© EDP Sciences, 2023

1 Introduction

Main technological part of magnetron sputtering (MS) is a target that serves as a source for film/coating material. During operation erosion zone, a ring-shaped groove is formed on the target surface. Over time the groove broadens somewhat and becomes deeper. As soon as the thickness of the target in the groove becomes ∼15–20% of its initial thickness, the target is replaced with another one. Since almost the entire flux of film-forming particles (∼85 to 90%) is formed by the material sputtered from the groove, it is of interest to examine the composition and microstructure of the groove's bottom and side wall surfaces. Actually, just properties of these surfaces change that inevitably results on the properties of the films deposited from the target.

In principle, the target composition may change during its exploitation, especially if the target has complex composition, e.g., multicomponent alloy. However, the existing ideas about the mechanism of alloys sputtering assume that their composition remains unchanged during operation [1,2]. At that there exist original papers where experimentally proved the existence of the effect of preferential sputtering of one of the target components [3,4]. In view of the told above the questions arise − what really occurs with the composition and structure of the multicomponent alloy target during long term of its exploitation. There is huge amount of works devoted to various investigations of the influence of ion bombardment (IB) on the target surface. In these works the targets surface was modified by irradiation of various types of ions, starting from H⁺ and finishing Hg⁺. But practically in all these investigations the energy of ions varied in the range of 10–80 KeV [4–6]. While there are very few works is devoted to investigations of the structure of surfaces subjected to bombardment with ions with energies peculiar to MS processes (∼400 to 1000 eV) [7,8]. At the same time, the expectation that the surface microrelief formed under the action of ions with high energy can strongly differ from that created by ions with much smaller energies, is quite natural. Actually, the energy of ions with low energy dissipates within small subsurface volume, while the energy of high energetic ions dissipates within much larger volume. As a result, we can expect the appearance of completely different effects in the composition and microrelief of the target compared to those observed during the bombardment of targets with high-energy ions [4–6]. From the above consideration follows the goal of our work − to investigate the composition and microstructure of the grooves of the targets fabricated from CoCrCuFeNi, CoCrCuFeMnNi, AlCoCrCuFeNiV high entropy multicomponent alloys (HEAs) both before and after their service term.

2 Experimental

Targets (discs 55 mm in diam and 4 mm in thick) were made by turning from ingots smelted by vacuum-arc melting in a high-purity argon atmosphere. The ingots were remelted 5–6 times to homogenize the composition after what they were cooled at a rate of ∼50 K/s.

The technological conditions at what the target sputtering and film depositions were carried out are as follows. Working gas (Ar) pressure P_Ar = 1 × 10⁻² ÷ 5 × 10⁻³ torr; discharge current I_d = 0.1 ÷ 0.5 A; discharge voltage U_d = 400 ÷ 700 V; groove (race track) square S ≈ 5–7 cm²; q_in ≈ 10 W × cm⁻² for moderate and q_in ≈ 50 W × cm⁻² for strong IB, where q_in is the flux density of ions incident on the groove. The service life of each of the targets depended on their composition and ranged from 32 to 40 hours.

Chemical composition of the targets and coatings was studied by electron energy loss spectroscopy (EELS), and X-ray photon spectroscopy (XPS) analyses using JSM-6U90LV and CAMEBAX S300 scanning microscopes equipped with attachment for chemical analysis. Note that we did not find any noticeable difference in the results obtained by both methods. The microstructure of the samples was studied by the scanning electron microscopy (SEM) in two modes − scanning electron image (SEI) and backscattered electron image (BEI) using the facility above. The targets were utilized for the deposition of HEA films with correspondent composition. Important to note that during operation the targets were cooling as usual in such type processes. For further details of the films properties and technology of their deposition we refer the reader to [9,10].

