Method for emissivity corrected pyrometry.

Abstract

The present invention relates to a method for coating a substrate. E,n ) and reflectance value (U R,n ) is measured using a pyrometer for substrate temperature control. Due to slight differences in the wavelengths of the two pyrometers, the emissivity value (U E,n The actual temperature is determined by the reflectance value (U R,n ) cannot be optically compensated for by using two values ​​(U E,n ) and (U R,n ) curve's angular frequency (ω E , ω R ) differ slightly from one another, leading to oscillations in the actual temperature value. To counter this phenomenon, the present invention proposes the use of values ​​corrected by a numerical time transformation rather than the use of measured values.

Description

[Technical field]

[0001] The present invention relates to a method for coating a substrate with at least one layer. E,n U R,n} is measured multiple times in succession with at least one optical measurement device on the layer during deposition, each of which corresponds to a thermal radiation output measured at a first wavelength, emissivity value U E,n and a reflectance value U measured at a second wavelength that differs at most slightly from the first wavelength. R,n The temperature value of the substrate T i is the combination of measurements {U E,n U R,n} The emissivity value U E,n and the reflectance value U R,n Each of these has an angular frequency ω over time t. E , ω R and lies on a curve that oscillates at an angular frequency ω E , ω R The ratio of is slightly different from 1. The temperature value T i is preferably the actual value T act and using it, the substrate temperature is adjusted to a set value T set The temperature of the substrate is controlled using a temperature controller.

[0002] The invention further relates to an apparatus comprising a computer device programmable and programmed such that the correction values ​​are calculated. [Background technology]

[0003] US Pat. No. 5,399,633 forms the technical background of the present invention. The method of emissivity-corrected pyrometry described therein, also called reflectivity- or reflectance-corrected or emissivity-compensated, allows non-contact optical temperature measurement during thin film deposition with unknown and continuously changing optical properties of the measurement object. The pyrometry method for non-contact temperature measurement utilizes the equation of the thermal radiation emitted by a hot measurement object and the temperature of the measurement object. This equation is described by the known Planck radiation equation and, in fact, except for the emissivity of the object, is uniquely captured by a corresponding prior calibration. The measurement object can be any optically accessible surface of a process chamber for temperature monitoring or control. In the present invention, the measurement object is in particular a substrate or a surface of a substrate in a process chamber during a deposition process for manufacturing a semiconductor layer structure with different approximately stoichiometric compounds of group III (Al, Ga, In) and nitrogen. US Patent No. 5,399,633 describes a method in which a temperature value for a substrate temperature is determined from a combination of measurements, each including an emissivity value and a reflectance value. The two sinusoidal measurement curves are slightly out of phase. Using a numerical time transformation, in which the minimum and maximum values ​​of the curves are determined at different times, a transformed curve is formed and used to calculate the temperature value instead of the emissivity and reflectance values.

[0004] The state of the art is also illustrated by non-patent documents 1 to 3, which are hereafter referred to as Breiland 2003.

[0005] The state of the art also includes US Pat. No. 5,399,633, which describes a related type of device.

[0006] US Pat. No. 5,399,633 describes a method by which the radiation and the reflected radiation from a substrate can be measured during the deposition of a layer.

[0007] US Pat. No. 5,399,633 describes emissivity corrected pyrometry for minimization of residual vibrations.

[0008] Known methods of emissivity correction are based on determining the missing unknown emissivity by measuring the surface reflectance of the object being measured. The emissivity is determined using Kirchhoff's law as ε=1-ρ for opaque substrates. The detection wavelength of the pyrometer is selected such that the selected substrate (silicon in this case) is opaque at wavelengths at typical operating temperatures (T=600-1200°C), e.g. in the range of 800nm-1000nm. For the method to work with sufficient accuracy, the reflectance is measured at exactly the same wavelength as the thermal radiation. The light required for this is provided by a laser. However, the light required for this can also be generated by a diode. This can be more suitable than using a laser, if the filter width is finite. In fact, the pyrometer does not have a definite measurement wavelength, but a wavelength interval (approximately ±10 nm, but it can also be narrower or wider). The thermal radiation output is therefore measured at the first wavelength. The reflectance is measured at the second wavelength. Ideally, the two light wavelengths are identical, but for technical reasons they differ slightly, in particular the deviation is no more than 1%, 2%, 5% or at most 10% from the light wavelength. In some cases it may be more. The light source should preferably have a spectrum that fits the Planck distribution formula. This interval width and the centroid wavelengths of the radiation and the reflectance measurement must match as closely as possible. The reflectance is measured by emitting light of a given wavelength at the sensor position and reflecting the perpendicularly incident light from the wafer surface, as far as possible at the same position as the pyrometer measurement. The reflectance is determined from the measured signal strength of the reflected light using pre-calibration. In practice, the thermal emission and reflectance of an object are often measured alternately rather than simultaneously, so that the reflectance measurement does not interfere with the measurement of thermal emission. See also the "lock-in technology" described in US Pat. No. 5,399,992 on this subject.

