EP4193378A1

EP4193378A1 - Coating apparatus and coating method having divided pulses

Info

Publication number: EP4193378A1
Application number: EP21777418.1A
Authority: EP
Inventors: Walter May; Werner Kölker; Stephan Bolz
Original assignee: Cemecon AG
Current assignee: Cemecon AG
Priority date: 2020-09-15
Filing date: 2021-09-06
Publication date: 2023-06-14
Also published as: WO2022058193A1; CN116438623A; DE102020124032A1; JP2023540624A; US20230340658A1

Abstract

The invention relates to a coating method and to a coating apparatus for coating a body (40). A magnetron-cathode (22a, 22b, 22c, 22d) having a target (24a, 24b, 24c, 24h) is arranged in the vacuum chamber (12). Electrical power is supplied to the magnetron-cathode (22a, 22b, 22c, 22d) such that a plasma is generated and the target (24a, 24b, 24c, 24h) is atomized in order to deposit a coating (44) on the body (40). The electrical power is periodically supplied within a period duration T according to the HIPIMS method as a cathode pulse (60), wherein each cathode pulse (60) comprises at least two cathode sub-pulses (62) and an intermediate cathode sub-pulse break (64). In order to be able to deposit coatings (44) having favorable properties, using the chopped HIP IMS method in a particularly favorable manner, a bias voltage is applied to the substrate (40) to be coated with bias voltage pulses (66), wherein each bias voltage pulse (66) comprises at least two bias sub-pulses (68) and an intermediate bias sub-pulse break (70).

Description

description

Split pulse coating apparatus and method

The invention relates to coating methods and coating devices. In particular, the invention relates to methods and devices for coating a body by means of cathode sputtering, electrical power being supplied to at least one cathode in accordance with the HIPIMS method.

It is known to provide bodies or parts of bodies with a surface coating to improve the mechanical or chemical properties. It is known, in particular for tools and components that are subject to wear and tear, to provide functional surfaces with coatings.

Hard material layers in particular are known as coatings. In addition to CVD processes, PVD coating processes in particular are known for the formation of thin coatings, in particular the processes of cathode sputtering.

In the field of cathode sputtering, the HIPIMS process (High Power Impulse Magnetron Sputtering) is of particular interest. In contrast to conventional cathode sputtering, in which the cathodes are operated with constant electrical power, e.g. using direct current, with HIPIMS short electrical pulses of high voltage are applied to the cathode and very high peak powers are achieved. The result is a significantly increased ionization of the plasma compared to conventional cathode sputtering, which results in advantageous properties of the coatings produced with it.

In current further developments of the HIPIMS method, it has been proposed to subdivide the electrical pulses supplied to the cathodes, ie to apply a sequence of shorter pulses instead of a continuous cathode pulse, which are referred to here as cathode sub-pulses are denoted and in their sequence form a divided cathode pulse. This method is known as "chopped HIPIMS" or also as DOMS (deep oscillation magnetron sputtering).

For example, Barker et al, "Modified high power impulse magnetron sputtering process for increased deposition rate of titanium", Journal of Vacuum Science & Technology A: Vacuum, Surfaces and Films 31, 060604 (2013) discloses the deposition of titanium layers using a modified HIPIMS method in which HIPIMS pulses are divided into a pulse sequence. Pulse sequences with a duration of 100 ps are applied at a frequency of 200 Hz and an average power of 0.75 kW in various sequences with 4 or 2 micropulses of different durations and with different switch-off times. A significant effect of the split HIPIMS pulses on the deposition rate is reported, which is above that of conventional HIPIMS methods.

Barker et al, "An investigation of c-HiPIMPS discharges during titanium deposition", Surface & Coatings Technology 258 (2014) 631 - 638 report on titanium layers deposited by means of chopped HIPIMS using pulse sequences with 4 or 8 micropulses. A significantly increased deposition rate is reported with a longer delay between micropulses, which is explained by the ionization of the plasma.

EP 2587518 Ai discloses a device for producing at least essentially hydrogen-free ta-C layers on substrates (workpieces) made of metal or ceramic materials. A vacuum chamber is connected to a vacuum pump and to a source of inert gas. A support device is provided for the substrates (workpieces). At least one graphite cathode with an associated magnet assembly forming a magnetron serves as a source of carbon material. A bias power supply is used to apply a negative bias to the substrate. A cathode power supply is connected to the graphite cathode and an associated anode and is adapted to transmit high power pulse sequences at (preferably programmable) time intervals. Each high-power pulse sequence comprises a series of high-frequency direct current pulses adapted to be applied to the at least one graphite cathode, optionally after a build-up phase. EP 3457428 Ai discloses a method and apparatus for processing a substrate in a semiconductor processing system. The method begins with the initialization of a pulse lock controller coupled between a pulse RF bias generator and a HIPIM generator. A first timing signal is sent from the pulse synchronization controller to the pulse RF bias generator and the HIPIM generator. A sputtering target and an RF electrode arranged in a substrate support are excited based on the first timing signal. The target and electrode are turned off based on an end of the timing signal. A second timing signal is sent from the pulse synchronization controller to the pulse RF bias generator and the electrode is energized and de-energized without energizing the target in response to the second timing signal.

US 2008/0135400 Ai discloses a device for sputtering a target to produce a coating on a substrate. The device includes a magnetron having a cathode and an anode. A power supply is connected to the magnetron and at least one capacitor is connected to the power supply. The device also includes an inductor operatively connected to the capacitor. A first switch operatively connects the power supply to the magnetron to charge the magnetron according to a first pulse. A second switch is connected to discharge the magnetron according to a second pulse.

It can be regarded as an object to propose methods and devices with which coatings with advantageous properties can be deposited in a particularly favorable manner using the chopped HIPIMS method.

To achieve the object, a coating method according to claim 1 and a coating device according to claim 19 are proposed, in which a bias voltage with a special pulse shape is applied to the substrate to be coated.

Dependent claims relate to respective advantageous embodiments of the invention.

According to the invention, a coating method or a coating device proposed for coating a body, also referred to as a substrate. The substrate can in particular be a tool. A large number of individual bodies are usually coated at the same time.

According to the invention, a vacuum chamber is provided with at least one magnetron cathode with a target, in which the body (substrate) is arranged. Applying electrical power to the magnetron cathode generates a plasma, sputters the target and deposits a coating on the body by depositing components of the plasma on the substrate surface. According to the HIPIMS method, the electrical power is supplied in pulses in the form of cathode pulses, periodically with a period T. According to the chopped HIP IMS variant of the method, each cathode pulse is divided into at least two cathode sub-pulses, between which there is a cathode sub-pulse pause, in which no or an electrical power that is significantly reduced compared to the cathode subpulses is applied. The cathode pulses thus form sequences of cathode sub-pulses spaced apart in time.

