacing between both substrates is constant in this region and is between 1 mm and 0 mm, preferably between 100 μm and 0 μm, even more preferably between 10 μm and 0 μm, most preferably between 1 μm and 0 μm. In a special embodiment according to the invention it is also possible to bring the two substrates closer together in the area d. However, it should be noted that in this case, sufficient time still has to be left for the actual bonding process. Furthermore, in the temperature saturation region d, the temperature difference ΔT between the lower substrate and the upper substrate remains constant. In this case, the variation of the temperature difference ΔT is less than 4K, preferably less than 3K, even more preferably less than 2K, most preferably less than 1K, and most preferably less than 0.1K. In particular, the temperature difference ΔT can be adjusted accurately and reproducibly by the choice of the thermal resistance and/or the heat source, especially the heater in the lower substrate holder, and/or the choice of the heat sink, especially the cooling fluid.

In particular, the period t1 in which the interval d3 is constant and the constant temperature T4o occurs is longer than 5 seconds, preferably longer than 10 seconds, more preferably longer than 15 seconds, still more preferably longer than 20 seconds. Long, most preferably longer than 40 seconds. This advantageously leaves sufficient time for the bonding process.

Furthermore, inter alia the period t1, the interval d3, and/or the constant temperature T4o may be determined before the first step, inter alia by experience, preferably taking into account the temperature of the second substrate, and It is determined by taking into consideration the material of the substrate holder, the heat conductor, and/or the substrate, and/or the approach speed. Therefore, it is particularly advantageous to define or calibrate the method before the first step so that it is possible to determine the optimal parameters of the method.

The bonding process, in particular the fusion bonding process, requires a period t2 which is not more than the length of the period t1. A further important feature according to the invention is that the bonding process is preferably carried out within the temperature saturation region d at a given temperature T4o. This has the advantage that the bonding process can be carried out without changing the temperature of the first substrate, which avoids or at least reduces the runout errors described above. It will be possible.

In the subsequent cooling zone e, the upper first substrate is cooled, in particular exponentially.

Finally, in the region f following this, a constant saturation temperature occurs which is higher than the initial temperature of the upper first substrate in the first section a before the approaching process. However, this saturation temperature is generally lower than the temperature of the lower second substrate or substrate holder. It is also conceivable to carry out the bonding process in region f at temperature T6o.

Preferably, before using the method according to the invention, all the required physical parameters are obtained, which allow to obtain an accurate judgment regarding the temperature-time diagram. By varying the physical parameters, it is possible during the actual bonding process to ensure that a temperature-time distribution is created which allows optimal bonding of both substrates and in particular also a corresponding throughput. It is necessary to change the method. By using a corresponding heat conductor according to the invention with a corresponding thermal mass, a correct cooling fluid, a correct cooling fluid pressure, a correct cooling fluid flow rate, a correct proximity distribution, etc., the saturation temperature T4o in the area d, the area d The period t1 as well as all other desired regions of the temperature/time diagram can be adjusted accordingly.

Once the system was calibrated for temperature-time behavior, it was also ensured that the upper first substrate had a well-defined temperature at a well-defined time, and that this temperature was reached. It is also ensured from the beginning that, among other things, a well-defined time is available for carrying out the actual bonding process by bending and/or releasing the locking caused by the vacuum, among other things. The fact that the bonding is possible already in the early part of the area d provides two important features which are the basis of the invention. First, the bonding can be started early, which greatly increases throughput. Second, it ensures that the substrate has a very constant temperature within a well-defined period. This allows the invention to completely avoid the runout problem, which is quite well known in the prior art. It is ensured that both substrates have a substantially constant temperature during the region d and that the temperature of the region d remains substantially unchanged during the bonding process. In this connection, it should be explicitly mentioned again that the above-mentioned constant temperature condition does not mean that both substrates have to have the same temperature. It may be highly desirable to preheat at least one of the two substrates to a relatively high temperature or to cool to a relatively low temperature, by forcing the thermal expansion to be as desired. The desired compulsory board size is adjusted, and only then does the matching of both functional units of both boards. However, according to the present invention, these once adjusted temperatures are kept constant during the bonding process.

