WO2013127891A1

WO2013127891A1 - Method and device for controlling the surface temperature of a susceptor of a substrate coating apparatus

Info

Publication number: WO2013127891A1
Application number: PCT/EP2013/053986
Authority: WO
Inventors: Ralf Leiers; Markus LÜNENBÜRGER; Gerhard Karl Strauch; Bernd Schineller; Karl-Heinz BÜCHEL
Original assignee: Aixtron Se
Priority date: 2012-03-01
Filing date: 2013-02-28
Publication date: 2013-09-06
Also published as: TWI571528B; CN104204291B; TW201346061A; CN104204291A; DE102012101717A1; DE112013001238A5

Abstract

The invention relates to a method for treating at least one substrate (105, 106, 107) in a process chamber (101) of a reactor housing, wherein the one or more substrates (105, 106, 107) are laid on a susceptor (108) that can be heated by means of heating elements (109, 110, 111), wherein by means of the heating elements (109, 110, 111), spatially associated zones of the susceptor (108) are heated, with which respective surfaces zones (112, 113, 113', 114) of the side of the susceptor (108) facing the process chamber (101) are associated, wherein temperatures of the surface zones (112, 113, 113', 114) and/or the at least one substrate (105, 106, 107) arranged there are measured at a plurality of measurement points by means of optical measuring sensors (1 to 35), and the measured values determined by means of the sensors (1 to 35) are fed to a control apparatus (115, 116, 117, 122) by means of which the heating power of the heating elements (109, 110, 110', 111) is controlled. According to the invention, in order to optimize the temperature control, respective combinations of measured temperature values are used to control the heating power of the heating elements (109, 110, 110', 111).

Description

Method and device for controlling the surface temperature of a susceptor of a substrate coating device

The invention relates to a method for treating at least one substrate in a process chamber of a reactor housing, wherein the one or more substrate is placed on a susceptor which can be heated from below with heating elements, wherein spatially associated zones of the susceptor are heated with the heating elements. Surface zones of the side of the susceptor facing the process chamber are uniquely assigned to the heating elements, temperatures of the surface zones or the at least one substrate arranged there being measured at a plurality of measuring points by means of optical measuring sensors, and the measured values ascertained with the sensors are fed to a control device , with which the heating power of the heating elements is regulated.

The invention further relates to an apparatus for treating at least one substrate having a reactor housing and a process chamber arranged therein, which has a susceptor, for receiving the at least one substrate, a plurality of heating elements arranged below the susceptor and a plurality of temperature sensors, each at one Measuring point to provide a temperature reading of the surface of the susceptor or a substrate arranged there, with a control device to which the measured values are supplied and which controls the heating elements measured with the measurement points functionally associated with the respective heating element.

DE 10 2004 007 984 A1 describes a CVD reactor with a process chamber arranged in a reactor housing. The bottom of the process chamber is formed by a susceptor which carries the substrates to be treated, in particular to be coated. The process chamber ceiling is covered by a gas inlet member formed having inlet openings through which the process gases can enter the process chamber. Below the susceptor, a heater is arranged to heat the susceptor to the treatment temperature. The surface temperature of the susceptor is measured by means of a multiplicity of temperature measuring sensors.

No. 6,492,625 B1 describes a device for thermal treatment, in particular for coating substrates resting on a susceptor, the susceptor being heated from below. Below the susceptor are several heating elements that can be controlled individually. Each heater is assigned a controller that receives actual values of the surface temperature of the susceptor. The actual values are determined with optical measuring sensors. Each heating zone is functionally assigned to several measuring sensors.

It is apparent from EP 1 481 117 Bl that the temperature profile on the surface of the susceptor facing the process chamber on which the substrates lie is of great importance for the quality of the layers deposited on the substrates. In particular, it is desired to influence the lateral temperature profile in such a way that the lateral temperature gradient is as low as possible. The temperature on the susceptor surface should preferably have the same value at all locations on the susceptor.

DE 10 2007 023 970 A1 describes a susceptor with a multiplicity of hexagonally arranged pockets for accommodating in each case one substrate. Usually, the substrate surfaces or the layers deposited on the substrate surfaces have different optical properties, such as the degree of absorption or emissivity, than the surface of the surrounding susceptor. In a coating process, not all available receiving pockets for the substrates need to be equipped with substrates. It is also take into account that only a selection of the available receiving pockets is equipped with substrates to be treated.

No. 6,706,541 B1 describes an apparatus for carrying out a CVD method with an automatic process control unit that is capable of controlling the temperatures of a plurality of surface zones. Substrates are to be coated with the device described there. Sensor elements are provided which observe the layer thickness of the substrates during layer growth. These measured values are input data for the control device.

US 2003/0038112 A1 describes a method for stabilizing a plasma in a process chamber of a plasma reactor. This purpose is served by a control system which uses a multiplicity of measured values determined with optical sensors.

US 2006/0027169 A1 describes a method with which a temperature profile monitoring is performed on the surface of a substrate holder. It uses a controller that receives readings from temperature sensors that detect the temperatures of heating zones.

No. 5,782,974 describes a temperature measuring system in which the temperature of a rear side of a susceptor is determined pyrometrically. US 5,970,214 describes an apparatus for heat treatment of semiconductor substrates with a plurality of photosensitive sensors, which determine a surface temperature of the substrate. The measured values of the sensors are fed to a controller, which controls lamps. US Pat. No. 6,079,874 describes a device with which the surface temperature of a substrate can be measured at various locations. By means of a controller, a heating device is regulated. The controller uses the measured values supplied by the pyrometers for regulation.

