EP3453058A1

EP3453058A1 - Method for the production of layers of reram memories, and use of an implantation device

Info

Publication number: EP3453058A1
Application number: EP17721514.2A
Authority: EP
Inventors: René BOROWSKI; Won Joo Kim; Vikas Rana; Rainer Waser
Original assignee: Forschungszentrum Juelich GmbH
Current assignee: Forschungszentrum Juelich GmbH
Priority date: 2016-05-04
Filing date: 2017-03-31
Publication date: 2019-03-13
Also published as: US20190115533A1; WO2017190719A1; DE102016005537A1; JP2019517131A; CN109155363A

Abstract

The invention relates to a method for producing layers of ReRAM memories and the use of an implantation device. According to the invention, in order to produce ReRAM memories, TMO layers are applied to an electrode in a desired sequence, a process during which at least one TMO layer is bombarded with ions, e.g. oxygen ions, by means of an ion implantation device such that ions are imported into said TMO layer.

Description

Method for producing layers of ReRAM memories

and using an implanter

The invention relates to a method for producing layers of ReRAM memories and to the use of an implanter.

According to the state of the art, non-volatile data memories exist, such as flash memories, with which information can be permanently stored. However, for these supplies, limits of miniaturization are achieved, which creates the need for nonvolatile storage media that are smaller in size. Because the currently available charge-based memories (ex-flash memories) will have reached the physical limits of miniaturization in the near future, new memory concepts are needed.

ReRAM memories are composed of two opposite electrodes, between which transition metal oxide (TMO) layers are stacked. TMO layers can have an active, electrically conductive layer region or be continuously active or electrically conductive or have a passive, non-electrically conductive layer region or be passive or electrically insulating. In particular, Resistive Random Access Memory (ReRAM ^* s) among other non-volatile memory forms are considered as successors to the current charge-based memory cells. However, there are some technical challenges to overcome, such as damage to the device during application of the necessarily high forming voltage, which can destroy adjacent memory cells in the passive array by undesirably high current flashovers, a narrow read margin (lon / loff) in low voltage operation, and an erzie - Lende good efficiency of the components below 20 nm size.

It is known that control of the oxygen ion (vacancy) profile within the transition metal oxide (TMO) layer is the key to improving ReRAM performance. Thus, numerous approaches have been made to methods for making the TMO layers by means of deposition methods and element doping. A special, active oxide switching layer is known whose material has been pretreated specifically with element doping and in which nano-crystals are embedded embedded within a special switching layer (nc-TiO 2). This structure is located between two metallic electrodes. Such ReRAM's are known from the documents US8569172, US8546781, US8441835, US2013 / 0089949, US2014 / 0302659, US8791444, US2013 / 0187116,

US2015 / 0034898, US8835890, US8487290 and US8907313. Corresponding ReRAM's can also be obtained by thermal processes in which TMO layers can be produced by thermal annealing at 200 ° C to 800 ° C with various gases, which have desired properties.

The first method requires a forming step, which is energy intensive because high voltages must be generated, as well as device components on the ReRAMs that enable formation, viz. Voltage regulators and voltage generators that must be precisely manufactured, but only used once for formation. This consumes material for ReRam ^' s components and leads to costs.

The second method takes place at high temperatures, also consumes energy during thermal annealing (200 ° C ~ 800 ° C) with various gases. This requires one

Heat treatment step (200 ° C ~ 800 ° C) to the process gas (for example, NH ₃ , N ₂ ,

0 ₂ , 0 ₃ , H ₂ O, Cl ₂ , Ar, H ₂ , N ₂ O, SiH ₄ , CF ₄ ) within the active switching oxide layer.

Better heat treatment at high temperature should be avoided as much as possible, because it damages CMOS performance. And besides, it is still difficult here, the oxygen ions or nitrogen ions and defects in place and

Amount to control. Thermal processes of this type are disclosed in US 2013/0336041 and US 8913418.