3 Results

Before presenting the results of our study, let us consider how the form of the target changed during its operation. An annular groove (sometimes they call it a race-track) appears on the target surface during its operation, see Figure 1. The groove appears due to focusing of ions by circular magnetic field formed by permanent magnets located in the magnetron unit under the target. The shape and dimensions of the magnetic field define the diameter and the width of the groove. The cross section of the groove is a rounded angle, the sides of which are the side walls of the groove, and its bottom is located between them, Figure 1. With time the groove becomes deeper and some wider and the target should be changed when the deepness of the groove reaches ∼3.5 mm for the targets of 4 mm in thick.

An estimation of the groove's area at the end of the lifetime of the target shows that it is ∼25% of the total target surface area. At that, the target thickness in close vicinity of the groove decreases not more than 0.5 mm, while its thickness away from the groove remains practically unchanged. This means that these areas are practically not subjected to ion bombardment (IB), and almost ∼75% of the total ion current falls on the groove. That is, the ion current is distributed over the target area very non-uniformly and its maximum falls on the bottom of the groove, its side walls are subjected to moderate IB, while the remaining target area is almost not under the IB. As a consequence, maximum changes experience the structure and composition of the bottom of the groove and its side walls, while the properties of the rest part of the target remain practically unchanged. Obviously, it is the composition and microstructure of the walls and bottom of the groove are of interest for research, since they are the sources of the sputtered material for the film. Therefore, we have studied in detail the composition and microstructure of the bottom and sides of the groove, and everything that will be discussed in the article refers specifically to these parts of the groove.

Now let us turn to the results of investigation. As it follows from the table in Figure 1 the composition of the target at the end of its service (the composition of the bottom of the groove) has become noticeably depleted in copper. To check how this effect affected the composition of the films deposited at the beginning and at the end of the target lifetime, the composition of these films was studied, which results are presented in Table 1. The films obtained both at the beginning of the target's service life and at the end of it are enriched with copper, while the tendency to decrease in the concentration of copper both in the target and in the films remains, Table 1. At the same time, the concentration of the rest elements changes not as much both in the target and in the films.

Initially surface microstructure of all investigated targets does not fundamentally differ. The general views of microstructures and compositions are presented in Figure 2 and in the tables on sides. Results of detailed studies of microstructures and compositions of these targets are presented in Figure 3 and in the tables. The main particularity of the initial microstructure of targets is that they consist of two parts − a matrix (dark background) and inclusions (light strips of various shapes and widths) imbedded in the matrix, Figure 2.

Initially chemical composition of the targets on macroscopic level was uniform and close to the equiatomic (tables in Fig. 2). At the same time, the composition of these targets on microscopic level is quite non-uniform. Namely, the composition of inclusions and matrices in these targets strongly differ (tables in Fig. 3). As it is seen from the tables, the inclusions mainly consist of Cu and Al (or Cu and Mn for another target) while the composition of the matrix in both targets mainly consists of the base elements Fe, Co, Cr, Ni. Emphasize that the described effects are characteristic of both the composition and microstructure of all studied targets prior to their operation.

The microstructures of the grooves surfaces subjected to IB with different intensity are shown in Figures 4 and 5. The microrelief of the side walls of the grooves subjected to moderate IB characterized by cellular structure crossed in various directions by trenches, Figures 4a–4c. In Figure 4a two areas are clearly distinguished. The upper left corner is the side wall of the groove, while the lower right corner is the groove's bottom, Figure 5. The side wall of the groove is formed from a cellular structure with the shape of cells close to rectangular.

The side wall of the groove merging with its bottom is formed by cells with smoothed edges. At the very bottom of the groove the cells already disappear, and the structure formed instead resembles the waves solidified during the movement, Figure 5.

Another important observation follows from a comparison of the initial target microstructure, Figure 6a, with that appeared at moderate IB, Figure 6b. Analyzing these micrographs one can note similarity in size, shapes and distribution of inclusions observed on the original surface of the target (Fig. 6a) and trenches formed on the side walls of the groove under moderate IB (Fig. 6b).

The microreliefs of the bottoms of the grooves formed under strong IB are presented in Figures 7a–7c. These surfaces resemble the surface of the bottom of the groove in Figure 5 and remind the view of incoming and overlapping soft waves solidified during the movement.

