RU2426193C1 - Method of depositing platinum layers onto substrate - Google Patents

Abstract

FIELD: chemistry. ^ SUBSTANCE: method of depositing platinum layers onto a substrate involves pre-formation of an intermediate adhesion layer from a mixture of platinum and silicon dioxide nanocrystals on a silicon oxide and/or nitride surface. The intermediate adhesion layer with thickness 1-30 nm can be formed via simultaneous magnetron sputtering using magnetrons with platinum and silicon dioxide targets, respectively. ^ EFFECT: high quality of elements, processes, reliability during prolonged use, adhesion of the deposited layers to the substrate. ^ 8 cl, 3 dwg

Description

The invention can be used in the fields of electronic technology, for example, in the processes of forming film elements of microelectronic devices, in particular for the manufacture of semiconductor gas sensors.

The principle of operation of semiconductor gas sensors is based on a change in the conductivity of the semiconductor sensitive layer under the action of chemisorbed gas at grain boundaries or on the surface of metal oxide semiconductor nanoparticles. Using semiconductor gas sensors, it is possible to determine the concentration of donor gases (various combustible gases, including methane, propane, organic solvents, CO, ammonia, etc.) or acceptor gases (ozone, nitrogen oxides, chlorine, fluorine, etc.). The detection threshold for various gases is different and is approximately 0.1 ppm for CO, ozone, nitrogen oxides; a few ppm for ammonia; about 10 ppm for methane, etc. The upper limit at which it is reasonable to use semiconductor sensors is a concentration of about 0.5 LEL (lower concentration limit of flame propagation). In order to provide a sensor response time of the order of several seconds, the sensor must be heated to a high temperature lying in the range from about 250 (when detecting hydrogen and ethanol) to 500 ° C (methane).

Typically, the sensitive layer is made of nanocrystalline metal oxides (SnO ₂ , ZnO, In ₂ O ₃ , etc.) with a specific surface area of ~ 50-200 m ² / g (which corresponds to a particle size of ~ 5-20 nm). These materials have a fairly high stability at the operating temperature of the sensor. The surface of the nanoparticles can be doped with clusters of catalytic metals, which provide a certain selectivity of the oxidation process and improve the selectivity of the sensor.

Another type of microelectronic sensors - thermocatalytic sensors - measure the heat generated by the oxidation of combustible gases on a catalyst deposited on a platinum heater. In this case, nanoparticles of oxides of aluminum, zirconium, hafnium, and some other metals with palladium and platinum nanoparticles deposited on their surface are used as a catalyst. Heat generated by a chemical reaction is measured as the temperature change of a platinum heater. Thermocatalytic sensors are used to detect relatively high concentrations of combustible gases close to LEL.

Gas sensors seem to be simple devices, but the technology of their production has concentrated all the modern achievements of the science of materials and microelectronics, since a competitive sensor should be characterized by high stability and selectivity, should work for several years at temperatures up to 500 ° C, while consuming no more than several tens of milliwatts of power for heating. In addition, the drift of the sensor parameters should be small, it is desirable that it could work without recalibration for several years.

The most important element of the gas sensor, designed for stand-alone and wireless devices, are microheaters made using micromachine technology (MEMS technology - microelectromechanical systems).

Regarding the use of silicon technology.

In this device, the membrane is stretched on a rigid frame made of silicon. It has two roles: it serves as a supporting element for the functional elements of the sensor (heater, thermometer, sensitive layer, etc.) and, on the other hand, a thin membrane having low thermal conductivity in the plane is an insulating element for a sensitive layer heated to a high temperature . Such thermal insulation is necessary to minimize the power consumed by the sensor at operating temperature.

Historically, microheaters have been associated with thick-film gas sensors. Currently, the most famous sensors are manufactured by this technology by Japanese companies Figaro Inc. and FIS. The Figaro TGS 2610TM sensors (www.Figaro.co.jp) have a ruthenium dioxide heater. The dimensions of the chip are 1.5 × 1.5 × 0.5 mm.

The chip is mounted in a TO-5 package; it is suspended on 40 micron Pt / W alloy wires. The size of the chip is not optimal, so it consumes a fairly high power when detecting methane - about 280 mW. The power consumed by the FIS SP-11TM type methane sensor (2 × 2 × 0.3 mm) is even higher, about 400 mW. The choice of sensor topology is determined by the packaging technology, since almost half of the sensor cost is associated with this operation. The high power consumption and the associated difficulties in the manufacture of explosion-proof equipment that use these sensors are pushing researchers to look for ways to improve sensor technology.