[0009] Two different calibration steps are required to measure the temperature accurately. The calibration run allows the determination of the calibration parameters that are included in the calculation of the temperature from the measurement signal. It involves a radiometric calibration with a blackbody radiation source (blackbody oven, special standard source) that establishes the link between the signal intensity and the measured temperature. The use of the calibration parameters thus determined during the measurement allows to determine the “raw temperature” that is not corrected for the effects of the unknown emissivity. A separate calibration step is used to determine the calibration parameters, by which each measured reflectance signal is assigned a reflectance value in the interval 0 to 1. This calibration step is performed on a substrate of known reflectivity (or emissivity in case of opaque substrates), such as silicon, immediately after a desorption process step (removal of the native oxide) on an uncontaminated surface at a known temperature and before the start of the layer deposition.

[0010] When depositing thin films with constant growth rate, if the described known method of emissivity correction is not used, sinusoidally oscillating temperature measurements are observed, which are related to interference effects (Fabry-Perot oscillations) in transparent thin films. In the specific case of MOCVD of GaN or AlGaN on silicon or other materials at temperatures in the range of 950 °C to 1100 °C, the oscillations are up to ± 30 °C. The aim of this procedure is to limit the temperature oscillations to less than ± 2 °C, more preferably to ± 1 °C. This problem can occur not only in the above mentioned material systems, but also in other material systems, e.g. other III-V and II-IV compound semiconductors.

[0011] If the temperature measurement methods described in the prior art are carried out as described, a number of errors arise, as described below. All of these sources of error lead to an incomplete or artificially excessive emissivity correction. An incorrect emissivity correction appears as residual temperature oscillations (residual oscillations) whose amplitude is larger than the desired degree of error. The patent document 2 cited at the beginning already presents one approach intended to minimize this source of error.

[0012] The GaN on Silicon (AlGaN) material system has been found to be particularly susceptible to the described error sources since the measured reflectivity R oscillates between values ​​close to 0 and 0.5 depending on the values ​​of the refractive index of the layer and wafer materials and on the junction of the transparent layer with the opaque substrate. Other material systems, for example the AlGaInP / GaAs system, can also be affected by this problem. It should be noted that in this case the impact is not as large.

[0013] The observed errors may be due to the following real-world errors: - During reflectance calibration, the exact value of the reflectance of the calibration object is unknown, so the reflectance values ​​used in the calibration do not match the physical reflectance, and the reflectance values ​​used for emissivity correction are inaccurate. - Errors in the adjustment and setup of the measurement optics. Due to scattering at layer boundaries in semiconductor stacks during reflectance measurements, part of the actually reflected light is not recorded. - Radiation is scattered from hot surfaces in the process chamber, reaching the measurement head via multiple reflections on the process chamber walls and the wafer surface.

[0014] In the manufacture of electronic components, such as transistors in circuits for power conversion or high frequency amplification, in which the material pairs described above or elsewhere are used, in an embodiment of the known method on which the invention is based, the measured wafer temperature is used for temperature control in a closed control loop, which significantly impairs the control and reproducibility of the deposition process and the yield rate of usable components per wafer. The temperature control adjusts the heating device so that the measured temperature always corresponds to a specific set point. The physical temperature then oscillates around the residual amplitude of the insufficiently corrected temperature oscillations, which are measurement artifacts. The component comprises a multilayer structure deposited on a substrate having a first portion and a second portion. A transition layer, in particular AlGaN, and a buffer layer, in particular made of GaN, are deposited on the first portion. An AlGaN barrier layer is deposited on the GaN buffer layer such that a two-dimensional electron gas is formed in the region of the layer boundary between the GaN layer and the AlGaN barrier layer. The impairment of reproducibility is especially related to the fact that component structures typically consist of a sequence of functional blocks consisting of a thin AlN seed layer on a Si substrate, a thick buffer layer sequence, and a relatively thin but temperature-sensitive barrier layer of AlGaN or AlInN. At the edge of the buffer layer, depending on the random phase position of the residual measured temperature oscillations, the deviation of the physical temperature from the target value has different values ​​from run to run or wafer to wafer, which changes the composition of the barrier layer, which is important for the function of the component, to different values.

