Fabrication of Single‐Crystalline CoCrFeNi Thin Films by DC Magnetron Sputtering: A Route to Surface Studies of High‐Entropy Alloys

High‐entropy alloys (HEAs) with their almost limitless number of possible compositions have raised widespread attention in material science. Next to wear and corrosion resistive coatings, their application as tunable electrocatalysts has recently moved into the focus. On the other hand, fundamental properties of HEA surfaces like atomic and electronic structure, surface segregation and diffusion as well as adsorption on HEA surfaces are barely explored. The lack of research is caused by the limited availability of single‐crystalline samples. In the present work, the epitaxial growth of face centered cubic (fcc) CoCrFeNi films on MgO(100) is reported. Their characterization by X‐ray diffraction (XRD), energy dispersive X‐ray spectroscopy (EDX), and transmission electron microscopy (TEM) demonstrates that the layers with a homogeneous and close to equimolar elemental composition are oriented in [100] direction and aligned with the substrate to which they form an abrupt interface. X‐ray photoelectron spectroscopy (XPS), low‐energy electron diffraction (LEED), and angle‐resolved photoelectron spectroscopy are employed to study chemical composition and atomic and electronic structure of CoCrFeNi(100). It is demonstrated that epitaxially grown HEA films have the potential to fill the sample gap, allowing for fundamental studies of properties of and processes on well‐defined HEA surfaces over the full compositional space.


Introduction
Starting with the groundbreaking work of Cantor et al. [1] and Yeh et al., [2] the material system of high entropy alloys (HEAs), sometimes also referred to as complex concentrated alloys (CCAs) or multiple principal element alloys (MPEAs), have gained significant research interest due to their engineering and application potential. [3] The basic idea of a high-entropy alloy is that the mixing entropy ΔS mix is large enough so that the change in the Gibbs free energy ΔG mix = ΔH mix − TΔS mix for forming a solid solution from the elements becomes favorable over the formation of intermetallic phases. [4][5][6] Herein, ΔH mix is the reaction enthalpy, which characterizes the bonding of the atoms in the alloy. According to the literature (see, e.g., [4][5][6] and references therein), the definition of a HEA appears not very stringent. Frequently used criteria are the number of elements n (typically n ⩾ 4), the composition (concentration of each element between 5% and 35%) and the mixing entropy (ΔS mix ⩾ 1.5R), where R = 8.314 JK −1 mol −1 is the ideal gas constant. According to Yeh [4] definitions of HEAs "are guidelines, not laws".
Since the introduction of HEAs, a large number of studies were published focusing on the investigation and modification of mechanical properties. HEAs have been observed to exhibit comparatively high hardness values [7][8][9] and exceptional yield strength. [10,11] Those properties were also found to reveal severe changes depending on the fabrication methods and post treatments applied. [12,13] For the so-called Cantor alloy CoCrFeMnNi an exceptional tensile strength and fracture resistance was reported which increases when the temperature is lowered from room temperature to liquid nitrogen temperature (77 K). [14] Furthermore, a good resistance against wear [15,16] and corrosion [17][18][19] was reported. Due to their promising irradiation resistance, [20][21][22] some HEAs are under consideration as novel materials for radiation exposed building parts in, for example, nuclear reactors. Also in the fields of magnetism [23,24] and superconductivity [25,26] there are reports regarding highly interesting HEA systems. HEAs have also been suggested as effective electrocatalysts. [27][28][29] In particular, it was suggested that their (bulk truncated) surfaces offer a large number of structurally and compositionally different sites [28] with different selectivity and activity. It was further proposed that HEAs offer the possibility to tailor active sites by a proper selection of elements and composition, thus leading to optimized catalysts for different chemical reactions including, for example, hydrogen evolution reaction, oxygen evolution reaction, or carbon dioxide reduction reaction. [27] Despite the interest in HEAs as corrosion-and wear-resistant coatings and more recently as optimized catalysts, little is known about the physical and chemical properties of their surfaces like, for example, composition, atomic structure or electronic band structure. To our knowledge, only a single paper exists which experimentally addresses the atomic structure of two low-index surfaces of a HEA. [30] Certainly, there must be room for the possibility that nature has something else in mind than a bulk truncated surface. Surface reconstruction, for example, could occur due to surface energy minimization, thereby also changing available adsorption sites. Furthermore, the chemical composition of a HEA surface can be influenced by temperature and chemical potential of the constituents. Whereas bulk diffusion in HEAs is debated controversially, [31,32] surface diffusion remains to be explored. So far one can only speculate how these properties are influenced by the chemical environment, for example, by a vapor phase or liquid phase above the surface. Corrosion and catalysis involve subsequent steps of adsorption, dissociation, surface reaction and desorption. These processes need to be studied on an atomic scale in order to fully understand the behavior of HEAs. [33,34] These few examples clearly show that there is a yet unsatisfied demand for surface investigations of HEAs.
