Method for depositing a iii-v layer sequence containing a thin layer, and device configured therefor

Abstract

The invention relates to a method for determining the layer composition (x) of a thin AlXGa1-XN layer deposited on a substrate (7), wherein an optical variable, in particular the reflection (R) of light having a light wavelength, is measured by an optical measuring unit (13), wherein the thickness of the thin layer (27) is significantly smaller than the light wavelength, wherein one or more subsequent layers (28) are deposited onto the thin layer (27) and a series of measurement values of the optical variable, in particular of the reflection (R), is recorded temporally successively during the deposition of the subsequent layers (28), and wherein the layer composition (x) of the thin layer (27) is determined by evaluating a temporal profile of a measurement curve formed from the measurement values. For this purpose, the measurement curve is examined in regard to characteristic shapes of its profile and characteristic value tuples are formed from the temporal positions (31, 32, 33) of the characteristic shapes and the values (34, 35, 36) of the measurement curve at the temporal positions of the characteristic shapes and the characteristic value tuples are compared with comparison tuples.

Description

Description Method for depositing a thin layer containing III-V layer sequence and device adapted for this purpose field of technology

[0001] The invention relates to a method and a device for depositing multiple layers on a substrate, wherein the method includes determining the property of a thin layer in the layer sequence. During the deposition of the layer sequence, the reflection of light at a specific wavelength is measured using an optical measuring device. The thickness of the thin layer is significantly less than the wavelength of light, so that a direct measurement of the layer composition and layer thicknesses is not possible. During the deposition of subsequent layers deposited on the thin layer, successive measured values of the reflectance and emissivity are obtained. Since the layer sequence is transparent to light at this wavelength and the light is reflected at interfaces between layers in the layer sequence, a measurement curve obtained from the measured values has an oscillating shape.Based on the curve, a layer property can only be determined if the layer is sufficiently thick, with the layer thickness being greater than half the wavelength. State of the art

[0002] DE 10 2022101 809 Al and DE 10 2018 106481 Al describe a method in which reflectance and emissivity values are measured for temperature measurement.

[0003] JP 3876787 B2 shows a method for depositing an AlGaN Layer on a GaN substrate, wherein the GaN layer is illuminated with light from a light is illuminated for a certain length of time and the reflection is measured to determine the layer thickness of the GaN layer.

[0004] US 2002 / 0113971 A1 describes a method for determining characteristic properties during the deposition of III-V layers by evaluating a measurement curve formed from optical measurement values.

[0005] US 10,658,251 B2 describes an apparatus and a method for depositing an AIN layer on a SiC substrate, wherein reflectance values are determined during deposition.

[0006] DE 10 2015 108 878 B3 describes a semiconductor component and a method for manufacturing a semiconductor component, wherein the composition of a layer consisting of more than two elements is determined by means of reflectometry.

[0007] US 2015 / 0203966 A1 describes a method for determining the layer thickness of a thin silicon-containing layer deposited on a substrate, wherein further, in particular thin, layers are deposited onto the thin layer and, during the deposition of these subsequent layers, the reflection of light at a light wavelength is measured using an optical measuring device. A series of measured reflection values are recorded sequentially. The temporal progression of a measurement curve formed from the measured values is evaluated in order to determine the layer thickness of the thin layer. For this purpose, a mathematical model is used with which the temporal progression of a measurement curve is numerically calculated. The measured curve is compared with curves calculated according to the mathematical model in order to determine the layer thickness. Summary of the invention

[0008] The invention is based on the object of specifying a method with which layer properties of a III-V layer, in particular the layer composition of a thin III-V layer, which is deposited in particular on a buffer layer structure, can be determined during the growth of the subsequent layers in order to obtain information as early as possible as to whether the layer composition of the thin layer lies within a tolerance range in order to make a decision as to whether the deposited layer sequence can be used for the further processing of a semiconductor component.

[0009] The problem is solved by the invention specified in the claims. The subclaims represent not only advantageous developments of the technical teaching specified in the subordinate claims, but also independent solutions to the problem.