It is therefore the object of the invention to provide a process for the preparation of ReRAM ^'s are available, which saves energy and electrical components, whereby costs are reduced and process gases can be saved. Formations and associated damage to the ReRAM memory or thermal processes should be avoided. The CMOS performance should not be reduced. In particular, it should be possible to control oxygen ions or nitrogen ions and defects in the TMO layers with respect to location and quantity. Further miniaturization of the ReRAMs should be made possible. The reading tolerance should be increased in low-voltage operation. The efficiency of the components below 20 nm should be increased. It is intended to enable the use of a device with which the method according to the invention can be carried out, wherein the objects mentioned for the method are achieved.

Starting from the preamble of claim 1 and the independent claim, the object is achieved with the features stated in the characterizing part of these claims.

With the method according to the invention and the use of the device according to the invention, it is now possible to save energy and material and thus to reduce costs. The CMOS performance is not reduced. It is possible in the TMO layers Controlling oxygen and nitrogen ions or other ions in terms of location and quantity. The inventive method allows the miniaturization of ReRAM's. The efficiency of components below 20 nm is increased.

Advantageous developments are specified in the subclaims. In the following, the invention is described in its general form without this having to be construed restrictively.

The ReRAM memories according to the invention have an upper and a lower electrode.

The upper and lower electrodes may be made of the same material, which is referred to as a symmetrical structure or of different materials, which is referred to as an asymmetric structure.

In the symmetrical structure, the electrodes may both be made of Pt, Ti, Ta, TiN, Hf, Al or W.

For the asymmetric structure, the material combination PtTi, Pt / a, Pt / TiN, Pt / Hf, Pt / Al, Pt / W, Ti / Pt, Ti / a, Ti / TiN, Ti / Hf , Ti / Al, Ti / W, Ta / Pt, Ta / TiN, Ta / Hf, Ta / Al, Ta / W, TiN / Pt, TiN / Ti, TiN / Ta, TiN / Hf, TiN / Al, TiN / W, Hf / Pt, Hf / Ti, Hf / Ta, Hf / TiN, Hf / Al, Hf / W, Al / Pt, Al / Ti, Al / Ta, Al / TiN, Al / Hf, Al / W , W / Pt, Wm, W / Ta, WmN, W / Hf, or W / Al, wherein the first element is the material for the lower electrode and the second element is the material for the upper electrode.

The upper and lower electrodes may have layer thicknesses of 25 nm to 100 nm, preferably 30 nm to 50 nm.

According to the method of the invention, at least one TMO layer is applied to the lower electrode in each case. In this case, all TMO layers known to the person skilled in the art can be applied. For example, for the TMO layers, materials consisting of a component of the group consisting of hafnium oxide, tungsten oxide, aluminum oxide, aluminum oxide nitride, titanium oxide, tantalum oxide, nickel oxide, niobium oxide, magnesium oxide, cobalt oxide, germanium oxide, molydine oxide, silicon oxide, silicon nitride, tin oxide, zirconium oxide, cerium oxide, Zinc oxide, copper oxide, strontium titanate can be applied.

The stoichiometry of the composition of the sputtered TMO layers proves to be not always well defined in practice. They can contain oxygen and / or nitrogen atoms. For example, Al ₂ O ₃ , HfO _x , HfO ₂ , HfO, HfO _x N _y , Hf _n O _x , W ₂ 0 ₃ , W0 ₃ , Τίθχ _, Ti0 ₂ , Ti ₂ 0 ₃ , TaO _x , Ta ₂ 0 _5, TaON, NiO, Nb ₂ 0 _5, MgO, CoO, W _n O _x, Ti _n O _x, Ta _n O _x, Sn0 _2, Zr _x O _y ZrO, GeOx, Ce0 _2, ZnO, WO, Cu0 ₂ , SrTiO ₃ , MoO, AIO _x N _y , Al _n O _x , Si _x N _y as TMO- Layer be broken up. The inventive method is basically not limited to certain materials for TMO layers.