Optimization of thick-film sensors was undertaken in [1]. The sensor was manufactured using the original technology without the use of a substrate. All thick film layers were obtained by screen printing on the sacrificial layer (the layer removed from under the working layer after manufacturing the latter), which was etched after production. Particular attention was paid to the long-term stability of the heater material. It is well known that commercial resistive pastes based on ruthenium dioxide or platinum cannot be used for the manufacture of sensors. The first ones contain lead glasses that interact with RuO ₂ , which leads to a drift of the heater resistance at high temperature. Platinum-based pastes have too low resistivity, this requires the use of a meander-shaped heater and thus increases the size and power of the sensor. Materials based on RuO ₂ and Pt [1] composites are free from these disadvantages. The material based on RuO ₂ has a resistance of 6-20 Ohms / square and can be used up to a temperature of 450 ° C; Pt-based material has a resistance of 2-6 ohms / square and is used up to 600 ° C. The resistance drift of the heater with both materials does not exceed 3% per year. A sufficiently high TCS (temperature coefficient of resistance) allows the use of heaters as temperature sensors to stabilize the temperature of the sensor. The power of the sensor at a temperature of 450 ° C is approximately 220 mW. The thermal response time is about 1 s, the sensor can withstand more than 10 million on-off cycles without damage.

A further decrease in the power consumed by semiconductor and thermocatalytic power sensors is not possible within the framework of thick-film technology. For thick-film sensors, the limitation is related to the need to weld wire contacts to the chip — in [1], the sizes of the contact pads and the chip were comparable.

This means that any further reduction in the power consumed by semiconductor and thermocatalytic power sensors is impossible without the use of micromachined approaches in the design of devices.

Such a sensor contains a thin dielectric film made of several layers of silicon oxide and nitride. This film carries a sensitive layer and insulates it from a silicon frame at room temperature. The application of this scheme allows us to solve the problem that cannot be solved in thick-film technology: you can separate the micro-heater and contact pads and minimize the heated area and power consumed by the sensor. There are several problems that have not been completely resolved to date in silicon micromachining technology, that is, the technology for manufacturing microelectromechanical devices using silicon technology.

Problems

1. Stability of the membrane at high temperature.

This problem is associated with very high mechanical stresses in silicon oxide and nitride, which are used to make the membrane. These stresses have opposite signs, and the use of a multilayer structure can reduce the average voltage. Nevertheless, thermal fatigue of the membrane ultimately leads to its destruction even at a constant temperature.

2. The stability of the material of the heater and its adhesion to the membrane material at high temperature.

3. Stability of silicon nitride at high temperature in a humid atmosphere.

4. Adhesion of the sensitive layer to the membrane material.

Over the past decades, various research groups have tried to solve these problems, however, a completely satisfactory solution, which allows the production of sensors, has not yet been found.

The authors of [2] fabricated a micromachine system containing not only a gas sensor on a thin dielectric membrane, but also a microelectronic circuit made using CMOS (Complementary Metal-Oxide-Semiconductor) technology / CMOS: Complementary-metal-oxide semiconductor - technology for constructing electronic circuits /, intended for reading data, signal processing and communication with the device. The heater is made of polycrystalline silicon, the diameter of the membrane is approximately 500 microns, the size of the heater is 300 microns. These dimensions make it possible to manufacture a sensor with an energy efficiency of about 7 ° C per milliwatt (an increase in the heater power by 1 mW leads to a temperature increase of 7 degrees. That is, this is the slope of the curve of the temperature of the heater as a function of its power), that is, the power consumed by the sensor equal to about 60 mW at 450 ° C. The authors do not provide information on the long-term stability of the system and on the differences in the reliability of the sensor and the electronic circuit. Another issue that is not clarified in the work is the possibility of technological annealing of the integrated circuit together with the applied sensitive layer. Usually, annealing at 700-800 ° C is necessary, which the CMOS circuit can hardly withstand. Also unclear about the cost of such a product.

Platinum heaters for thermocatalytic sensors were proposed in [3].

In this work, the authors investigated the ratio of the size of the membrane and the heater to the thermal characteristics of the sensor. The dimensions of the heater varied from 200 × 200 μm to 570 × 570 μm with a fixed membrane size of 1 × 1 and 2 × 2 mm. The power of a microheater of size (200 × 200 μm) is 120 mW at 450 ° C. The authors of [3] report the stability of the heater for 1000 hours under constant and pulsating heating, but do not indicate either the drift value or the technology for applying a platinum heater.