In any of the methods according to the invention described above, the substrate can be pretreated and/or posttreated. As the pretreatment, the following can be considered in particular.
● cleaning, especially by chemical processes, especially by liquid, especially by
● By water ○ By physical process, especially ・By sputtering, especially
● Especially by Aeon
○By plasma activation
● by uncharged particles ● grinding ● polishing ● alignment, especially ○ mechanical alignment and/or ○ optical alignment ● deposition

The following can be considered as post-processing, among others.
● cleaning, especially by chemical processes, especially by liquid, especially by
● By water ○ By physical process, especially ・By sputtering, especially
● By ion
● By non-charged particles ● Grinding ● Polishing ● Inspection, especially ○ Bonding interface inspection, especially ・Regarding defects (voids) ・Alignment error
● Regarding runout error ● Heat treatment, especially ○ Inside furnace ○ Heater plate ● Above all, re-peeling of substrate by the method described in WO 2013/091714 (WO2013/091714A1)

Embodiments according to the invention make it possible in particular to compensate for runout errors known in the prior art. Therefore, in some cases, the alignment accuracy was sufficiently minimized for the purpose of peeling the substrates from each other again by a special method, particularly by the method described in WO 2013/091714 (WO2013/091714A1). It is especially important to inspect the bonding interface after bonding both substrates to confirm This avoids the loss of both substrates or the entire substrate stack and allows the substrates to be realigned and bonded to each other if desired.

The alignment accuracy achievable by the device according to the invention or the process according to the invention is better than 100 μm, preferably better than 10 μm, even more preferred better than 500 nm, most preferred better than 200 nm, Most preferably, it is better than 100 nm. The alignment accuracy is, in particular, the same everywhere in the substrate stack, which is a crucial and characteristic feature of the successful compensation of runout errors. In this case, the standard deviation of the alignment accuracy, calculated by averaging all the alignment errors of the substrate stack, is less than 1 μm, preferably less than 500 nm, even more preferably less than 250 nm, most preferably less than 100 nm, especially most preferably Is less than 50 nm.

After the bonding process according to the invention, the substrate is optionally heat-treated if the result of the test, which is optionally but preferably performed, is positive. This heat treatment is necessary especially for fusion bonded substrates. In this case, the heat treatment produces a permanent bond between the two substrates which can no longer be dissociated. If a heat treatment of the substrate is no longer required after the bonding process according to the invention, the heat treatment is accordingly omitted.

In the method according to the invention, the bonding of both substrates is carried out by deforming the first upper substrate, in particular in region d. The deformation is preferably carried out centrally by the deformation elements already described. The advantage of the first process according to the invention lies in particular in the throughput. Since the bonding process has already been carried out in section d and there is no need to wait for the cooling of the upper first substrate, the throughput (ie the number of substrates that can be processed per unit time using an embodiment according to the invention) is It can be increased compared to technology. The cooling of the upper first substrate is a process which is adapted mainly to the ambient temperature and/or to a predetermined ambient temperature by the lower second substrate or the lower second substrate holder.

In another process according to the invention, the bonding of both substrates is carried out by deforming the upper first substrate in the region f in particular. The deformation is preferably carried out centrally by the deformation elements already described.

The temperatures T4o, T6o can be varied and optimally adjusted by the substrate holder according to the invention, inter alia by thermal mass, cooling elements and devices, cooling processes, cooling fluids, etc.

Further advantages, features and details of the invention will become apparent from the following description of a preferred embodiment, based on the drawings.

