Laser Metal Deposition of an AlCoCrFeNiTi 0.5 High-Entropy Alloy Coating on a Ti6Al4V Substrate: Microstructure and Oxidation Behavior

: Ti6Al4V has been recognized as an attractive material, due to its combination of low density and favorable mechanical properties. However, its insu ﬃ cient oxidation resistance has limited the high-temperature application. In this work, an AlCoCrFeNiTi 0.5 high-entropy alloy (HEA) coating was fabricated on a Ti6Al4V substrate using laser metal deposition (LMD). The microstructure and isothermal oxidation behaviors were investigated. The microstructure of as-deposited HEA exhibited a Fe, Cr-rich A2 phase and an Al, Ni, Ti-enriched B2 phase. Its hardness was approximately 2.1 times higher than that of the substrate. The oxidation testing at 700 ◦ C and 800 ◦ C suggested that the HEA coating has better oxidation resistance than the Ti6Al4V substrate. The oxide scales of the Ti6Al4V substrate were mainly composed of TiO 2 , while continuous Al 2 O 3 and Cr 2 O 3 were formed in the HEA coatings and could be attributed to oxidation resistance improvement. This work provides an approach to mitigate the oxidation resistance of Ti6Al4V and explore the applicability of the HEA in a high-temperature environment.


Introduction
Ti6Al4V has been an important and versatile titanium alloy currently used in petrochemical, automotive, power generation, and biomedical industries. Indeed, this alloy possesses a desirable combination of properties, such as high melting point, superior corrosion resistance, good biocompatibility, and weldability [1][2][3]. However, Ti6Al4V alloy is very active at elevated temperature, which results in being easily oxidized, as well as its insufficient oxidation resistance [4,5]. At present, the permitted maximal service temperature of Ti6Al4V alloy still does not exceed 600 • C [1,4]. The improvement of its high-temperature oxidation property can be solved by the surface modification process [3][4][5][6].
As a promising surface modification technology, laser metal deposition (LMD) has introduced a number of capabilities unparalleled by conventional process [3,[6][7][8][9][10][11]. LMD achieves layer-by-layer fabrication of near net-shaped deposition onto the substrate by introducing a powder stream into a laser beam. In addition to the geometry freedom, precise tailoring of compositions and microstructure can be achieved to produce highly specialized coatings. On the basis of these advantages, many efforts have been made to improve the oxidation resistance of Ti6Al4V alloy. In particular, Liu et al. prepared a TiN/Ti 3 Al composite coating on the Ti6Al4V substrate [12]. The isothermal oxidation results indicated  44-145 µm particle size distribution. The powder mixture (total: 276.5 g) was prepared by carefully weighing the powders using a weighing balance with an accuracy of ±0.1 mg. The mixed powders were homogenized in a Turbula mixer (Glen Mills Inc., Clifton, NJ, USA) for 1 h. The experimental setup consisted of an IPG 1 kW continuous wave YAG fiber laser (IPG Photonics, Oxford, MA, USA), a numerical control working table, and a vibration powder feeder (Powder Motion Labs, Rolla, MO, USA), as shown in Figure 1. The Grade 5 Ti6Al4V bar stock was used as the substrates (dimensions 75 mm × 12 mm × 6 mm) and cleaned with acetone to remove the dirt and oil before the experiment. The deposition process was performed in a sealed, controlled environment purged with a continuous flow of argon gas. A pre-heating was undertaken to minimize the thermal stress between the deposit and the Ti6Al4V substrate. The powders were delivered to the laser beam by an argon jet with a flow rate of 3 g/min. The deposits with a thin-wall structure were fabricated at the transverse laser speed of 200 mm/s, 2 mm laser beam size, 0.5 mm layer thickness, and the powers of 700 W for the first two layers, 600 W for the next two layers and 500 W for further layers.
The microstructure was characterized by a Helios Nanolab 600 (Thermal Fisher Scientific, Waltham, WA, USA) scanning electron microscope (SEM). The energy dispersive spectroscopy (EDS) data were collected and analyzed in the factory standardization manner (Oxford AZtec version 4.2). The electron backscatter diffraction (EBSD) scan was acquired using an Oxford HKL system with a step size of 2 µm, and data analysis was performed on HKL Channel 5 software