Currently, the most interesting results on minimizing heater power were obtained in [4, 5]. The micro-heater in the form of a bridge was optimized in terms of power consumption [4]. The microbridge was obtained by etching silicon on the front side, in contrast to the devices that were discussed earlier. At 450 ° C, the microheater consumes about 18 mW, which is currently a world record. Nevertheless, taking into account the very small dimensions of the heater (100 × 100 μm), the specific power is not so small, this can affect the stability of the sensor. This technology has several disadvantages: the membrane is too close to the bottom of the cavity (200 μm) and is connected by a needle to this bottom. Unfortunately, the authors do not provide data on the stability, yield, and reliability of the sensors presented. The thermal inertia of the presented sensors is very small: the thermal response time is only 3 ms.

Systematic studies of the topology of micromachined heaters were performed in [6–8]. The topology was optimized using the ISE SOLIDIS program (ISE AG, Zurich (Switzerland)) taking into account the results of [1]. To obtain the minimum power, the membrane should be 8-10 times larger than the heater.

The starting material for the manufacture of microheaters was a 4-inch silicon plate of p-type orientation (100), polished on both sides. 500 nm thermal silicon oxide was grown on a wafer, then a layer of stoichiometric Si ₃ N ₄ 150 nm thick was deposited using LPCVD, followed by 500 nm oxide of TEOS (tetraethoxysilane). On the reverse side, a window was removed to release the membrane. A platinum heater was deposited by electron beam evaporation of platinum; a 5 nm thick Cr, Ti, or Ta layer was used as an adhesive layer.

The use of adhesive layers of these metals has not yet yielded the desired result. The fact is that for the formation of a sensitive layer, the chip must be heated to 700-800 ° C for 15-60 minutes, and its operating temperature can reach 500 ° C. At this temperature, the metal of the adhesive layer is oxidized, which leads to peeling of platinum.

In the manufacture of semiconductor and thermocatalytic gas sensors, platinum layers 100-200 nm thick obtained by magnetron sputtering are used as heaters. The use of platinum for these purposes is due to the fact that no other material (for example, nichrome, metal silicides, polycrystalline silicon, etc.) has sufficient oxidation stability when used in the form of a thin film during long-term operation over several years at a sensor operating temperature of 400-500 ° C.

The disadvantage of platinum is its poor adhesion to the surface of silicon dioxide, which is used for the manufacture of micromachined membranes of low-power sensors of wireless gas alarm systems.

The use of adhesive sublayers 10-20 nm thick, which are usual for microelectronics, which are applied under a platinum layer, are made of titanium, tantalum, chromium and the like metals, does not give a positive effect, since the sensor’s operating temperature is 400-500 ° С. At this temperature, the adhesive layer is gradually oxidized, which leads to a loss of adhesion of platinum.

The proposed solution includes the creation of adhesive layers made of a mixture of nanostructured silicon dioxide and platinum, and provides, in particular, the stability of the heaters in the operating temperature range of 400-500 ° C.

Known technical solution: RU 2365403, IPC B01D 67/00; B01D 71/02; B82B 1/00. Application 2008130031/15, 07/21/2008. METHOD FOR PRODUCING A GAS-PERMEABLE MEMBRANE.

The specified solution includes fragmentary deposition on both surfaces of a single-crystal silicon wafer of metal coatings chemically inert in solutions of hydrofluoric acid with an oxidizing agent. The plate is annealed under conditions that ensure the formation of an ohmic contact between the deposited metal and silicon, and then the pore formation process in silicon is carried out by treating the plate in a solution containing hydrofluoric acid, an oxidizing agent, a substance that helps to restore the oxidizing agent on the metal surface, and a surfactant.

However, there are such disadvantages as the complexity of the processes, insufficient accuracy and reliability of manufacture, insufficient adhesion of the applied layers to the surface of the substrate, reliability during long service life.

The technical solution is also known: SU 1195849, IPC 5 H01L 21/28. Application 3737793/25, 05/07/1984. METHOD FOR CREATING MULTILAYERED CONTACT SYSTEMS.