In a cell, all TMO layers can be made of the same material, preferably a component from the group of materials mentioned in the preceding paragraph. In multi-layer cells having at least two TMO layers, at least two TMO layers may be made of different materials, preferably materials from the group of materials mentioned in the preceding paragraph.

The stoichiometric ratios of metal to oxygen may also be broken so that they do not correspond to a composition resulting from the charge ratios of the ions.

A TMO layer may have a thickness of 1.5 nm-10 nm in the case of a thin TMO layer.

A thick layer may be between> 10 nm, for example 10.1 nm and 40 nm.

Embodiments are typical in which only one layer is applied between the lower and upper electrodes. This layer may, for example, have a thickness of 1.5 nm-40 nm.

It is also possible for thin and thick layers to be applied in a layer sequence to the lower electrode.

If at least two TMO layers are applied to the lower electrode, they are preferably thin and each have a thickness of 5 nm-10 nm.

The number of TMO layers is freely selectable and may for example be between one and five TMO layers.

The thickness of the individual TMO layers in a layer sequence is also freely selectable.

Each individual TMO layer can be applied to the lower electrode by methods known to those skilled in the art.

For example, the TMO layers can be sputtered with a reactive PVD (physical vapor deposition) method. Alternatively, reactive ALD methods (atomic layer deposition method) or reactive CVD methods (chemical vapor deposition method) can be used. A reactive process is understood to mean a process in which the applied metal layer is oxidized with oxygen. According to the invention, in at least one TMO layer, foreign ions, for example oxygen ions, for example 0 ⁺ , 0 ²⁺ or nitrogen ions, for example N ⁺ , N ₂ ⁺ , are introduced into the TMO layer by an ion implantation process. Ion implantation is a process of introducing impurities (in the form of ions) that are shot at the layer to be implanted. Furthermore, the implantation of elements with cations of the elements Li, Be, B, C, F, Ne, Na, Mg, Al, Si, P, S, Cl and Ar is possible. The charge of the cations depends on the conditions under which they are formed. As a result, little or no ions are left in the corresponding TMO layer, but rather oxygen, nitrogen atoms or atoms of another element. In principle, all known ion implantation methods and ion implantation devices can be used.

Which and how many TMO layers are implanted with oxygen ions, nitrogen ions or other ions is the freedom of design left to the skilled person. It can be implanted in one, several or all TMO layers with oxygen ions, nitrogen ions or other ions.

In a ReRAM memory, either all TMO layers treated by ion implantation can be implanted with oxygen ions, or all TMO layers treated by ion implantation can be implanted with nitrogen ions or other ions, such that a ReRAM memory exclusively contains oxygen-implanted layers exclusively nitrogen-implanted or exclusively with contains another ion-implanted layers.

Within a ReRAM memory, however, there may also be TMO layers in which various elements are implanted.

For the ion implantation, various methods are known which can be used in principle.

An implanter, for example an ion gun, can be used with which the oxygen ions, nitrogen ions or other ions are accelerated.

In this embodiment, the TMO layer to be treated is in a vacuum chamber and is bombarded with oxygen ions, nitrogen ions or other ions. The energy for these ions is preferably 0.5 keV to 200 keV. At these energies, the oxygen, nitrogen ions or other ions reach velocities that are one Ion implantation of the TMO layer without causing excessive removal of TMO molecules result.

In this case, an ion density of 10 ⁸ ions per cm ² to 10 ¹⁸ ions per cm ² can be achieved.

The gas flows of oxygen, nitrogen or other gases which are preferably to be used are 1 sccm (standard cubic centimeter = 1 cm ³ at T = 0 ° and p = 1013.25 hPa) and 100 sccm.

In ion implantation procedures, it is important that the ion fluxes be directed at the TMO substrate. For this purpose, so-called ion optics are used, with which it is possible to direct ion currents on surfaces. In the process, a plasma is initially generated which contains the desired ions. The plasma can be the cations of o.g.