Characterization
After the deposition, the specimens were sectioned by electrical discharge machining (EDM) (Hansvedt Industries Inc., Rantoul, IL, USA) and prepared by standard metallographic procedures. They were polished with 320-1200 grit SiC grinding paper, followed by 9 µm, 3 µm, 1 µm diamond solutions with a final step of 0.05 µm colloidal silica suspension. The HEA sample was etched with aqua regia ( The microstructure was characterized by a Helios Nanolab 600 (Thermal Fisher Scientific, Waltham, WA, USA) scanning electron microscope (SEM). The energy dispersive spectroscopy (EDS) data were collected and analyzed in the factory standardization manner (Oxford AZtec version 4.2). The electron backscatter diffraction (EBSD) scan was acquired using an Oxford HKL system with a step size of 2 µm, and data analysis was performed on HKL Channel 5 software (Oxford Instruments, Abingdon, UK). The grain was determined by a misorientation angle of 10 • , and grain size was measured using the line intercept method. X-ray diffraction (XRD) profiling was performed to determine the phase constituents in the samples. The phases were identified by Phillips X'Pert diffractometer (Amsterdam, The Netherlands) using Cu-Kα radiation at 45 kV/40 mA with a scanning step of 0.05 • and a scanning range from 20 • to 90 • . The Vickers hardness was measured using Struers Duramin hardness equipment (Struers Inc, Cleveland, OH, USA) at a 9.8 N load and a duration of 10 s. The reported hardness results were the average of ten indentations.

Oxidation Tests
The HEA and Ti6Al4V specimens for oxidation tests with dimensions of 10 mm (length) × 10 mm (width) × 2 mm (height) were prepared using EDM. Sample surfaces were subsequently polished using 320-1200 grit SiC grinding paper and cleaned with acetone. Oxidation tests were carried out in an electric furnace (DT-29-RSA, Deltech, Denver, CO, USA) under atmospheric pressure at 700 • C and 800 • C for 45 h. The samples were heated from room temperature to the target temperatures at a heating rate of 10 • C/min. The weight gain was measured before and after at specified intervals using an analytic balance (AG204, Mettler Toledo, Columbus, OH, USA) with an accuracy of 0.1 mg. The oxidized samples were characterized by XRD and SEM, as described above.

Results and Discussion
3.1. Microstructures of the As-Deposited HEA Figure 2 illustrates the XRD patterns of as-deposited AlCoCrFeNiTi 0.5 HEA and the Ti6Al4V substrate. The α-Ti phase was detected in the Ti6Al4V (Figure 2a). Two BCC phases were observed in the HEA (Figure 2b), and they are ordered a BCC structure (B2) phase and a disordered BCC structure (A2) phase. Through careful analysis of the standard PDF database, the two BCC phases were identified as Al-Ni and Fe-Cr phases, which is in accordance with the similar alloys synthesized using arc-melting [23,30] and casting [33]. Phillips X'Pert diffractometer (Amsterdam, The Netherlands) using Cu-Kα radiation at 45 kV/40 mA with a scanning step of 0.05° and a scanning range from 20° to 90°. The Vickers hardness was measured using Struers Duramin hardness equipment (Struers Inc, Cleveland, OH, USA) at a 9.8 N load and a duration of 10 s. The reported hardness results were the average of ten indentations.

