WO2018085268A1

WO2018085268A1 - Layer sequence with alternating acoustic impedance, acoustic component having the layer sequence, and method for production

Info

Publication number: WO2018085268A1
Application number: PCT/US2017/059281
Authority: WO
Inventors: Maximilian SCHIEK; Fabian Florek
Original assignee: Snaptrack, Inc.
Priority date: 2016-11-07
Filing date: 2017-10-31
Publication date: 2018-05-11
Also published as: DE102016121220B3

Abstract

A layer sequence is proposed that comprises first and second layers that are deposited atop one another in alternation, wherein the first layers (SC₁, SC₂) comprise SiC₁ _-xH_x and the second layers (SOC₁, SOC₂) comprise SiOC(H).

Description

LAYER SEQUENCE WITH ALTERNATING ACOUSTIC IMPEDANCE, ACOUSTIC COMPONENT HAVING THE LAYER SEQUENCE, AND METHOD FOR

PRODUCTION

CROSS-REFERENCE TO RELATED APPLICATION This application claims priority to German Patent Application No. 102016121220.2, filed November 7, 2016, which is expressly incorporated herein by reference in its entirety.

Description

Bulk wave resonators are, for example, used as basic elements in microacoustic filters that respectively comprise an interconnection of multiple such resonators. Such filters are widely used as frequency filters for wireless communication and information transfer. Bulk wave resonators typically have a sandwich structure in which a layer of a piezoelectric material is deposited onto a substrate, embedded between two electrode layers.

In order to achieve a sufficient electroacoustic coupling and a high quality factor for the resonators, it must be ensured that the acoustic energy remains within the sandwich and cannot escape via the substrate or the environment.

In order to prevent an escape of acoustic energy from the resonators in the direction of the substrate, either a cavity or a Bragg mirror may be arranged between resonator and substrate. Resonators having a cavity are referred to as being a membrane-type FBAR (= film bulk acoustic resonator) , in which the high impedance jump from the electrode surface to the air within the cavity ensures a complete reflection of acoustic waves. Alternatively, a Bragg mirror made up of alternating layers of high and low acoustic impedance may be realized between substrate and sandwich. If the layers are applied in a controlled thickness that corresponds to a quarter-wavelength of the acoustic wave, the acoustic energy may also be completely reflected back into the sandwich with such a mirror. Such components are also designated as SMR resonators [sic] (= Solidly Mounted Resonator) . Tungsten is typically used as a material with high acoustic impedance, whereas S1O₂ is typically used as a material with low

acoustic impedance.

While membrane-type FBARs exhibit a very high quality, the mirror reflectivity of a Bragg mirror is determined by the impedance jump between the alternating layers, and

additionally is also dependent on the number of mirror layer pairs or by the number of impedance jumps between directly successive mirror layers.

Tungsten has, for example, a longitudinal acoustic impedance of approximately 100 MRayls, in contrast to which S1O₂ possesses an impedance of only approximately 13 MRayls. While it is barely possible to replace tungsten with a material having higher acoustic impedance, it appears to be thoroughly possible to find additional materials having lower acoustic impedance. For example, it has been proposed to use SiOC (silicon oxycarbide) as a low-impedance material in Bragg mirrors for bulk acoustic wave resonators. Alternating layers of SiOC and tungsten were thereby used for the Bragg mirror. While tungsten may be deposited in a sputtering system, for example, a chemical vapor deposition (CVD) is required for the deposition of S1O₂ or SiOC. To produce a Bragg mirror, this requires a repeated changing of the system,

corresponding to the number of different mirror layers that are to be deposited. If a filter made up of SMR resonators is produced using tungsten as a mirror layer having relatively high impedance, individual resonators are typically generated next to one another on a common substrate, structured, and thereby are integrated in connection with one another to form a filter. In this instance, tungsten must always also be structured in order to prevent an electromagnetic coupling between adjacent resonators via an otherwise common, underlying mirror layer comprising tungsten. Such a coupling is to be avoided in order to not degrade the selectivity of the filter.

It is the object of the present invention to specify a layer sequence of layers having alternating high and low acoustic impedance from which a Bragg mirror can be realized. An additional object is to specify an improved microacoustic component having an improved Bragg mirror, and a method which facilitates the manufacturing of the Bragg mirror.

At least a partial object is achieved by a layer sequence having the features of claim 1. Advantageous embodiments of the invention and a method for producing the layer sequence are to be learned from additional claims.