[0015] The following theoretical corrections of this measurement method to compensate for the above-mentioned error sources are given in the state of the art and are based on mathematically deducible facts, such that the effects of a series of error sources are effectively compensated for by an additional correction value γ, and the residual vibrations can theoretically be reduced to zero.

[0016] A similar method is described in US Pat. No. 5,399,633.

[0017] The starting point of the invention is the equation between the measurement signal detected by a pyrometer due to the thermal radiation of the wafer surface and the temperature of the wafer surface, in which the emissivity of the wafer surface, which differs from 1 due to changes in its physical and optical properties during layer growth, is taken into account.

[0018] This equation is described by Planck's radiation law in the Wien approximation, and is written as follows:

number

[0019] The measurement signals of thermal radiation UE and reflectivity UR of the wafer surface are recorded as close as possible in time and space. The signal corresponding to the thermal radiation is the radiation intensity, which is captured by a detector in the pyrometer and converted into a temperature measurement by the calibration parameters A and B, which can be determined by a blackbody calibration performed before the start-up. The reflectivity signal is generated by measuring the intensity of an optical signal, whose wavelength is as close as possible to the thermal radiation measurement. This optical signal is emitted by the measurement unit and reflected from the reflecting wafer surface to the detector. Due to the optical thin film effect (Fabry-Perot effect), the reflectivity of the wafer surface exhibits an approximately sinusoidal variation over time during the deposition of the thin film. With typical growth rates of 0.5-5 μm / h in GaN-on-Si processes and a wavelength of 950 nm used in the measurements, the duration of the oscillation period is up to 10 minutes.

[0020] From equations (1) and (1a), the following equation is derived:

number

number

[0021] In a particular configuration of the method for a Planetary Reactor® with several single wafers, the pyrometer is fixedly mounted on an optical window at the top of the process chamber pointing at a point on the surface of the substrate carrier. For thermal averaging, the substrate rotates slowly around the center of the reactor during coating. A typical rotation period is about 12 seconds, which corresponds to 5 revolutions per minute. However, the rotation speed can be faster or slower. The combination of measurement signals UE and UR is therefore recorded at a specific location on the wafer of interest for the measurement every 12 seconds. Figures 1 and 2 show the configuration used. The measurement location 13 may be a measurement area over which several measurements are made. The combination of measurement signals UE and UR corresponds to an average value over this area. The measurement location 13 or its measurement area may be on each wafer 7 and may be located in the center of the wafer, at the edge of the wafer or anywhere in between.

[0022] It is only possible to carry out a rough correction of the emissivity values ​​using the above-mentioned method. In fact, it is technically impossible to measure the reflected and radiometric signals at exactly the same wavelength. The wavelengths differ from each other by a few nanometers or even only 1 nanometer, for example depending on the precision of the dexterity and the tolerances of the filters. Among other things, the wavelength of the emissivity value curve depends on the temperature. The period of the oscillation curve, one of the emissivity values ​​and the other of the reflectivity values, is about 300 seconds. The difference in period is of the order of about 0.1 seconds. It can be up to 0.1 seconds, 0.2 seconds, 0.5 seconds or even up to 1 second. This path difference causes the temperature determined using equation 3 to oscillate with residual oscillations as shown diagrammatically in FIG. 3. However, the above-cited patent application WO 2005 / 023366 already presents a method for reducing or eliminating the residual oscillations. To do this, however, it is necessary to measure one complete oscillation of the curve. [Prior art documents] [Patent documents]

[0023] [Patent Document 1] U.S. Patent No. 6,398,406 [Patent Document 2] JP 2017-017251 A [Patent Document 3] DE 10 2018 106 481 A1 [Patent Document 4] German Patent No. 44 19 476 [Patent Document 5] DE 10 2020 111 293 A1 [Patent Document 6] DE 10 2020 126 597 A1 [Non-patent literature]

[0024] [Non-Patent Document 1] WG Breiland, Technical Report SAND2003-1868, June 2003. [Non-Patent Document 2] internet<URL:https: / / www.osti.gov / biblio / 820889> [Non-Patent Document 3] internet<URL:https: / / prod-ng.sandia.gov / techlib-noauth / access-control.cgi / 2003 / 031868.pdf> Summary of the Invention [Problem to be solved by the invention]

[0025] The problem that the invention aims to solve is to improve the above-mentioned method with regard to the accuracy of the determination of the correction values. [Means for solving the problem]

[0026] This problem is solved by the invention as set forth in the claims, the dependent claims present independent solutions to the problem as well as advantageous further developments of the invention presented in the main claim.