The coating device according to the invention comprises a vacuum chamber with a receptacle for the body and with at least one magnetron cathode with a target and an electrical cathode power supply and a control device. The cathode power supply is designed to supply electrical power to the magnetron cathode according to the HIPIMS method, so that cathode pulses are supplied periodically within the period T for generating the plasma and sputtering the target. The control device is designed to control the electrical cathode power supply so that individual, some or all parameters of the electrical power supplied to the magnetron cathode can be controlled, in particular power, frequency, pulse shape and/or pulse sequence, etc. The control device is designed for driving the cathode electrical power supply according to such a sequence that each cathode pulse applied to the cathode comprises at least two cathode sub-pulses and a cathode sub-pulse pause therebetween. The control device can in particular be a programmable controller with which, in addition to the cathode power supply, other functions of the coating device can also be controlled in a time-dependent manner according to a coating program, in particular a bias power supply and/or the supply of process and/or reactive gases. The advantage of the sequences of cathode sub-pulses provided according to the chopped HIPIMS method is that the risk of electric arc discharges forming is reduced or arc discharges which form are terminated by a subsequent sub-pulse pause. In addition, an even higher ionization of the plasma can be achieved through the sequences of cathode sub-pulses than with continuous HIPIMS cathode pulses.

The cathode pulses are applied as voltage pulses, so that the following descriptions and information on duration, pulse shape, sequence, etc. always refer to the voltage at the magnetron cathode. The resulting current and thus the electrical power depend on the conditions in the plasma, in particular the ionization.

The cathode pulses can include different sequences of cathode sub-pulses. These are preferably voltage pulses. Sequences of, for example, 2-10 cathode sub-pulses per cathode pulse, preferably 3-8 cathode sub-pulses, more preferably 4-6 cathode sub-pulses, have proven to be advantageous. The duration of the cathode sub-pulses can generally be, for example, 4-80 ps, preferably 5-50 ps or 6-35 ps, particularly preferably 8-25 ps. As will be explained below, the cathode sub-pulses can always have the same duration, although sequences within which cathode sub-pulses have different durations are preferred.

Within the respective sequences of cathode subpulses there are respective cathode subpulse pauses between the cathode subpulses. The duration of the cathode sub-pulse pause(s) is preferably shorter on average than the duration of the cathode sub-pulses. For example, the duration of the cathode sub-pulse pause(s) can be 2-30 ps, preferably 5-25 ps, particularly preferably 8-20 ps. The duration of the cathode sub-pulse pauses within the sequence can be the same or different.

According to the invention, a bias voltage is applied to the body in the form of bias voltage pulses which are periodic within the period. Each bias voltage pulse is subdivided, ie comprises at least two bias sub-pulses and one in between lying bias subpulse pause.

The bias voltage applied to the substrate is negative, so that positive ions of the plasma are accelerated towards the substrate surface. In contrast to the known use of a DC voltage, the bias voltage according to the invention is applied in pulse form, also with sequences of individual bias sub-pulses, each with bias sub-pulse pause(s) in between. The frequency or period of the pulse train of the bias voltage is the same as that of the electrical power or voltage at the cathode.

Such a specification of a time profile of the bias voltage that is synchronized with the time profile of the voltage at the cathode results in particular in control options with regard to a selection of the components of the plasma that form the coating. As has been shown, the composition of the ions in the plasma is time-dependent when the cathode is operated in the chopped HIP IMS process, i.e. with cathode sub-pulses spaced apart in time, i.e. different ions are present in the plasma at different points in time. Matching the time profile of the bias voltage to this results in the possibility of a targeted selection of types of ions that are integrated into the layer.

This relates in particular to the possibility of influencing the respective amount of metal and gas ions in the structure of the coating through suitable timing. It could thus be observed that the amount of gas ions contained in the plasma, in particular ions of the process gas, for example argon, has a time profile that differs from the amount of metal ions from the start of a cathode sub-pulse. By suitably selecting the duration and start time of the bias sub-pulses relative to the cathode sub-pulses, the respective amount of gas and metal ions in the coating can be adjusted in a targeted manner. In particular, the amount of argon used as process gas has a decisive influence on the properties of the coating: It has been shown that with comparatively high argon proportions, coatings can be achieved with very high hardness and high residual stresses, ie hard and brittle layers, while Layers without argon or with a lower proportion are rather ductile in comparison. The use according to the invention of a “chopped bias”, ie a subdivided bias voltage, also supports an advantage of the chopped HIPIMS method, namely the suppression of arc discharges (Ares). By subdividing the cathode pulses into at least two cathode sub-pulses with a cathode sub-pulse pause in between, the formation of arc discharges is inhibited - due to the power being switched off in the cathode sub-pulse pause, any arc that may occur cannot fully develop and thus causes hardly any damage to the system (e.g . Targets) and especially on the coating of the body. It is assumed that this effect is additionally supported if the bias voltage pulses are also subdivided and have at least one bias sub-pulse pause per period. In the bias sub-pulse pause, the electrical power directly on the body is also reduced or completely switched off, so that the formation of arcs that may be incipient is inhibited and extinguished. This applies not only when the bias sub-pulse pauses overlap or are completely congruent with the cathode sub-pulse pauses, but also in the case of a time delay.

Advantageous developments of the invention relate in particular to the timing of bias voltage pulses and cathode pulses or of bias sub-pulses and cathode sub-pulses.

At least one of the bias sub-pulses preferably starts with a delay time after the start of one of the cathode sub-pulses, more preferably several or all bias sub-pulses can each start with a delay time after the start of an associated cathode sub-pulse. For example, the delay times of several or all bias sub-pulses can be the same for a respective associated cathode sub-pulse.

The respective sequences of bias sub-pulses and cathode sub-pulses can be coordinated with one another in terms of time, but the sequences do not have to be congruent and can also have, for example, a different number and duration of the respective sub-pulses. For example, the time interval between two bias sub-pulses can correspond to the time interval between two cathode sub-pulses, at least essentially (ie deviations preferably less than +/-10%). The time interval should preferably be measured between the respective start of bias subpulses and cathode subpulses. senior The bias sub-pulses and cathode sub-pulses considered in the comparison can follow one another directly in time, or further bias sub-pulses or cathode sub-pulses can be arranged in between. One, several or all bias sub-pulse pauses can preferably have at least essentially the same length as the respective associated cathode pulse pauses.

The number of bias sub-pulses can be equal to or different from the number of cathode sub-pulses. In some preferred embodiments, the number of bias sub-pulses may be less than the number of cathode sub-pulses, i.e. there is not an associated bias sub-pulse for each of the cathode sub-pulses. For example, bias sub-pulses can only be provided specifically associated with cathode sub-pulses of higher power, while no associated bias sub-pulse is provided for one or more cathode sub-pulses of lower power.

The duration of the bias sub-pulses and/or the bias sub-pulse pauses can preferably be in the same interval as the duration of the cathode pulses or cathode sub-pulse pauses. The delay time between the start of a cathode sub-pulse and the start of an associated bias sub-pulse can be, for example, 5-200 ps, preferably 10-150 ps, particularly preferably 10-60 ps.

It is possible that the bias voltage pulses at least partially overlap with the cathode pulses, i.e. at least part of each bias voltage pulse can be present at the same time as at least part of one of the cathode pulses. However, this is not necessary in all cases; in some applications it has also proven to be advantageous to delay the bias voltage pulses in relation to the cathode pulses to such an extent that there is actually no longer any temporal overlapping.

The sequences of cathode sub-pulses and bias sub-pulses can be coordinated with one another; in particular, each cathode sub-pulse can be assigned a bias sub-pulse, which preferably begins with a time delay and, for example, has at least essentially the same duration. Thus, the number of bias sub-pulses can be equal to the number of cathode sub-pulses. However, it can also be preferred if the number of bias sub-pulses is less than the number of cathode sub-pulses, so that, for example, not every cathode sub-pulse but only some of the cathode sub-pulses can be assigned a bias sub-pulse of, for example, at least essentially the same duration, for example with a time delay in each case. This can be particularly preferred in the case of cathode sub-pulses of different durations: If, for example, a rather low peak power is achieved in short cathode sub-pulses, it can be advantageous not to provide an assigned bias sub-pulse for this and thus to accelerate the ones occurring at the lower peak power ions on the body to be coated.