US Pat. No. 5,871,805 describes a CVD apparatus in which the temperature of a susceptor, on which a substrate rests, is regulated by means of a regulating device. US 6,034,357 describes an apparatus for determining the surface temperature of a substrate in a process chamber using temperature sensors that interact with a controller that uses a correction factor to control a lamp heater. The temperature profile of the susceptor depends not only on the degree of loading of the susceptor with substrates, but also on other process parameters such as total gas pressure in the process chamber, the chemical composition of the gases that are introduced into the process chamber for treating the substrates, the material of the susceptor, the type of the substrate and the aging state of the susceptor, in particular its coating, from.

In a device for treating semiconductor substrates in a process chamber or in such a method, the heating elements are located on one side of a susceptor. The heating elements lie directly below the surface zones directly heated by them. In a susceptor constructed rotationally symmetrically, the surface zones and their associated heating elements are arranged on ring zones which are adjacent to one another. The temperature sensors are located on the side of the susceptor opposite the heating elements. The introduced by a heating element in the susceptor heating power not only heats the their associated surface zone. As a result of heat transfer mechanisms within the susceptor, ie essentially the heat conduction and heat radiation of the heating element to other surface areas of the susceptor, the heating power of an individual heating element not only influences the temperature of the associated surface zone, but also influences the temperatures of all surface zones. The immediately adjacent surface zones are affected the most and the furthest away the surface zones. The measuring sensors thus provide mutually coupled temperature measured values.

The invention has for its object to further optimize the generic method and the generic device in terms of temperature control. The object is achieved by the invention specified in the claims.

The claims indicate various variants of the generic method or of the generic device in which the heating power fed into a heating element not only determines the measured value supplied to the respective heating element or a surface zone assigned to the individual surface zone associated with the heating element. Instead, a combination of the measured values of a plurality of temperature measuring sensors is used. In a first variant, it is provided that the control takes place with varying combinations of measured values. While in the prior art each control device is functionally firmly connected to their associated actual encoders in the form of temperature sensors, the invention pursues the concept of making this functional linkage variable. Not all of the available measured values or temperature Measuring sensors to be used for the scheme, but only an individual selection thereof. The selection is a combination of measurements that depends on the operating parameters. Operating parameters affecting the quality of the combination include the target surface zone temperatures, the total gas pressure in the process chamber, the chemical composition of the gas phase in the process chamber, the susceptor material, the type of substrates to be coated, and the susceptor population with the substrates and / or the aging state of the susceptor. The device used to carry out the method has a susceptor, which preferably has the shape of a circular disk and which can be driven in rotation about its axis of symmetry. The gas inlet member disposed above the susceptor may be in the form of a shower head. As is known from the prior art, the openings of the shower head can be used as an optical channel, through which the temperature measuring sensors arranged above the openings receive in particular optical (pyrometric) information about the surface of the susceptor. It is provided a plurality of radially arranged sensors, wherein the individual temperature sensors may have an equal distance from each other. Each temperature sensor preferably determines optically / pyrometrically the surface temperature of the susceptor at a location below it. As the susceptor is rotated, these measuring points travel on circular paths over the susceptor and also cover the substrate surfaces. In the gas inlet member, a gas mixture is fed in a known manner. The gas inlet member may include a plurality of chambers so that various gas mixtures are introduced into the process chamber separately from one another. In a coating process, for example in a MOCVD process, organometallic compounds of the II. Or III. Main group initiated in the process chamber. A component of the V. or VI. Main group is introduced in the form of a hydride in the process chamber. The process gases decompose py- rolytically such that layers are deposited on the substrates. The layers depend essentially on the gas composition. The layer composition also depends strongly on the surface temperature of the substrate. The surface temperature of the substrate depends not only on the heating powers of the arranged below the susceptor heating elements. The surface temperature also depends on other growth parameters, which in particular affect the heat dissipation from the substrate surface. These are the aforementioned process parameters. If the height of the process chamber can be varied, then the heat flow and thus the temperature distribution on the surface of the susceptor also depends on the height of the process chamber. Although the individual heating zones are locally assigned surface zones of the susceptor whose surface temperature is significantly influenced by the particular underlying heating elements. However, it has been shown that adjacent surface zones are also influenced to a considerable extent by temperature. This influence depends on the operating parameters. It is thus of advantage if the temperature measuring sensors used according to the invention for controlling control the surface temperature of the susceptor at different points, depending on the set of operating parameter set. With the method according to the invention, it is possible to locally vary the measuring points used for regulation, without having to intervene in the structural design of the sensor field. Of a large number of available temperature sensors, each of which measures only the temperature at a measuring point, a selection, which may possibly only be limited to a single temperature sensor, is used. In the simplest case, when changing the operating parameters and the temperature sensor used for control is changed. Preferably, however, each are qualitatively and quantitatively different combinations of temperature sensors, which are used. The combinations of the measured values used for control can be determined on the one hand by the number of Deten or unused measuring points of the respective surface, on the other hand, but also by their weighting based on the respective surface zone differ. Thus, for example, it is possible to use only the measuring sensors arranged at the edge of the zone for temperature control of one of a plurality of radial surface zones, or alternatively to use only the temperature sensors arranged in the middle of the zone. Furthermore, it is possible according to the invention to use for controlling the heating element of a heating zone with temperature sensors, which are spatially associated with an adjacent heating zone. In a preferred embodiment, the heating zones are rotationally symmetrical about the center of rotation, wherein the heating zones are adjacent to each other in the radial direction. They are thus arranged concentrically to one another. It is also possible that the measured value of individual temperature sensors is used by a plurality of control devices. It is also possible to weight the contribution of a single temperature sensor to the control. The weighting can be between zero and one. Which sensors are used for certain operating parameters and which sensors are disregarded for control is the result of preliminary tests or computer-aided simulation calculations. It is essential that mutually different operating parameters are each assigned a different combination of the measured values used in the control.