[0027] First, essentially the emissivity and reflectivity values ​​are not used to calculate the actual value of the substrate surface temperature, but rather the emissivity value U E,i or reflectance value U R,i Convert value U from either * i are then used in place of the emissivity or reflectance values ​​in the calculation of the actual values. The transformation may be a modification of the emissivity or reflectance values. By means of the transformation, the curve of emissivity or reflectance values ​​can be mapped onto a curve oscillating at different angular frequencies in time. The transformation can be performed in such a way that after the transformation, two curves are available, either the transformed emissivity curve and the reflectance curve, or the transformed reflectance curve and the emissivity curve, which have the same angular frequency. The curve of the transformed measurement value is thus more or less stretched or compressed in time, which can be said to be a temporal transformation of the measurement value. In particular, the measurement value itself is transformed. The transformation value U * i To determine , consecutive measurements taken at different times, for example one measurement taken immediately after another, may be used. According to a first aspect of the invention, the measured values ​​used to calculate the conversion value include the current measured value, i.e. the most recently measured value in each case. Preferably, this value and the value measured immediately before it are used. However, in addition to the current measurement, the measurement taken immediately prior to the previous measurement may also be used, and in fact several measurements taken prior to the current measurement may be used. Alternatively or in combination with the first aspect of the invention, it is provided that in each case a value of the slope of the curve of the measured emissivity or reflectance values ​​is calculated from at least two measured values, this slope value being used in the calculation of the conversion value. A further development of the state of the art according to the invention makes it possible to calculate updated transformation values ​​at any time. A conversion factor may be provided to be used for the conversion. The conversion factor may be determined in a preliminary test. The conversion factor may be the ratio of the angular frequencies of the curves of the reflectance and / or emissivity values ​​oscillating in time. The conversion factor may be provided to be greater than 1, but essentially only slightly greater than 1. On the other hand, depending on the formulation of the conversion, the conversion factor may be less than 1, but essentially only slightly less than 1. The value of the ratio is, for example, 1+10 -2 and 1-10 -6 Range between or 1+10 -6 and 1+10 -2 In a preferred variant of the invention, to calculate the actual temperature, instead of the emissivity or reflectance values, time-transformed emissivity or reflectance values ​​are used. For the time conversion, it is further preferred that a transformed time is formed which consists of the untransformed time, a conversion factor a and, optionally, a phase factor b such that

number

number

number

number

number

[0028] Using the method according to the invention, the conversion factor a, which is preferably determined in a preliminary test, can be used to calculate the actual temperature with progressively smaller residual vibrations in each case using the current measured value and at least one previously recorded measured value.

[0029] One or more preliminary tests are carried out to determine the conversion factor. In these preliminary tests, a layer is deposited on a substrate with the same or similar process parameters as those for carrying out the method having the above-mentioned characteristics. A combination of measurements is thereby obtained, one emissivity value and one reflectance value, which is stored. An optimization is then carried out, in which the experimental value of the conversion factor a is varied using the stored measurements and the above-mentioned formula until the residual oscillation of the temperature curve plotted over time is minimized, i.e. for example until the area integral under the residual oscillation curve is minimized.

[0030] However, the conversion factor can also be determined during the same "run", for example if other layers, in particular buffer layers, are deposited prior to the deposition of the layer.

[0031] Furthermore, it is also possible to adapt the conversion coefficients continuously, which can in particular be performed by depositing several layers one on top of the other on the substrate.

[0032] This transformation is preferably formed by a differential quotient and makes use of the time derivative of the measurement curve plotted over time.

number

[0033] The slope of the transformed measurement curve can be calculated from the slope of the untransformed measurement curve by means of a conversion factor. The differential quotient can be calculated from successively measured measurement points. The differential quotient is preferably calculated using successive measurement points. However, these do not necessarily have to be the last two. Other measurement points can also be used, especially if there are not enough time points measured from the transformed section. However, measurements that are not taken one after the other, but are derived from measurements that are more widely separated, can also be used to calculate the differential quotient. This applies in particular when the transformed measurement is outside the interval between the two measurements used to determine the differential quotient. The measurement interval can then be extended backwards in time until the transformed measurement falls within the measurement interval of the measurements used to determine the differential quotient.

[0034] The temperature value calculated by the above-mentioned method represents the surface temperature of the substrate. The temperature of the substrate is preferably regulated by means of a temperature control device, the control circuit used for this being provided with a value against which the actual value is controlled. Preferably, the temperature value calculated by means of the conversion according to the above-mentioned method is used as the actual value.