Oxidation Tests
The HEA and Ti6Al4V specimens for oxidation tests with dimensions of 10 mm (length) × 10 mm (width) × 2 mm (height) were prepared using EDM. Sample surfaces were subsequently polished using 320-1200 grit SiC grinding paper and cleaned with acetone. Oxidation tests were carried out in an electric furnace (DT-29-RSA, Deltech, Denver, CO, USA) under atmospheric pressure at 700 °C and 800 °C for 45 h. The samples were heated from room temperature to the target temperatures at a heating rate of 10 °C/min. The weight gain was measured before and after at specified intervals using an analytic balance (AG204, Mettler Toledo, Columbus, OH, USA) with an accuracy of 0.1 mg. The oxidized samples were characterized by XRD and SEM, as described above. Figure 2 illustrates the XRD patterns of as-deposited AlCoCrFeNiTi0.5 HEA and the Ti6Al4V substrate. The α-Ti phase was detected in the Ti6Al4V (Figure 2a). Two BCC phases were observed in the HEA (Figure 2b), and they are ordered a BCC structure (B2) phase and a disordered BCC structure (A2) phase. Through careful analysis of the standard PDF database, the two BCC phases were identified as Al-Ni and Fe-Cr phases, which is in accordance with the similar alloys synthesized using arc-melting [23,30] and casting [33].  Figure 3a presents the interface between the HEA coating and the Ti6Al4V substrate, which can be seen as a good metallurgical bond without crack. The EDS line scan was used to characterize the elemental evolution from the Ti6Al4V substrate to the HEA, and the quantitative results are shown in Figure 3b. The EDS measured compositions of the Ti6Al4V substrate (Ti: ~84-86 atom%, Al: ~10-11 atom%, V: ~3-4 atom%) did not deviate from the nominal compositions of Grade 5 Ti6Al4V. Since the Ti6Al4V was mixed with the HEA layer, the elements of Al (~12-14 atom %), Co (~10-12 atom%), Cr (~11-13 atom%), Fe (~9-11 atom%), Ni (~11-14 atom%) and Ti (~36-39 atom%) were detected. A small amount of V (~1-3 atom%) was detected in the HEA layer, due to the dilution. According to the equilibrium Fe-V, Fe-Ti phase diagrams, the Fe-V or Fe-Ti intermetallic phases  Figure 3a presents the interface between the HEA coating and the Ti6Al4V substrate, which can be seen as a good metallurgical bond without crack. The EDS line scan was used to characterize the elemental evolution from the Ti6Al4V substrate to the HEA, and the quantitative results are shown in Figure 3b. The EDS measured compositions of the Ti6Al4V substrate (Ti:~84-86 atom%, Al:~10-11 atom%, V:~3-4 atom%) did not deviate from the nominal compositions of Grade 5 Ti6Al4V. Since the Ti6Al4V was mixed with the HEA layer, the elements of Al (~12-14 atom %), Co (~10-12 atom%), Cr (~11-13 atom%), Fe (~9-11 atom%), Ni (~11-14 atom%) and Ti (~36-39 atom%) were detected. A small amount of V (~1-3 atom%) was detected in the HEA layer, due to the dilution. According to the equilibrium Fe-V, Fe-Ti phase diagrams, the Fe-V or Fe-Ti intermetallic phases might be formed when the content of V was in a range of 30-65 atom% or the content of Fe was above 50 atom% [34][35][36]. From the elemental analysis above, the contents of V and Fe were low (~10 atom% or below), thus the HEA layer had a low risk of forming those intermetallic phases. Moreover, the XRD patterns obtained did not show Fe-V or Fe-Ti intermetallic phases.