Contain substances. The ions of the plasma are subjected to an acceleration which is directed to the surface of the TMO layer. For this purpose, anode grid can be used. By way of example, it is possible to use 1 to 3 anode lattices, which are preferably arranged perpendicularly or substantially perpendicular to the ionic current in relation to the target direction of the ions. Decisive here is that the ion beam is directed so that it hits as homogeneously as possible with respect to its cross-sectional area on the surface, so that a uniform implantation takes place. In this case, the ion current in an embodiment with two anode gratings passes through an output voltage which directs the ion current in one direction and then an acceleration voltage which accelerates the ion current to the desired kinetic energy. The ion streams thus directed can then impinge on the uppermost TMO layer.

It is customary that the accelerated oxygen, nitrogen ions or other ions are at least partially neutralized by a neutralizer, so that no positive charge cloud forms when the ions strike in the vicinity of the TMO layer to be implanted, the penetration of ions or oxygen or nitrogen atoms in the TMO layer to be implanted prevented or reduced. For this purpose, an electron source, for example a cathode attached laterally over the TMO layer, may be present above the TMO layer, which emits electrons into the region of the surface of the TMO layer. There are ion-implantation devices known in the art that can be used. These are known implanters or with appropriate process control, for example as described above, reactive ion beam etchers (RIBE).

With the ion implantation, at least one layer of the aforementioned TMO materials can be treated. The number of TMO layers whose chemical composition Composition and their sequence, as well as the individual layer thicknesses can be chosen freely.

Various embodiments of ion implantation into a TMO layer are possible.

A TMO layer can over its entire layer thickness with the o.g. Be implanted or implanted ions or elements.

However, a TMO layer can also be implanted or implanted with ions or elements only up to a certain penetration depth.

The penetration depth of the ions into the TMO layer to be treated can be controlled via the kinetic energy. However, it also depends on the material to be treated for implantation of the TMO layer. Thus, the atomic mass of the cations of the TMO layer and thus the space filling of the atoms and the lattice properties of the TMO layer have an influence on the penetration depth of the atoms of the elements to be implanted. The experimental parameters for the implantation must therefore be selected by the person skilled in the art according to the desired results. This is within the skill of the art. The implanted part of a TMO layer is the active, electrically conductive layer of the TMO layer. The non-implanted portion of the TMO layer is the passive electrically insulating portion of the TMO layer.

A cell consisting of a lower electrode of at least one TMO layer and an upper electrode may be differently composed. Within a cell, entire TMO layers, or at least one, can be active, ie electrically conductive, so that they are suitable as a switching function, or a layer region, according to the penetration depth of the import, can be active and thus electrically conductive.

Within a cell, however, at least one TMO layer can also be completely passive and thus electrically insulating, so that no ions have been imported into this TMO layer.

How exactly the sequence and composition of TMO layers in a cell is designed depends on the desired property of the cell and is freely configurable.

The TMO layer on the lower and / or upper electrode should be passive at least on the side facing the electrode. But it can also be passive over its entire layer thickness. The TMO layer adjacent to the electrodes must not be an insulator, however, the resistance must be such that no short circuit occurs.

In particular, thin TMO layers can be implanted in the specified parameter ranges of kinetic energy in the range of 0.5 keV to 200 keV with oxygen, nitrogen ions or other ions.

If a thin layer has a thickness of 1.5 nm-10 nm, for example, the implantation can take place up to half the thickness. The depth of implantation can be chosen freely as needed and depends on the implantation energy.

It is also possible to bombard thick TMO layers with a thickness of> 10 nm-40 nm with higher energies, the entire layer thickness being implanted with oxygen, nitrogen ions or other ions.

The penetration depth of the import of the elements requires the storage duration, the switching speed, the efficiency or the duration of the storage capacity of the cell.

The figures show examples of the method according to the invention and the cells thus obtained, as well as experimental data.

It shows:

Hg. 1a: The cell resulting from the process steps.

FIG. 1b: Experimental comparison data for a cell according to FIG. 1a.