[0035] The invention also relates to an apparatus for carrying out the method. The apparatus comprises two optical measuring devices capable of measuring emissivity and reflectivity values. This is preferably one measuring device functionally incorporating the two optical measuring devices, which alternately measures emissivity or reflectivity values. The apparatus comprises a computer device for calculating an actual value of the surface temperature of the substrate or of a layer deposited on the substrate. The computer device is configured to form a converted value from either the emissivity or reflectivity values ​​using a conversion according to the method described above, and to use this instead of the emissivity or reflectivity values ​​to calculate the actual value.

[0036] The invention further relates to a CVD reactor comprising a temperature control device for controlling the temperature of the substrate, for example a heating device, and a device for determining the actual value of the substrate temperature as described above. The CVD reactor may comprise a gas inlet member, by means of which a process gas can be fed into the process chamber. The process gas, which may be an organometallic compound from main group III or a hydride from main group V, may be fed into the process chamber together with an inert gas, for example a noble gas or hydrogen. However, a process gas comprising compounds from main groups II and IV or a compound from main group IV may also be used. The one or more substrates are arranged on a susceptor forming the floor of the process chamber. Temperature control of the substrates is ensured by heating of the susceptor. A heating device is provided which may be arranged preferably below the susceptor for this purpose. The heating device is controlled by a computer device with reference to the actual value calculated from the reflectance and emissivity values ​​by the method described above.

[0037] The above method may be modified so that the time conversion is optionally performed on reflectance or emissivity values, and the conversion coefficients may be prefixed with either positive or negative numbers. The time transformation may use either linear or non-linear interpolation. Linear interpolation is advantageous in that only two measurement points are needed to determine a value that lies between those two measurement points. Higher levels of interpolation require more measurement points. The method is particularly related to temperature measurements using transformations. However, the invention is particularly advantageous for calculating temperatures that are used as actual values ​​in closed control loops. [Brief description of the drawings]

[0038] Exemplary embodiments of the present invention will now be described with reference to the drawings. [Figure 1] FIG. 1 shows a schematic cross-sectional view of an apparatus for carrying out the method. [Diagram 2] FIG. 2 is a detail along line II-II above the susceptor 4 on which the substrate 7 in FIG. 1 is placed, and a schematic diagram of measurement points 13 at which the emissivity values ​​UE,i and reflectivity values ​​UR,i can be measured using an emissivity measuring device 11 and a reflectivity measuring device 12. [Diagram 3] Figure 3 shows a schematic diagram of the time progression of a measurement curve UR ​​of several reflectance measurements recorded over time and a measurement curve UE of several emissivity measurements recorded over time, the curves being normalized for the purposes of clarity. The angular frequency ωE of the measurement curve of the emissivity values ​​is slightly greater than the angular frequency ωR of the measurement curve of the reflectance values. The result is the oscillation of the sum of the two normalized curves, shown as a dotted line and characterizing the calculated temperature progression. [Figure 4] FIG. 4 shows a detail of region IV of FIG. 3, where a somewhat larger ratio of the two angular frequencies has been used for purposes of clarity, and the dotted line represents the transformed curve of reflectance values ​​rather than the sum of the two curves. [Diagram 5] FIG. 5 shows a detail of region V in FIG. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0039] The CVD reactor shown in Figures 1 and 2 includes a reactor housing 1, a heating device 5 disposed therein, a susceptor 4 disposed above the heating device 5, and a gas inlet member 2 for introducing, for example, TMGa, TMAI, NH3, AsH3, PH3, and H2. The susceptor 4 is rotated about a vertical rotation axis a by means of a rotation drive 14. For this purpose, one side of the rotation shaft 9 is connected to the rotation drive 14 and the other side is connected to the underside of the susceptor 5.

[0040] The substrate 7 is placed on the horizontal surface of the susceptor 5 facing away from the heating device 5. A substrate holder 6 is provided, on which the substrate 7 is placed. The substrate 7 is located radially outward of the rotation axis a, and its position is held by the substrate holder.

[0041] Two measuring devices may be provided. The emissivity measuring device 12 may be formed by a pyrometer. The reflectivity measuring device 11 may also be formed by a pyrometer. A beam splitter 10 may be provided, by which the input beam is split between the two measuring devices 12, 11. The beam path is incident on the substrate 7 at a measuring point 13. Figure 2 shows that the measuring point 13 moves over all substrates 7 during the rotation of the susceptor 4.

[0042] However, the two measuring devices 11, 12 may also be combined into one measuring device.

[0043] FIG. 3 shows the reflectance measurement value U R,i The solid line shows an interpolated measurement curve through several measurement points, not shown separately. The oscillations are due to reflections at the interface layers. The dashed lines show several emissivities U measured with the measuring device 12. E,i The two curves correspond to the curves interpolated through measurements of ω. Here too the oscillations are due to reflections at the layer interfaces. The two curves can vary in size as a result of the tolerances of the filters used, e.g. their angular frequency ω R and ω Ehave slightly different path differences from each other. As a result of this deviation, the sum of the two curves oscillates, which is represented by the dotted line and is qualitatively equal to the calculated temperature course.