Microstructures of the As-Deposited HEA
Crystals 2020, 10, x FOR PEER REVIEW 5 of 17 The microstructural details of the laser deposited HEA at different magnifications are shown in Figures 3c,d. The equiaxed grains were delineated by intergranular phases (Figure 3c). The dendritic structure can be observed within the equiaxed grains. The EDS was used to analyze the element distribution, and the results are listed in Table 2 and Figures 4a-g. The dark contrast phase was enriched in Al, Ni, Co, and Ti, and the high level of Fe and Cr concentration was detected in the bright contrast phase. As learned from Table 2, the chemical compositions of Fe and Cr were ~45 atom% in the bright contrast phase while were ~16 atom% in the dark contrast phase. In combination with the phase identification, the dark contrasted phases rich in Al, Ni, and Ti are B2 phase while the bright contrast phases are Fe, Cr enriched A2 phase. A similar microstructure has been observed, and it was attributed to the spinodal decomposition of B2 dendrites into B2 and A2 coexisting phases [20,23,37].  The microstructural details of the laser deposited HEA at different magnifications are shown in Figure 3c,d. The equiaxed grains were delineated by intergranular phases (Figure 3c). The dendritic structure can be observed within the equiaxed grains. The EDS was used to analyze the element distribution, and the results are listed in Table 2 and Figure 4a-g. The dark contrast phase was enriched in Al, Ni, Co, and Ti, and the high level of Fe and Cr concentration was detected in the bright contrast phase. As learned from Table 2, the chemical compositions of Fe and Cr were~45 atom% in the bright contrast phase while were~16 atom% in the dark contrast phase. In combination with the phase identification, the dark contrasted phases rich in Al, Ni, and Ti are B2 phase while the bright contrast phases are Fe, Cr enriched A2 phase. A similar microstructure has been observed, and it was attributed to the spinodal decomposition of B2 dendrites into B2 and A2 coexisting phases [20,23,37].    The chemical mixing enthalpies of element pairs in AlCoCrFeNiTi0.5 alloy are tabulated in Table  3. The phase formation and element segregation are determined by the mixing enthalpy among the constituent metallic elements. It is shown that Al, Ni, and Co has high negative mixing enthalpy with Ti, and they are liable to generate the B2 phase. For example, the mixing enthalpy between Al-Ti, Al-Ni, and Al-Co are −30 kJ/mol, −22 kJ/mol, and −19 kJ/mol, respectively. Furthermore, Cr and Fe tend to form the A2 phase as they exhibit low mixing enthalpy close to zero.
The EBSD inverse pole figure (IPF) image of a region of 1180 µm by 1056 µm taken in the XZ-plane of the as-deposited AlCoCrFeNiTi0.5 alloy is given in Figure 5a. Each individual grain in the IPF image was color-coded based on the relationship between its crystallographic orientation and the building direction, which was vertically upwards (Z direction indicated in Figure 5a). There was no obvious preferred crystallographic texture developed in the alloy because the grains were The chemical mixing enthalpies of element pairs in AlCoCrFeNiTi 0.5 alloy are tabulated in Table 3. The phase formation and element segregation are determined by the mixing enthalpy among the constituent metallic elements. It is shown that Al, Ni, and Co has high negative mixing enthalpy with Ti, and they are liable to generate the B2 phase. For example, the mixing enthalpy between Al-Ti, Al-Ni, and Al-Co are −30 kJ/mol, −22 kJ/mol, and −19 kJ/mol, respectively. Furthermore, Cr and Fe tend to form the A2 phase as they exhibit low mixing enthalpy close to zero.
The EBSD inverse pole figure (IPF) image of a region of 1180 µm by 1056 µm taken in the XZ-plane of the as-deposited AlCoCrFeNiTi 0.5 alloy is given in Figure 5a. Each individual grain in the IPF image was color-coded based on the relationship between its crystallographic orientation and the building direction, which was vertically upwards (Z direction indicated in Figure 5a). There was no obvious preferred crystallographic texture developed in the alloy because the grains were randomly color-coded. It revealed the equiaxed grains from the IPF image. Moreover, the histogram of the grain size measurement is illustrated in Figure 5b. The grain size measured by the vertical intercept was approximately 20.3 µm and was 21.1 µm along with the horizontal intercept. It has been acknowledged that, for a specific alloy, the temperature gradient G and the solidification rate R determine the solidification microstructure during the laser process. As the advancing laser moved away from the substrate, the melt pool retreated, and the solidification from moved upward, which led to the temperature field to low G and high R and the formation of the equiaxed grains. A recent study demonstrated that B2-structured dendrites were frequently fragmented, providing profuse effective nucleation sites, and therefore, promoted equiaxed grain formation in the AlCoCrFeNiTi 0.5 HEA [37].  Figure 6 presents the microhardness of AlCoCrFeNiTi0.5 (Al1.0Ti0.5) HEA and Ti6Al4V (Ti64) substrate. Those of AlCoCrFeNi (Al1.0), CrCuFeNi2 (Al0), Al0.75CrCuFeNi2 (Al0.75), and AISI 304 stainless steel substrate (304 SS) alloys reported in our previous work are shown for comparison [17,18]. The addition of Ti into the AlCoCrFeNi system enhanced the microhardness from 418 HV (Al1.0) to 634 HV (Al1.0Ti0.5), which was about 2.1 times that of the Ti6Al4V substrate. Since Ti has a larger atomic radius than Al (as in Table 3), it would increase lattice distortion and improve the effect of solid solution strengthening. Our previous study reported that Al0 and Al0.75 possessed FCC phase structures, making them less resistant towards localized plastic deformation, thereby exhibiting low hardness (170 HV and 290 HV, respectively). The BCC phase consisted of a decreased number of dislocation slip systems compared to the FCC phase, which could explain the high hardness of AlCoCrFeNiTi0.5 HEA in this work.   [17,18]. The addition of Ti into the AlCoCrFeNi system enhanced the microhardness from 418 HV (Al1.0) to 634 HV (Al1.0Ti0.5), which was about 2.1 times that of the Ti6Al4V substrate. Since Ti has a larger atomic radius than Al (as in Table 3), it would increase lattice distortion and improve the effect of solid solution strengthening. Our previous study reported that Al0 and Al0.75 possessed FCC phase structures, making them less resistant towards localized plastic deformation, thereby exhibiting low hardness (170 HV and 290 HV, respectively). The BCC phase consisted of a decreased number of dislocation slip systems compared to the FCC phase, which could explain the high hardness of AlCoCrFeNiTi 0.5 HEA in this work.  Figure 7 displays the isothermal oxidation results of the AlCoCrFeNiTi0.5 HEA and Ti6Al4V at different temperatures (700 °C and 800 °C), over a period of 45 h in the air atmosphere. As illustrated in Figure 7a, the weight gains (∆ ) of the HEA were found to be 0.23 mg/cm 2 and 0.61 mg/cm 2 at 700 °C and 800 °C, respectively. For the Ti6Al4V alloy, its mass gain was measured at 5.87 mg/cm 2 at 700 °C, and its curve was recorded only within 20 h due to the scale spalling at 800 °C. The oxidation curves were observed to follow the parabolic form. The results of the parabolic rate law plot are shown in Figure 7b-d. Here, the parabolic rate constant, can be considered as a measure of the oxidation resistance, and it was calculated using the Equation (1).