The lack of activity in the field of surface science of HEAs is to a great extent caused by the limited availability of proper samples. Typical experiments for the investigation of structural, chemical, electronic and vibronic properties of surfaces require single crystals with surfaces oriented in certain low-index directions and with an area of approximately a few mm 2 in size. Such samples are usually cut from larger single crystal ingots obtained, for example, by the Czochralski method as demonstrated by Ledieu et al. [30] In the latter work, the (110) and (320) surfaces of Fe 23.8 Co 22.8 Cr 22.3 Mn 12.0 Ni 19.1 were investigated. Us-ing scanning tunneling microscopy (STM), low-energy electron diffraction (LEED) and X-ray photoelectron spectroscopy (XPS) new and interesting insights into structure and composition on an atomic scale of these surfaces were revealed. Unfortunately, the growth of HEA single crystals is not widely performed and appropriate samples are difficult to obtain.
The deposition of crystalline thin films from HEAs on suitable substrates is an alternative that should certainly be considered to overcome the above mentioned bottleneck. Different techniques have been employed successfully to grow thin films of high entropy alloys. [9,[35][36][37][38] Depending on the chosen method, a film thickness ranging from few nm up to mm can be obtained. Among the reports on thin film growth, magnetron sputtering from disc shaped targets appears to be by far the most commonly used method, see for example the recent work from Zandejas Medina et al. [39,40] or by Rao et al. [41,42] Magnetron sputtering is a very versatile technique which allows high deposition rates. Recently, we have carried out a study of the growth of CoCrFeNi by DC magnetron sputtering [43] in which we compared the chemical and structural properties of the films grown from two different targets. Both targets were prepared by spark plasma sintering (SPS). While one target was made from a powder of the equimolar HEA alloy, the other target was made from an equimolar mixture of pure elemental powders. The structural and chemical properties of the films grown from the two targets were virtually identical. X-ray diffraction (XRD) revealed a single phase solid solution with an fcc structure with a lattice constant of a = 3.575 Å. X-ray fluorescence and XPS indicated a chemical composition very close the intended equimolar composition and in good agreement with the composition of the targets, which was determined in the same way.
In the present work, we investigate the growth of singlecrystalline thin films of a CoCrFeNi high entropy alloy on singlecrystalline MgO(100) substrates. MgO crystallizes in the cubic rock-salt structure with a lattice constant of 4.212 Å . [44] Basically, the rock-salt structure is a fcc structure with a two-ion basis containing one Mg 2 + ion and one O 2 − ion. It can also bee seen as made up from two fcc sublattices, one for the Mg 2 + ions and one for the O 2 − ions. The cubic MgO(100) surface should enable heteroepitaxial growth of CoCrFeNi. As will be shown below, the single-phase films in fcc structure are of high crystalline order and well oriented with respect to the substrate. As intended by the choice of substrate, the surfaces of the CoCrFeNi films are oriented in [100] direction. Furthermore, by using different experimental techniques, we demonstrate that these films are suitable for studying structural and electronic properties of their surfaces. The present work thus opens a route to the investigation of surface properties of HEAs for which bulk crystals are barely available.