[0044] The temperature T can be calculated by the above-mentioned formula 3 using the above-mentioned formulas 1 and 2.

[0045] Figure 4 shows the reflectance values ​​U plotted over time t. R,i The larger angular frequency ω of the measured curve R Therefore, this curve corresponds to the emissivity value U E,i According to the invention, the transformed curve is transformed into the reflectance measurement U by a suitable transformation. R,i and is represented by the dotted line. For the calculation of temperature, the converted measurement value U * R,i is used, and the measurement time t i In Figure 4, we show the combination of two such measurements and the single transformed measurement U * R,i are shown, and the indices i are assigned the values ​​1 and 2. The transformed curve U * R Angular frequency ω * R is the emissivity value U E,i The angular frequency ω of the curve E It has the same value or approximately the same value.

[0046] Figure 5 shows the transformed measurements U * R,i It will be helpful to explain in detail how to determine the angular frequency ω. In fact, the curves shown in Figs. 3 and 4 are not strictly sinusoidal. They are shown as sinusoidal in Figs. 3, 4 and 5 only for the sake of illustration. Nevertheless, the curves are periodic, so that in preliminary tests it was possible to determine the angular frequency ω R and ω E can be determined by layer deposition. From these two angular frequencies the following ratio is formed:

number

number

[0047] The converted measurement U * R,i The calculation of is done using two reflectance measurements U taken at times t1 and t2 in the manner of a Taylor series truncated after the first order term. R,1 and U R,2 The reflectance measurement U R,2 A quantity is subtracted from the slope U of the transformed reflectance curve. *’ R,2 and at2-t2, the difference between the transformed and untransformed time.

number

[0048] Gradient U * R,2 is the reflectance measurement value U using the conversion coefficient a. R,i can be obtained from the slope of the untransformed curve.

number

number

number

number

[0049] In the exemplary embodiment (see FIG. 5), the converted measurement U * R,2 is the sum of two measurements U taken at times t1 and t2, one immediately after the other. R,1 and U R,2 The converted measurement value U at time t1 is located at the bottom between R,1 Under other conditions, such as when the bottom of the measurement U measured at time t2 is located below the bottom of the measurement U R,2 Measurements recorded before time t1 (eg, time t0) are used to form a differential quotient from which a gradient triangle is formed, with which a transformed measurement is formed.

[0050] The conversion factor may be slightly temperature dependent.

[0051] The foregoing is intended to serve as an illustration of the invention as generally described herein, which furthers the state of the art in at least the following combination of features, each with its own advantages, and in particular may combine any two, some or all of the combinations of features described.

[0052] The converted value U * i is the emissivity value U E,i and / or the reflectance value U R,i and a temperature value T i In the calculation of the emissivity value U E,i or reflectance value U R,i is used instead of

[0053] The converted value U * i is the emissivity value U E,i or reflectance value U R,i The method according to claim 1, wherein the time domain is formed by any one of the numerical time transformations.

[0054] Emissivity values ​​U measured at different times E,i or reflectance value U R,i At least two measurements of * i The method according to claim 1, wherein the

[0055] Conversion U i (t) → U * i (t * i ) is determined in a preliminary test, and / or the conversion factor a is determined based on the angular frequency ω E , ω R The method according to claim 1, wherein the ratio of

[0056] During the deposition of the layer, the value U * i is the combination of measurements {U E,n U R,n}, the expression U * i =U i -(1-1 / a)·t i ·U' i is determined according to t i is the time t from the start of the layer deposition, optionally corrected by a phase offset, and U i is the time t i Emissivity value U at E,i or reflectance value U R,i is a measurement of either ’ i is time t i The measured emissivity value U E,i or reflectance value U R,i is the slope of the curve passing through * i is the measurement value U to calculate the temperature E,i , U R,i is used instead of

[0057] Gradient U ’ i is the differential quotient of two measurements taken successively in time. (U i -U i-n ) / (t i -t i-n ) The method according to claim 1, wherein the calculation is performed by forming

[0058] Temperature value T i is the following formula T i =B / (ln(U E,i )-ln(A·(1-α·U R,i ))) where A, B and α are the calibration parameters and the measurement U E,i , U R,i One of them is the time-converted value U * i wherein:

[0059] Angular frequency ω of the progression of the curveE is the reflectance value U R,i The angular frequency ω of the progression of the curve R If the emissivity value U is greater than E,i is time transformed.