Fig. 1 General view of AlCoCrCuFeNiV target sputtered during 40 h. The initial composition of the target and at the end of its service (bottom of the groove) are presented in the table on the right.

Fig. 2 Microstructure of initial surfaces of AlCoCrCuFeNiV and CoCrCuFeMnNi targets (BEI mode). Compositions in macro-level are presented in tables on the sides of micrographs.

Fig. 3 Enlarged images of microstructures of AlCoCrCuFeNiV and CoCrCuFeMnNi targets (BEI mode). Corresponding compositions of matrices and inclusions presented in tables on the sides of micrographs.

Fig. 4 Microstructure of the side walls of the grooves (areas of moderate IB): (а) CoCrCuFeNi; (b) AlCoCrCuFeNiV; (c) AlCoCrCuFeMnNiV alloys.

Fig. 5 Enlarged image of the microstructure of the groove. The smooth merging of the side wall of the groove (upper left corner) with its bottom (lower right corner) is clearly seen. CoCrCuFeNi alloy.

Fig. 6 Microstructure of the surface of the same target: (a) Initial surface of the target; (b) Side wall of the target (moderate IB) after the service term. AlCoCrCuFeNiV alloy.

Fig. 7 Microstructure of the bottoms of the grooves (area of strong IB): (а) CoCrCuFeNi; (b) AlCoCrCuFeNiV; (c) AlCoCrCuFeMnNiV alloys.

4 Discussion

4.1 Composition

The effect of alteration of the target composition during its exploitation is somewhat unexpected in view of widespread conception that the sputtering of alloys occurs without compositional changes [1,2]. Accepting the mentioned conception we should solve the raised contradiction between this conception and our results demonstrating clear changes that have occurred in the composition of multicomponent (HEA) targets during their service term. Note that the same effect we observed for all investigated targets.

The key hint for understanding the reason for the revealed effect is the inhomogeneity of the composition and microstructure of the initial target surface. Initially smooth and plain target surface looked like a matrix containing basic Co, Cr, Fe, Ni elements presented in all target compositions, and micro-inclusions highly enriched in Cu, tables in Figure 3. Under long term of IB initially smooth microstructure turned to highly developed microrelief in the groove of the target with clearly observing trenches which dispositions coincide with those of inclusions, Figure 6. Remind that the origin of the inclusions is the absence of the chemical bonding (positive enthalpy of mixing) between Cu and other components of the alloy target, Table 2.

This occurs because despite of the decreasing of copper concentration in the target with time its sputtering occurs at the same rate which is highest compared to other target components. As a result, the films deposited from such target will be always enriched in copper.

Another fact that must be taken into account is that the sputtering yield of Cu is about 1.5–2 times higher compared to other components of the alloy, Table 3. Accounting for the high rate of sputtering of inclusions mostly composed of copper weakly bound with the matrix, it becomes clear that instead of easily sputtered inclusions the trenches should easily form. Just this effect we observe looking at the microstructure of the side walls of the targets grooves (areas of moderate IB), Figure 4. Based on this reasoning it is possible to explain continuous decrease of the copper content in the target during operation. Since the distribution of inclusions enriched in copper is uniform over the entire target's volume, they present on the target surface during the whole service term of the target. It means that Cu will be continuously sputtered from the target surface along with other alloy components. However because the sputtering rate of Cu is higher compared to other elements, the target will continuously depleting in Cu. As a result, its concentration will be lower at the end of the target's service term than it was at the beginning. In the same manner the copper concentration changes in films deposited from the target during its service term. It decreases from ∼13 at% at the beginning of the term and finishing at ∼9 at% at its end, Table 1. This is expecting effect, because the lower the copper content in the target, the lower it will be in the film. At that, the concentration of copper in the films is always somewhat higher than that in the target throughout its entire service life, Table 1.

The effect of preferential sputtering in alloy targets revealed in our experiments has quite different nature compared to well-known effect of preferential sputtering of some elements from the alloy or compound [3,4]. The latter effect we observe considering the composition of the film deposited under severe IB (ion energy, E_i = 200 eV), Table 1. Because of intensive ion bombardment this film becomes strongly depleted in copper due to its preferential sputtering from the growing film. This is a good example of known phenomenon of preferential sputtering of certain elements from alloys. Although in the case of the IB of a growing film this effect has important specific features considered in detail in references [9,10].