FIG. 3 is a schematic cross-sectional view, out of scale, of a first embodiment of a substrate holder according to the present invention. FIG. 6 is a schematic cross-sectional schematic view of a second embodiment of a substrate holder according to the present invention, not to scale. FIG. 6 is a schematic cross-sectional view, out of scale, of a third embodiment of a substrate holder according to the present invention. FIG. 6 is a schematic cross-sectional view, out of scale, of a fourth embodiment of a substrate holder according to the present invention. FIG. 9 is a schematic cross-sectional view, out of scale, of a fifth embodiment of a substrate holder according to the present invention. FIG. 6a is a schematic cross-sectional view of the first step of the method according to the invention in non-scale, FIG. 6b is a schematic cross-sectional view of the second step in non-scale, and FIG. Figure 6d is a schematic cross-sectional view of the third step in non-scale, Figure 6d is a schematic cross-sectional view of the fourth step in non-scale, and Figure 6e is a schematic cross-section in the fifth step in non-scale. FIG. 6f is a schematic cross-sectional view, out of scale, of the sixth step, FIG. 6f. FIG. 7a is a schematic diagram of a first temperature/time diagram and interval/time diagram, and FIG. 7b is a schematic diagram of a second temperature/time diagram and interval/time diagram. FIG. 6 is a schematic diagram of possible overlay errors. It is a schematic diagram of a thermal equivalent circuit.

In the drawings, the same reference numerals are given to the same components and components having the same function.

FIG. 1 shows a first embodiment according to the invention of a substrate holder 1 having a position fixing part 4 and a heat conductor 2. The position fixing part 4 has a position fixing element 5, in particular a vacuum passage, and even more preferably an individually controllable vacuum passage, by means of which the first substrate 11 (not shown) is fixed on the surface. It can be fixed in position at 4o. The heat conductor 2 preferably has a plurality of ribs 3, which can emit heat via a rib surface 3o of the rib 3 to a fluid not shown. The heat conductor 2 is coupled to the position fixing portion 4 via the boundary surface 6.

FIG. 2 shows a second preferred embodiment according to the invention of a substrate holder 1'according to the invention, which substrate holder 1'has a heat conductor 2'which simultaneously functions as a position fixing part. In other words, the heat conductor 2 ′ and the position fixing portion are formed as one piece, that is, integrally, unlike the embodiment of FIG. 1. This eliminates the interface between the fixed part and the heat conductor 2 ′, so that it is advantageous to move the first substrate 11 (not shown) to the fluid (not shown) flowing around the ribs 3. There are no thermal barriers that hinder the transport of heat.

FIG. 3 shows an even more preferred third embodiment according to the invention of a substrate holder 1″ according to the invention, which substrate holder 1″ has a hole 7 in a heat conductor 2″. The holes 7 are intended to allow deformation elements 8, notably the spikes, of the substrate 11, not shown, to access the back side 11o, not shown. Since the present embodiment is the same as the embodiment of FIG. 2 in other points, the description of FIG. 2 is referred to.

FIG. 4 shows a fourth embodiment according to the invention of a substrate holder 1′″ according to the invention, which substrate holder 1′″ is the backside (not shown) of a first substrate 11 (not shown). In order to minimize contact with (not shown), in addition to the features described in FIG. 3, there is also a recess 9 in the fixing surface 4o. This minimization serves to avoid contamination of the substrate by the fixing surface 4o, in particular metal contamination. This minimization further helps to avoid local deformation of the substrate by particles. The recess 9 can be filled with a fluid of high heat capacity and/or thermal conductivity in order to increase the thermal coupling.

FIG. 5 shows a fifth embodiment according to the invention of a substrate holder 1 ^{IV according to} the invention, which substrate substrate 1 minimizes contact with the back side, not shown, of a first substrate 11, not shown. In addition to the features described in FIG. 4, there are projections and/or recesses 9 filled with needles and/or posts 10 in order to ensure that the first substrate 11 is supported almost entirely. .. This minimization also serves to avoid contamination, especially metal contamination. The recess 9 can be filled with a fluid of high heat capacity and/or thermal conductivity in order to increase the thermal coupling.

FIG. 6a shows a first step of an exemplary method according to the invention, in which the first upper substrate 11 is initially spaced from the second lower substrate 11′ by a distance d1. positioned. This process step is carried out in the region a of the corresponding Tt diagram already defined above. The two substrates 11, 11' are brought close to each other, at which time the thermal effect of the lower second substrate 11' or the lower substrate holder 14 on the upper first substrate 11 has already been explained above. It is excluded to the maximum by the relatively large spacing.