[0060] Temperature value T i However, for the control loop of the temperature controller for controlling the temperature of the substrate, the actual value T act and the substrate temperature is set to the T set The method according to claim 1, wherein the

[0061] The conversion factor a is determined during a previously performed deposition of a layer on the substrate, and during the deposition, the combination of measurements {U E,n U R,n} are recorded multiple times in succession, each with an emissivity value U E,i and reflectance value U R,i and whereby a periodic compensation curve is generated.

[0062] First, a layer is deposited on a substrate to determine the conversion factor a using the growth parameters {U E,n U R,n} and then storing the stored emissivity values ​​U E,i From or stored reflectance value U R,i Convert value from U * i is formed and this converted value is used to calculate the emissivity value U E,i or reflectance value U R,i wherein the test value is varied until the amplitude of the residual oscillations of the curve of temperature values ​​T calculated according to any one of claims 1 to 9 and plotted over time is minimized.

[0063] A computer device calculates the emissivity value U using a transformation according to any one of claims 1 to 11. E,i or reflectance value U R,i Convert value U from either *i and a temperature value T i In the calculation of the emissivity value U E,i or reflectance value U R,i An apparatus configured to use these instead of.

[0064] The temperature value T of the substrate temperature according to claim 12 i The CVD reactor is characterized by an apparatus for determining

[0065] All disclosed features are essential to the invention (both for themselves and in combination with one another). The disclosure of the present application includes in its entirety the disclosure content of the relevant / attached priority documents (copies and earlier applications), also with a view to incorporating the features of these documents into the claims of the present application. The dependent claims are characterized by an independent, inventive further development of the prior art, even without the features of the claims cited, in particular in order to file a divisional application on the basis of these claims. The invention specified in each claim may additionally have one or more features specified in the preceding description, in particular those given reference signs and / or specified in the explanation of the signs. The present invention also relates in particular to embodiments in which individual ones of the features mentioned in the preceding description are not implemented, insofar as they are obviously unnecessary for the respective purpose of use or can be replaced by other means having the same technical effect. [Explanation of symbols]

[0066] 1 Reactor housing 2 Gas inlet member 3 Gas supply lines 4 Susceptor 5 Heating device 6 Substrate holder 6' Gas Cushion 7 Substrate 8 Process Chamber 9 Rotation Axis 10 Beam splitter 11 Reflectance measuring device 12 Emissivity measuring device 13 Measurement points 14 Rotational drive unit 15. Computer Devices 21 Multilayer structure 22 Substrate 30 Barrier Layer 31 Cover Layer α Calibration parameter Gamma correction value Gamma i Measurement interval correction value λ wavelength A Calibration parameters B. Calibration parameters U E Emissivity Value U R Reflectance Value U E,n Emissivity Value U R,n Reflectance Value T M Measurement temperature t i Time i Measurement index {U E,n U R,n} Combination of measurements T act Actual Value ω E , ω R Angular Frequency U R,i Reflectance Value U E,i Emissivity Value U * i , U * E,i Conversion Value U ’ i , U *’ i , U *’ R,i Time derivative W * i Conversion Value

Claims

1. 1. A method for coating a substrate with at least one layer, comprising the steps of: During deposition of the layer, a combination of measurements ({U Ｅ，ｎ U Ｒ，ｎ }) are successively determined a number of times, each of which corresponds to an emissivity value (U Ｅ，ｎ ) and the reflectance value (U Ｒ，ｎ ) and The temperature value of the substrate temperature (T ｉ ) is the combination of measurements ({U Ｅ，ｎ U Ｒ，ｎ }) is calculated from The emissivity value (U Ｅ，ｎ ) and the reflectance value (U Ｒ，ｎ ) each vibrates at an angular frequency (ω Ｅ , ω Ｒ ) and the angular frequency (ω Ｅ , ω Ｒ ) ratio is slightly different from 1, Emissivity values ​​(U Ｅ，ｎ ) or reflectance value (U Ｒ，ｎ ) are used to obtain the converted values ​​(U ＊ ｉ ) is the emissivity value (U Ｅ，ｎ ) and / or the reflectance value (U Ｒ，ｎ ) and said temperature value (T ｉ In calculating the emissivity value (U Ｅ，ｉ ) or the reflectance value (U Ｒ，ｉ ) is used instead of Measurement value (U Ｅ，ｎ Or U Ｒ，ｎ ) is a current measurement, and / or The measured emissivity value (U Ｅ，ｎ ) or the reflectance value (U Ｒ，ｎ ) is the slope of the curve of the measured value (U Ｅ，ｉ , U Ｒ，ｉ ) and the converted values ​​(U ＊ ｉ ) is the gradient (U' ｉ ) values.