Results and Discussion
Although the overall scope of the present work is to establish HEA thin films as suitable systems for surface investigations it is important to also characterize and understand their properties below the surface. Only in this way one can establish that the thin films can bee seen as thin slices of HEA crystals. Consequently, this section is divided into two parts. In the first part (Section 2.1) we will investigate the "bulk" properties of the CoCrFeNi films www.advancedsciencenews.com www.advmat.de grown on MgO(100) substrates. Section 2.2 is devoted to an analysis of the surface of the thin films.  (200) and (400) reflections of both, the MgO substrate with rock salt structure and the CoCrFeNi layer with fcc structure on top. From the position of the CoCrFeNi reflections, the lattice constant has been estimated to be 3.575 Å which agrees well with literature data. [43,45,46] The absence of additional reflections of CoCrFeNi like, for example, the (111) or (220) reflection [43] indicates that the films are very well oriented. Obviously, the MgO(100) substrate supports the preferential growth of fcc-CoCrFeNi in [100] direction. Pole figure measurements shown in Figure S1 (Supporting Information) confirm the azimuthal alignment of the CoCrFeNi layer with respect to the MgO substrate. Figure 2a displays a typical scanning electron microscopy (SEM) image in top-view geometry of a CoCrFeNi film on MgO(100). The SEM image is free of any structural features. In comparison, CoCrFeNi films deposited on Si or stainless steel substrates showed a grainy structure typical for polycrystalline films. [43] Thus the SEM image in Figure 2a points toward a significantly improved crystalline order obtained on MgO(100) substrates. In addition, the contrastless image indicates a homogeneous secondary electron yield which would be expected for a homogeneous elemental composition. The latter is strongly supported by the corresponding EDX maps compiled in Figure 2b for all four HEA elements, which show no evidence for elemental segregation. The composition (atomic percentage) obtained from the EDX maps is (26.6 ± 1.7) % Co, (23.8 ± 2.3) % Cr, (24.6 ± 2.0) % Fe, and (24.9 ± 2.2) % Ni, which is very close to the targeted equimolar composition. The almost equimolar composition and homogeneous elemental distribution was observed so far for all CoCrFeNi films deposited in our studies, including polycrystalline films on Si and stainless steel substrates [43] and highly crystalline films on MgO(100) substrates. Figure 3a. The interface between the HEA and the MgO substrate is clearly discernible. Furthermore, the capping layer from Pt deposited on top of CoCrFeNi film is visible in the top of the image. A thin bright line just below the capping layer signals an oxide layer formed on top of the HEA film. The formation of the oxide layer can be understood as follows. The sample transfer from the magnetron sputtering setup to the focused ion beam (FIB) device was carried out in air. As all elements in the CoCrFeNi film are known to readily oxidize in contact with air, [34,47] the formation of a native surface oxide layer of a few nanometers is unavoidable. This oxide layer persists during the TEM lamella fabrication and thus can be seen in the cross-sectional electron microscopy image as well as in the EDX profiles shown in Figure 3c, which are discussed below. In the HEA film shown in Figure 3a, planar defects are contained (stacking faults and twin boundaries, for details please see Supporting Information). The planar defects are visible as apparent lines in the image, which are inclined to the interface at an angle of ≈±55°. Figure 3b depicts EDX maps for the six elements Co, Cr, Fe, Ni, Mg, and O. The EDX maps show a very homogeneous distributions of the elements in the HEA film and an abrupt interface to the substrate, indicating negligible diffusion across the interface. Figure 3c shows the horizontally averaged distribution of the elements as a function of vertical position along the dashed line in Figure 3a. The high Pt concentration close to zero position is due to the capping layer. A peak in the O concentration at a position of around 5 nm, that is, just below the capping layer, indicates the native oxide layer on top of the HEA film mentioned above. The interface between the HEA and the substrate is clearly visible at a position of around 260 nm, where the Mg and O concentrations show an abrupt increase. At the same position, the four elements of the HEA show an equally abrupt decrease. In the region between the substrate and the superficial oxide layer, the four elements of the HEA show a distribution which is constant over the whole film thickness and almost equal for all four elements. The signals due to Pt and Mg are negligible in that area. On the other hand, the signal due to O is significantly larger, with values of the order of 5 at.%. We believe that this is due to a post-preparation degradation of the vertical surfaces of the TEM lammellae, which also were transported in air. Taking only the four HEA elements into account, the composition (atomic percentage) obtained from the EDX maps is (25.5 ± 0.8) % Co, (25.2 ± 0.5) % Cr, (23.1 ± 0.7) % Fe, and (26.1 ± 0.8) % Ni. This agrees well with the composition determined from the EDX experiments carried out in the SEM setup. The latter was measured on a different sample which shows the reproducibility of our depositon process. At this point we should mention, that one sample investigated by STEM showed two Cr grains with sizes of approximately 150 nm. The reason for the formation of these grains is unclear as yet.  exemplarily marked by white arrows in Figure 4b (for details please see Supporting Information). Besides these defects, which may be initiated at the interface in combination with the low stacking fault energy of CoCrFeNi, [48] the crystalline structure appears well ordered. This is confirmed by the nano-beam diffraction patterns shown in Figure 4c, which were acquired in the substrate (left) and HEA (right) region of the sample. MgO crystallizes in the rock-salt structure, which is a fcc lattice with a twoatomic basis. The CoCrFeNi high-entropy alloy has a fcc struc-ture with a one-atomic basis. The diffraction patterns also indicate that the crystal lattices are aligned with respect to each other confirming heteroepitaxy.