In the subsequent step, both substrates 11, 11' are brought closer to each other up to a distance d2. At this point, the system is already located in the region b defined above, ie in the so-called coarse approach region, in which region b already has, among other things, the first radiation on the upper side due to the heat radiation of the lower substrate 11 ′. A relatively slight heating of the substrate 11 is performed.

In a subsequent step, both substrates 11, 11' are brought closer to each other to a well-defined distance d3 as already explained above. At this point, the system is already located in the region c defined above, ie in the so-called near-proximity region, in which the abruptness of the upper first substrate 11 is caused, inter alia, by thermal radiation and thermal convection. Heating is performed.

In the subsequent step of FIG. 6d, the bonding process of both substrates 11, 11' is performed. Both the substrates 11 and 11' are positioned at a constant distance d3. At this point the substrates 11, 11' are already located in the region d defined above, i.e. in the so-called bonding region, in which the temperature T4o is constant over the period t1.

In the subsequent step of FIG. 6e, the cooling of the substrate 11 and/or 11' is carried out in the region e already defined above. The cooling itself is in the process of adapting the temperature of the upper first substrate 11 to ambient temperature, in particular the temperature of the ambient atmosphere and/or the temperature of the lower second substrate 11 ′ or the lower substrate holder 14. is there. However, at this point, the bonding of the two substrates 11, 11' has already been carried out, inter alia by prebonding.

It is omitted to illustrate the region f already defined above with further figures. Because it is not possible to gain important knowledge from here. As already disclosed in the description of the specification, it is also possible to carry out the bonding process in a constant temperature region in the region f.

FIG. 7a shows the temperature-time already explained above with the six characteristic temperature regions a, b, c, d, e, f already defined above, shown on the upper horizontal axis. A diagram is shown. On the lower horizontal axis the time t is shown in seconds, and on the left vertical axis the temperature T is plotted in Kelvin. On the right vertical axis is plotted the unscaled distance d(au) between both substrates 11 and 11'. Furthermore, four temperature graphs 12, 12′, 12″, and 12′″ are entered. The temperature graph 12 represents the temperature of the first substrate 11. The temperature graph 12' represents the temperature of the heat conductors 2, 2', 2'', 2''', 2 ^IV which more or less corresponds to the temperature Tk of the cooling fluid. This temperature substantially matches the temperature T1o of the upper substrate 11 before the two substrates 11 and 11' are brought close to each other. The temperature graph 12″ represents the temperature of the second substrate 11′. The temperature graph 12″′ represents the temperature of the lower substrate holder 14. If the thermal coupling between the second substrate 11' and the lower substrate holder 14 is large enough, these two temperatures will be approximately the same.

Furthermore, a distance graph 13 representing the distance d between the two substrates 11 and 11' is also shown. The interval graph 13 should be interpreted exclusively as a symbol and in fact shows a relatively gentle transition from region c to region d, since the substrate has to be negatively accelerated or braked. Become. In particular, the substrate can also change its velocity during the approach phase. The temperature difference ΔT between the temperature of the lower substrate and the temperature of the upper substrate in the temperature saturation region d is due to the thermal resistance and/or due to the heat source, especially the heater in the lower sample holder 14, and/or It can be adjusted accurately and reproducibly by means of a heat sink, in particular a cooling fluid.

The course of the temperature graph 12 and the interval graph 13 during the implementation of the exemplary method according to the invention is as follows: time at the start of the method, i.e. in the region marked a (so-called temperature region a). On the leftmost side of the scale, the two substrates 11, 11' are brought closer together, which reduces the distance d between the substrates 11, 11'. At the start of the method, the distance between the two substrates 11, 11' is d1 and this distance d1 is continuously reduced. In the temperature region a, the temperature of the first or upper substrate 11 is substantially constant and is T1o.

The temperature region a is followed by the temperature region b in terms of time, and in this temperature region b, the temperature of the substrate 11 rises relatively slightly (temperature curve section T2o), while the temperature of the substrates 11, 11' is increased. The distance d between them is further reduced.