A layer sequence is specified that comprises first and second layers that are deposited atop one another in alternation. The first layers thereby comprise silicon carbide SiCi-_xH_x. The second layers comprise silicon oxycarbide SiOC (H) . In such a layer sequence, the first and second layers have starkly different acoustic impedances. While the first layers comprising silicon carbide have an acoustic impedance of approximately 90 MRayls each, the second layers comprising silicon oxycarbide have the cited low acoustic impedance of only approximately 3 to 4 MRayls. In this way, an impedance jump can be realized that is approximately as large as the impedance jump of conventional Bragg mirror layers that had up to this point in time been realized from tungsten and Si0₂.

The layer sequence furthermore has the advantage that both materials can be deposited by means of CVD. Moreover, the materials of both layers are chemically related and may be produced from the same or similar precursors in the CVD process. It is therefore possible to produce the two layers in the same CVD reactor, and one after another in a direct chronological sequence. This significantly reduces the process cost since no apparatus change is required for deposition, and in particular no repeated transport of the substrate to be coated back and forth between the two

reactors. The method can therefore also be implemented in a shorter amount of time, and with better monitoring in its entire scope.

The layer sequence according to the invention is therefore very well suited to realize from this a Bragg mirror for microacoustic applications.

In one embodiment, first and second layers are adjusted to be semiconducting or dielectric. This has the advantage that, in particular for the production of a Bragg mirror, no

additional insulation measures are necessary, and in

particular no structuring of electrically conductive layers to avoid electromagnetic coupling between acoustic components that are arranged next to one another. A semiconducting or dielectric adjustment of the first layers is achieved via a doping during the deposition. P-dopings or n-dopings are possible, p-dopings may be produced via

admixture of Be, B, Al or Ga, whereas an n-doping may be generated via addition of N or P. The electrical properties of the first and second layers can thus be monitored in a simple manner, and in particular can be adjusted to be electrically insulating.

In light of the noted advantages of the layer sequence according to the invention, also within the scope of the invention is a microacoustic component in which a layer sequence made up of alternating first and second layers is deposited onto a substrate having a flat main surface. A stack made up of a first electrode layer, a piezoelectric layer and a second electrode layer is applied via the layer sequence. All layers together form an SMR-type BAW resonator, wherein the layer sequence serves as an acoustic Bragg mirror .

In an advantageous embodiment of the invention, all first layers of the layer sequence are adjusted to be

semiconductive or dielectric. Moreover, the layer sequence may advantageously be deposited over a large area onto the substrate and be laterally unstructured. In this context, "over a large area" means that the footprint of the layer sequence is larger than is necessary for a single component. Rather, an unstructured layer sequence may serve as a common Bragg mirror for a plurality of resonators deposited onto the same substrate. The layer sequence may therefore serve as a Bragg mirror for a series of microacoustic components that are in particular interconnected to form a filter. Via the use of exclusively dielectric or semiconducting layers in the layer sequence, it is possible to leave these laterally unstructured below the cited interconnection of microacoustic components without needing to fear an

electromagnetic coupling of adjacent microacoustic

components. The use of a large-area Bragg mirror below an interconnection of multiple individual microacoustic

resonators furthermore has the advantage that the entire structuring cost for the Bragg mirror is canceled, such that the component may be produced cost-effectively.

An unstructured mirror layer that is arranged over a large area below a plurality of microacoustic resonators

furthermore has the advantage that a large, flat surface is provided for generation of the resonators, on which surface the resonators can grow simply and at higher quality since the layer sequence has no structures and only flat layers.

Even if individual layers of an interconnection of multiple microacoustic resonators requires a structuring of the individual layers, overall the number of structuring steps is starkly reduced.

According to one exemplary embodiment, the layer sequence has 2.5 to four layer pairs made up of a respective first and second layer. This thus means five to eight individual layers. The uppermost layer of the layer sequence is

advantageously a second layer that has a low acoustic

impedance, and thus enables an additional high impedance jump to the first electrode layer, which preferably comprises tungsten. Aluminum nitride A1N may then be used as a

piezoelectric layer. A Bragg mirror has a frequency-dependent reflection behavior. A specific frequency is maximally reflected if all layers of the layer sequence respectively have a layer thickness that corresponds to quarter-wavelength of the cited frequency. The wavelength of that acoustic wave that propagates within the material of the layer is thereby taken as a measurement.