According to a preferred embodiment, the development of the plasma over time and the formation and course of different types of ions can be considered for the respective sequence of cathode sub-pulses in order to specify an optimized sequence of bias sub-pulses. For example, gas and metal ions can be considered separately and a sequence of bias subpulses can be selected in such a way that the desired types of ions are preferably used to form the coating. These are particularly preferably metal ions. For example, it can be observed that during and after the cathode pulses there is a time-dependent course of the occurrence of different types of ions, namely gas and metal ions. If one considers a type of metal ion (whereby different types of metal ions differ from each other in terms of the respective metal and different types can also be distinguished by the degree of ionization), then in many cases a time course can be determined for this that has a maximum or several has maxima. It is then preferred to select the sequence of bias sub-pulses in such a way that at least one of the bias sub-pulses is present during at least one maximum. If the time curve has a number of maxima, two or more bias subpulses are preferably applied in such a way that they are present during at least one of the maxima. More preferably, a sub-pulse pause can be arranged during a minimum of the time profile lying between two maxima for at least one type of metal ions. It is thus possible to select suitable sequences of bias sub-pulses in such a way that the desired types of ions are selected in accordance with the timing of the occurrence of metal ions and are used in preference to other ions, in particular gas ions, for the build-up of the coating.

The invention can be supplemented and developed in various aspects, e.g. differentiated by an aspect relating to advantageous sequences of cathode subpulses. Licher duration, an aspect relating to chopped HIPIMS method with a short period or high frequency and an aspect relating to an advantageous system technology with a capacity and a charging device. While each of these aspects offers individual advantages on its own, the combination of two or more of the aforementioned aspects proves to be particularly advantageous.

According to one aspect as a development of the invention, at least two of the cathode sub-pulses differ in terms of their duration. The cathode pulses include at least a first and a second cathode sub-pulse, each lasting at least 8 ps. The first and the second cathode sub-pulse differ in their duration.

The terms "first" and "second" cathode sub-pulse are generally to be understood as meaning that the second cathode sub-pulse within the sequence of cathode sub-pulses occurs later than the first cathode sub-pulse, without the first and second cathode sub-pulses necessarily having to follow one another directly and regardless of whether the first cathode sub-pulse is the first cathode sub-pulse in time within the sequence of cathode sub-pulses. However, said first cathode sub-pulse can preferably also be the first, i.e. earliest, cathode sub-pulse within the sequence and said second cathode sub-pulse can be the cathode sub-pulse within the sequence immediately following the subsequent cathode sub-pulse pause.

The plasma can be influenced in a targeted manner by such a variation in the duration of the cathode sub-pulses. Because although the applied voltage is at or near zero in the cathode subpulse pause separating the cathode subpulses, it has surprisingly been shown that the plasma behaves differently in successive cathode subpulses, for example due to remaining ionization lasting over the cathode subpulse pause. The proposed variation in the duration of the cathode sub-pulses makes it possible to use such effects in the plasma and, in particular, to specify optimized sequences of cathode sub-pulses.

For example, a pre-ionization can be achieved by one or more shorter cathode sub-pulses and in one or more temporally - directly or at a distance - subsequent, longer cathode subpulses, the pre-ionization can be used to achieve a particularly high peak power.

According to a preferred embodiment, said first cathode sub-pulse is the first cathode sub-pulse in time of the cathode pulse, i.e. the sequence of cathode sub-pulses. Said second cathode sub-pulse follows later in time within the sequence, it being possible for it to follow directly on the first cathode sub-pulse or for further cathode sub-pulses to be arranged chronologically between the first and second cathode sub-pulses. The first cathode sub-pulse is preferably shorter than the second cathode sub-pulse. For example, the duration of the second cathode sub-pulse can be no% to 600% of the duration of the first cathode sub-pulse, preferably 150% to 400%, particularly preferably 200%-300%. With a sequence of such cathode sub-pulses, a particularly high peak power can be achieved during the second cathode sub-pulse.

In preferred embodiments, the second cathode sub-pulse can last for example at least 15 ps, preferably at least 20 ps, particularly preferably at least 25 ps. The first cathode sub-pulse can last for example a maximum of 25 ps, preferably a maximum of 20 ps, particularly preferably a maximum of 15 ps.

In sequences with more than two cathode sub-pulses, it is preferred that the pulse durations of consecutive cathode sub-pulses increase monotonically, i.e. always last the same or longer in the time sequence from one cathode sub-pulse to the next. According to a preferred embodiment, the cathode pulse can include a third cathode sub-pulse, which follows the first and the second cathode sub-pulse and lasts at least as long as the first and at least as long as the second cathode sub-pulse.

According to one aspect as a development of the invention, the chopped HIPIMS method is carried out with an unusually short period of at most 1.5 ms, ie an unusually high frequency of the pulses at the magnetron cathode of at least 667 Hz. Conventional implementations of the HIPIMS method assume a significantly lower frequency or longer period. The inventors have recognized that the chopped HIPIMS method with greatly increased peak powers seeing that the short period durations are also possible and advantageous. Even shorter period durations of preferably 1.25 ms or less (corresponds to a frequency of at least 800 Hz) or 1 ms or less (corresponds to a frequency of at least 1 kHz) are possible. In some implementations even shorter periods of 0.2 - 0.6 ms (1.7 - 5 kHz) are applicable.

At the thus unusually high frequencies, the power is released within a very short period of time, but with an exceptionally high electrical peak power, which results in high ionization. The main advantage of the high frequency is a high deposition rate, i.e. faster layer build-up compared to lower frequencies. This makes a decisive contribution to the economics of the process and the device when used industrially.

The cathode pulse itself, i.e. the total duration of the sequence of cathode sub-pulses and thus the time during which electrical power is present in the form of cathode sub-pulses, preferably takes up only a small part of the period, e.g. less than half the period and particularly preferably a maximum of one third the period duration. This is particularly advantageous when the cathode pulses are fed from a charged capacitance. The capacitance is thus discharged during the individual cathode sub-pulses over the entire duration of the cathode pulses and is recharged during the remaining period duration. By limiting the duration of the cathode pulses, sufficient time can be made available for charging the capacities. The duration of the cathode pulse is preferably in the range of 30 ps-400 ps, particularly preferably 8ops-3oops.

According to one aspect as a development of the invention, the electrical power, which is supplied to the cathode in the form of the cathode sub-pulses, is provided from at least one charged capacitance of a cathode power supply. The cathode power supply comprises a capacitor, preferably provided in the form of a plurality of individual capacitors connected in parallel (capacitor bank), and a charging device therefor.

Such a structure has proven to be particularly suitable for electrical power in accordance with to provide the chopped HIPIMS method. The entire electrical power of all cathode subpulses can preferably be supplied from the same capacity, the electrical connection to the cathode being closed by means of a switch during the duration of the cathode subpulses and disconnected during the cathode subpulse pause(s) and during the remaining period duration. The switch can preferably be controlled by the control device, for example it can be an IGBT. The charging device can supply electrical power to the capacitor at least during the period outside of the cathode pulses in order to charge it, preferably also during the cathode sub-pulse pause(s). Particularly preferably, the charging device can remain permanently connected to the capacitor, so that it is connected in parallel, for example, during the cathode sub-pulses.