The temperature region c is followed by a temperature region c in terms of time, and the temperature of the substrate 11 increases relatively greatly in this temperature region c as compared with the temperature region b (temperature curve section T3o). The distance d between 11, 11' is further reduced. At the end of the temperature region c, a substantially constant final distance d is realized between the substrates 11, 11'.

The temperature region c is followed by the temperature region d, in which the interval d remains constant and the temperature T4o of the first substrate 11 is substantially constant. The same applies for the temperature difference ΔT between the lower substrate 11 ′ and the upper substrate 11. This constant temperature T4o is maintained for the time t1. In particular, it should be pointed out that the transition from the temperature region c (so-called close proximity region c) to the temperature region d (so-called bonding region d) is rapidly performed.

The temperature region d is followed by the temperature region e, in which the temperature of the substrate 11 decreases (temperature curve section T5o), while the distance d remains substantially constant. In the subsequent temperature region f, the temperature of the substrate 11 is substantially constant (see the temperature curve section T6o).

FIG. 7b shows another temperature-time diagram with the six characteristic temperature regions a, b, c', d', e, f already defined above. The spacing graph 13 is the same as the spacing graph of Figure 7a. Since the temperature graph 12 corresponds to the temperature graph of FIG. 7a in the temperature regions a, b, c, f, the description of FIG. 7a is referred to for these regions. Differences from FIG. 7a become apparent in regions c′ and d′ when compared with regions c and d in FIG. 7a. In this example, the transition from the close proximity region c'to the bonding region d'is not abruptly performed as in FIG. 7a but is continuously and gently performed.

FIG. ~VII. In Fig. 1C, there are shown a plurality of overlay errors already explained or defined above between the upper structure 15 of the upper substrate 11 and the lower structure 15' of the lower substrate 11', these At least some of these can be avoided using the present invention. Some overlay errors are known as run-out errors.

Figure 8-I. The overlay error is a non-matching overlay of the upper structure 15 and the lower structure 15' and is a typical result of runout error. The structures 15 and 15' have the same shape, but do not match. The causes of this type of error are (i) that the structures 15, 15' on the substrates 11, 11' are fundamentally inaccurately manufactured, and/or (ii) especially before the bonding of the substrates 11. , 11' due to the strain of the structures 15, 15' and/or (iii) during the bonding, especially of the substrates 11, 11' due to the strain of the structures 15, 15'. Yet another possibility is that the two substrates 11, 11' are totally displaced from each other. However, in this case, it is due to the basic alignment problem of the overall alignment of the two substrates with respect to each other and has little to do with the concept of runout.

Figure 8-II. Shows yet another overlay error of the two structures 15 and 15' rotating with respect to each other. The rotation of the two structures 15 and 15' with respect to one another is shown exaggeratedly and in practice only occurs a few degrees, in particular a few tens of degrees. Such rotation may occur because (i) both structures 15, 15' are not properly manufactured on both substrates 11, 11' and/or (ii) before the bonding process. , 15' in the vicinity of a local strain, and correspondingly in particular a local rotation of both structures 15, 15', and/or (iii) during the bonding process. In particular, local distortions occur in the vicinity of the bodies 15, 15′, and correspondingly particularly locally local rotations of the two structures 15, 15′. Yet another possibility is that the two substrates 11, 11' are generally twisted with respect to each other. In this case, 8-II. It should be seen that the overlay error of the form is increasing radially, especially from the inside to the outside, at a plurality of positions between the two substrates 11, 11'.

Figure 8-III. ~ 8-VII. Overlay error is primarily due to (i) scaling errors caused by inaccurate manufacturing, and/or (ii) scaling errors caused by distortion of structures 15, 15' prior to bonding, especially due to distortion of substrates 11, 11'. , And/or (iii) scaling errors caused by the distortion of the structures 15, 15 ′, especially due to the distortion of the substrates 11, 11 ′ during bonding. These are typically not called runout errors.