Converted into length units, a quarter-wavelength is of different thickness for different materials. A layer

thickness ratio of approximately 1:2 thereby results for the cited first and second layers, which moreover is, however, dependent on the crystalline modification of the two

respective layer materials.

It may also be advantageous for a portion of first and second layers to deviate somewhat from a layer thickness that corresponds to exactly a quarter-wavelength, in order to achieve a broader reflection behavior. This may in particular be advantageous when the resonance frequencies of the

respective resonators differ in an interconnection of

different resonators, such that the Bragg mirror shows a sufficiently good reflection behavior for different resonance frequencies of the microacoustic resonators interconnected with one another.

To produce a layer sequence according to the invention, all first and second layers may be deposited in succession or, respectively, in alternation on a substrate in one and the same CVD reactor. For the CVD process, gaseous precursors are used that contain at least silicon and carbon for production of the first layers. Compounds may be used that contain both, or different compounds may be used that contain either silicon or carbon. The deposition then arrives on the

substrate at a sufficiently high surface temperature while a low pressure is advantageously set in the CVD reactor. With precursors containing Si and C, first layers made of silicon carbide may be deposited.

To produce the second layers, oxygen or an oxygen-containing compound is additionally introduced into the CVD reactor while the silicon- and carbon-containing precursors continue to be supplied to the reactor chamber. While the oxygen supply is opened upon transition from the deposition of a first layer to the deposition of a second layer, it is closed again upon transition from a second layer to a first layer, or the introduction of oxygen-containing precursors is stopped. Overall, five to eight first and second layers may be deposited atop one another in alternation for a Bragg mirror .

To generate a frequency-selective Bragg mirror, a layer thickness monitoring is required during the deposition of every single one of first and second layers. A good layer thickness monitoring is achieved with a low deposition rate. The deposition speed may thereby be increased, and thus adjusted, via higher gas pressure, higher temperature and stronger gas flow. It may be advantageous to reduce the deposition speed upon transitioning from the deposition of a first layer to the deposition of a second layer.

Depending on the growth conditions, the modification of the growing layers may also be adjusted. For example, silicon carbide can be deposited well in polycrystalline form. Under better-controlled conditions, epitactic silicon carbide layers can even be generated. However, a polycrystalline silicon carbide layer is entirely sufficient for use as a first layer in a Bragg mirror. First layers of silicon carbide may be deposited with

different precursors in a typical hydrogen-containing

atmosphere. For example, silane S1H₄ mixed with methane CH₄ is suitable. Also suitable are akylsilanes and

alkylchlorsilanes , for example methytrichlorsilane or

trimethylsilane, which are present in gaseous form under process conditions. Higher alkyl silanes or silanes having more or fewer alkyl groups may also be used. Hydrocarbons may be used in conjunction with silanes or disilane, for example

In addition to oxygen, in particular nitrous oxide N₂0, nitric oxide NO, ozone O3 or also just ambient air may serve as an additional precursor or an oxygen-containing component that is used in the CVD reactor to generate the second layer.

In the CVD reactor, upon transition from the first layer to the second layer, it is also possible not only to supply an additional component but also to change the silicon- or carbon-containing source. For example, it is also possible to use an alkyl siloxane as a precursor simultaneously

containing Si and 0, possibly in combination with additional precursor gases containing C and 0.

The invention will be explained in greater detail below with reference to exemplary embodiments and the associated

figures. The Figures are only schematically executed and are not true to scale, such that individual parts may be depicted enlarged, and no absolute or relative size information is to be learned from the Figures.

Figure 1 shows a layer sequence applied onto a substrate, Figure 2 shows a microacoustic component,

Figure 3 shows a microacoustic component having two

interconnected resonators,

Figure 4 shows a microacoustic component having two

resonators that are interconnected in a different manner,

Figure 5 shows a workflow diagram of the production method of the layer sequence and of the microacoustic component,

Figure 6 shows an interconnection, known per se, of

microacoustic resonators to form a filter.

Figure 1 shows a schematic cross section through a layer sequence SF according to the invention that, depending on the production, is applied onto the surface of a substrate SU. In the presented exemplary embodiment, the layer sequence comprises two first layers SCI, SC2 that, alternating with two second layers SOC1, SOC2, form the layer sequence SF. Depending on the desired application, the layer thicknesses of all first layers SC and the layer thicknesses of all second layers SOC are respectively identical. In the layer sequence SF, it is also possible to use different layer thicknesses within the group of first layers and/or within the group of second layers.