According to a preferred embodiment, the charging device can be controlled with constant power, i.e. designed such that the time-averaged power supplied to the capacitor (or during the pulse, if necessary, to a parallel connection of capacitor and chamber) is set to a fixed value. This has proven to be particularly stable compared to alternative concepts such as setting a fixed voltage to which the capacitance is charged.

The various aspects and developments of the invention are each individually advantageous, but especially in combination. Some preferred embodiments are given below, which can be used with each individual aspect, but also with combinations of any two, three or all four aspects.

According to a development, the cathode pulse is a voltage pulse with a peak value of 600-1200 V. While the individual values can vary depending on the design, this range of the voltage value has proven to be preferred.

The cathode sub-pulses can have different curves and pulse shapes, for example as triangular pulses. The cathode pulses are preferably at least essentially rectangular pulses or trapezoidal pulses, ie they have a time profile with a steep rising and falling flank and an approximately linear profile in between. In particular in the preferred case of feeding the cathode pulses from a charged capacitance, the course of a discharge curve corresponds on closer inspection, although such a short section of the discharge curve is selected so that the course of a linear course between an initial voltage (after the steep rising edge) and an end voltage (before the steep falling edge) preferably no longer deviates by more than 20%, particularly preferably by no more than 10% . It has proven to be advantageous that the progression between the initial voltage and the final voltage is at least essentially constant, which is understood here to mean that the voltage value does not change by more than preferably a maximum of 25%, particularly preferably a maximum of 20%. Such pulses are essentially viewed here as square-wave pulses.

In preferred embodiments, the peak power of at least one of the cathode sub-pulses is at least 50 kW. More preferably, significantly higher peak powers can be achieved, for example more than 100 kW, 200 kW or higher, particularly preferably 300 kW or higher. It has been shown that the level of the maximum peak power of cathode sub-pulses has a significant influence on the ionization of metals in the target material. Studies show, for example, that with titanium target material the ratio of ionized to non-ionized particles (Ti+/Ti) at maximum peak power of a cathode sub-pulse of almost 500 kW is almost twice as high as when the highest peak power of the cathode sub-pulse is 150 kW .

The peak power achieved during the cathode sub-pulses preferably increases within the time sequence of the cathode pulse. According to a preferred embodiment, the cathode pulse can, for example, comprise at least a first cathode sub-pulse and a second, chronologically subsequent cathode sub-pulse. The second cathode sub-pulse can immediately follow the first cathode sub-pulse (after an intervening cathode sub-pulse pause), or further cathode sub-pulses can be arranged in between. It is then preferred that the peak power during the second cathode sub-pulse is at least 30% higher than during the first cathode sub-pulse. In this way, particularly high peak power can be achieved overall.

The invention can be advantageously used with a wide variety of target materials and combinations of materials. The target can include both metallic and non-metallic materials. At least one of the components of the target is preferably selected from the group comprising the materials from groups 4 - 6 of the periodic table tems as well as boron, carbon, silicon, yttrium and aluminum. All components of the target are preferably selected from the group mentioned. With the method according to the invention and the device according to the invention, coatings can be produced from different material systems, which are formed by the target components and, if necessary, components that are supplied in gaseous form. In particular, in addition to a process gas, a reactive gas such as nitrogen, a carbon-containing gas or oxygen can also be used.

The bias voltage applied to the body to be coated is preferably pulsed with bias voltage pulses which are synchronous with the cathode pulses, i.e. are applied with the same frequency and a fixed phase relationship. It is preferred that their position and duration is selected in such a way that specific types of ions are selected. The position and/or duration can be selected depending on a profile of the occurrence of at least one type of metal ion during and after the cathode pulses, which has at least one maximum. It is then preferred that the bias pulse is present during this maximum.

Embodiments of the invention are described in more detail below with reference to drawings. show:

FIG. 1 shows a schematic representation of a plan view of a coating system as an embodiment of the invention; FIG.

2 shows an indexable insert as the body to be coated;

3 shows a schematic sectional view of a coating on a substrate material;

4 is a circuit diagram of one embodiment of a HIPIMS power supply;

5a, 5b diagrammatically illustrated time curves of a cathode voltage and a bias voltage over time;

6a, 6b measured curves over time of cathode voltage and cathode current as well as quantity and types of ions for a first exemplary embodiment;

7a, 7b measured time curves of cathode voltage and cathode current as well as quantity and types of ions for a second embodiment game;

8a, 8b measured time curves of cathode voltage and cathode current as well as bias voltage and bias current according to a third embodiment;

9a, 9b measured time curves of cathode voltage and cathode current as well as bias voltage and bias current according to a fourth embodiment;

10a, 10b show diagrams of measured time curves of electrical quantities according to a fifth and sixth example;

na, nb diagrams of measured time curves of electrical variables according to a seventh and eighth example;

12 shows a schematic representation of the alignment of a body to be coated with respect to a target, a cathode;

13a, 13b SEM images of the fracture morphology of a coating on the flank face and rake face of a body according to a comparative example;

14a, 14b SEM images of the fracture morphology of a coating on the flank face and rake face of a body according to an embodiment.

In Fig. 1 components of an embodiment of a PVD coating system 10 are shown schematically. An interior 20 of a vacuum chamber 12 can be evacuated through a vent 14 to create a vacuum. A process gas, preferably a noble gas or a mixture of different noble gases, for example argon and/or krypton, can be admitted through an access 16 . A reactive gas such as nitrogen can be admitted through an access 18 . In alternative embodiments, the inlets 16, 18 can also be replaced by a common inlet for a process gas and a reactive gas.

In an exemplary configuration, four magnetron cathodes 22a, 22b, 22c, 22d are arranged in the interior 20 of the vacuum chamber 12, each with plate-shaped sputtering targets 24a, 24b, 24c, 24d. The magnetron cathodes 24a, 24b, 24c, 24d are each connected to a controllable electrical HIPIMS power supply 26a, 26b, 26c, 26d, by means of which an electrical voltage relative to the electrically conductive wall of the Vacuum chamber 12 can be created, as will be explained in detail below.

The arrangement of four magnetron cathodes 22a, 22b, 22c, 22d shown in FIG. 1 and their interconnection are to be understood as examples. In alternative embodiments, other electrode configurations can be provided, e.g. only one magnetron cathode 22, two, three or more than four. In addition or as an alternative to the magnetron cathodes 22a, 22b, 22c, 22d shown, which are each connected to HIPIMS power supplies 26a, 26b, 26c, 2öd, cathodes of other types can also be provided, e.g. DC magnetron cathodes which connected to DC power supplies. The magnetron cathodes can alternatively be electrically connected differently to the wall of the vacuum chamber 12 as shown by way of example, for example to a separate anode (not shown).

The power supplies 26a, 26b, 26c, 20d are shown only schematically in FIG. 4 shows a simplified circuit diagram of a HIPIMS power supply 26a. This comprises, as a capacitance 48, a capacitor bank made up of a number of capacitors connected in parallel, a charging device 46 for the capacitance 48 and a controllable switch 36, which is preferably in the form of an IGBT.