2. The time transformation (U ｉ (t) → U ＊ ｉ (t ＊ ｉ ) the conversion factor (a) used in) is determined in preliminary tests, and / or The conversion coefficient (a) is the angular frequency (ω Ｅ , ω Ｒ 2. The method of claim 1, wherein the ratio of

3. During the deposition of the layer, the value (U ＊ ｉ ) is expressed as U ＊ ｉ =U ｉ -(1-1 / a)・t ｉ ・U' ｉ The combination of measurements ({U Ｅ，ｎ U Ｒ，ｎ }) are determined for each of t ｉ is the time (t) from the start of deposition of the layer, optionally corrected by a phase offset; U ｉ is the time (t ｉ The emissivity value (U Ｅ，ｎ ) or the reflectance value (U Ｒ，ｎ ) is a measurement of a is a transform coefficient, and U' ｉ is the time (t ｉ The measured emissivity value (U Ｅ，ｎ ) or the reflectance value (U Ｒ，ｎ ), and the value (U ＊ ｉ ) is used to calculate the temperature Ｅ，ｎ , U Ｒ，ｎ 2. The method of claim 1, wherein the .alpha.-

4. The gradient (U′ ｉ ) is the differential quotient of two measurements taken in succession in time. (U ｉ -U ｉ－ｎ ) / (t ｉ -t ｉ－ｎ ) The method according to any one of claims 1 to 3, characterized in that the calculation is performed by forming:

5. The temperature value (T ｉ ) is expressed as follows: 4 ｉ =B / (lnn(5 Ｅ，ｉ )-ln(A・(1-α・U Ｒ，ｉ ))) is calculated according to (A, B and α) are the calibration parameters and the measured value (U Ｅ，ｉ , U Ｒ，ｉ ) is the time-converted value (U ＊ ｉ 5. The method according to claim 1, wherein the first and second radicals are replaced by the second radicals.

6. The emissivity value (U Ｅ，ｉ ) of the progression of the curve, said angular frequency (ω Ｅ ) is the reflectance value (U Ｒ，ｉ ) of the progression of the curve, said angular frequency (ω Ｒ ), the emissivity value (U Ｅ，ｉ ) otherwise the reflectance value (U Ｒ，ｉ 6. The method according to claim 1, wherein the time is converted.

7. The temperature value (T ｉ ) is the actual value (T ａｃｔ ) and the substrate temperature is set to the set value (T ｓｅｔ 7. The method according to claim 1, wherein the concentration of the oxidized carbon dioxide is adjusted to 1.

8. The conversion factor (a) is determined during a previously performed deposition of a layer on a substrate, and the combination of measurements ({U Ｅ，ｎ U Ｒ，ｎ }) are recorded multiple times in succession, each with an emissivity value (U Ｅ，ｉ ) and reflectance value (U Ｒ，ｉ 8. The method according to claim 2, further comprising:

9. To determine the conversion factor (a) using the first growth parameters, a layer is deposited on a substrate and a combination of measurements ({U Ｅ，ｎ U Ｒ，ｎ }) and then measuring and storing the stored emissivity values ​​(U Ｅ，ｉ ) or the stored reflectance value (U Ｒ，ｉ ) to the converted value (U ＊ ｉ ) is formed and this converted value is used in the calculation of the temperature value (T) Ｅ，ｉ ) or the reflectance value (U Ｒ，ｉ 9. The method according to claim 2, wherein the test value is used instead of the temperature (T) value calculated according to claim 1 and the test value is varied until the amplitude of the residual oscillations of the curve of temperature values ​​(T) calculated according to claim 1 and plotted over time is minimized.

10. Emissivity value (U Ｅ，ｉ a first optical measurement device for measuring the reflectance value (U Ｒ，ｉ a second optical measuring device for measuring the temperature value (T ｉ and a computer device for calculating The computer device converts the emissivity values ​​(U Ｅ，ｉ ) or the reflectance value (U Ｒ，ｉ ) to the conversion value (U ＊ ｉ ) and a temperature value (T ｉ In calculating the emissivity value (U Ｅ，ｉ ) or the reflectance value (U Ｒ，ｉ ) in place of the

11. 1. A CVD reactor having a temperature control device for controlling a temperature of a substrate, comprising: The temperature value of the substrate temperature (T ｉ 4. A CVD reactor comprising: a device for determining a temperature at which the CVD reactor

12. A method, an apparatus or a CVD reactor characterized by one or more of the features characterized by any of claims 1 to 12.

Images