The essence of the method lies in the fact that when creating multilayer contact systems consisting of a base layer of an oxidizable metal and a base layer containing at least one non-oxidizable metal, including ion etching of the base layer in a mixture of oxygen and an inert gas through a protective layer of oxidizable metal and chemical etching of the base and protective layers, before chemical etching, the oxidized layer is removed from the surface of the base and protective layers by ion etching in an inert gas environment.

Removing the oxidized layer eliminates the large raster of the elements of the pattern during chemical etching of the base and protective layers, which ensures high fidelity of reproduction of elements when creating multilayer contact systems. Removing the oxidized layer by ion etching in an inert gas environment allows this operation to be carried out in a single technological process for creating a multilayer contact system, including ion etching of the main layer, which eliminates undesirable contamination with extraneous impurities of the processed plates.

The disadvantage of this method is the relatively high complexity and complexity of the processes, insufficient accuracy and reliability of manufacture, insufficient adhesion of the applied layers to the substrate (main layer), reliability for long periods of operation.

A technical solution is also known: RU 2110112, IPC 6 H01L 21/033. Application 96112802/25, 06/18/1996. METHOD FOR FORMING FILM ELEMENTS BASED ON PLATINUM.

The essence of this solution: when forming film elements based on platinum, a metal oxide adhesive layer and a platinum layer are applied onto a substrate with an oxygen-free dielectric coating, photolithographic configuration of the elements is performed. The adhesive layer is applied by spraying titanium in an oxygen-containing plasma, starting at a pressure ratio P _O2 / p = 0.3-0.4 and ending at P _O2 / p <0.2, where P _O2 is the oxygen pressure, p is the total pressure of the mixture. After the platinum layer is deposited, annealing is performed in vacuum until platinum dioxide is reduced to platinum and lower oxides, and after the formation of the configuration of the elements, the substrate is annealed in an oxygen-containing medium.

The disadvantage of this method is the relatively large number of materials used, including refractory, which necessitates the use of multifunctional and multi-position devices for their application, insufficient adhesion of the applied layers to the substrate, reliability during long service life. The specified solution can be considered as a prototype to the declared.

The technical task is to improve the quality of the elements, the manufacturability of the processes, reliability for long periods of operation, the adhesion of the applied layers to the substrate, which ensures the achieved technical result in combination and interdependence.

The specified provides the proposed combination of essential features.

A method of applying platinum layers on a substrate, comprising pre-forming on the surface of silicon oxide and / or silicon nitride an intermediate adhesive layer, wherein the adhesive layer is formed from a mixture of silicon dioxide and platinum nanocrystals,

wherein

- the adhesive layer is formed with a thickness of 1-30 nanometers;

- the adhesive layer is formed by simultaneous magnetron sputtering from two magnetrons with targets of platinum and silicon dioxide, respectively;

- the adhesive layer is formed by magnetron sputtering using a platinum target, part of the surface of which is covered with a silicon dioxide plate;

- after the formation of the adhesive layer from the surface of the target remove the plate of silicon dioxide, then apply the bulk of the platinum layer by magnetron sputtering of platinum;

- after the formation of the adhesive layer with targets of platinum and silicon dioxide, respectively, turn off the magnetron with the target of silicon dioxide and then apply the main part of the platinum layer;

- the platinum layer is applied in a mixture of oxygen and an inert gas containing 5-50 volume% oxygen;

- after applying the main part of the platinum layer, stabilizing annealing is performed, for which the substrate is placed in an oven, the temperature is increased at a speed of no more than 200 ° C per hour to 650 ° C, then it is kept at 650 ° C for at least 2 hours until complete decomposition platinum oxides, then the temperature is increased at a speed of no more than 200 ° C per hour to 850 ° C, then it is kept at 850 ° C for 2 hours and the temperature is lowered to room temperature at a speed of no more than 200 ° C per hour.

The proposed solution is illustrated by illustrations.

Figure 1 shows a cross section of an article consisting of a silicon dioxide supported on a silicon substrate 1 coated with a silicon dioxide layer 2, an adhesive layer 3 composed of a mixture of silicon dioxide and platinum nanocrystals and a platinum layer 4.