In the illustrated embodiment, the charging device 46 is a power regulated voltage source connected in parallel with the capacitor 48 . Capacitor 48 is connected through switch 36 to output terminals 28 of HIPIMS power supply 26a. With the switch 36 open, the charging device 46 charges the capacitance 48 . With the switch closed, the charge stored in the capacitor 48 is provided at the output terminals 28, i.e. fed to the magnetron cathode 22a (Fig. 1), in addition to which the parallel connected charger 46 also provides power. The switch 36 is controlled by a control 50 of the system 10 via the control 38 .

The magnetron cathodes 22a, 22b, 22c, 22d are aligned with their sputtering targets 24a, 24b, 24c, 24d in the direction of the center of the vacuum chamber 12. There is a rotatable substrate table 30 on which a number of rotatable substrate carriers 32 for carrying substrates 40, ie bodies to be coated, are arranged. In the example shown, indexable inserts 40 are charged as substrates, as shown in FIG. the Substrates 40 consist of a substrate material 42, for example in the case of the indexable inserts 40 of WC/Co sintered material. This is merely an example of a substrate 40 to be coated; alternatively, components or tools of different shapes and different materials can be coated.

The substrates 40 are electrically connected to the substrate table 30 via the substrate carriers 32 . A controllable bias power supply 34 is connected to the substrate table 30, by means of which an electrical bias voltage VB can be applied to the substrate table 30 and thus to the substrates 40 relative to the wall of the vacuum chamber 12. This results in an ion current in the plasma, which can be measured as a bias current IB at the connection of the substrate table. The connection of the bias power supply to the wall of the vacuum chamber is to be understood as an example; alternatively, this can also be connected to a separate anode.

The HIMPS power supplies 26a, 26b, 26c, 20d, bias power supply 34 and pumps (not shown) at the inlets and outlets 14, 16, 18 are each connected to the central controller 50 of the system 10. The central controller 50 is programmable so that all parameters of the pretreatment and coating processes taking place in the interior 20 of the vacuum chamber 12 are controlled by the controller 50 according to a fixed and stored time-dependent control program. The control program can be changed and the controller 50 can store several different control programs that can be called up selectively. As far as sequences and settings in the operation of the system 10 are referred to below, these are specified in a control program that is executed by the controller 50 .

During operation of the system 10--as specified by the respective control program running in the controller 50--a vacuum is first generated in the interior 20 of the vacuum chamber 12 and process gas, preferably argon, is admitted. Then the substrates 40 are pretreated by means of ion etching. During etching, the bias power supply 34 is driven to provide a high (negative) bias voltage VB which accelerates (positive) ions onto the substrate. Gas and/or metal ions can be used for etching and there are different processes possible.

In a preferred embodiment, first a gas ion etching takes place without the operation of the magnetron cathodes 20 in the HIPIMS operating mode. Gas ion etching can be carried out as DC etching by applying a bias voltage IB as a direct voltage, for example in the range from -100 to -400 V, or alternatively as MF etching (bias voltage IB -100 to -700V). In DC etching, the electrons are generated by a hollow cathode (not shown) and removed at an anode (not shown). The substrate table 30 lies between the anode and hollow cathode. In MF etching, an MF power supply (not shown) is responsible for generating the electrons, which in turn ionize the gas. These electrons are dissipated at the wall of the vacuum chamber 12 (operating ground).

After gas ion etching, metal ion etching can preferably be used to further improve adhesion. In this case, for example, one or two of the magnetron cathodes 22a-d (equipped, for example, with targets made of Cr, Ti, V) are operated in the HiPIMS operating mode with such a high peak power that the donor material is ionized. In addition, a bias voltage VB is applied either as a direct voltage (DC) or in a pulsed manner, pulses of the bias voltage VB being able to be synchronized with the cathode pulses. The bias voltage VB is preferably between -300 to -1200V.

A coating 44 (FIG. 3) is then deposited on the substrate surface thus pretreated. To do this, the HIMPS power supplies 26a, 26b, 26c, 26d and the bias power supply 34 are driven to provide appropriate bias voltage VB and cathode voltages Vc; Examples of this are explained below. The inlets 16, 18 are controlled to supply process gas (argon) and, if necessary, reactive gas (e.g. nitrogen). A plasma is thus generated in the interior 20 of the vacuum chamber 12, under which the targets 24a are sputtered. Positive ions of the plasma are accelerated towards the surface of the substrate materials 42 of the substrates 40 by the negative bias voltage VB and form the coating 44 there (FIG. 3).

In this case, the electric power from the HIMPS power supplies 26a, 26b, 26c, 2od to the magnetron cathodes 22a, 22b, 22c, 22d according to the HIPIMS method in periodic dic pulses supplied. The electrical power is supplied according to the chopped HIPIMS method, e.g. as shown schematically in Fig. 5a, 5b for the course of the voltage Vci at the first magnetron cathode 22a, in the form of cathode pulses 60 with a pulse duration P, which in turn are subdivided into - In the example shown in Figure 5a three and in Fig. 5b four - cathode subpulses 62 with pulse durations Pi, P ₂ , P ₃ , P4. Between the cathode sub-pulses 62 there are cathode sub-pulse pauses 64 with durations Zi, Z ₂ Z ₃ , in which no voltage is present. The cathode pulses 60 are periodic with a frequency f or a period T. After the end of the cathode pulses 60 there is no voltage for the remainder of the period T either.

In the example of FIG. 5a, the sequence of cathode pulses 60 comprises three cathode sub-pulses 62 each having the same duration Pi, P ₂ , P ₃ . In the example of FIG. 5b another sequence of cathode sub-pulses 62a, 62b is shown with initially two shorter cathode sub-pulses 62a and then two longer cathode sub-pulses 62b.

The pulse sequences of the voltage Vci at the magnetron cathode shown in FIGS. 5a, 5b are specified by the controller 50 by driving the HIPIMS power supplies 20a-d, specifically by suitably driving the switches 36 in the HIPIMS power supplies 20a-d. It should be pointed out here that the representation in FIGS. 5a, 5b is, for the sake of simplicity, idealized rectangular pulse shapes, while actual voltage curves, as discussed below and shown separately, deviate from this. In particular, the real pulse shapes show a certain drop over the duration of the cathode sub-pulses in the form of a discharge curve of the capacitance 48, but with a certain recharging by the charging device 46 during the cathode sub-pulse pauses 64.

The bias voltage VB can have different time curves during the coating period, depending on the design. In particular, the bias voltage VB can be applied as a DC voltage or in a pulsed manner. Examples of preferred embodiments are shown schematically in Figures 5a, 5b.

In the example of FIG. 5a, the bias voltage VB is applied in the form of three bias pulses 66 with a total duration B, the bias pulses 66 having the same frequency or period. the duration T are periodic like the cathode pulses 6o. Like the cathode pulses 6o, the bias pulses 66 are also subdivided, namely into bias sub-pulses 68 of the respective duration Bi, B ₂ , B ₃ with bias sub-pulse pauses 70 of the respective duration ZBI, ZB ₂ in between. The sequence of bias sub-pulses 68 within each bias pulse 66 corresponds to the sequence of cathode sub-pulses 64 within each cathode pulse 60, ie each cathode sub-pulse 64 is assigned a bias sub-pulse 68 of the same duration, but offset in time, namely delayed by a delay time TD.