FIG. 9 shows a partial schematic cross-sectional view, not to scale, of a substrate holder according to the invention with the equivalent circuit diagram described above for the thermal resistances Rth1 to Rth8. Above all, the thermal resistances Rth1 to Rth3 should be minimized in order to be able to maximize the heat transfer from the lower substrate holder 14 with the heating device (not shown) to the lower substrate 11'. Is. This allows the present invention to efficiently and quickly heat the lower substrate 11'. Furthermore, with a series of minimum thermal resistances, a very rapid change of the temperature T1u of the lower substrate 11' can be realized.

The thermal resistance Rth4 should be maximum according to the invention. In the purely theoretical ideal case where the thermal resistance Rth4 is infinite, the amount of heat does not reach the upper substrate 11 from the lower substrate 11'. Due to the finiteness of the thermal resistance Rth4, a small amount of heat that cannot be ignored always reaches the upper substrate 11 from the lower substrate 11'. The thermal resistance Rth4 can be adjusted relatively easily and accurately by selecting a vacuum or a special mixture between the two substrates 11 and 11'.

According to the invention, the thermal resistances Rth5 to Rth8 are also minimized in order to be able to maximize the heat transfer between the cooling fluid, in particular the atmosphere, and the upper substrate 11 and thus the efficiency. Should do. During the bonding process in the temperature saturation region d, the upper temperature T4o or the temperature difference ΔT between the temperature T4o of the upper substrate 11 and the temperature T1u of the lower substrate 11′ can be accurately measured as desired. Adjusting in a reproducible and reproducible manner is crucial according to the invention. This is in particular according to the invention: (i) selecting at least one of the thermal resistances Rth1 to Rth8 as desired and/or (ii) especially in the lower substrate holder 14. This is achieved by adjusting the lower temperatures T1u-Tp with a heating device and/or (iii) by adjusting the upper temperatures T1o-Tk with a cooling fluid according to the invention.

1, 1', 1'', 1''', 1 ^IV substrate holder 2, 2', 2'', 2''', 2 ^IV heat conductor 3 rib 3o rib surface 4 position fixing portion 4o position fixing surface 5 Position fixing element 6 Boundary surface 7 Hole 8 Deformation element 9 Recess/Dimple/Dent 10 Protrusion/Needle 11, 11' Substrate 12, 12', 12'', 12''' Temperature graph 13 Interval graph 14 Bottom Substrate holder 15, 15' Structure d1, d2, d3 Substrate spacing t1 Period T1o, T2o, T3o Temperature/temperature curve section T4o, T5o Temperature/temperature curve section Tp Substrate holder temperature Tw Thermal conductor temperature Tk Cooling Fluid temperature a, b, c, c', d, d', e, f Temperature range

Claims (19)