The operating parameters, which are input as an input to the selection device, can also act directly on the control devices. For example, the control characteristic values can be input as additional input variables, that is to say, for example, for proportional-integral-differential controllers, the proportional component, the integral component and / or the differential component. On the other hand, it is also possible for the selection device to determine these characteristic values on the basis of the process parameters, for example from a table stored in the selection device. A further aspect of the invention is concerned with the problem that the temperature measured values supplied by the measured values are coupled to one another from the respective heating element to the susceptor due to the heat transfer mechanisms. Basically, each heating element affects the surface temperature of each surface zone of the susceptor. According to the invention, characteristic temperatures are determined which are each assigned to a surface zone. Each characteristic temperature may be an average value, in particular a weighted mean value of the temperature measured values of a multiplicity of temperature sensors. The structure of the device according to the invention substantially corresponds to the structure of the device described above. A number of measuring sensors are arranged one behind the other in the radial direction above the rotationally drivable susceptor. The number of temperature measuring sensors can be significantly greater than the number of separately heatable surface zones. However, it is sufficient if each of these separately heatable surface zones is assigned only one sensor. In a preferred variant of the invention, a multiplicity of temperature measured values is obtained during the rotation of the susceptor, so that after a rotation a complete lateral temperature profile is available. The temperature profile consists of a grid-like field distribution, the temperature of each measuring field being known. The measuring fields are distributed uniformly over the surface of the susceptor in the radial direction and in the circumferential direction. They thus include fields that lie on substrate surfaces or lie on areas of the susceptor that are not covered by substrates. The characteristic temperatures can be obtained by considering only those measuring fields which lie on a substrate or not on a substrate in the case of a certain surface zone. Depending on the field size or the position of the field, the contribution of a single field to determine the mean value, ie the characteristic temperature, weighted. The characteristic temperatures determined in this way are fed to a control device. The number of characteristic temperatures preferably corresponds to the number of heating elements or the number of surface zones. The control device contains a decoupling device, which effectively decouples the characteristic temperature measured values coupled with each other. The control device thus provides, so to speak, decoupled control signals in order to supply the heating elements with heating power. The control device has an input, which receives a characteristic temperature measured value as input data for each individual surface zone. The control device has an output which supplies a control signal for each heating element individually assigned to a surface zone, which control signal determines the heating power to be supplied to the heating element. According to the invention, the control signals are converted from coupled values, each converted value having a contribution of a plurality of characteristic temperatures, or each converted value containing a contribution of a plurality of characteristic temperatures. In a variant of the invention it is provided that each characteristic temperature is supplied to an individual controller as an input variable. The number of regulators preferably corresponds to the number of temperature-regulating surface zones. The controllers provide first, mutually coupled values. These values are converted by a decoupling device. For this purpose, the decoupling device uses, for example, a decoupling matrix, which is applied to the first values, so that second values are calculated, which are then, as it were, decoupled from one another. The second values are amplified by an amplifier and assigned to the heating elements of the surface zones as heat output control values. The decoupling device is capable of converting first values individually associated with a characteristic temperature but influenced by a plurality of heating elements into second values. Every second value is individually assigned to a heating element. Its height corresponds to the heating power of the heating element. With the decoupling device, the second values with the first values that every other value contains a contribution of multiple first values. Just as the first values (characteristic temperatures) via the heat input of the heating elements into the susceptor have a contribution of several second values (heating powers), the second values (heating powers) each have contributions of several first values (characteristic temperatures). The decoupling device compensates for the coupling caused by the controlled system. The second values (heating power values) are thus the result of a compensation of the coupling of the first values (characteristic temperature measured values). The controlled system is formed by the respective controller, an amplifier, the heating elements, the heated susceptor and the temperature sensors. According to the invention, this controlled system is supplemented by the decoupling element. The decoupling element together with the regulators, the amplifiers, the heating elements, the susceptor and the temperature sensors is element of the control loop. From the point of view of the controller, the decoupling element is regarded as belonging to the controlled system and compensates for the couplings within the heating elements and the susceptor. With a perfectly designed decoupling element, the independent controllers work with a controlled system that is extended by the decoupling element and whose inner coupling of the heating zones to the outside, ie not visible to the controller. This improves the control behavior and simplifies the tuning of the controllers. To determine the decoupling matrix used in the decoupling device, a gain matrix is determined. For this purpose, the contributions by which the heating element of a certain surface zone influences the characteristic temperatures of all surface zones are determined in preliminary tests or in model calculations. For example, in an arrangement consisting of four surface zones, one of the four heating elements influences the temperature of all four surface zones, the surface zone directly associated with the heating element being most strongly influenced and the surface zone furthest away from the heating element being least affected. Consequently, the diagonal elements The gain matrix has the largest values, while the diagonally most distant matrix elements have the lowest values. Since each of the four surface zones provides four matrix entries, the gain matrix in the example consists of 4 × 4 matrix elements. An inversion of the amplification matrix forms the decoupling matrix. By matrix multiplication of the first values with the decoupling matrix, the second values are formed. In a further development of the invention, the characteristic temperatures can be determined by repeated recording of thermal images. The thermal images can be recorded with the above-mentioned measuring sensors, wherein the measuring sensors arranged in a radial line provide a three-dimensional thermal image of the surface. For this purpose, the susceptor is rotated below the measuring sensors, which may be photodiodes. But it is also possible to optically record the thermal image with a lens system. The method of determining the characteristic temperatures by means of a thermal image is carried out by the following process steps: recording a thermal image, evaluating the thermal image, whereby the characteristic temperatures are calculated, forwarding the characteristic temperatures as temperature actual values to the respective temperature controller, calculation of the heating power under consideration - transmission of operating parameters, adjustment of heating power, recording of the next thermal image. In a development of the invention, it is proposed that the selection of the measuring points or of the measuring sensors used for the regulation takes place via a neural network. Two-stage neural networks can be used. Each measuring sensor, ie each measuring diode, is in each case connected to a node of the hidden layer of the neural network. Each point of this layer is then connected to all the output nodes of the neural network which serve as the input circuit for the subsequent control. This has the advantage that not only a measuring sensor is selected as such, but also an optimal weighting of the measuring sensors takes place with each other. The system can be be learned in learning mode. For this purpose, the system is set to a constant temperature. The neural network is told which temperature has been set. The learning sequence can contain up to 100 different profiles. The profiles can be designed such that the measuring points are located only on the substrates, only on the free surface areas of the susceptor or on both zones. The neural network can interact with a one-dimensional, that is a linear measuring sensor matrix. However, it can also interact with a two-dimensional measuring sensor matrix. Thus, temperature images are processed as an input variable. In this case, not only a number of nodes exist in the hidden layer but an entire area of nodes. Each measuring sensor can be connected to a node of the neural network. The weighting factors can be in the range of zero and one. Embodiments of the invention will be discussed below with reference to accompanying drawings. Shown schematically are the cross section through a process chamber of a