In the example of FIG. 5b, the bias voltage VB is applied in the form of only two bias sub-pulses 66 within each period duration T. FIG. The sequence of the bias sub-pulses 66 corresponds only in part to the sequence of the cathode sub-pulses 64a, 64b, since a bias sub-pulse 68 is not assigned to each cathode sub-pulse 64: no bias sub-pulse is assigned to the two first, shorter cathode sub-pulses 64a, while the longer, third and fourth cathode sub-pulses associated bias subpulses 68 of the same duration, each delayed by a delay time TD, follow.

The pulse sequences of the bias voltage VB shown in FIGS. 5a, 5b are also shown in an idealized form. Like the pulse sequences of the cathode voltage Vci, they are specified by the controller 50 by suitably driving the bias power supply 34 . This also includes a controllable switch (not shown) through which, depending on the design, either a capacitor bank is selectively connected to the output, as is the case with the HIPIMS power supply 26a, or alternatively a direct voltage source is connected directly.

Using a first example (FIGS. 6a, 6b) and a second example (FIGS. 7a, 7b), it is shown below how the composition of the coating 44 is influenced by suitable synchronization of the bias pulses 66 and bias sub-pulses 68 with the cathode pulses 60 and cathode sub-pulses 62 can:

In the first example, Figure 6b shows an example of a magnetron cathode with a target made of titanium and with the supply of nitrogen as a reactive gas and argon as a process gas for some types of ions formed in the plasma, the respective time profile after application of a cathode pulse 60 with the in Fig. 6a with two cathode subpulses 62 with a duration Pi, P ₂ of 20 ps each with a cathode subpulse pause 64 with a duration Zi of 30 ps between the cathode subpulses 62.

As shown, the plasma contains various species of metal and gas ions, although the respective time courses are only shown for four types of ions: Ti ⁺ , Ti ²⁺ , N ₂ ⁺ and N ⁺ . These show different time courses. Thus, the Ti ⁺ metal ions show a pronounced maximum at about 120 ps after the start of the cathode pulse 60. This is particularly remarkable since the total duration P of the cathode pulse 60 is only 70 ns.

As shown schematically, a period of time 72 can be defined in which the Ti ⁺ metal ions predominate.

The positive gas and metal ions of the plasma are accelerated by the negative bias voltage VB in the direction of the surface of the substrate material 42 and thus become part of the coating 44 that is deposited there. In the case of a DC bias, ie a continuous DC voltage as the bias Voltage VB, all ions for the layer structure are selected without distinction. With a pulsed timing of the bias voltage VB, which is synchronous with the timing of the voltage Vc at the magnetron cathodes 22a-d, as shown in Fig. 5a, 5b, by a suitable choice of the timing synchronization, ie the delay time TD and the Duration of the respective bias sub-pulses 68, a selection can be made from the gas and metal ions present in the plasma at different points in time.

For example, in the first example of FIGS. 6a, 6b, the bias voltage VB can only be applied during the time period 72 in which Ti ⁺ ions reach a maximum. In this case, the bias voltage VB can be applied in the form of continuous bias pulses 66 that are not divided into bias sub-pulses 68 .

In the second example of FIGS. 7a, 7b, the conditions are shown according to the representation of the first example, also for a magnetron cathode 22a with a target made of titanium and with the supply of nitrogen as the reactive gas and argon as the process gas. As shown in Fig. 7a, the cathode pulse 60 in the second example comprises two cathode sub-pulses 62 of different durations Pi=10 ps and P ₂ =20 ps with a cathode sub-pulse pause 64 a duration Zi of 18 ps between the cathode sub-pulses 62.

As shown in FIG. 7b, the sequence of cathode sub-pulses 62 shown in FIG. 7a shows time profiles of the different types of ions Ti ⁺ , Ti ²⁺ , N ₂ ⁺ and N ⁺ which deviate from one another. The time courses differ from the time courses shown in FIG. 6b for the sequence of cathode sub-pulses 62 shown in FIG. ^6a . In FIG intervening minimum approx. at t=50ps. Here, too, the second, higher maximum after the end of the cathode pulse is at t=48ps.

In contrast to the first example according to FIG. 6b, two separate time periods 72 can thus be defined in FIG. 7b, in which the Ti ⁺ metal ions predominate.

Accordingly, in the second example of Fig. 7a, 7b, the bias voltage Vßbspw. are applied in the form of two bias sub-pulses 68, which are present during the time segments 72, with a bias sub-pulse pause 70 in the range of approximately 40-70 ps. A high proportion of Ti ⁺ ions in the coating 44 that forms is thus ensured by the targeted selection of a suitable sequence of bias subpulses 68 .

Thus, by means of the temporal synchronization (duration Bi, B ₂ , ... B _n of the bias sub-pulses 68 and the start time of the bias sub-pulses 68, e.g. calculated from the beginning of the cathode pulse 60, or in the case of the assignment of cathode sub-pulses 62 and bias sub-pulses 68 by the respective delay time TD) the composition of the coating 44 is specified and the layer properties are significantly influenced. In particular when considering the proportion of the argon process gas in the coating 44 (although the time course of Ar ions is not shown in the examples in FIGS. 6b, 7b), the properties can be suitable, in particular with regard to internal stresses in the coating 44 and its hardness be specified. A high proportion of argon results in a coating 44 with high internal stresses and high hardness. A coating 44 with a lower argon content tends to be ductile and has significantly lower residual stresses.

Further examples of pulse sequences for cathode pulses 60 or sequences of cathode sub-pulses 62 are shown and explained below. In the third example (FIGS. 8a, 8b) and in the fourth example (FIGS. 9a, 9b), four magnetron cathodes 22a, 22b, 22c, 22d are operated in the HIPIMS method with a frequency of 2 kHz, ie at T=5 oops. The electrical power averaged over time is 12 kW per cathode, ie a total of 48 kW. The magnetron cathodes 22a-d are equipped with targets 24a-d made of titanium, aluminum and silicon, for example two titanium-silicon targets and two titanium-aluminum targets. Argon is used as the process gas and nitrogen as the reactive gas.

The layer 44 deposited as the substrate 40 on indexable inserts is in each case an Al-Ti-Si-N layer.

The third and fourth examples differ by different sequences of cathode sub-pulses 62 (FIGS. 8a, 9a) and different sequences of bias sub-pulses 68 (FIGS. 8b, 9b).

In the third example, the sequence of cathode sub-pulses 62 includes three cathode sub-pulses 62, the third of which is longer than the first two. The respective durations of the cathode sub-pulses 62 and cathode sub-pulse pauses 64 are

Pi= 30ps,

Zi=25hp,

Z ₂ =30ps, resulting in a pulse length P of l ops within the period T of 500 LIS.

The pulse shape of the cathode sub-pulses 62 is essentially rectangular with a voltage Vci of approx. 680 V, which drops slightly over the course of the pulses is considered to be rectangular. As shown, the cathode current Ici ramps up, respectively, to about 75 A during the first cathode sub-pulse 62, to about 145 A during the second cathode sub-pulse 62, and to a peak value of about 220 A during the third cathode sub-pulse 62. It is thus proved favorable that the third cathode sub-pulse 62 is significantly longer than the first and the second cathode sub-pulse 62, so that the peak power achieved there is approximately 120 kW.