FIG. 2a: another cell resulting from the method steps. FIG.

FIG. 2b: Experimental comparative data for a cell according to FIG. 2a.

FIG. 2c: Experimental comparison data for a cell according to FIG. 2a.

Fig. 3: Examples of cells with layer sequences, which were prepared by the method according to the invention.

Figure 1 a shows on the left side a lower electrode 1, on which a TMO layer 2 is located, act on the oxygen ions 3, which are shown in the TMO layer as circles and oxygen vacancies 4, as branched arms are shown lead. In addition, a further stage of the cell produced by the method according to the invention is shown in the middle, which has a further TMO layer 5 which has not been treated with ion implantation. On the right side of the completed cell is shown, in which the upper electrode 6 is applied. FIG. 1b shows experimental comparative data between cells produced by the method according to the invention according to FIG. 1a and cells according to the prior art.

On the left side of the figure, a graph is shown in which the resistance of a cell in ohms against the reset voltage V _ReS et is plotted. The curve 1 shows the dependence of the resistance of the cell according to the invention as a function of the applied reset voltage V _rese t. Curve 2 shows the dependence of the resistance of a cell according to the prior art t- from the reset voltage V _Rese It can be seen that the cell according to the invention produced at lower values for the restoring voltage V _reset, over the prior art has higher bandwidth of the resistor. Below are the curves 3 and 4 shown. In this case, curve 3 denotes the data for a cell according to the invention for the low-resistance state and curve 4 the data for the low-resistance state of a cell according to the prior art.

On the right side of FIG. 1b, the relationship between the resistance for state 1 to state 0 for various reset voltages V _Rese t in volts is shown. On the left side of this figure, the value range for cells produced by the method according to the invention is shown, on the right side of the range of values for cells is shown according to the prior art. The size of the box shows the distribution of the measured values between 25% and 75% of the measured values. The horizontal line in the boxes shows the median value. The extension lines of the boxes mean 5% to 95% of the measurements. By comparison, it can be seen that the cells produced by the method according to the invention have a higher quotient for the ratio between the values of the resistance for the states 1 and 0 at the same reset voltages V _Reset t. In Figure 2a, the same components of the cell have the same reference numerals. The left part of the figure is identical to FIG. 1a. The middle part of the figure shows a second TMO layer 5 treated with injected oxygen ions 3.

FIG. 2b shows in the left part of the figure experimental comparison data between cells produced by the method according to the invention according to FIG. 2a and cells according to the prior art for every 50 cells which have been measured. The forming stress in this representation is the abscissa and the Weibul distribution function in% is the ordinate. In the graph, curve 1 corresponds to a cell made in accordance with the invention, and curve 2 corresponds to a cell according to the prior art. It shows for these cells the percentage of switched cells at a given forming voltage V. On the right side of FIG. 2b, the number of cells which have reached a certain resistance is shown as a function of the initial resistance. Here, the abscissa is the initial resistance of a cell in ohms and the ordinate is the Weibul distribution in%. Curve 1 again denotes the values for the cell according to the invention. Curve 2 shows the values for cells for the example in Figure 2a according to the prior art. It becomes apparent that no forming voltage is required for a cell produced according to the invention.

Figure 2c shows comparative experimental data between cells prepared by the method according to the invention according to Figure 2a and cells according to the prior art. On the left side of the figure, a graph is shown in which the resistance of a cell in ohms is plotted against the reset voltage V _Re set. Curve 1 shows the dependence of the high-impedance resistor of the cell according to the invention as a function of the applied reset voltage V _Reset . Curve 2 shows the dependence of the high-impedance resistance of a cell according to the prior art on the reset voltage V _Reset t. Below these are curves 3 and 4. In this case, curve 3 denotes the data for a cell according to the invention for the low-resistance state and curve 4 the data for the low-resistance state of a cell according to the prior art.