MOCVD reactor with a total of thirty five temperature sensors, each measuring the surface temperature at a measuring point on the

Determining the susceptor, the measuring points having mutually different radial distances from the center of rotation of the susceptor 108, a plan view of the susceptor 108 with indicated coaxially arranged heating zones 109, 110, 111,

3 shows the influence of the heating elements on the surface along a line III-III in Fig. 2, 4 is a view according to FIG. 1, wherein a first combination of temperature sensors 1 - 35 is used for temperature control,

5 is a view according to FIG. 4, wherein a second combination of temperature sensors 1 - 35 is used for temperature control,

Fig. 6 is a view according to FIG. 1, wherein a third combination of

7 is a schematic view of a representation according to FIG. 1 of a further exemplary embodiment, FIG.

FIG. 8 schematically shows the plan view of a susceptor and the arrangement of the surface zones or the surface area of a respective surface zone used to determine characteristic temperatures. FIG

9 shows a representation similar to FIG. 3 for determining a gain matrix K.

Fig. 1 shows schematically the cross section through a process chamber. The bottom of the process chamber 101 is formed by a susceptor 108 which is rotatable about an axis of rotation 120. Below the susceptor 108 are in concentric arrangement three heating zones 109, 110, 111. The heating zone 109 is located below the center of the susceptor 108 and is surrounded by the heating zone 110 annular. The latter is in turn surrounded annularly by the outermost heating zone 111. The heating zones 109, 110, 111 are formed by infrared heating elements or RF heating elements and are capable of heating the surface of the susceptor 108 in three surface zones 112, 113, 114. FIG. 2 shows, in FIGS. 1, 4, 5 and 6, receptacle pockets 119 not shown for the sake of clarity and arranged in a circle around the center of rotation for receiving a respective substrate 105, 106, 107. The substrates 105, 106, 107 are located thus with different radial distance away from the axis of rotation 120th

The ceiling of the process chamber 101 running parallel to the extension direction of the susceptor 108 is formed by a gas inlet element 103 in the form of a showerhead. The latter is shown only schematically. It has a multiplicity of sieve-like arranged openings 104, through which process gases fed into a gas distribution chamber of the shower head 103 can enter the process chamber 101. The process gases may be organometallic compounds of elements of the III. or II. main group as well as hydrides of the V. or VI. Main group act. In addition, a carrier gas, for example hydrogen, or another inert gas can be fed into the process chamber. The process gases decompose pyrolytically on the surface of the substrates 105, 106, 107 in order to deposit a layer there. Above the gas outlet openings 104 is a sensor arrangement 102 with optical temperature sensors 1 to 35. The optical temperature sensors 1 to 35 are arranged such that they measure, for example pyrolytically, the temperature at an individually assigned measuring point, the individual measuring points having different radial distances from the axis of rotation - have 120. As a result of the rotation of the susceptor 108 about the axis of rotation 120, the measuring points travel on concentric circles across the surface of the susceptor 108 and over the surfaces of the substrates 105, 106, 107 thereon. Via a data line 121, the temperature sensors 1 to 35 are connected to a selection electronics 118. This selection electronics 118 links the measured values coming from the sensor arrangement 102 with control devices 115, 116, 117. Each of the three heating elements 109, 110, 111 is individually assigned a control device 115, 116, 117. As a target value, the respective control device 115, 116, 117 receives temperatures to which the surface zones 112, 113, 114 are to be regulated. As actual values, the control devices 115, 116, 117 receive measured values determined by the temperature sensors 1 to 35. However, the control devices 115, 116, 117 do not receive all the temperature measured values, but only those of a selection of the entirety of the

Temperature sensors 1 to 35 measured values. These are the numbers schematically entered in the rectangles 115, 116, 117 symbolizing the control devices. The selection electronics 118 receives an input variable P. This input variable P contains information about the operating parameters of the respective method carried out in the process chamber. These operating parameters include i.a. the target temperatures of the surface zones 112, 113, 114, the total pressure in the process chamber 101, the chemical composition of the gas phase in the process chamber 101, ie the type of process gases used, the material of the susceptor 108, for example. Graphite or coated graphite, the Type of substrate, so its crystalline property and crystalline composition, the placement of the susceptor 108 with substrates, so the distribution of the substrates on the receiving pockets 119, unless all Aufnahmeta- see 110 are equipped with substrates and / or the aging state of the susceptor 108, For example, the number of production steps that the susceptor has behind.