As shown in FIG. 8b, the associated sequence of the bias voltage VB has three bias sub-pulses 68, the durations of which essentially correspond to those of the respective associated cathode sub-pulses 62, which, however, are present with a delay time TD of approximately 20 ps.

The resulting bias current IB reaches a peak value of approximately 15 A during the first bias sub-pulse 68, a peak value of approximately 33 A during the second bias sub-pulse 68 and a peak value of over 60 A during the third, longest bias sub-pulse 68, which shows that a high number of ions are accelerated to the substrate 42.

In the third example, the coating 44 grows at a deposition rate of 1.9 μm/h.

As a measure of the ionization, the proportion of ionized titanium atoms relative to the number of non-ionized titanium atoms in the plasma can be measured using OES. In the third example, Ti ⁺ /Ti has a very high ionization ratio of about 1.2. A comparative example with non-divided HIPIMS pulses under otherwise identical conditions only achieves a ratio of 0.93 here.

In the fourth example, the sequence of cathode sub-pulses 62 also includes three cathode sub-pulses 62, as in the third example, of which the first is even shorter and the third is even longer compared to the third example, resulting in a sequence of increasing cathode sub-pulse durations:

Pi= 20pS,

P2 = _25pS ,

P ₃ = 70 hp, Zi=25hp,

Z ₂ =30ps.

Here, too, the total pulse length P is t ops . As in the third example, the pulse shapes are essentially rectangular (with a somewhat greater drop during the third cathode sub-pulse 62). The cathode current Ici, which rises in a ramp-like manner, reaches only approx. 25 A during the first cathode sub-pulse 62, approx. 100 A during the second cathode sub-pulse 62 and approx achieved.

The bias voltage VB also has three bias sub-pulses 68 in the fourth example (Fig. 9b), the durations of bias sub-pulses 68 and cathode sub-pulses 62 as well as bias sub-pulse pauses 70 and cathode sub-pulse pauses 64 being essentially the same, although the bias sub-pulses 68 have a delay time TD of approx. 25ps are delayed. The bias current IB reaches a peak value of only about 5 A during the first bias sub-pulse 68, a peak value of about 25 A during the second bias sub-pulse 68 and a peak value of about 65 A during the third bias sub-pulse 68. The deposition rate is also 1, 9 um/h as in the third example.

In the fourth example, a Ti ⁺ /Ti ratio of 1.57 results, and thus again increased ionization compared to the third example.

The coating 44 produced in this way on the substrate material 42 is shown in FIGS. 14a, 14b in comparison to a coating 44 (FIGS. 13a, 13b) according to a comparative example with a non-divided HIPIMS pulse, but otherwise the same conditions. The coatings 44 are each shown on a free surface 52, which, as sketched in FIG.

14a shows the coating 44 on the free surface 52 of the body 40, produced by the special chopped HIPIMS method with the sequence of cathode sub-pulses 62 according to FIGS. 9a, 9b. Relative to the comparative example shown in Fig. 13a, the coating shows 44 a much finer structure.

Also on the rake face 54, which is not aligned in the direction of the target 24b (FIG. 12), the coating 44, as can be seen from FIG. 14b, shows a significantly finer structure than the comparative example (FIG. 13b). It is also evident that in the comparative example according to FIG. 13b, the coating 44 has an oblique growth direction, shown here by a white arrow as an example, which is not orthogonal to the substrate surface but obliquely aligned in the direction of the target 24b, while as can be seen in FIG. 14b In the example of the chopped HIP IMS method with a special sequence of cathode sub-pulses 62 according to FIGS. 9a, 9b, no such oblique growth direction can be seen, but the coating 44 can be seen growing orthogonally to the substrate surface.

It is believed that this smoother growth, as well as the finer structure, is a result of the higher ionization.

The coatings 44 created according to the fourth exemplary embodiment show particularly favorable properties, in particular on a coated tool. In machining tests with indexable inserts, which were coated on the one hand according to the above comparative example (continuous, undivided HIPIMS pulses) and on the other hand with the sequence of cathode sub-pulses 62 and bias sub-pulses 68 according to the fourth example above, a 6 pm thick coating showed 44 for the fourth Example a 20% longer tool life when machining X6CrNiMoTii7-i2-2 compared to the comparative example.

Further examples of sequences of cathode sub-pulses 62 and bias sub-pulses 68 are shown below.

Thus, first of all, in a comparison of a fifth example according to FIG. 10a and a sixth example according to FIG. 10b, different types of synchronization of the time profile of the bias voltage VB are shown. In both examples, the bias voltage VB is divided into bias sub-pulses 68 that are synchronous with the cathode sub-pulses 62 . However, in the fifth example (Fig. 10a) the synchronization is without delay time (TD = o ps), while in the sixth example (Fig. 10b) there is a delay time (TD = 10 ps): 10a shows a measured profile of the voltage Vci at the first magnetron cathode 22a, the current Ici to the first magnetron cathode 22a, the bias voltage VB and the bias current IB for a sequence of two consecutive cathode subpulses 62 with an intermediate cathode sub-pulse pause 64.

The frequency is 1000 Hz (T=i ms). The pulse durations of the cathode sub-pulses 62 and cathode sub-pulse pause 64 are Pi=20ps, P ₂ =20ps, Zi=iops. The pulse duration of the cathode pulse 60 is 50 ps (corresponds to 5% of the period duration T). The bias voltage VB is periodic with the same frequency and includes two bias sub-pulses 66 and a bias sub-pulse pause 70 with Bi=20ps, B ₂ = 20ps, ZBi=iops, which are synchronous in time to the cathode sub-pulses 62 with the same duration and without delay (TD = o ps ).

In the example of FIG. 10a, the charging device 46 of the HIPIMS power supply 26a is regulated to a constant power of 3000 W. The capacitance 48 of the HIPIMS power supply 26a is charged to approximately 700 volts at the beginning of the first cathode sub-pulse 62 . As shown, the cathode sub-pulses 62 have a substantially rectangular shape with little overshoot and a slight drop in voltage Vci over the duration of the cathode sub-pulses 62.

During the first cathode sub-pulse 62, the cathode current Ici shows a curve that rises slightly to a peak value of approximately 150 A, while during the second cathode sub-pulse 62 it rises significantly more sharply to a peak value of approximately 300 A. Thus, in the first cathode sub-pulse 62 a peak power value (peak power) of approximately 100 kW is achieved, and in the second cathode sub-pulse 62 a peak power of approximately 200 kW is achieved. This results in high ionization of the plasma and, in particular during the second bias sub-pulse 62, in a large number of charge carriers in the plasma that hit the substrate 40, which can be read from the significantly increasing bias current IB.

The example of FIG. 10a thus shows a fifth exemplary embodiment with cathode sub-pulses 62 of the same length and exactly (without delay) synchronized bias voltage VB of the same pulse sequence. The relatively high peak power achieved and the significant increase in the currents Ici and IB during the second sub-pulses 62, 68 are demonstrated here. For comparison with FIG. 10a, FIG. 10b shows a delay time TD of 10 ps between the beginning of the respective cathode sub-pulse 62 and bias sub-pulse 68 as a sixth exemplary embodiment with otherwise the same parameters Bias subpulse 68 increases faster because more ions are already available.