Concluding this subsection we note that new type of preferential sputtering effect has been revealed. The effect is especially pronounced for the alloys containing: (i) the second phase in form of micro-inclusions consisting of component(s) with high positive enthalpy of mixing with other alloy components and (ii) which sputtering yield(s) noticeably higher than that of the other alloy components.

4.2 Microstructure

Heterogeneity of the microstructure of high entropy alloys (HEA) composed of matrices with embedded inclusions was noted in several references [12–14]. The reason for such heterogeneity of HEAs containing Co, Cr, Fe, Ni as a matrix, and Cu + (Mn or Al) as a base for the inclusions is the bonding energy between alloy atoms. Using method of secondary ion mass spectrometry (SIMS) it was shown, that minimal bonding energy in (CoCrCuFeMn)_14.3Ni_28.6 alloy possess Mn and Cu atoms [15]. Another approach to evaluate bonding energy is to compare the enthalpies of mixing that release during interaction of the alloy atoms (the heat released during chemical interaction of the elements). If it is positive, it means that atoms do not interact with each other, and vice versa. In Table 2 [11] the enthalpy of mixing for atomic pairs of the alloy elements is presented. Accounting for the data from Table 2, we see that copper does not form any alloys (besides aluminum) with other alloy components. Close to this the behavior is the manganese. At the same time the enthalpy of mixing between pairs of such alloy components as Cr, Co, Fe, Ni, is highly negative. This means that these elements strongly bound with each other. Therefore, it is natural to expect that the microstructure of alloys of this type will contain two separate phases, one of which exists in the form of inclusions containing “foreign” elements, and the second is a matrix containing components “related” to each other, just what we observe in our targets.

To explain the reasons for changes of the target microstructure we remind that the distribution of the ion current across the groove's square is not uniform. As a result the microrelief of the groove should be also non-uniform. This effect is clearly observed in Figures 8a, 8b where the microrelief of the side wall and the bottom of the groove formed on AlCoCrCuFeNiV target is presented.

The microrelief of the bottom of the groove (strong IB zone) is formed of densely packed round-like shape elements, merging due to increased fluidity and wettability, Figure 8a. Such features indicate to that the temperature in the bottom of the groove was high enough for the surface melting. Another sign of the surface melting are the microreliefs in Figures 7a–7c, which micrographs were taken from the bottoms of grooves formed on the targets of different alloys. The microrelief in these micrographs resembles melted and frozen waves rolling on one another. This is another proof that during ion bombardment the temperature in the target groove is so high that surface melting of the alloy occurs. Direct measurements of the temperature of targets of various metals during their magnetron sputtering [16] confirm these observations.

Let us now consider the reasons for formation of unusual elements (micro-figures) that form the microrelief of the surface of the groove under the ion bombardment.

The processes occurring on the surface subjected to the IB are thermally-stimulated mechanical stresses, increased mobility of dislocations, acceleration of recrystallization processes, variations in composition of near-surface layers. It is expected that these processes should result in changes in the microrelief of surfaces of various materials [4–8]. Discussing the mechanisms of formation and evolution of cones, pyramids, etch pits on the ion irradiated surfaces, authors of [17] conclude that the reasons for these effects are not yet fully clear. Emphasize that majority of the works devoted to the above problems dealt with high energy ion fluxes (tens of keV). At the same time, in very few works the processes of modification of the microstructure of the targets occurring under bombardment by ions with energies peculiar to magnetron sputtering (∼400 ÷ 1000 eV) are considered.

At that one may assume that the processes occurring in a target irradiated by ions with substantially different energies should be noticeably different. The energy of high-energy ions is dissipated deeply inside the target. Whereas low-energetic ions lose their energy within a near-surface layer which average thickness is about several atomic layers, that is, within ∼20 to 50 Å [16].