Depending on these operating parameters P, the selection electronics 18 determine the combination of the measured values used for the control. In the simplest Case, which is not shown in the figures, only a single temperature sensor is used to control the heating element 109, which is arranged above the surface zone 112, so for example. One of the temperature sensors 1 to 12. Analogously, to control the heating element 110, a single temperature sensor 13 to 23, which is located above the surface zone 113. Analogously, a single temperature sensor 23 to 35 arranged above the surface zone 114 is used to control the heating element 111. In addition, however, it is also possible to use a plurality of further temperature sensors, wherein it is essential that the individuality of the temperature measuring sensors used also change as the operating parameters P change. If, for example, the coating process is carried out at a higher temperature, the heat flow within the process chamber or within the susceptor 108 changes, so that the control-relevant surface temperature must be measured at another surface location. This is done by changing the relevant temperature sensor 1 to 35.

In the embodiment shown in FIG. 4, for example, to control the temperature of the surface zone 112 of the controller 115, only the temperature sensors 2 to 11, in the embodiment shown in FIG. 5, only the sensors 1 to 10 and in the in the Fig. 6 illustrated embodiment, only the sensors 3 to 11 used. In the embodiment shown in FIG. 4, only a selection of the available measurement values is used by the control device 116 to control the surface temperature 113, namely the measured values of the temperature sensors 14, 15, 16, 17, 18, 19, 21, 22 In the exemplary embodiment illustrated in FIG. 5, the measured values of the temperature sensors 12 to 21 and in the exemplary embodiment illustrated in FIG. 6 are the measured values of the temperature measuring sensors 12 and 15 to 24. The control device 117 , which is associated with the surface zone 114, that is, the heating element In the exemplary embodiment shown in FIG. 4, only the measured values of the temperature measuring sensors 25 to 33 are used and, in the exemplary embodiment illustrated in FIG. 5, only the measured values of the temperature sensors 25 to 34 and in that shown in FIG Embodiment, only the measured values of the temperature sensors 26 to 35th

The combinations shown in FIGS. 4 to 6 are merely examples. It is also possible that, for example, only the measured value of every second or every third measuring sensor can be used, or that only measuring sensors 1, 11, 12, 13, 22, 23, 24, 34, 35 are used, that is to say measuring sensors corresponding to

Edge of the respective surface zone 112, 113, 114 are assigned. It is also conceivable to use only the sensors 6, 7, 18, 19, 28, 29, ie those temperature measuring sensors which are assigned to the central region of each surface zone 112, 113, 114.

FIG. 3 shows schematically the influence of the individual heating elements 109, 110, 111 on the temperature profile over a diagonal line over the susceptor. With the curve A, the influence of the central heating element 109 is shown. The heating element 109 not only affects the temperature in the central area of the susceptor, but also, but less so, the temperature in the periphery. This also applies to the influence of the heating element 110, which is shown by B in FIG. The heating element 110 not only influences the temperature in the radially middle region of the susceptor, that is, in the surface zone 113, but also the temperatures in the adjacent surface zones 112, 114. The curve C represents the influence of the radially outermost heating element 111 on the surface temperature. This heating element 111 also influences the temperature in the adjacent surface zone 113.

The quantitative course of the curves A, B, C depends on the process parameters mentioned above. Through the different combinations of measured values, the regulation takes into account the quantitative differences.

In the exemplary embodiments described above, the measured values of individual sensors are either taken into account or not taken into account. However, it is also possible to use the measured values of individual temperature measuring sensors for controlling mutually different heating elements 109, 110, 111, for example, the measured values of the temperature sensors 12, 13 or 23, 24 can each be used by two control devices 115, 116, 117 become. It is also possible to use the individual measured values weighted for the control, for example with a weighting factor between zero and one.

FIG. 7 schematically shows a cross section through a process chamber, as shown in FIGS. 1, 4, 5 and 6. The temperature sensors are indicated here only symbolically. They provide characteristic temperatures Ti, T ₂ , T ₃ to T _n . Each characteristic temperature Ti to T _n is individually assigned to a surface zone 112, 113, 113 ', 114. The characteristic temperatures Ti to T _n can be determined with a sensor arrangement, as shown in FIGS. 1, 4, 5 and 6. The determination of the characteristic temperatures Ti to T _n will be discussed further below. The device shown in FIG. 7 has a control device 122 which comprises regulators 115, 116, 116 ', 117, a decoupling device 113 and an amplifier 124. The control device 122 supplies from the characteristic measured values Ti to T _n control data Pi, P ₂ , P ₃ to P _n , with which in each case one heating element 109, 110, 110 ', 111 is actuated.