Different sequences of cathode sub-pulses 62 are shown below in a comparison of a seventh example according to FIG. 11a and an eighth example according to FIG. 11b. In both examples, sequences of five cathode sub-pulses 62 are applied. However, in the seventh example (FIG. 11a), the first cathode subpulses 62 in time are shorter than the later cathode subpulses 62, while in the eighth example (FIG. 11b), the earlier cathode subpulses 62 are longer than the later cathode subpulses 62. The two examples are characterized by the following parameters:

In FIGS. 8a, 8b, the bias voltage VB is in each case periodically with the same frequency and pulse sequence synchronous in time with the cathode sub-pulses 62 with the same duration and without a delay time (TD=0 ps).

In the example of FIG. na, only low peak powers of 27 kW and 68 kW initially occur during the first two cathode sub-pulses 62 . During the subsequent, longer cathode sub-pulses 62, however, the peak powers then rise sharply to 179 kW and 190 kW, whereupon only 86 kW are achieved in the last cathode sub-pulse 62.

In the example of FIG. 11b, the peak power increases only during the first three cathode sub-pulses 62 with values of 64 kW and 156 kW, while the peak power is lower in subsequent cathode sub-pulses. Overall, therefore, lower peak power values are achieved than in the example of FIG. 11a.

Thus, sequences with initially shorter cathode sub-pulses 62 and subsequently longer cathode sub-pulses 62 have proven to be favorable.

Claims

- 32 - Claims

1. Coating method for coating a body (40), in which the body (40) is arranged in a vacuum chamber (12), wherein at least one magnetron cathode (22a, 22b, 22c, 22d) with a target (24a, 24b, 24c, 24d) is arranged in the vacuum chamber, the magnetron cathode (22a, 22b, 22c, 22d) is supplied with electrical power so that a plasma is generated and the target (24a, 24b, 24c, 24d) is sputtered to form a Deposit coating on the body (40), wherein the electrical power according to the HIPIMS method as cathode pulses (60) is supplied periodically within a period (T), each cathode pulse (60) having at least two cathode sub-pulses (62) and one in between cathode sub-pulse pause (64), wherein bias voltage pulses (66) are applied to the body periodically within the period duration (T), each bias voltage pulse (66) comprising at least two bias sub-pulses (68) and an intervening bias sub-pulse pause (70).

2. The coating method of claim 1, wherein at least one of the bias sub-pulses (68) begins after the start of one of the cathode sub-pulses (62) with a delay time (TD).

3. Coating method according to claim 1 or 2, wherein the number of bias sub-pulses (68) is equal to or less than the number of cathode sub-pulses (62).

4. Coating method according to one of the preceding claims, wherein metal ions are formed in the plasma by the cathode sub-pulses (62), wherein the amount of metal ions of one type has a time profile from the start of a cathode pulse (60) which has at least one maximum, and wherein at least a bias subpulse (68) is applied during the maximum. - 33 -

5. Coating method according to one of the preceding claims, in which the cathode pulses (60) comprise at least a first and a second cathode sub-pulse (62), the first and the second cathode sub-pulse (62) differing in their duration and the first and the second Cathode sub-pulse (62) each last at least 8 ps.

6. Coating method according to claim 5, wherein the first cathode sub-pulse (62) is the first cathode sub-pulse (62) of the cathode pulse (60), and wherein the first cathode sub-pulse (62) is shorter than the second cathode sub-pulse (62).

7. The coating method according to claim 6, wherein the second cathode sub-pulse (62) lasts at least 15 ps, preferably 20 ps, particularly preferably 25 ps.

8. The coating method according to claim 6 or 7, wherein the first cathode sub-pulse (62) lasts a maximum of 25 ps, preferably a maximum of 20 ps, particularly preferably a maximum of 15 ps.

9. The coating method according to any one of claims 5 - 8, wherein the cathode pulse (60) comprises a third cathode sub-pulse (62) which follows the first and the second cathode sub-pulse in time, the third cathode sub-pulse (62) being at least as long as the first and at least as long as the second cathode sub-pulse (62) lasts.

10. The coating method as claimed in one of the preceding claims, in which the cathode pulses (60) are periodic within a period (T), the period (T) being a maximum of 1.5 ms.

11. The coating method according to claim 10, wherein the cathode pulse (60) lasts less than half the period (T), preferably a maximum of one third of the period (T).

12. Coating method according to one of the preceding claims, in which the cathode power supply (26a, 26b, 26c, 2öd) comprises a charged capacitance (48) and a charging device (46) therefor, so that the magnetron cathode (22a, 22b, 22c, 22d) supplied electric power from the charged Capacity (48) is provided. Coating method according to Claim 12, in which the charging device (46) is regulated to have a constant power. A coating method according to any one of the preceding claims, wherein the cathode pulse (60) is a voltage pulse having a peak value of 600V to 1200V. Coating method according to one of the preceding claims, wherein at least one of the cathode sub-pulses (62) is at least essentially a rectangular pulse or at least has a trapezoidal profile. A coating method according to any one of the preceding claims, wherein the peak power of at least one of the cathode sub-pulses (62) reaches at least 50kW. Coating method according to one of the preceding claims, wherein the cathode pulse (60) comprises at least a first and a second cathode sub-pulse (62), wherein the second cathode sub-pulse (62) follows the first cathode sub-pulse (62) in time, and wherein the peak value of the the electrical power supplied to the magnetron cathode (22a, 22b, 22c, 22d) is at least 30% higher during the second cathode sub-pulse (62) than during the first cathode sub-pulse (62). Coating method according to one of the preceding claims, wherein the target (24a, 24b, 24c, 24d) comprises materials from groups 4-6 of the periodic table, boron, carbon, silicon, yttrium and/or aluminum. Coating device for coating a body (40), having a vacuum chamber (12) with a receptacle (30, 32) for the body (40) and at least one magnetron cathode (22a, 22b, 22c, 22d) having a target (24a, 24b, 24c, 24d), a cathode power supply (26a, 26b, 26c, 26d) for supplying electrical power to the magnetron cathode ( 22a, 22b, 22c, 22d) according to the HIPIMS method as cathode pulses (60) periodically within a period (T) for generating a plasma and sputtering the target (24a, 24b, 24c, 24d), a control device (50), the is designed to control the cathode power supply (26a, 26b, 26c, 2öd) for supplying electrical power to the magnetron cathode (22a, 22b, 22c, 22d) in such a way that each cathode pulse (60) has at least two cathode sub-pulses (62) and an intermediate cathode sub-pulse pause (64), and a bias power supply (34) for supplying electrical power to the body (40) in the form of bias voltage pulses (68), the control device (50) being further developed to control the bias power supply (34) so that at the Body (40) bias voltage pulses (66) are applied periodically within the period (T), each bias voltage pulse (66) comprising at least two bias subpulses (68) and an intervening bias subpulse pause (70). Coating device according to claim 19, wherein the cathode pulses (60) comprise a first and a second cathode sub-pulse (62), wherein the first and the second cathode sub-pulse (62) differ in their duration and wherein the first and the second cathode sub-pulse (62) respectively last at least 8ps. Coating device according to Claim 19 or 20, in which the cathode pulses (60) are periodic within a period (T), the period (T) being a maximum of 1.5 ms. Coating apparatus according to any of claims 19 - 21, wherein the cathode power supply (26a, 26b, 26c, 26d) comprises a charged capacitance (48) and a charging device (46) therefor such that that of the magnetron - 36 -

Cathode (22a, 22b, 22c, 22d) supplied electrical power from the charged capacitance (48) is provided.