[0010] Firstly and essentially, it is proposed that in the manufacture of semiconductor components the manufacturing process is terminated if a layer property of an active layer sequence deposited on a buffer layer or a buffer layer sequence has values that lie outside a tolerance range, wherein the active layer sequence has at least one thin layer that has a property that cannot be optically determined by direct observation of the growth of the thin layer and cannot be determined in any other way after the deposition of further layers on the active layer sequence. According to the invention, the property is determined by indirect observation of the growth, namely during the deposition of one or more subsequent layers deposited on the thin layer. This is carried out using an optical measuring device with which in particular the reflectance is measured.To measure the thickness of a deposited layer, a pyrometer can be used to measure the emissivity of the deposited layer. determined, as described, for example, in DE 102020126597 A1. To correct the measured emissivities, the reflectance of the layer or another optical parameter, for example a band gap within the layer, is also preferably measured using a pyrometer. In the method according to the invention, during the deposition of one or more subsequent layers, the surface of the layer currently being deposited is illuminated with a light source, for example a laser. The reflected light is measured with the pyrometer. Measurement values are obtained. A measurement curve is created from measurement values obtained successively at short time intervals. The measurement curve has characteristic shapes, for example one or more minima or maxima. This measurement curve is evaluated and compared with reference curves obtained by model calculations or determined in preliminary tests.The comparison curves represent measured values corresponding to thin layers that exhibit a specific property, where the property is a layer composition. However, the property can also be a layer thickness. The comparison curve from the model calculation or a preliminary test that shows the greatest degree of agreement with the measurement curve provides the specific value of the desired property. According to a first aspect of the invention, the manufacturing process of a semiconductor component is aborted or an external message is transmitted if the determined value of the property lies outside a tolerance range.

[0011] The method according to the invention can also be used to determine several layer properties simultaneously, for example, the layer composition and the layer thickness. The determination of parameters such as layer thickness and layer composition can also be determined by an optimization task. In this case, the previously determined mo- Arbitrary input parameters are fed into the model. In an optimization phase, these input parameters are changed, and the model's output values are compared with the actual measured values, for example, by determining the squared deviation. If the deviation between the model's output values and the actual measured values is below a certain limit, the corresponding input parameters are considered the parameters to be determined.

[0012] According to a first aspect of the invention, a comparison with concrete numerical values is carried out to determine the value of the at least one property. For this purpose, characteristic value tuples are formed. The value tuples are formed from characteristic shapes of the course of the measurement curve. The measurement curve is examined, for example, for inflection points, minima, or maxima of its course. It can be provided that a value of the measurement curve at a characteristic shape, for example the value of the measurement curve at an inflection point, a minimum, or a maximum, is an element of the value tuple. It can also be provided that a value of the measurement curve that has been measured at an offset that marks the beginning of the deposition of a first subsequent layer deposited onto the thin layer, is an element of the value tuple.Furthermore, it can be provided that the temporal position of the characteristic shape, for example, an inflection point, a minimum, or a maximum, is an element of the value tuple. It can also be provided that one or more amplitude values are determined, for example, by determining a difference between a minimum value and a maximum value of the measurement curve, and this is at least one amplitude value element of the value tuple.