Surface Properties of CoCrFeNi Films on MgO(100)
We start the discussion of the surface properties by XPS investigations. Due to the unavoidable sample transfer through air a   contamination of the surface, that is, oxidation and adsorption of contaminants, has to be expected and can be verified within an XPS survey scan of the pristine state as shown in Figure 5 (bottom curve). This measurement shows clear signals of the four alloy elements Co, Cr, Fe and Ni as well as strong signals for O and C. Surface preparation of single crystal metal surfaces in ultra-high vacuum (UHV) ist typically performed by Ar + ion bombardment (Ar ion sputtering) followed by annealing to elevated temperatures in order to restore crystalline order. For that purpose, a proper preparation scheme was developed which is described in Section 4. By applying this surface preparation procedure a complete removal of oxides and other contaminations is demonstrated (top curve in Figure 5). Furthermore, the XPS survey scans do not show signals associated with Mg or, after the UHV preparation, oxygen. This indicates the absence of diffusion between substrate and HEA thin film which is in good agreement with the TEM/EDX results.
In order to gain insights into the chemical state of the the HEA elements, individual core level spectra were measured before and after the UHV preparation. Typical results of such measurements are shown in Figure 6 . It can be seen, that in the pristine state all the four metals suffer from oxidation as indicated by the marked oxide energy ranges, respectively. After the UHV preparation by Ar ion sputtering and annealing, no oxide signals can be identified in either of the core level measurements in agreement with the survey spectra discussed before. This is also in agreement with the STEM experiments (see Section 2.1), which showed only a thin oxide layer on top of the HEA thin film. In the prepared sample state some features gain visibility in the core level spectra. These are the well-known 6 eV satellite of Ni (859 eV), [49] the Fe L 3 M 23 M 45 Auger-emission (748 eV) and the Co L 3 M 23 M 45 Auger line (713 eV), [50] respectively. For all the elements present in the HEA, no significant shift of the core level binding energies with respect to pure element values is observed (see, e.g., [51] or appendix section in Ref. [50]).
The surface composition of the HEA thin films can be approximated from the integrated intensity of the core levels. The measured intensity I X for element X has to be corrected for the corresponding photoionization cross section value X taken from Scofield, [52] the spectrometer specific transmission function T(E kin, X ) as well as the electron inelastic mean free path (E kin, X ), where E kin, X is the kinetic energy of the photoelectrons from element X. Using the equation  [53] In this study it was shown that the oxidation tendency in this HEA is the strongest for Cr, followed by Fe and Co and is rather weak for Ni. From theoretical investigations it is demonstrated that oxygen  atoms migrate toward Cr-and Fe-rich sites and away from Ni. We conclude that the favored Cr-O bonding behavior is the origin for the Cr enrichment we measured in the pristine state of the thin films and for the strong Cr signal decrease following the surface treatment after which no oxygen is present. It deserves to be men-tioned that repetitions of the UHV preparation procedure did not lead to significant further changes in the composition obtained by XPS. Representative electron diffraction patterns obtained after the UHV preparation step are shown in Figure 7 for different 486 Å −1 and fits well to the distance of the first order diffraction spots from the image centre (red arrows). If a longer annealing time (> 90 min) is used for preparing the Ar + sputtered HEA samples we observe a (2×2) surface reconstruction (see Figure 7c,d). The origin of this reconstruction is not clear as of yet. So far, we can only conclude by virtue of the XPS results, that no adsorbates are present on the surface. Therefore, a reconstruction driven by adsorbed atoms or molecules is to be ruled out.