The energy of ions of this range is dissipated in two major processes − in heating of the near-surface layer, and in sputtering. Most of the energy of ions dissipates as a heat that results in temperature rise on the irradiated surface and formation of a liquid-like layer on it [16]. At these conditions a microrelief forms mainly by the processes of: (i) sputtering, (ii) recrystallization both on the surface and in the near-surface layer, and by (iii) heat-stimulated transport of atoms.

The sputtering in our case manifests in the form of preferential sputtering. The latter process affects both on local changes in the composition and on the microstructure − a specific microrelief in the form of trenches is formed on the target surface. The heating of the near-surface layer by IB stimulates processes of recrystallization and heating-enhanced transport of atoms by diffusion. These effects cause the nucleation and growth of elements of various shapes on the irradiated surface, Figures 8a, 8b. Clearly, the shapes of the elements forming the microrelief depend on the target material. This is observed in Figures 4 and 7 where the microreliefs of multicomponent alloys with different compositions are presented.

The reasons for this effect may be as follows. Since the nearest environment of atoms in different alloys is different it is clear that the energy of interatomic bonding is also different. At that, it is obvious that the interatomic bonding energy directly influences the melting temperature, sputtering yields and other thermo-physical parameters of this alloy. That in turn, affects the shape of the elements forming the microrelief of the target surface.

To demonstrate the role of material properties in formation of the relief both on micro- and macro-levels we present Figures 9 and 10. In these photos the microrelief of the surface of the side wall of the groove formed under moderate IB and the macrorelief of the earth surface created by atmospheric effects are presented. As has been shown the microrelief formed by the ion bombardment is a result of three major factors − heating, sputtering and physical-chemical properties of an alloy. The weathering figures have appeared on the earth due to the action of several factors − wind, temperature, humidity, and due to specific properties of the rocks from which the figures are composed.

Thus, it is clear that the weathering figures in Figure 9b are composed of hard rocks, while that in Figure 10b are composed of the mixture of rocks with different hardness. Obviously, the empty cells in the weathering figures in Figure 10b were filled with rocks with lower density or lower hardness (the inclusions in the hard rock) than the remained network (hard rock matrix) in which the lower density rocks were have been built in. Another important factor that contributes to the destruction of the rock and the formation of a weathering figure from it (Fig. 10b) is the weak bonding between the rocks that form the rock. It is appropriate to recall that the main factor which led to formation of the trenches on AlCoCrCuFeNiV target under the ion bombardment (Fig. 4b) or cells on AlCoCrCuFeMnNiV target (Figs. 4c, 9a) is also a weak bonding between the two phases − the matrix and inclusions.

Comparing the micro- and macroreliefs one may conclude that they form under the influence of similar physical factors. Actually, the factors that form the microrelief on a target are the ion bombardment, the temperature, and the physics-chemical properties of the target material. While the factors promoting formation of the weathering figures are the wind − an analogue of ion bombardment, the temperature and physic-chemical properties of the rock. Truly, “in the small as in the big”.

Fig. 8 Microstructure of different areas of the groove: (a) bottom; (b) side wall. AlCoCrCuFeNiV alloy.

Fig. 9 (a) Micro-world: figures formed under moderate IB on the side wall of the groove. AlCoCrCuFeNiV alloy; (b) macro-world: weathering figures consisting of hard rocks formed under the influence of the wind and temperature.

Fig. 10 (a) Micro-world: figures formed under moderate IB on the side wall of the groove. AlCoCrCuFeMnNiV alloy; (b) macro-world: weathering figures consisting of mixed rocks under the influence of the wind and temperature.

5 Conclusions

The structure and composition of the grooves naturally forming under the ion bombardment on the targets used for MS have been investigated at the beginning and at the end of the targets operation. The target materials were CoCrCuFeNi, AlCoCrCuFeNiV, AlCoCrCuFeMnNiV high entropy alloys. It was revealed that before the service life the microstructure of the surface of all targets was smooth and flat and composed of matrix and inclusions homogenously distributed in the matrix. The matrices contain basic elements (Co, Cr, Fe, Ni) while the inclusions mainly consist of 70–80% of copper, aluminum or manganese. The composition of all targets, being inhomogeneous at the microscopic level, is homogeneous and equiatomic at the macroscopic level.