[0013] In an analogous manner, for example, comparison curves are determined using the model calculation or the preliminary tests and comparison tuples are created from these formed by determining value tuples in the same way as described above. A large number of comparison tuples can be formed, with each of the comparison tuples representing the value of a layer property, for example the value of a layer composition or even a layer thickness. After determining a value tuple, this can be compared with the comparison tuples. The value tuple that has the greatest match with the determined value tuple is sought. For this purpose, a sum of the squares of the differences of the respective elements can be formed, for example. The value tuple found in this way then provides the desired property, which can then be compared with a target value in order to decide whether the manufacturing process should be discontinued.With the method according to the invention, the measurement curves and the comparison curves are not directly compared with one another, for example by attempting to find the comparison curve for a measurement curve that apparently comes closest to the course of the measurement curve. In principle, the measurement curves and the comparison curves have similar courses, i.e. inflection points, maximum values or minimum values at similar points, which are referred to as characteristic shapes. According to the invention, these characteristic shapes are identified and their position on the time axis is identified. This results in a time value. The measurement curve also provides measured values or values can be derived from the measurement curve. Pairs are formed from these values and the assigned time values. Each measurement curve can thus be characterized by several pairs of values.The measured value pairs are compared with a previously created set of comparison pairs, and the comparison pair (comparison tuples) that most closely matches the measured value pair (value tuples) is determined. Each of the value tuples is assigned a property of the layer. Furthermore, each of the value tuples can also be assigned a variety of layer properties. This can be done through selection. The comparison tuple can thus be assigned the property that was previously assigned to the comparison tuple. The comparison of the value tuples with the comparison tuple can be performed using a numerical operation; for example, distances can be defined. From the number of comparison tuples, the tuple that is closest to the value tuple is then determined. The property of the separated layer is then the property that was assigned to the associated comparison tuple when the comparison curves were created.

[0014] If the method according to the invention is intended to determine only one property of the thin layer, a one-dimensional array of comparison tuples can be used, each of the comparison tuples corresponding to a specific value of the layer thickness. If the method according to the invention is intended to determine two layer properties simultaneously, for example the layer thickness and the layer composition, a two-dimensional array of comparison tuples can be used, each comparison tuple corresponding to a pair of properties, for example a layer thickness and a layer composition. It can also be provided that a model, for example in the form of a neural network, is used to determine the comparison tuples. A backward optimization can be carried out, with which the comparison tuple that comes closest to the value tuple is found step by step.The layer properties assigned to this comparison tuple are then evaluated.

[0015] According to a second aspect of the invention, the thin layer is an AlGaN layer or, more generally, a III-V layer, and the property of the aluminum component or the Al / Ga ratio of this layer. This III-V layer is preferably deposited on a III-V buffer layer and, in particular, deposited onto a GaN layer. A buffer layer sequence formed in this way is preferably deposited onto a Si(III) layer. A III-V subsequent layer can be deposited onto the thin layer, which is in particular a barrier layer of an active layer sequence of a component. The subsequent layer can be a GaN layer. A further active layer of the active layer sequence can also be arranged between the barrier layer and the buffer layer sequence. According to the invention, the measurement curve described above is recorded during the deposition of the layer or layers deposited onto the thin active layer.

[0016] The layer thickness of the thin layer is less than 50 nm and preferably less than 20 nm and in particular lies in a range between 8 and 14 nm.

[0017] The invention further relates to a method for determining one or more optical properties, wherein the measurement curve is examined for characteristic shapes of its course, and characteristic value tuples are formed from the temporal positions and the values of the measurement curve at the temporal positions of the characteristic shapes. However, a set of curves (or a plurality of points) can also be used to determine a model (such as a neural network).

[0018] The invention further relates to a device for depositing layers, in particular III-V layers, comprising a susceptor arranged in a reactor housing, which can be heated by a heating device. One or more storage locations, for example in the form of rotatable substrate holders, are provided on the susceptor for storing at least one substrate each. A gas inlet device can be used to introduce gas into a process chamber of theA process gas is fed into the reactor housing. The process gas can contain organometallic compounds of elements from main group III and hydrides of elements from main group V. The process gas is fed into the process chamber together with a carrier gas, for example hydrogen. Optical measuring devices, for example pyrometers, are provided with which the optical properties of the layers deposited on the substrate can be measured. A programmable computing device is also provided. This forms a control device for controlling a heating device for heating the susceptor. The control device can also be used to feed the process gases into the process chamber using valves and a mass flow controller. Furthermore, the computing device is designed to interact with the optical measuring devices.The computing device can in particular evaluate the optical measured values in the manner described above.

[0019] According to a further embodiment, a band edge measurement can also be performed to determine substrate properties. The band edge measurement is performed analogously to the previously described reflectance measurement using optical measuring devices. Here, the band gap is measured. As described above, an optical quantity and a measurement curve are determined and evaluated analogously.