Oxidation Kinetics
where ∆ is the mass gain (mg/cm 2 ), A represents the unit area, is the parabolic rate constant in mg 2 cm −4 h −1 , t is the oxidation time (h), and C is a constant value. In addition, the coefficient determination R 2 values (>0.95) indicate that the parabolic model fits well with the observations in Figure 7b-d. The values were determined as 0.8501 (Ti6Al4V at 700 °C), 0.0011 (HEA at 700 °C) and 0.0077 mg 2 cm −4 h −1 (HEA at 800 °C). For discussion in Section 3.3, the values were converted to the unit of g 2 cm −4 s −1 , i.e., 2.36 × 10 −10 (Ti6Al4V at 700 °C), 3.06 × 10 −13 (HEA at 700 °C), and 2.14 × 10 −12 (HEA at 800 °C). Hence, better oxidation resistance was observed with the HEA as its low weight gain and parabolic rate constant.  Figure 7 displays the isothermal oxidation results of the AlCoCrFeNiTi 0.5 HEA and Ti6Al4V at different temperatures (700 • C and 800 • C), over a period of 45 h in the air atmosphere. As illustrated in Figure 7a, the weight gains (∆W) of the HEA were found to be 0.23 mg/cm 2 and 0.61 mg/cm 2 at 700 • C and 800 • C, respectively. For the Ti6Al4V alloy, its mass gain was measured at 5.87 mg/cm 2 at 700 • C, and its curve was recorded only within 20 h due to the scale spalling at 800 • C. The oxidation curves were observed to follow the parabolic form. The results of the parabolic rate law plot are shown in Figure 7b-d. Here, the parabolic rate constant, K p can be considered as a measure of the oxidation resistance, and it was calculated using the Equation (1).

Oxidation Kinetics
where ∆W is the mass gain (mg/cm 2 ), A represents the unit area, K p is the parabolic rate constant in mg 2 cm −4 h −1 , t is the oxidation time (h), and C is a constant value. In addition, the coefficient determination R 2 values (>0.95) indicate that the parabolic model fits well with the observations in Figure 7b-d. The K p values were determined as 0.8501 (Ti6Al4V at 700 • C), 0.0011 (HEA at 700 • C) and 0.0077 mg 2 cm −4 h −1 (HEA at 800 • C). For discussion in Section 3.3, the K p values were converted to the unit of g 2 cm −4 s −1 , i.e., 2.36 × 10 −10 (Ti6Al4V at 700 • C), 3.06 × 10 −13 (HEA at 700 • C), and 2.14 × 10 −12 (HEA at 800 • C). Hence, better oxidation resistance was observed with the HEA as its low weight gain and parabolic rate constant.  Figure 7 displays the isothermal oxidation results of the AlCoCrFeNiTi0.5 HEA and Ti6Al4V at different temperatures (700 °C and 800 °C), over a period of 45 h in the air atmosphere. As illustrated in Figure 7a, the weight gains (∆ ) of the HEA were found to be 0.23 mg/cm 2 and 0.61 mg/cm 2 at 700 °C and 800 °C, respectively. For the Ti6Al4V alloy, its mass gain was measured at 5.87 mg/cm 2 at 700 °C, and its curve was recorded only within 20 h due to the scale spalling at 800 °C. The oxidation curves were observed to follow the parabolic form. The results of the parabolic rate law plot are shown in Figure 7b-d. Here, the parabolic rate constant, can be considered as a measure of the oxidation resistance, and it was calculated using the Equation (1).