The substrate SU may be an arbitrary carrier that is a match for the thermal load due to the layer deposition process. The substrate is preferably electrically insulating and has a low coefficient of thermal expansion, or a coefficient of thermal expansion that is adapted to the layer sequence. Ceramics or crystalline substrates are well suited, and moreover

substrates made from glass or semiconductor material.

Figure 2 shows an application of the layer sequence for a microacoustic component according to the invention. The microacoustic component in turn has, on a substrate SU, a layer sequence SF as depicted in Figure 1 that, however, is designed as a Bragg mirror BS . In the presented exemplary embodiment, the Bragg mirror BS has five layers, wherein first and second layers are arranged alternating one another another. A second silicon oxycarbide layer SOC that

represents a layer of low acoustic impedance is respectively provided as lowermost and uppermost layer. Arranged between two respective silicon oxycarbide-containing second layers SOC is a respective first layer SCI, SC2 comprising silicon carbide .

Applied directly over the Bragg mirror BS is a microacoustic resonator AR that comprises a first electrode layer ESI made from a heavy metal, preferably tungsten; a piezoelectric layer PS, for example an aluminum nitride layer; and a second electrode layer ES2 made from a metal with good electrical conductivity, for example aluminum. The acoustic resonator AR has a center frequency corresponding to its resonance

frequency, which is essentially dependent on the layer thickness of the piezoelectric layer PS. Depending on this resonance frequency, the first and second layers of the Bragg mirror BS are designed with a layer thickness of respectively approximately a ¾ wavelength at the cited resonance

frequency .

If an alternating electrical signal is now applied between the two electrode layers ESI, ES2, for example at an RF frequency, the acoustic resonator AR may begin to resonate at its layer thickness-dependent eigenfrequency, wherein a bulk oscillation is excited. As a dominant mode, a longitudinal wave is typically generated that travels and oscillates parallel to the normal, thus vertically relative to the layer planes of the microacoustic resonator AR. Acoustic waves that might propagate in the direction of the substrate are

reflected back into the resonator by the Bragg mirror BS .

Figure 3 shows an additional microacoustic component

according to the invention in which two stacks that

respectively form an acoustic resonator AR1 and AR2 are arranged next to one another on a Bragg mirror and are connected with one another. Given a component having multiple acoustic resonators, according to the invention a continuous Bragg mirror BS made up of electrically insulating first and second layers may be used. In the Figure, the depiction of the first and second layers is omitted.

For two acoustic resonators AR1, AR2 that are connected with one another, a continuous first electrode layer ESI may be applied directly over the Bragg mirror BS . Applied over this is a piezoelectric layer PS that may remain unstructured, or that is structured in the individual resonators.

Applied over the piezoelectric layer PS is the second

electrode layer ES2, and said piezoelectric layer PS is structured into two individual electrode sections for the first and second acoustic resonator.

The first acoustic resonator AR1 then comprises the overlap region from the second electrode layer ES2 of the first resonator with the piezoelectric layer PS and the first electrode layer ESI. Accordingly, the second acoustic

resonator AR2 comprises the volume region in which the area of the second electrode layer ES2 ^λ that is associated with the second resonator AR2 overlaps with the piezoelectric layer PS and the first electrode layer ESI. The two

resonators are electrically connected in series across the continuous first electrode layer ESI. An electrical

connection may then take place via the two subsurfaces of the second electrode layer ES2, ES2'.

Figure 4 shows an alternative interconnection of two acoustic resonators AR1 and AR2 in which the first electrode layer ESI is already structured in the corresponding electrode surfaces ESI and ESI', which are respectively associated with one of the two resonators. The piezoelectric layer PS applied over this may be applied continuously over both subsurfaces of the first electrode layer ESI. The second electrode layer ES2 may likewise be applied continuously over the piezoelectric PS. In this way, the two acoustic resonators AR1, AR2 are also electrically connected in series with one another.

The contacting of this series connection then takes place via the subsurfaces ESI, ESI' of the first electrode layer ESI [sic] . However, it is also possible to additionally contact the second electrode layer ES2, wherein then both resonators may alternatively also be electrically connected in parallel.