Figure 2 shows schematically the operations according to claim 3 of the formula - applying a mixture of silicon dioxide and platinum nanocrystals on the surface of a silicon substrate coated with silicon dioxide 2 and fixed in the substrate holder 5 by means of two magnetrons 7, one of which is supplied with a constant voltage (-U) and is equipped with a target of platinum 8, and the second is powered by an alternating high-frequency voltage (~ U) and is equipped with a target of silicon dioxide 9. The application of the adhesive layer 3, consisting of a mixture of platinum nanocrystals and silicon dioxide, occurs when simultaneously spraying from both magnetrons 7, after applying the adhesive layer, one of the magnetrons is turned off and the platinum layer is deposited by sputtering the target with a second magnetron. The spraying of platinum and silicon dioxide occurs under the action of bombardment by argon ions (Ar ⁺ ) when gaseous argon (Ar) is fed into a vacuum chamber containing all of the above elements.

Figure 3 shows schematically the operation according to claim 4 of the formula — depositing a mixture of silicon dioxide and platinum nanocrystals on the surface of a silicon substrate coated with silicon dioxide 2 and fixed in the substrate holder 5 by means of a magnetron 7 using a target made of platinum 8, part of which is covered by a plate 10 made of silicon dioxide. The spraying of platinum and silicon dioxide occurs under the action of bombardment by argon ions (Ar ⁺ ) when gaseous argon (Ar) is fed into a vacuum chamber containing all of the above elements.

The proposal is illustrated graphically.

Figure 1 is a cross section of the product.

Figure 2 - schematic representation of the process in accordance with paragraph 3 of the formula.

Figure 3 is a schematic representation of a process in accordance with claim 4 of the formula.

The positions in figure 1-3 are indicated:

1 - silicon wafer;

2 - a layer of silicon dioxide;

3 - adhesive layer of a mixture of nanocrystals of silicon dioxide and platinum;

4 - a layer of platinum;

5 - substrate holder;

6 - substrate *, including a silicon wafer 1 and a layer of silicon dioxide 2;

7 - magnetron;

8 - a platinum target placed on one of the magnetrons 7;

9 - a target of silicon dioxide placed on the second magnetron 7;

10 - a plate of silicon dioxide (for example, an annular), covering the part of the platinum target 8;

11 - annular zone of sputtering of a platinum target 8;

Ar ⁺ - argon ions bombarding a platinum target 8.

* Explanations:

Substrate 6 shown in FIGS. 2 and 3 includes the silicon wafer 1 indicated in FIG. 1 and a silicon dioxide layer 2.

In accordance with the explanatory graphic materials, the proposed set of essential features can be specified by the corresponding positions in the figures as follows.

A method of depositing platinum layers 4 on a silicon wafer 1 coated with a layer of silicon dioxide 2, comprising pre-forming an intermediate adhesive layer 3, wherein the adhesive layer is formed from a mixture of silicon dioxide and platinum nanocrystals,

wherein

- the adhesive layer 3 is formed by simultaneous magnetron sputtering from two magnetrons 7 with targets of platinum 8 and silicon dioxide 9, respectively;

- the adhesive layer 3 is formed by magnetron sputtering using a platinum target 8, part of the surface of which is covered by a silicon dioxide ring plate 10;

- after the formation of the adhesive layer 3 from the surface of the target 8 remove the annular plate of silicon dioxide 10, then apply the main part of the platinum layer 4 by magnetron sputtering of platinum;

- after the formation of the adhesive layer 3 by simultaneous magnetron sputtering of targets from platinum and silicon dioxide from two magnetrons 7, turn off the magnetron 7 with the target from silicon dioxide 9, then the main part 4 of the platinum layer is applied;

- the formation of the platinum layer 4 is carried out in a mixture of oxygen and an inert gas, for example (Ar) containing 5-50 volume% oxygen,

The implementation of the proposed method is as follows.

Two magnetrons 7 with a platinum target 8 and a target of silicon dioxide 9 (Fig. 2) or a magnetron 7 (Fig. 3) with a target of platinum 8, part of which is closed by an annular plate of silicon dioxide 10, are placed in a sealed, evacuated volume.

A mixture of oxygen and an inert gas, for example (Ar), is supplied into the indicated volume, an alternating and / or constant voltage U is supplied to the magnetron.

Simultaneous spraying of platinum and silicon dioxide forms an adhesive layer 3, consisting of a mixture of platinum nanocrystals and silicon dioxide.

After the formation of the adhesive layer, the magnetron 7 with the target of silicon dioxide 9 is turned off and the main part of the platinum layer is formed (4).

In the case of using one magnetron (Fig. 3), after the formation of the adhesive layer 3, the magnetron is turned off, the vacuum chamber is depressurized, it is filled with air, the silicon dioxide ring plate is removed from the surface of the platinum target 10. Then the chamber is sealed and vacuum, the magnetron 7 is turned on and formation is carried out the main part of the platinum layer 4.