The heating elements 109, 110, 110 ', 111 are below a susceptor 108 which is rotatable about a rotation axis 120. The heating elements 110, 110 ', 111 are in concentric arrangement around a central heating element 109. The heating elements 109 to 111 are the already mentioned surface zones 112 to 114 assigned locally. This means that the surface zones 112 to 114 lie in concentric arrangement above the heating elements 109 to 111. Again above the surface zones 112 to 114 are the temperature measuring sensors, which in the exemplary embodiment are designed as optical sensors arranged on a strip. In the simplest case, however, it is sufficient if, for each surface zone 112 to 114, an individual temperature measuring sensor is provided, which supplies the characteristic temperature Ti to T _n .

For each surface zone 112 to 114 there exists a regulator 115, 116, 116 ', 117 assigned to it individually. As input values, the regulators 115 to 117 receive the characteristic temperature Ti to T _n .

Due to the thermal radiation of the heating elements 109 to 111 not only in the direction of its associated surface zone 112 to 114, but also to adjacent surface zones, and the heat conduction within the example of graphite, molybdenum or other thermally conductive material existing susceptor 108, as well as due to a convective Heat transport within the process chamber, each heating element 109 to 111 affects the characteristic temperature Ti to T _{n of} each surface zone 112 to 114. Accordingly, the first values supplied by the regulators 115 to 117 are U ^' i, U ₂ , U'3 to U ^' _n coupled with each other. The first values U ^' i to U _n are converted by the decoupling device 123 into decoupled second values Ui, U2, U3, U _n , which are amplified in the amplifier 124. The decoupling carried out within the decoupling device 123 has the consequence that the first value U ^' i to U ^' _n delivered by each of the regulators 115 to 117 influences each second value Ui to U _n delivered by the decoupling device 123. In the amplifier 124, the second values Ui to U _{n are} only amplified, so that control values Pi to P _{n are provided} for the individual heating elements 109 to 111. The decoupling device is used to improve the control behavior of the described multivariable system. The decoupling of the first values takes place by means of a decoupling matrix L. In order to determine this decoupling matrix L, a thermal image of a susceptor calculated in model calculations or determined in preliminary experiments is used. FIG. 8 shows the plan view of a susceptor which carries in the center a substrate 106 and six further substrates 107 arranged in an arrangement which surrounds the center annularly. The surface zones 112, 113, 113 ', 114 are shown in FIG. 8 with ring zones bounded by dashed lines. Individual hatched areas A, B, C, D of the surface zones 112 to 114 form areas which are used to determine the characteristic temperature Ti, T2.

The figure 8 also shows a grid-like arrangement. Each field of the polar grid corresponds to a temperature reading obtained from one of the temperature measurement sensors 1 through 35 during one revolution of the susceptor 108. During one revolution of the susceptor 108, the measuring sensors 1 to 35 thus provide a large number of angle-dependent measurement data which leads to the mentioned thermal image. A plurality of thermal images are recorded, in which the heating elements 109 to 111 are supplied with different heating powers.

FIG. 9 shows a diagram obtained from such thermal images, on the abscissa of which, starting from the center of rotation 120, a radial R, with reference to the surface of the susceptor, has been removed. On the ordinate a gain factor F is removed, which corresponds essentially to a temperature. The graduation on the abscissa indicates a temperature sensor. With the areas A, B, C, D, the surface areas of the radially adjacent surface zones 112, 113, 113 'and 114 are shown. The reference numerals 212, 213, 213 'and 214 are above the Um- fang averaged temperature curves. The curve 212 indicates the influence of the heating element 109 on all surface zones 112 to 114. The curve 212 reflects the influence of the heating element 110 on all surface zones. The curve 213 'shows the influence of the heating element 110' on the surface temperature of all surface zones and the curve 214 shows the influence of the heating element 111 on all surface zones. The bars labeled K (1, 1) to K (4, 4) represent the matrix elements of the gain matrix K. The gain matrix K is derived from the transfer factor diagram shown in FIG. The surface areas A, B, C, D have been selected taking into account the geometry of the heating elements and the thermal image. The elements of the gain matrix K are the average of the curves within the intervals A, B, C, D.

As can be seen from FIG. 8, the surface areas A, B, C of the surface zones 111, 113, 113 'lie exclusively on the surfaces occupied by the substrates 106 to 107. The surface area D of the surface zone 114, on the other hand, lies on the area of the susceptor 108 that is not covered by the substrates. In the calculation of the matrix elements K (1, 1) to K (4, 4), taking into account a weight of the contributions of all within the intervals Turning surface according to the number of measuring fields.

By averaging results, for example, for the embodiment, the following matrix K

9.94 4.55 1.49 0.61

3.00 7.64 3.06 1.13

0.50 3.17 6.30 3.26

0.18 1.80 5.39 5.13 By inverting this matrix K L = K " ¹ one obtains a decoupling matrix L

0.124 -0.078 0.014 -0.007

-0.057 0.205 -0117 0.036

0.023 -0.134 0.428 -0.245

-0.009 0.071 -0.409 0.440

With this inversion matrix L, the decoupled values Ui to U _n can be obtained from the coupled first values U'i to U'n by matrix multiplication:

Ui = 0.124 ^• U'i -0.078 ^• U ₂ +0.014. , U ₃ -0.007, U ₄

U ₂ = -0.057. , U ^'i +0.205. , u ₂ -0.117, u ₃ +0.036. , u ₄ u ₃ = 0.023. , U'i -0.134, u ₂ +0.428. , u ₃ -0.245, u ₄ u ₄ = -0.009. , U'i +0,071. , u ₂ +0.409. , u ₃ +0.440. , U ₄ The Ui to U ₄ correspond to the heating power, which are supplied to one of the heating elements 109 to 111. It can be seen that the heating power of each heating element 109 to 111 contains a contribution of each regulator 115 to 117 or each characteristic temperature Ti to T _n . The decoupling device converts the actuating signals U'i to U ' _n into second values Ui to U _n . The result is a compensation of the coupling of the characteristic temperature measured values.