In a substrate holder (1, 1', 1'', 1''', 1 ^IV ) having a position fixing surface (4o) for holding a substrate (11, 11'),
The substrate holders (1, 1', 1'', 1''', 1 ^IV ) are heat conductors (2, 2', 2'', for discharging heat from the position fixing surface (4o). 2″′,2 ^IV ),
A substrate holder (1, 1′, 1″, 1′″, 1 ^IV ) characterized by the above. The heat conductors (2, 2′, 2″, 2′″, 2 ^IV ) dissipate the heat, in particular on the side of the heat conductor opposite the fixing surface (4o). With ribs (3) for
The substrate holder (1, 1', 1'', 1''', 1 ^IV ) according to claim 1. The ribs (3) are arranged perpendicular to the position fixing surface (4o) and/or parallel to each other,
The substrate holder (1, 1', 1'', 1''', 1 ^IV ) according to claim 2. The heat conductors (2, 2', 2'', 2''', 2 ^IV ) are also configured to supply heat to the position fixing surface (4o),
The substrate holder (1, 1', 1'', 1''', 1 ^IV ) according to any one of claims 1 to 3. The position fixing surface (4o) is integrally formed with the heat conductor (2, 2′, 2″, 2′″, 2 ^IV ),
The substrate holder (1, 1', 1'', 1''', 1 ^IV ) according to any one of claims 1 to 4. The substrate holder (1,1′,1″,1′″,1 ^IV ) has a deformation element (8) for deforming the substrate (11,11′),
The substrate holder (1, 1', 1'', 1''', 1 ^IV ) according to any one of claims 1 to 5. The deformation element (8) is arranged in the center of the substrate holder (1, 1′, 1″, 1′″, 1 ^IV ),
The substrate holder (1, 1', 1'', 1''', 1 ^IV ) according to claim 6. The deformation element (8) is configured such that the substrate (11, 11') is deformable in a direction away from the position fixing surface (4o),
The substrate holder (1, 1', 1'', 1''', 1 ^IV ) according to claim 6 or 7. Positioning elements (5) for fixing the substrate (11, 11 ′) are arranged inside, on and/or on the fixing surface (4 o ),
The substrate holder (1, 1', 1'', 1''', 1 ^IV ) according to any one of claims 1 to 8. The position fixing element (5) is at least partially a vacuum passage,
The substrate holder (1, 1', 1'', 1''', 1 ^IV ) according to claim 9. The specific heat capacity of the heat conductor (2,2′,2″,2′″,2 ^IV ) is larger than 0.1 kJ/(kg·K), preferably 0.5 kJ/(kg·K). Larger, even more preferably larger than 1 kJ/(kg·K), most preferably larger than 10 kJ/(kg·K), and most preferably larger than 20 kJ/(kg·K),
The substrate holder (1, 1', 1'', 1''', 1 ^IV ) according to any one of claims 1 to 10. In an apparatus for bonding a first substrate (11) to a second substrate (11'),
At least one substrate holder (1, 1', 1'' according to any one of claims 1 to 11 for holding at least one of the two substrates (11, 11'). , 1′″,1 ^IV ), especially with the upper substrate holder (1,1′,1″,1′″,1 ^IV ),
apparatus. Substrate holder (1, 1', 1'', 1', 1', 1 ^IV, as an upper substrate holder (1, 1', 1'', 1''', 1 ^IV ) according to claim 1. Use of 1″′, 1 ^IV ). In a method for bonding a first substrate (11) to a second substrate (11'),
In the first step, the substrates (11, 11 ′) are brought close to each other, thereby increasing the temperature (T 2 o, T 3 o) of the first substrate (11 ),
In the second step, the substrates (11, 11') are stopped from approaching each other, and the interval (d3) between the substrates (11, 11') is maintained constant, and the interval (d3) is constant. In the case of, a constant temperature (T4o) is generated in the first substrate (11) for at least a predetermined period (t1),
In the third step, both the substrates (11, 11') are at least temporarily bonded to each other within the period (t1) in which the first substrate (11) has a constant temperature (T4o).
Method. The interval (d3) when the constant temperature (T4o) is occurring over the period (t1) is set between 1 mm and 0 mm, preferably between 100 μm and 0 μm, and even more preferably between 10 μm and 0 μm. Most preferably between 1 μm and 0 μm,
The method according to claim 14. The period (t1) in which the interval (d3) is constant and the constant temperature (T4o) is generated is longer than 5 seconds, preferably longer than 10 seconds, more preferably longer than 15 seconds, and even more preferably still. Is longer than 20 seconds, most preferably longer than 25 seconds,
The method according to claim 14 or 15. The time period (t1), the interval (d3), and/or the constant temperature (T4o) are set before the first step, especially by experience, preferably at the temperature of the second substrate (11′). And/or the substrate holder (1,1′,1″,1′″,1 ^IV ), the thermal conductor (2,2′,2″,2″′,2). ^IV ) and/or considering the material of the substrate (11, 11') and/or considering the approach speed,
The method according to any one of claims 14 to 16. In a product having a first substrate (11) and a second substrate (11′), especially a substrate stack,
The substrates (11, 11') are bonded to each other by the method according to any one of claims 14 to 17,
Product. Substrate holder (1,1) according to any one of claims 1 to 11 for holding a substrate (11,11') during the implementation of the method according to any one of claims 14 to 17. Use of', 1'', 1''', 1 ^IV ).

Images