On the right side of Figure 2c, the relationship between the resistance for state 1 to state 0 for various reset voltages in volts is shown. On the left side of this figure, the value range for cells produced by the method according to the invention is shown, on the right side of the range of values for cells according to the prior art is shown. The size of the box shows the distribution of the measured values between 25% and 75% of the measured values. The horizontal line in the boxes shows the median value. The extension lines of the boxes mean 5% to 95% of the measurements.

Figure 3 shows three examples of cells made according to the invention. It includes a third TMO layer 7.

Example:

The experiments were carried out in an ion beam etching plant of Oxford (Oxford lonfab 300 plus). This facility has not been modified or rebuilt. The basic function of the etching system is the physical dry etching process. The surface of the substrate is etched by the bombardment of ions. The bombardment leads to sputtering of the substrate material; the expired processes are similar to those at Sputtering, which is not counted among the dry etching. In principle, it is an "atomic sandblaster" when it is presented in a greatly simplified manner.

It has been found that, as a side effect of ion beam etching, implantation of ions, for example oxygen ions or nitrogen ions, can take place with the aid of the oxygen ion beam or the nitrogen ion beam with low energy.

The etching is carried out in a high-vacuum chamber in order to prevent interactions such as scattering of the particle beam with the residual gas atoms of the vacuum. The ion beam has a diameter of 150 mm and is coherent in order to achieve the same ion density everywhere on the substrate and thus a homogeneous etching effect. The etched profile obtained is anisotropic (directional).

To generate the ion beam, first a gas (eg Ar, O ₂ or N ₂ ) is excited in a high-frequency alternating field whose frequency is typically 13.56 MHz. The ions thus generated are simultaneously accelerated and brought to coherence in the direction of the substrate by means of an ion optic, in this case two gratings. In order to keep the free, mean path length of the ions as large as possible, the etching process takes place depending on the gas flows used at about 4 x 10 ^" mbar.

In order to prevent a positive space charge from being formed on a non-conductive substrate by the incident ions, the ion beam is penetrated by an electron beam after passing through the ion optics, in this case the grating, perpendicular to the direction of movement. The electrons neutralize the positively charged gas ions in such a way that they can not give off any positive space charge when they strike the substrate. If no neutralization took place, then a positive electric field would build up very quickly over the substrate and distract the positively charged ions. The kinetic energy of the gas atoms is also converted into thermal energy upon impact. In order to protect the substrates from the thermal influences, a cooling, for example with helium gas, can be carried out on the back of the substrate.

The 1st TMO layer (Ta205 deposited on the surface of the lower electrode) already has an oxygen ion implantation. However, the 2nd TMO layer (Ta205, disposed between the 1st TMO layer and the upper electrode) has NOT received any oxygen ion implantation, see Fig. 1 (a). So there are two TMO layers between the upper and the lower electrode. A significant improvement in reading tolerance (Roff / Ron) will be found in the test components (with oxygen-ion implant only in the 1st TMO layer) observed over all RESET voltage conditions of -1, 0 V to -1, 6 V, see Figure 1 (b).

If both the 1st TMO layer and the 2nd TMO layer have already received oxygen ion implantation, see Fig. 2 (a), no further forming step is needed while the reference components (double TMO layer WITHOUT oxygen ion implantation) more than +1, 8 V forming voltage required, see Fig.2 (b).

All of the measured test components have low resistance states (~ 1.4 k ohms) before the forming step, while the reference components show a fairly high resistance (> 12 G ohms). The oxygen vacancies intentionally created by oxygen ion implantation cause the good conducting path between the upper and lower ones

Electrode before molding. Therefore, the component with the double ion implant TMO layer without forming step could have a lower resistance state. Fig. 2 (c) shows the comparison of the electronic performance between the test and reference components. There is no drop in the observed behavior in the test components. Based on these experimental results, we can obtain various types of benefits depending on the location of the oxygen ion implantation. Depending on the application of the ReRAM components, we can selectively install the position of the oxygen ions and defects within a specific TMO layer. FIG. 3 shows examples of variations of the components by means of oxygen ion implantation, as a result of which the individual layers produced were obtained with oxygen ions and oxygen vacancies specifically controlled in place and in quantity.