For contacting of the underlying first electrode layer ESI, or the corresponding partial surfaces of the first electrode layer ESI, in the connection region the overlying layer areas may advantageously be removed in order to enable an

electrical connection of the first electrode layer ESI from above . Figure 5 shows a workflow diagram having seven process stages, with which a microacoustic component together with a layer sequence may be produced.

In process stage 1, a second layer comprising silicon oxycarbide is applied onto a substrate. For this, a CVD reactor is used which enables purely a CVD deposition or a plasma-assisted CVD deposition, also called PECVD. The substrate, for example a silicon substrate, is introduced into the reactor chamber. The reactor chamber is evacuated and the substrate is heated to a suitable deposition

temperature. For example, approximately 400 °C is sufficient in PECVD deposition, whereas a CVD process requires 600 to 800 °C. A gas flow of the precursor gases is subsequently set, wherein silane SiH₄ and methane CH₄ may be used, for example. The gas flow is set so that a low gas pressure may be maintained, approximately 1 - 10 torr in the exemplary embodiment .

The first layer is grown in thickness until the desired laye thickness is achieved, which corresponds to a quarter- wavelength in the instance of the microacoustic component.

In process stage 2, in addition to the precursor gases for the first layer, oxygen or an oxygen-containing compound is now introduced into the reactor chamber. This leads to the situation that the growing silicon carbide oxidizes and, as result, a silicon oxycarbide layer grows. To grow this layer it is possible to optimize the deposition conditions, such a pressure and temperature and/or gas flow, so that a silicon oxycarbide layer may grow with high layer uniformity and approximately stoichiometric composition. If, after these first two process stages, the Bragg mirror already has a sufficient reflection via the already deposited mirror layers (first layer and second layer), the third process stage is begun.

In a deviation from this order of process stages, in a first upstream stage a second layer comprising silicon oxycarbide may be applied as a lowermost layer onto the substrate. The process subsequently proceeds further with process stage 1, and subsequently with stages 2 through 7 as presented in Figure 7.

If the reflection with the already deposited mirror layers is still too low, the process is begun again with process stage 1. Together with the second process stage, an additional pair of mirror layers is deposited, and possibly a third pair, until the required or desired reflection of the Bragg mirror is achieved.

If this is the case, in process stage 4 the deposition of the first electrode layer is begun. This may include a CVD deposition of a tungsten layer, or sputtering of a tungsten layer, wherein the first electrode layer is advantageously likewise generated as a λ/4-layer.

To improve the conductivity, the first electrode layer ESI may be enhanced by an additional partial layer of a material with better electrical conductivity, for example aluminum or Al (Cu) . This is then preferably provided on the side of the respective electrode layer that faces away from the piezo- layer, which also applies to the later, second electrode layer . After production of the first electrode layer ESI, this is structured in order to enable an interconnection of multiple resonators across common electrode surfaces for the first electrode layer, corresponding to the desired design of the microacoustic component. For this, individual electrode surfaces are defined, likewise the connections between adjacent resonators or resonators to be connected.

In process stage 5, a piezoelectric layer is deposited over the entire surface, which likewise may take place by means of CVD or sputtering according to known methods . The

piezoelectric layer requires no structuring. At the edge regions of the substrate, the uppermost mirror layer or the first electrode layer may be uncovered for the application of additional unit components, or for the production of vias. It is also possible to thus achieve a lateral energy barrier.

In process stage 6, the second electrode layer is applied, for example via vapor deposition or sputtering of a metal with good conductivity, for example aluminum.

After the application of the second electrode layer, this is also structured in order to define the upper electrodes of the individual acoustic resonators, and to enable a

connection with adjacent resonators. A lift-off technique may be used for this.

In process stage 7, the resonator is finished by applying additional insulating layers. In particular, trimming layers are applied onto such resonators, the resonance frequency of which trimming layers is still to be adapted. Alternatively, the frequency adaptation of individual resonators is achieved via introduction of additional layers into the stack, and via corresponding structuring of these additional layers.

With the aid of the trimming layer, it is achieved that different resonance frequencies can be set for serial and parallel resonators in a ladder-type connection.

Figure 6 shows as an example a ladder-type structure that is known per se, in which serial acoustic resonators SR are connected in series in a signal line between a signal input SE and a signal output SA. Parallel branches PZ in which respectively at least one parallel resonator PR is arranged branch off to ground from circuit nodes before, after or between the microacoustic resonators. Series circuits made up of multiple acoustic resonators PR may also be arranged in the parallel branches PZ . The series resonators SR may also comprise a series circuit of multiple individual resonators.