The inventive device is characterized preferably by a disposed in the controlled system of the control device decoupling device 123 from which to U of the first coupled values U'i 'generates _n second values Ui to U _n, each of a heating element 109, 110, 110' , 111 individually orderly heat output, wherein the second values Ui to U _n contain weighted contributions of the first coupled values U'i to U ' _n , wherein the weighting compensates for the coupling. It is thus a weighting which as a result provides the compensation of the coupling of the characteristic temperature measured values.

The decoupling device is effectively a stage upstream or downstream of the control device, which supplies control values by suitable combination of the characteristic temperature measured values, so that a change of a first value U'i to U ' _n essentially only changes the surface temperature of the surface zone 112 assigned to it , 113, 113 ', 114 and thus causes only a change in the characteristic temperature associated with it.

The characteristic temperatures are coupled controlled variables of a complex controlled system, from whose deviation from assigned reference variables (setpoint temperatures) a control device gains actuating variables in the form of heating outputs of the heating elements. According to the invention, the couplings of the controlled variables are largely compensated with a decoupling device.

All disclosed features are essential to the invention. The disclosure content of the associated / attached priority documents (copy of the prior application) is hereby also incorporated in full in the disclosure of the application, also for the purpose of including features of these documents in claims of the present application. The subclaims characterize in their optionally sibling version independent inventive developments of the prior art, in particular to make on the basis of these claims divisional applications. LIST OF REFERENCE NUMBERS

1 - 35 Temperature measuring sensor

101 process chamber

102 sensor arrangement

103 showerhead

104 gas outlet opening

105 substrate

106 substrate

107 substrate

108 susceptor

109 heating element

110 heating element

110 'heating element

111 heating element

112 surface zone

113 surface zone

113 'surface zone

114 surface zone

115 controller

116 controllers

116 'controller

117 controller

118 selection electronics

119 receiving pockets

120 axis of rotation

121 data line

122 control device

123 decoupling device

124 amplifiers 212 temperature curve

213 Temperature curve 213 'Temperature curve

214 temperature curve

F gain factor

K coupling matrix

L decoupling matrix P operating parameters

R radial

Claims

Method for treating at least one substrate (105, 106, 107) in a process chamber (101) of a reactor housing, wherein the one or more substrates (105, 106, 107) are mounted on a heating element (109, 110,

110 ', 111) susceptor (108) is placed, wherein the heating elements (109, 110, 111) spatially associated zones of the susceptor (108) are heated, each surface zones (112, 113, 113', 114) of the Temperatures of the surface zones (112, 113, 113 ', 114) and / or of the at least one substrate arranged thereon () 105, 106, 107) are measured, and the measured values determined with the sensors (1 to 35) are fed to a control device (115, 116, 117, 122) with which the heating power of the heating elements (109, 110, 110 ', 111 ) is regulated, characterized in that for controlling the heating power of the heating elements (109, 110, 110 ', 111) is used in each case a combination of temperature measured values. 2. Apparatus for treating at least one substrate (105, 106, 107) with a reactor housing and a process chamber (101) arranged therein, which has a susceptor (108) for receiving the at least one substrate (105, 106, 107) a plurality of heating elements (109, 110, 110 ', 111) for heating associated surface zones (112, 113, 113', 114) of the susceptor and a plurality of temperature sensors (1 to 35), each at a measuring point Temperature measurement of the surface of the susceptor (108) or arranged there substrate (105, 106, 107), wherein the heating elements (109, 110, 110 ', 111) of a control device (115, 116, 117, 122) with controlled heating power supplied to the control device (115, 116, 117, 122), the temperature measurement values are supplied, characterized in that the control device (115, 116, 117, 122) for controlling the temperature of each of the surface zones (112, 113, 113 ', 114) is a combination of temperature measurement values of a plurality of temperature sensors (1 to 35).

Method according to Claim 1 or in particular according thereto, characterized in that the combination of temperature measured values are assigned to one or more operating parameters (P), the operating parameters (P) being selected from: the setpoint temperatures of the surface zones (112, 113, 113 ', 114), the total gas pressure in the process chamber (101), the chemical composition of the gas phase in the process chamber (101), the material of the susceptor (108), the type of substrate (105, 106, 107), the mounting of the susceptor (108 ) with substrates (105, 106, 107) and / or the aging state of the susceptor (108).

Apparatus according to claim 2 or in particular, characterized by a selection device (118), the input variable (P) one or more operating parameters of: the target temperatures of the surface zones (112, 113, 113 ', 114), the total gas pressure in the process chamber (101 ), the chemical composition of the gas phase in the process chamber (101), the material of the susceptor (108), the nature of the substrate (105, 106, 107), the mounting of the susceptor (108) with substrates (105, 106, 107) and / or the aging state of the susceptor (108), and which, depending on the input variable (P), determines a combination of temperature measurement values to be used for regulation.

Method or device according to one or more of the preceding claims or in particular according thereto, characterized in that the combinations for controlling the measured values used the differences between the individuality or number of used or unused measuring points of the respective surface zone (112, 113, 113 ', 114) and / or that the combinations used to control the measured values used by each other by their weighting with respect to the respective surface zone (112 , 113, 113 ', 114), wherein the weighting of the values are between zero and one and / or that both on the surface of a substrate (105, 106, 107) and on one of the substrates (105, 106 , 107) measured values are used for the control and / or that the measured values are determined in an imaging method.