Since the basic principles of ion implantation (ion injection and void formation) are applicable to all types of TMO layers, such as e.g. HfOx, WOx, AlOx, etc., there is no limitation of TMO material selection for the ReRAM component design. This gives a great deal of flexibility to improve the performance of the components depending on the desired application.

Each single TMO layer could be made of the same TMO material (homo multi TMO) or different TMO material (hetero multi TMO).

Nitrogen ion implantation can be used instead of oxygen ion implantation. And also oxygen or nitrogen doping are possible as plasma-assisted process steps to install the oxygen or nitrogen ions in TMO layer can.

Claims

Patent claims

1. A method for producing layers of ReRAM memories in which

lower electrode (1) at least one TMO layer (1) is applied,

characterized,

that foreign atoms are implanted in at least one TMO layer by means of an ion implantation process.

2. The method according to claim 1,

characterized,

a TMO layer of a component from the group consisting of hafnium oxide, tungsten oxide, aluminum oxide, aluminum oxide nitride, titanium oxide, tantalum oxide, nickel oxide, niobium oxide, magnesium oxide, cobalt oxide, germanium oxide, molydine oxide, silicon oxide, silicon nitride, tin oxide, zirconium oxide, cerium oxide, zinc oxide, copper oxide, strontium titanate is applied.

3. The method according to claim 2,

characterized,

that a TMO layer of a component selected from the group Al ₂ 0 _3, HfO _x, Hf0 _2, HfO, HfO _x N _y, Hf _n O _x, W ₂ 0 _3, W0 _3, TiOx, Ti0 _2, Ti ₂ 0 ₃ , TaO _x , Ta ₂ 0 ₅ , TaON, NiO, Nb ₂ 0 ₅ , MgO, CoO, W _n O _x , TinOx _, Ta _n O _x , SnO ₂ , Zr _x O _y ZrO, GeOx, CeO ₂ , ZnO , WO, Cu0 ₂ , SrTi0 ₃ , MoO, AIO _x N _y , ΑΙ _π Οχ _, Si _x N _{y is} applied.

4. The method according to any one of claims 1 to 3,

characterized,

the TMO layer is applied by means of a reactive PVD method, a reactive ALD method or a reactive CVD method.

5. The method according to any one of claims 1 to 4,

characterized,

that a TMO layer ^{'is applied} with a layer thickness between 1, 5 nm to 40 nm.

6. The method according to claim 5,

characterized,

that a TMO layer is applied with a layer thickness between 1, 5 nm to 10 nm.

7. The method according to any one of claims 1 to 6,

characterized,

in that the elements introduced into the TMO layer by the implantation process comprise at least one component selected from the group consisting of oxygen, nitrogen, Li, Be, B, C, F, Ne, Na, Mg, Al, Si, P, S, Cl and Ar are.

8. The method according to any one of claims 1 to 7,

characterized,

that the implanted foreign atoms are introduced into the TMO layer with an energy in a range between 0.5 keV and 200 keV.

9. The method according to any one of claims 1 to 8,

characterized,

that an ion implanter, for example an ion gun, is used, the emerging ions being guided by an ion optics onto the TMO layer.

10. The method according to claim 9,

characterized,

the ions emerging from the ion implanter or the ion gun are at least partially neutralized before they enter the TMO layer.

1 1. A method according to any one of Ansprüel to 10,

characterized,

that the foreign atoms introduced into the TMO layer penetrate into the TMO layer at least over part of the layer thickness of the TMO layer.

12. The method according to any one of claims 1 to 10,

characterized,

that the foreign atoms introduced into the TMO layer penetrate into the TMO layer over the entire layer thickness of the TMO layer.

13. Use of an implanter, an ion gun or a reactive ion etcher for the implantation of ions in a TMO layer.

14. Use of the implanter, the reactive ion etcher or the ion gun according to claim 13 in a method according to one of claims 1 to 12.