[0020] According to a further aspect of the invention, the determined properties or at least one of the determined properties can be compared with a target value. A target value deviation is then calculated. The deposition process is carried out using process parameters specified by a recipe. These process parameters can include mass flows of reactive gases, temperatures, in particular susceptor temperatures, or the like. The process parameters influence the at least one property. business. The setpoint deviation can be used to vary the process parameters of a subsequent deposition process in such a way that the setpoint deviation is reduced. The determined property can thus be a parameter in a control loop to keep one or more properties, as described above, within a permissible range. Short description of the drawings

[0021] An embodiment of the invention is explained below with reference to the accompanying drawings. They show: Fig. 1 schematically shows a CVD reactor, Fig. 2 is a plan view of a susceptor 4 according to the section line II-II in Figure 1, Fig. 3 schematically shows an embodiment of a multilayer structure 21 deposited on a silicon substrate 4, Fig. 4 Measurement curves of the emissivity E and reflectivity R measured during the deposition of a multilayer structure 21 according to Figure 3, Fig. 5 schematically shows comparison curves a, b, c, d of temporal courses of the reflectivity R, created using model calculations, for comparison with a measurement curve during the deposition of subsequent layers 28 onto an active layer 27. Description of the embodiments

[0022] Figure 1 shows a cross-section through a CVD reactor according to the invention. The CVD reactor has a gas supply device (not shown) in which gas mixtures are provided by means of valves and mass flow meters. The gas mixtures can contain TMA1, TMGa, and NH3 as reactive gases. The reactive gases are conveyed with a carrier gas, which is preferably hydrogen, to a gas feed line 3 in a reactor housing 1. The reactive gases can be used to deposit semiconductor layers containing Al, Ga, and N on a substrate, in particular a Si substrate. In addition, the gas supply device can also provide doping gases in order to dope individual semiconductor layers, in particular active semiconductor layers.

[0023] The reactor housing 1 contains a gas inlet element 2 for feeding in the reactive gases. The gas inlet element 2 is arranged in the center of a process chamber 8. The process chamber 8 is bounded at the top by a process chamber ceiling 17. The process chamber 8 is bounded at the bottom by a susceptor 4 having pockets, in each of which a substrate holder 6 is arranged, which supports a substrate 7. Between the underside of the substrate holder 6 and the bottom of the pocket is a gas cushion 6', which is generated by a purge gas flowing into the pockets through a gas outlet opening (not shown) in the bottom of the pocket in order to keep the substrate holder 6 in a suspended position and rotate it about an axis. The susceptor 4 is heated by a heating device 5. The susceptor 4 can be rotated about a rotation axis 16 using a shaft 9.Several substrate holders 6 lie on the circular susceptor 7 and are arranged in a circle around the gas inlet element 2.

[0024] In another embodiment, the gas inlet element can be designed as a showerhead and have a gas outlet surface that extends over the entire surface of the process chamber ceiling 17. A plurality of gas outlet openings open into the gas outlet surface for the process gas to exit into the process chamber. Several substrates can be arranged in a regular arrangement on the susceptor 4. However, it is also contemplated that only a single substrate with a diameter of, for example, 300 mm or more can be placed on the susceptor.

[0025] An opening 18 is located in the process chamber ceiling 17, through which an optical path 19 passes, via which measured values can be measured at a measuring point 13 on the surface of the substrate 7. The optical path 19 extends via a beam splitter 10 to a reflectance value measuring device 11 and an emissivity value measuring device 12. With these two measuring devices, measured values of the reflectivity R and the emissivity E can be measured on the substrate surface. The measured values are fed to a computing device 15. The computing device is configured to control the aforementioned valves and mass flow controllers in order to deposit, for example, the multilayer structure 21 shown in Figure 3 on the substrates 7.Furthermore, the computing device 15 is configured to determine the method, explained further below, for determining one or more properties on a thin layer 27 deposited on the substrate 7 or on a layer structure 23, 24, 25, 26 deposited on the substrate 7. Furthermore, the computing device 15 is preferably configured to abort a coating process or to issue a message if the property does not lie within predetermined limits.