Angle-resolved photoelectron spectroscopy (ARPES) measurements of the occupied electronic bands of CoCrFeNi on MgO(100) were conducted along the high symmetry directions of the surface Brillouin zone, ΓX and ΓM. The electron distri-bution curves (EDCs) for the binding energy range from -0.5 to 8 eV are shown in Figure 8 a for the (1×1) surface and (c) the (2×2) reconstruction. The photoemission signal just below the Fermi energy is due to the d-bands which show a small dispersion compared to the s-p-bands. Hence, their comparatively high density of states leads to very intense signals. [54] In the higher binding energy regime (E B > 2 eV), s-p-states of strong dispersion and comparably weak intensity can be seen. [54] The EDCs of the (2×2) reconstructed surface show an additional state most prominent around the Γ-Point at E B ≈ 4.7 eV (arrow in Figure 8c). Since this state is only present when the surface shows the (2×2) reconstruction we assign this feature to an electronic surface state (ESS). For a better visualization of the electronic band dispersion the angle resolved photoemission data (intensity vs emission angle and binding energy) were converted into k-space and the 2 nd derivative is plotted in Figure 8b for the (1×1) surface and (d) for the (2×2) reconstructed case. In the latter, a clear dispersion of the ESS is visible along the ΓM direction with maxima in energy at Γ and M. Halfway along the ΓM direction at k || = 0.88 Å −1 , the dispersion of the ESS shows a minimum at E B = 5.8 eV. From this dispersion of the ESS, we conclude a periodicity half as large as for the CoCrFeNi SBZ which is in agreement with our LEED measurements. Along ΓX, the ESS dispersion is only clearly visible up to k || ≈ 0.5 Å −1 but appears to show a similar behavior as for ΓM. In general, we attribute the smeared out appearance of the bands to chemical disorder in consequence of the random element distribution laying within the nature of HEAs. This effect can also be seen in theoretical calculations for CoCrFeNi and even for binary alloys with random lattice occupation. [55]

Conclusion
We have demonstrated the epitaxial growth of well-ordered, single crystalline CoCrFeNi films on MgO(100) substrates by magnetron sputter deposition. XRD measurements unambiguously demonstrated that the layers are oriented in (100) direction. STEM and HRTEM experiments revealed that the crystal lattices of the CoCrFeNi films and the MgO substrates are aligned. Lattice mismatch leads to the formation of planar defects. A homogeneous element distribution with nearly equimolar composition in the films was observed by EDX.
In order to test the potential of the CoCrFeNi films for future experimental surface studies they were investigated by a variety of different surface techniques, that is, by XPS, LEED and ARPES.
To that end, a recipe was developed that allows the preparation of clean and well-ordered surfaces by Ar + ion sputtering and annealing. Compared to the bulk of the films, the (100) surface shows a tendency for an increased Cr and Ni content before surface cleaning. Afterward the Ni percentage was found to be even higher while the Cr content has decreased. LEED showed that, depending on the annealing time after Ar + ion bombardment, the surface is either unreconstructed or shows a (2×2) reconstruction for long annealing times. According to XPS the reconstruction is not caused by adsorbed species. ARPES experiments revealed, in addition to the transition metal d-and sp-bands, a surface state associated with the (2×2) reconstruction. The experimental observations clearly demonstrate that the high quality of our thin films permits the study of structural and electronic properties by well-established experimental techniques.
Given the results shown in this work it appears reasonable to conclude that epitaxial growth of HEA thin films on crystalline substrates by magnetron sputter deposition is promising for future experimental work related to the fundamental properties of high entropy alloy surfaces. Yet unexplored phenomena such as diffusion and/or segregation near the surface, surface reconstruction, surface electronic structure and adsorption on HEA surfaces can be studied without the need for bulk crystal growth. The high flexibility of the sputtering deposition process with regard to target and thin film composition as well as substrate material and crystal orientation thereby offers the opportunity to tune www.advancedsciencenews.com www.advmat.de the properties of HEA surfaces and thus to develop optimization routes for future catalysts. Besides that we can also imagine that high quality epitaxial films on insulating substrates such as MgO might be suitable for studying other fundamental properties of HEAs like, for example, electronic transport phenomena.

Experimental Section
Sputter Deposition: Thin films of CoCrFeNi were deposited via DC magnetron sputtering from spark plasma sintered targets. An alloy powder of gas atomized CoCrFeNi (Nanoval GmbH, Frankenblick, Germany) was thereby used as primary material. For details about the sintering process Ref. [43] was referred to. Commercial arc fusion grown MgO(100) crystals (CrysTec GmbH Kristalltechnologie, Berlin, Germany) were used as a substrate. After chemical cleaning with isopropanol in an ultrasonic bath for 10 min the substrates were annealed in UHV at 1000°C for 60min to dispose Mg(OH) 2 groups formed on the surface due to the hygroscopic nature of MgO. [56,57] To prevent recontamination, the substrates were transferred to the magnetron sputter setup right after annealing. The deposition was carried out in an INOVAP CF503 system (INOVAP GmbH, Radeberg, Germany) using an Ar atmosphere of 1.5 Pa with an Ar flow rate of 50 sccm. The deposition power and time were chosen to be 200 W and 30 min, respectively. An elevated substrate temperature of about 350°C has been applied during deposition. The distance between substrate and target was 160 mm which was expected to ensure a homogeneous film thickness.