Method or device according to one or more of the preceding claims or in particular according thereto, characterized in that each heating element (109, 110, 110 ', 111) is assigned individually a regulator (115, 116, 117, 122) which is selected by the selection device ( 118) receives a combination of measured values as selection values, the combinations belonging to different operating parameters (P) differing from one another, or the temperature measured values of each of the surface zones (112, 113, 113 ', 114) having individually assigned characteristic temperatures (Ti, T ₂ , T ₃ , T _n ), from which in the regulation of each of the surface zones (112, 113, 113 ', 114) individually assigned first values (U'i to U' _n ) are obtained which are decoupled from a decoupling device ( 123) are converted into second values (Ui to U _n ) in which the coupling of the first values (Ui to U _n ) is compensated and corresponds to the respective heating powers of the heating elements (109, 110, 110 ', 111) Echen.

Method or device according to one or more of the preceding claims or in particular according thereto, characterized in that for the selection of the temperature or the temperature used for the regulation Sensors (1 to 35) of the operating parameters (P) dependent contribution of the heating power of a heating element (109, 110, 110 ', 111) also on the temperature not this heating element (109, 110, 110', 111) associated surface zone (112 , 113, 113 ', 114) is used.

Method or device according to one or more of the preceding claims or in particular according thereto, characterized in that the combinations are determined in preliminary tests or by computer-aided simulation calculations, wherein as a convergence criterion a predetermined temperature profile, in particular a minimization of the lateral temperature gradient over the process chamber (101) facing Surface of the susceptor (108) is selected and / or that control parameters are determined by the use of neural networks.

Method or device according to one or more of the preceding claims or in particular according thereto, characterized in that a characteristic temperature (Ti, T ₂ , T ₃ , T _n ), which are each supplied to a controller (115, 116, 116 ', 117), which supplies first coupled values (U ^' i to U ^' _n ) which are used to compensate for the coupling of a decoupling device ( 123) in particular with a decoupling matrix (L) into second values (Ui to U _n ) are converted, each corresponding to a heating element (109, 110, 110 ', 111) individually supplied heating power.

Method for treating at least one substrate (105, 106, 107) in a process chamber (101) of a reactor housing, wherein the one or more substrate (105, 106, 107) can be heated on a heating element (109, 110, 110 ', 111) Susceptor (108) is placed, with the Heating elements (109, 110, 111) spatially associated zones of the susceptor (108) are heated, each of which surface zones (112, 113, 113 ', 114) of the process chamber (101) facing side of the susceptor (108) are assigned wherein for each surface zone (112, 113, 113 ', 114) a characteristic temperature (Ti, T ₂ , T ₃ , T _n ) is determined, which a control device (115, 116, 116', 117) for controlling the heating power of Heating elements (109, 110, 110 ', 111) is supplied, wherein the characteristic temperatures (Ti, T ₂ , T ₃ , T _n ) are coupled to each other, characterized in that in the control first coupled values (U ^' i to U 'n) are used, which are converted by a decoupling device (123) in particular with a decoupling matrix (L) to compensate for the coupling into second values (Ui to U _n ).

11. Apparatus for treating at least one substrate (105, 106, 107) with a reactor housing and a process chamber (101) arranged therein, which has a susceptor (108) for receiving the at least one substrate (105, 106, 107) a plurality of heating elements (109, 110, 110 ', 111) for heating corresponding surface zones (112, 113, 113', 114) of the susceptor, means being provided for each surface zone (112, 113, 113 ', 114) a characteristic temperature (Ti, T ₂ , T ₃ , T _n ) can be determined, which are fed to a control device for controlling the heating power of the heating elements (109, 110, 110 ', 111), wherein characteristic temperatures (Ti, T ₂ , T ₃ , T _n ) are coupled to each other, characterized by a decoupling device (123) for compensating the coupling from the first coupled values (U ^' i to U ^' _n ), in particular with a decoupling matrix (L), second values ( Ui to U _n ).

12. A method or device according to claim 10 or 11 or in particular according thereto, characterized in that the characteristic temperature each (Ti, T _{2 /} T 3, T _n ) in each case a controller (115, 116, 116 ', 117) are supplied, which supplies the first coupled values (Ui to U _n ).

13. Method or device according to one or more of the preceding claims or in particular according thereto, characterized in that for determining the decoupling matrix (L) in preliminary tests or by means of model calculations, a coupling matrix (K) is determined, the contribution of each heating element (109, 110, 110 ', 111) to the characteristic temperature (Ti to T _n ) of each surface zone (112, 113, 113', 114), the decoupling matrix (L) being in particular the inverted one

Coupling matrix (K) is.

14. A method or device according to one or more of the preceding claims or in particular according thereto, characterized in that the characteristic temperatures (Ti to T _n ) in each case the mean value of a plurality of surface areas of a selected surface zone (112, 113, 113 ', 114 ) measured temperature values, wherein the measured values are distributed three-dimensionally or two-dimensionally over the surface of the susceptor (108).

15. The method or device according to one or more of the preceding claims or in particular according thereto, characterized in that the surface portions in the circumferential direction spaced apart areas of the associated surface zone (112, 113, 113 ', 114).

16. Method or device according to one or more of the preceding claims or in particular according thereto, characterized in that the characteristic temperatures (Ti to T _n ) are determined by recording and evaluating one or more thermal images of the heated susceptor (108).