[0026] A light source (not shown), for example a laser, is provided, the light beam of which runs along the optical path in order to To illuminate measuring point 13 with reference light so that a reflectance measurement can be carried out.

[0027] Figure 3 shows a multilayer structure 21 that is deposited one after the other in a coating process in several consecutive coating steps. The multilayer structure 21 comprises a plurality of layers. First, a nucleation layer 23 made of AlN or another III-V material is deposited on a silicon substrate 7. A further layer 25, an AlGaN layer in the exemplary embodiment, is then deposited on the nucleation layer 23. This layer can also consist of several sublayers. The aluminum content of these sublayers can then decrease gradually. While the nucleation layer 23 is gallium-free, a layer 25 deposited on the AlGaN layer 24 is aluminum-free and preferably consists of GaN. However, the layer 25 can also consist of several sublayers and together form a buffer layer 25.

[0028] The deposition of the layer sequence 23, 24, 25 preferably takes place in a CVD reactor as previously described. In the same CVD reactor, an active layer sequence 26, 27, 28 can then be deposited, consisting of a first GaN layer 26, which has a layer thickness of 200 nm, for example. A barrier layer 27, consisting of AlGaN, is then deposited onto this layer 26. The thickness of the barrier layer 27 is less than 50 nm and in particular less than 20 nm. In the exemplary embodiment, the layer thickness is in the range between 8 and 14 nm. The layer thickness of the barrier layer 27 is so small that it cannot be measured optically in situ. The wavelength at which the measuring devices 11, 12 work is greater than 200 nm. The layer composition, i.e. the aluminum content x in the AlxGai-xN, can also be determined during the deposition of the layer and also immediately after layer 27 is deposited.

[0029] However, the layer properties of the barrier layer 27 are of great technological importance for the electronic component to be manufactured using the method, for example, a HEMT. This particularly applies to the layer composition.

[0030] At least one further active layer 28, in particular made of GaN, is then deposited onto the barrier layer 27. The layer thickness of the active layer 28 can be between 80 nm and 90 nm. While layer 26 can be n-doped, layer 28 can be p-doped. Further layers can be deposited onto layer 28. All of this takes place within the same process chamber 8 of the CVD reactor, directly one after the other, without removing the substrate 7 from the process chamber 8.

[0031] Figure 4 shows quasi-periodically fluctuating measured values of reflectance R and emissivity E. At least the barrier layer 27 and the subsequent layers 28 deposited immediately thereon are transparent to the light emitted by the light source of the optical measuring device, which impinges on the substrate 7 at the measuring point 13. The light can be reflected either by the boundary layer of the barrier layer 27 with the underlying layer 26, or by another boundary layer below the barrier layer 27. In addition, reflections occur at the boundary layer where the barrier layer 27 adjoins the subsequent layer 28, and reflections occur at the current surface of the subsequent layer 28, which changes during the coating process. As a result of these reflections and the associated interference, the measurement curves have a wave-like shape. They alternate between a minimum and a maximum. The two measurement curves of the re- The reflectivity R on the one hand and the emissivity E on the other hand have different signs or are 180° out of phase.

[0032] With the method according to the invention, a measurement curve, preferably of the reflectivity R, is recorded during the deposition of the subsequent layer 28. This is achieved by determining measured values at the measuring point 13 at short time intervals. A measurement curve is formed from the measured values, for example, by plotting the measured values over time. Such a measurement curve has a profile similar to the profile of the comparison curves a, b, c, d shown in Figure 5.