At the end of the service term spanned 30–40 hours on targets surfaces under the ion bombardment have appeared and developed grooves. The microrelief of the bottom and side walls of the grooves was different and depended on the intensity of the IB. In the area of strong IB the microrelief acquired features of a melted surface. In the area of moderate IB the microrelief was crossed in various directions by micro-trenches of various length and width. It was revealed that the shapes and distribution of these micro-trenches coincided with those of the inclusions observed on the original targets surfaces. It was also revealed that at the end of the service term the composition of the grooves became depleted with elements with minimal enthalpy of formation.

Analysis of the obtained results allows making some unexpected conclusions. (i) The presence of signs of melting on the surface of the targets indicates that the temperature of their surfaces reaches values close to the melting temperatures of the alloys under study; (ii) a new type of the effect of preferential sputtering has been revealed. The main condition for its existence is the presence in the target composition of the second phase in the form of micro-inclusions, which sputtering yield is noticeably higher than that of the components of the first phase; (iii) the formation of a specific microrelief on the surface of the targets upon bombardment with ions with energy typical for DC MS technology is a result of the combined action of two factors: high temperature and sputtering.

Author contribution statement

Shaginyan Leonid R.: The idea of the problem stated in the manuscript; the idea and general management of the experiments; manuscript preparation. Firstov Sergei A.: Discussion the results; co-writing the manuscript. Mironov Mikhail I.: Preparation and carrying out the experimental work. Krapivka Nicolay A.: Casting the alloys and preparation of the targets. Novichenko Viktor N.: Electron scanning microscopy of the targets. Kremenitsky Valery V.: Electron energy loss spectroscopy and X-ray photon spectroscopy for chemical analysis of targets.

References

All Tables

All Figures

	Fig. 1 General view of AlCoCrCuFeNiV target sputtered during 40 h. The initial composition of the target and at the end of its service (bottom of the groove) are presented in the table on the right.
In the text

	Fig. 2 Microstructure of initial surfaces of AlCoCrCuFeNiV and CoCrCuFeMnNi targets (BEI mode). Compositions in macro-level are presented in tables on the sides of micrographs.
In the text

	Fig. 3 Enlarged images of microstructures of AlCoCrCuFeNiV and CoCrCuFeMnNi targets (BEI mode). Corresponding compositions of matrices and inclusions presented in tables on the sides of micrographs.
In the text

	Fig. 4 Microstructure of the side walls of the grooves (areas of moderate IB): (а) CoCrCuFeNi; (b) AlCoCrCuFeNiV; (c) AlCoCrCuFeMnNiV alloys.
In the text

	Fig. 5 Enlarged image of the microstructure of the groove. The smooth merging of the side wall of the groove (upper left corner) with its bottom (lower right corner) is clearly seen. CoCrCuFeNi alloy.
In the text

	Fig. 6 Microstructure of the surface of the same target: (a) Initial surface of the target; (b) Side wall of the target (moderate IB) after the service term. AlCoCrCuFeNiV alloy.
In the text

	Fig. 7 Microstructure of the bottoms of the grooves (area of strong IB): (а) CoCrCuFeNi; (b) AlCoCrCuFeNiV; (c) AlCoCrCuFeMnNiV alloys.
In the text

	Fig. 8 Microstructure of different areas of the groove: (a) bottom; (b) side wall. AlCoCrCuFeNiV alloy.
In the text

	Fig. 9 (a) Micro-world: figures formed under moderate IB on the side wall of the groove. AlCoCrCuFeNiV alloy; (b) macro-world: weathering figures consisting of hard rocks formed under the influence of the wind and temperature.
In the text

	Fig. 10 (a) Micro-world: figures formed under moderate IB on the side wall of the groove. AlCoCrCuFeMnNiV alloy; (b) macro-world: weathering figures consisting of mixed rocks under the influence of the wind and temperature.
In the text