After the formation of the platinum layer, the vacuum chamber is depressurized, filled with air, the substrate 6 with the formed platinum layer 4 is removed from the substrate holder 5, and the obtained platinum layer is stabilized annealed.

Stabilizing annealing is carried out in air using the following mode. The substrate with a layer of platinum is placed in an oven, the temperature is increased at a speed of not more than 200 ° C per hour to 650 ° C, then it is kept at 650 ° C for at least 2 hours until the platinum oxides are completely decomposed, then the temperature is increased at a speed of no more 200 ° C per hour to 850 ° C, after which it is kept at 850 ° C for 2 hours and the temperature is lowered to room temperature at a rate of not more than 200 ° C per hour.

Information sources

1. J. H. Kim, J. S. Sung, A. A. Vasiliev, et al. Propane / butane semiconductor gas sensor with low power consumption. Sensors and Actuators, B 44, 1997, p. 452.

2. M.Graf, A.Gurlo, N. Barsan, U. Weimar, A. Hierlemann. Microfabricated gas sensor systems with sensitive nanocrystalline metal-oxide films. Journal of Nanoparticle Research, 2006, 8, p. 823.

3. S. M. Lee, D. C. Dyer, J. W. Gardner. Design and optimization of a high temperature silicon micro-hotplate for nanoporous palladium pellistors. Microelectronics Journal, v. 34, 2003, p. 115.

4. C. Ducsa, M. Adam, P. Furjes, et al. Explosion-proof monitoring of hydrocarbons by mechanically stabilized, integrable calorimetric microsensors. Sensors and Actuators B 95, 2003, p. 189.

5. P. Furjes, Cs. Ducso, I. Barsony et al. Thermal characterization of a direction dependent flow sensor. Sensors and Actuators, A, 115, 2004, p. 417.

6. Vasiliev A.A., Pisliakov A.V., Zen M., et al. Membrane - type Gas Sensor with Thick Film Sensing Layer: Optimization of Heat Losses. Eurosensors XIV, Denmark, 2000, p. 379.

7. D.Vincenzi, M.A. Butturi, V. Guidi, M.C. et al. Development of a low-power thick-film gas sensor deposited by screen-printing technique onto a micromachined hotplate. Sensors and Actuators, B 77, 2001, p. 95.

8. AAVasiliev, RGPavelko, X. Vilanova, et al. Micromachined thermocatalytic gas sensor with improved selectivity based on Pd / Pt doped YSZ material. 11 ^th International Meeting on Chemical Sensors, Brescia, Italy, 2006, p. 127.

Claims (8)

1. A method of applying platinum layers on a substrate, comprising pre-forming on the surface of silicon oxide and / or silicon nitride an intermediate adhesive layer, characterized in that the adhesive layer is formed from a mixture of silicon dioxide and platinum nanocrystals. 2. The method according to claim 1, characterized in that the adhesive layer is formed with a thickness of 1-30 nm. 3. The method according to claim 1, characterized in that the adhesive layer is formed by simultaneous magnetron sputtering from two magnetrons with targets of platinum and silicon dioxide, respectively. 4. The method according to claim 1, characterized in that the adhesive layer is formed by magnetron sputtering using a platinum target, part of the surface of which is covered with a silicon dioxide plate. 5. The method according to claim 4, characterized in that after the formation of the adhesive layer from the surface of the target remove the plate of silicon dioxide and then apply the bulk of the platinum layer by magnetron sputtering of platinum. 6. The method according to claim 3, characterized in that after the formation of the adhesive layer with targets of platinum and silicon dioxide, respectively, turn off the magnetron with the target of silicon dioxide and then carry out the application of the main part of the platinum layer. 7. The method according to claim 1, characterized in that the platinum layer is applied in a mixture of oxygen and an inert gas containing 5-50 vol.% Oxygen. 8. The method according to claim 5 or 6, characterized in that after applying the main part of the platinum layer, stabilizing annealing is performed, for which the substrate is placed in an oven, the temperature is increased at a speed of not more than 200 ° C per hour to 650 ° C, then it is held at 650 ° C for at least 2 hours until the platinum oxides are completely decomposed, then the temperature is increased at a rate of no more than 200 ° C per hour to 850 ° C, then it is kept at 850 ° C for 2 hours and the temperature is lowered to room temperature at a rate of not more than 200 ° C per hour.

Images