[0033] The measurement curve is examined by the computing device 15 for characteristic shapes of its course. For example, the value of an offset 35 is determined at a position 30 at which the barrier layer 27 is deposited and the subsequent layer 28 is deposited. During the deposition of the subsequent layer 28, for example, the position 31 of a first minimum and the value 36 of the first minimum are determined. Furthermore, the position 32 of a first maximum and the value 37 of a first maximum can be determined. Furthermore, the position 33 of a second minimum and the value 38 of the second minimum can be determined. Furthermore, it is possible to calculate a value 39 of a first amplitude by calculating a difference between the value 37 of the first maximum and the value 36 of the first minimum. The value 40 of a second amplitude can also be calculated by forming a difference between the value 37 of the first maximum and the value 38 of the second minimum. Each of the previously described values 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40 can be an element of a value tuple. The value tuple characterizes the measurement curve and provides a means of assigning a property of the barrier layer 27 to the measurement curve. The property is preferably the aluminum content x of the barrier layer 27. However, the property can also be the layer thickness of the barrier layer 27.

[0034] A value tuple can contain all of the previously described values 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40. However, it is also intended that a value tuple contains only some of the values 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40. Thus, a value tuple can contain the values of the offset 30, 35 and / or the values of the first minimum 31, 36 and / or the values of the first maximum 32, 37 and / or the values of the second minimum 33, 38. Likewise, the value tuple can also contain the value 39 of the first amplitude or the value 40 of the second amplitude.

[0035] To evaluate the measurement curve, comparison curves a, b, c, d are calculated in preliminary tests or using a model calculation, which are shown in Figure 5. Curve a characterizes a reflectance curve that is determined during the deposition of the subsequent layer 28 when the aluminum content is 18%. Curve b characterizes a reflectance curve with an aluminum content of 20%. Curve c characterizes a reflectance curve with an aluminum content of 22% and curve d a reflectance curve with an aluminum content of 24%.

[0036] From the previously described comparison curves a, b, c, d, the values 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40 are determined, as previously described using the measurement curve. A comparison tuple can be calculated from each of these comparison curves a, b, c, d.

[0037] The comparison tuples and the value tuples have the same structure, so they are comparable. A comparison between the value tuple and the comparison tuples can be made, for example, by forming a sum of the squares of the differences between the associated elements of the two tuples. This creates a number characterizing a distance. By comparing the value tuple with all comparison tuples, the comparison tuple with the smallest distance to the value tuple can be determined. The property, for example, aluminum content x, that is assigned to this value tuple is then considered the aluminum content of the barrier layer 27 deposited in the current coating process.

[0038] The layer thicknesses of the barrier layer 27 can also be determined in an analogous manner.

[0039] The property determined in this way can then be used to make a decision as to whether the layer structure produced up to that point should be further processed or whether it must be discarded.

[0040] The above statements serve to explain the inventions covered by the application as a whole, which also independently develop the state of the art at least by the following combinations of features, whereby two, several or all of these combinations of features can also be combined, namely:

[0041] A method which is characterized in that the measurement curve is examined for characteristic forms of its course and characteristic value tuples are formed from the temporal positions 31, 32, 33 of the characteristic forms and / or the values 34, 35, 36 of the measurement curve at the temporal positions 31, 32, 33 of the characteristic forms and the characteristic value tuples are compared with comparison tuples.

[0042] A method which is characterized in that the comparison tuples are determined by a model calculation or in preliminary tests, for which purpose comparison curves a, b, c, d are created, wherein it is particularly provided that a numerical model or a neural model is used in the model calculation.

[0043] A method which is characterized in that a value of an offset 35 is measured at the temporal position 30 of the start of the deposition of the one or more subsequent layers 28, that at least the temporal position 31, 32, 33 of an extremum of the measurement curve is determined and that the characteristic value tuple contains the position of the offset, the position of the extremum 31, 32, 33 and the value 35, 36, 37, 38 of the measurement curve at the positions of the offset and the extremum 31, 32, 33.

[0044] A method characterized in that the characteristic value tuple contains the temporal position 31 of a first minimum and the position 32 of a first maximum and the values 36, 37 of the measurement curve a, b, c, d at these positions 31, 32 and / or a first amplitude value formed from a difference between the values 36, 37.