SEM and EDX: An investigation of the samples bulk composition was carried out using energy dispersive X-ray spectroscopy (EDX). The EDX detector of type Bruker X Flash 1510 (Bruker Corporation, Billerica, MA, USA) was part of the Nova NanoSEM 200 electron microscope (FEI, Hillsboro, OR, USA) used to analyse the surface morphology. EDX mappings of the constituents main emission lines were used to reveal the lateral distribution of the elements and evaluated in the Bruker Esprit 1.8 Software.
XRD: Formation of crystalline phases was analysed by XRD experiments performed with a Rigaku SmartLab 9 kW with HyPix-3000 detector (Rigaku Corporation, Tokyo, Japan). The measurements were carried out in air using a parallel beam geometry with Cu K radiation ( = 1.5406 Å).
− 2 scans were measured in an angular range of 30°⩽ 2 ⩽ 130°. Pole figures were measured in in-plane configuration using the (200) diffraction signal of MgO and CoCrFeNi, respectively. Assignment of detected signals was done with help of the Crystallographic Open Database (COD) plug-in (see, e.g., [58] ) in the Rigaku SmartLab Studio II V 4.1.0.182 software.
TEM: Cross-section lamellae for TEM analysis were prepared using FIB milling in a Helios NanoLab 600i (FEI) dual beam scanning electron microscope. Sample transfer to the FIB and from the FIB to the TEM was carried out in air atmosphere. In the FIB, the sample surface was covered with Pt using both the electron and then the ion beam before cutting lamellae with the ion beam operating at 30 kV and 2.5 nA. The lamellae were transferred onto Cu grids using an OmniProbe system before applying the final milling. Acceleration voltage and ion beam current were reduced in four consecutive steps down to 5 kV/15 pA using stage tilts of 50.5°to 53.5°. High-resolution transmission electron microscopy (HR-TEM) and scanning transmission electron microscopy (STEM) were performed in an JEOL NEOARM 200F instrument to characterize the HEA films and the film-substrate interface. STEM EDX was used to map the concentration distribution in the HEA films and the MgO substrate. The acquired EDX spectra were subject to a background subtraction by post-processing with the software package Pathfinder 2.6 (ThermoFisher Scientific) using the Filter-Fit method for peak fitting. For noise reduction, the total counts for each pixel were enhanced by a 3 x 3 binning. For quantification, the Cliff-Lorimer correction method was used. Diffraction patterns from the HEA films and the substrate were recorded using nano-beam electron diffraction (NBED). Diffraction patterns were indexed using the rock-salt structure for MgO and the fcc structure for the HEA film using the software SingleCrystal 4.1.2.
Surface Preparation: To investigate the surface composition, band dispersion and electron diffraction patterns a transfer of the samples to an ultra-high vacuum (UHV) chamber was necessary which was unavoidably accompanied by a formation of native oxides. To get rid of the oxide components and restore crystallinity at the surface, a two-step UHV preparation was applied. First, the sample undergoes a normal incidence Ar + bombardment followed by a similar treatment under angles of incidence ±60°with respect to the surface, respectively. The latter was supposed to slightly flatten out the surface and reduce roughness after the high energy normal bombardment. Still, the surface usually does not show an electron diffraction pattern after this step which was expected to be due to an irregular surface structure. A subsequent annealing step at 700°C has turned out to lead to a crystalline surface structure.
LEED: The surface crystal structure has been investigated by means of LEED using an ErLEED 150 system (SPECS Surface Nano Analysis GmbH, Berlin, Germany). Usually, LEED images can only be measured on the CoCrFeNi thin films after the surface preparation described above has been applied in advance.
XPS: Before and after the UHV surface preparation, the surface stoichiometry of the HEA thin films was estimated from XPS measurements. Those were conducted using Al K radiation from a Specs XR50M X-ray source monochromatized with a Specs Focus 500 monochromator. All XPS measurements were taken at room temperature with a base pressure better than 5 × 10 −10 mbar. All survey spectra were measured using a pass energy of 50 eV while the core level data were taken at E Pass = 10 eV.
ARPES: The electronic band structure was investigated in angleresolved photoelectron spectroscopy (ARPES) experiments carried out using the monochromatic HeI emission line (21.218 eV) from a SPECS UVS 300 radiation source. For both XPS and ARPES, the photoelectrons were detected using a 2D CCD detector equipped to a Specs Phoibos 150 analyser (SPECS Surface Nano Analysis GmbH, Berlin, Germany). If not mentioned otherwise, all measurement were conducted at room temperature.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.