The shown ladder-type structure comprises three series resonators SRI, SR2, SR3 and three parallel branches PZ1, PZ2, PZ3 in which is respectively arranged a parallel resonator PR1, PR2, PR3. To improve the selectivity of the ladder-type filter, the structure may be supplemented with additional serial or parallel resonators. It is also possible to realize a ladder-type filter with fewer than the shown six resonators .

In a microacoustic component according to the invention, all resonators are generated on the same substrate and use the same Bragg mirror BS . However, it is also possible to produce variations in the layer sequence of the Bragg mirror between the resonators in order to achieve a frequency shift there, for example to suppress unwanted modes. Such variations can be realized by setting slightly different layer thicknesses of individual or multiple layers.

Due to the low residual conductivity, in particular of the silicon carbide layer, which may be achieved via doping, no interfering couplings of adjacent resonators occur. The formation of channels, similar to as in a MOS transistor, may likewise be prevented.

Here the invention has been described only using a few exemplary embodiments, and therefore is not limited to these. In particular, the deposition conditions, the precursors for first and second layers, and the composition of the other layers and structures of the microacoustic resonator may deviate from the presented method and exemplary embodiments. Since the corresponding individual process stages are

respectively known per se from the prior art, these also do not need to be further explained in detail. In the basic concept, the invention is comprised of the sequence of silicon carbide layers and silicon oxycarbide layers whose acoustic impedances exhibit a stark difference, and therefore may advantageously be used in Bragg mirrors for microacoustic resonators .

List of reference signs

AR BAW resonator

AR1, AR2 Individual resonators

BS Acoustic Bragg mirror

ESI First electrode layer

ES2 Second electrode layer

IS Isolation

PR Parallel resonator

PS Piezoelectric layer

PZ Parallel branch

SA Signal output

SC First layers SiCi-_xH_x

SE Signal input

SF Layer sequence

SOC Second layers SiOC (H)

SR Series resonator

SU Substrate

λ Acoustic wave velocity

Claims

1. A layer sequence comprising first and second layers that are deposited atop one another in alternation, wherein

the first layers comprise SiCi-_xH_x

the second layers comprise SiOC (H) .

2. The layer sequence according to claim 1,

in which all layers are adjusted to be semiconducting or dielectric.

3. The layer sequence according to claim 2,

in which all first layers are adjusted to be

semiconducting or dielectric via doping, wherein the doping [sic] a p-doping with Be, B, Al or Ga, or an n- doping with N or P.

4. A microacoustic component, comprising

a substrate having a flat primary surface onto which the following layers are deposited in succession and atop one another and form a layer stack

- a layer sequence according to any of the preceding claims ,

- a first electrode layer

- a first piezoelectric layer

- a second electrode layer

wherein all layers together form a BAW resonator in which the layer sequence serves as an acoustic Bragg mirror .

5. The microacoustic component according to the preceding claims , - in which in particular the first layers are adjusted to be semiconducting or dielectric,

- in which the layer sequence is deposited onto the substrate over a large area and laterally unstructured,

- in which the first electrode layer, the

piezoelectric layer and the second electrode layer are structured to form a plurality of individual resonators interconnected with one another, all of which are arranged on the same large-area layer sequence .

6. The microacoustic component according to any of the

preceding claims,

in which the layer sequence has 2.5 to 4 layer pairs made up of a respective first and second layer

in which the uppermost layer of the layer sequence is a second layer,

in which the first electrode layer comprises tungsten in which the piezoelectric layer comprises A1N.

7. The microacoustic component according to any of the

preceding claims,

in which the layer thicknesses of the layers in the layer sequence are adjusted, depending on the acoustic wave velocity transversal to the surface of the

respective layer and on the center frequency of the microacoustic component, so that the layer thickness corresponds to a quarter-wavelength.

8. A method to produce a layer sequence according to

claim 1, in which the layers of the layer sequence are all deposited onto a substrate in one and the same CVD reactor,

in which gaseous precursors for SiC are introduced into the CVD reactor for deposition of the first layer, in which oxygen is additionally introduced into the CVD reactor for the production of the second layers, in which the oxygen supply is either activated or deactivated for the deposition of the respective next different layer of the layer sequence,

in which overall 5 to 8 first and second layers are deposited in alternation.

9. The method according to claim 8,

in which the layer thickness is monitored during the deposition of each layer.