[0045] A method which is characterized in that the characteristic value tuple contains the temporal position 33 of a further extremum 33 and the value 38 of the measurement curve a, b, c, d at this position 33 or a second amplitude value obtained by subtraction.

[0046] A method characterized in that the one or more properties is a layer composition x of the thin layer 27.

[0047] A method characterized in that the thin layer 27 consists of elements of the III and V main groups and in particular is an AlxGai-xN layer.

[0048] A method characterized in that one or more properties of the thin layer 27 are determined according to the method according to claims 1 to 7.

[0049] A method which is characterized in that the deposition of the one or more subsequent layers 28 is interrupted or a message is issued if the determined properties x lie outside a predetermined tolerance range.

[0050] A device for depositing layers, in particular III-V layers, with a susceptor 4 arranged in a reactor housing 1, which can be heated by a heating device 5, wherein one or more storage locations 6 for storing at least one substrate 7 are provided on the susceptor 5, with a gas inlet element 2 arranged in the reactor housing 1 for feeding process gases into a process chamber 8, with an optical measuring device 10, 11, 12 for measuring an optical quantity E, R of the layers 23, 24, 25, 26, 27, 28 deposited on the substrate 4 and with a programmable computing device 15 for controlling the heating device 5 and the process gas and for processing measured values measured with the optical measuring device 10, 11, 12, characterized in that the computing device 15 is set up to carry out a method according to one of claims 1 to 9.

[0051] All disclosed features are essential to the invention (individually, but also in combination with each other). The disclosure of the application The disclosure content of the associated / attached priority documents (copy of the prior application) is hereby incorporated in its entirety, also for the purpose of incorporating features of these documents into the claims of the present application. The subclaims characterize, with their features, independent inventive developments of the prior art, even without the features of a referenced claim, in particular for filing divisional applications based on these claims. The invention specified in each claim may additionally have one or more of the features provided in the above description, in particular with reference numbers, and / or specified in the list of reference numbers.The invention also relates to designs in which individual features mentioned in the above description are not implemented, in particular insofar as they are clearly unnecessary for the respective intended use or can be replaced by other means having the same technical effect. List of reference symbols 1 reactor casing 2 gas inlet organ 30 position offset 3 Gas supply line 31 Position first minimum 4 Susceptor 32 Position first maximum 5 Heating device 33 Position second minimum 6 substrate holder 35 value offset 6' Gas cushion 36 value first minimum 7 Substrate 37 Value first maximum 8 Process chamber 38 value second minimum 9 shaft 39 first amplitude 10 beam splitter 40 second amplitude 11 Reflectance value- measuring device 12 Ernis sivity value- Measuring device a comparison curve 18% 13 Measuring point b comparison curve 20% 14 Rotary drive device c comparison curve 22% 15 Calculation device d comparison curve 24% 16 Rotation axis t time x property 21 Multilayer structure 23 Nucleation layer 24 Transition layer 25 Buffer layer E rnis sivity value 26 active layer R reflectance value 27 Barrier layer 28 active layer

Claims

Claims 1. A method for determining one or more properties of a thin layer (27) deposited on a substrate (7) or on a layer structure (23, 24, 25, 26) deposited on a substrate (7), wherein an optical quantity, in particular the reflection (R) of light with a light wavelength, is measured with an optical measuring device (13), wherein the thickness of the thin layer (27) is substantially less than the light wavelength, wherein one or more subsequent layers (28) are deposited onto the thin layer (27), and during the deposition of the subsequent layers (28), a series of measured values of the optical quantity, in particular the reflection (R), are recorded in temporal succession, wherein a measurement curve is generated by plotting the measured values along a time axis, and wherein the one or more properties (x) of the thin layer (27) are determined by evaluating a temporal progression of the measurement curve, characterized in thatthat the measurement curve is examined for predetermined characteristic forms of its course and characteristic value tuples are formed from the temporal positions (31, 32, 33) of the characteristic forms on the time axis and from the values (34, 35, 36) of the measurement curve at the temporal positions (31, 32, 33) of the characteristic forms and the characteristic value tuples are compared with a plurality of previously obtained comparison tuples, each of which is assigned a property (x), where the property whose associated comparison tuple is closest to the value tuple is selected.

2. Method according to claim 1, characterized in that the comparison tuples are determined by a model calculation or in preliminary tests, for which purpose comparison curves (a, b, c, d) are created, wherein it is particularly provided that a numerical model or a neural model is used in the model calculation.

3. Method according to one of the preceding claims, characterized in that a value of an offset (35) is measured at the temporal position (30) of the start of the deposition of the one or more subsequent layers (28), that at least the temporal position (31, 32, 33) of an extremum of the measurement curve is determined and that the characteristic value tuple contains the position of the offset, the position of the extremum (31, 32, 33) and the value (35, 36, 37, 38) of the measurement curve at the positions of the offset and the extremum (31, 32, 33).

4. Method according to one of the preceding claims, characterized in that the characteristic value tuple contains the temporal position (31) of a first minimum and the position (32) of a first maximum and the values (36, 37) of the measurement curve (a, b, c, d) at these positions (31, 32) and / or a first amplitude value formed from a difference between the values (36, 37).

5. Method according to claim 3 or 4, characterized in that the characteristic value tuple contains the temporal position (33) of a further extremum (33) and the value (38) of the measurement curve (a, b, c, d) at this po- sition (33) or a second amplitude value obtained by subtraction.

6. A method for determining one or more properties of a thin layer (27) deposited on a substrate (7) or on a layer structure (23, 24, 25, 26) deposited on a substrate (7), wherein an optical quantity, in particular the reflection (R) of light with a light wavelength, is measured with an optical measuring device (13), wherein the thickness of the thin layer (27) is substantially less than the light wavelength, wherein one or more subsequent layers (28) are deposited onto the thin layer (27), and during the deposition of the subsequent layers (28), a series of measured values of the optical quantity, in particular the reflection (R), are recorded in temporal succession, and wherein the one or more properties (x) of the thin layer (27) are determined by evaluating a temporal profile of a measurement curve formed from the measured values, in particular according to one of the preceding claims, characterized in thatthat the one or more properties is a layer composition (x) of the thin layer (27)., 7. Method according to the preamble of claim 1 or one of the preceding claims, characterized in that the thin layer (27) consists of elements of main groups III and V and is in particular an AlxGai-xN layer.

8. A method for depositing a plurality of layers (23, 24, 25, 26, 27, 28) on a substrate (4), wherein a layer sequence (23, 24, 25, 26) is first deposited on the substrate (4), a thin layer (27) is deposited onto the layer sequence, the layer thickness of which is at most 50 nm, and one or more subsequent layers (28) are deposited onto the thin layer (27), characterized in that one or more properties of the thin layer (27) are determined by the method according to claims 1 to 7.

9. Method according to claim 8, characterized in that the deposition of the one or more subsequent layers (28) is interrupted or a message is issued if the determined properties (x) lie outside a predetermined tolerance range.

10. Method according to one of the preceding claims, characterized in that a target value deviation of one of the determined properties from a target value is determined and in a subsequent deposition process process parameters are varied using the target value deviation.

11. Device for depositing layers, in particular III-V layers, with a susceptor (4) arranged in a reactor housing (1) which can be heated by a heating device (5), wherein one or more storage locations (6) for storing at least one substrate (7) are provided on the susceptor (5), with a gas inlet element (2) arranged in the reactor housing (1) for feeding process gases into a process chamber (8), with an optical measuring device (10, 11, 12) for measuring an optical variable (E, R) of the layers (23, 24, 25, 26, 27, 28) deposited on the substrate (4) and with a programmable computing device (15) for controlling the heating device (5) and the process gas and for processing measured values measured with the optical measuring device (10, 11, 12), characterized in that the computing device (15) is set up to carry out a method according to one of claims 1 to 9.

12. Device or method characterized by one or more of the characterizing features of one of the preceding claims.

Images