Oxidation Kinetics
where ∆ is the mass gain (mg/cm 2 ), A represents the unit area, is the parabolic rate constant in mg 2 cm −4 h −1 , t is the oxidation time (h), and C is a constant value. In addition, the coefficient determination R 2 values (>0.95) indicate that the parabolic model fits well with the observations in Figure 7b-d. The values were determined as 0.8501 (Ti6Al4V at 700 °C), 0.0011 (HEA at 700 °C) and 0.0077 mg 2 cm −4 h −1 (HEA at 800 °C). For discussion in Section 3.3, the values were converted to the unit of g 2 cm −4 s −1 , i.e., 2.36 × 10 −10 (Ti6Al4V at 700 °C), 3.06 × 10 −13 (HEA at 700 °C), and 2.14 × 10 −12 (HEA at 800 °C). Hence, better oxidation resistance was observed with the HEA as its low weight gain and parabolic rate constant.

Phase Analysis
XRD analyses were performed following oxidation tests, and the results are indicated in Figure  8. The strong diffraction peaks of TiO2 and weak diffraction peaks of Al2O3 were seen from the Ti6Al4V substrate, as in Figure 8a. This illustrated that the scale formed on the Ti6Al4V substrate was mostly composed of TiO2 and a small amount of Al2O3. The TiO2 was a poorly adherent and brittle scale, and Ti and O ions could diffuse through the porous oxides, which resulted in the fast oxidation kinetics. A thick oxide scale was formed and cracked due to the thermal stress at the elevated temperature [2].
For the AlCoCrFeNiTi0.5 alloy oxidized at 700 °C, Al2O3, together with B2 and A2 phases were detected, as shown in Figure 8b. When oxidized at 800 °C, the oxides were Cr2O3, TiO2, Al2O3, and spinel (mainly composed of NiCr2O4), as in Figure 8c. Besides, a Fe-Cr sigma phase with a tetragonal structure (P42/mmm, 136) was detected in AlCoCrFeNiTi0.5 alloy oxidized shown in Figure 8b. A similar phenomenon was described by Wang et al., in which the transformation occurred from BCC to sigma phase at 650 °C [38,39].

Cross-Sectional Morphology of Oxide Scales
The cross-sectional backscattered electron images and the elemental composition distribution of the Ti6Al4V and AlCoCrFeNiTi0.5 HEA oxidized at 700 °C and 800 °C are present in Figure 9. The oxide scales of the Ti6Al4V had a thickness of 25.54 ± 1.85 µm and they were loose, porous, and some cracks can be observed in Figure 9a. Its main composition was TiO2, as the content of Ti was ~25-28 atom% and ~47-60 atom% of O (as in Figure 9b). EDS mapping analysis of the cross-sectional HEA oxidized at 700 °C and 800 °C was performed to reveal the oxide scales better, and the results are

Phase Analysis
XRD analyses were performed following oxidation tests, and the results are indicated in Figure 8. The strong diffraction peaks of TiO 2 and weak diffraction peaks of Al 2 O 3 were seen from the Ti6Al4V substrate, as in Figure 8a. This illustrated that the scale formed on the Ti6Al4V substrate was mostly composed of TiO 2 and a small amount of Al 2 O 3 . The TiO 2 was a poorly adherent and brittle scale, and Ti and O ions could diffuse through the porous oxides, which resulted in the fast oxidation kinetics. A thick oxide scale was formed and cracked due to the thermal stress at the elevated temperature [2].

Phase Analysis
XRD analyses were performed following oxidation tests, and the results are indicated in Figure  8. The strong diffraction peaks of TiO2 and weak diffraction peaks of Al2O3 were seen from the Ti6Al4V substrate, as in Figure 8a. This illustrated that the scale formed on the Ti6Al4V substrate was mostly composed of TiO2 and a small amount of Al2O3. The TiO2 was a poorly adherent and brittle scale, and Ti and O ions could diffuse through the porous oxides, which resulted in the fast oxidation kinetics. A thick oxide scale was formed and cracked due to the thermal stress at the elevated temperature [2].
For the AlCoCrFeNiTi0.5 alloy oxidized at 700 °C, Al2O3, together with B2 and A2 phases were detected, as shown in Figure 8b. When oxidized at 800 °C, the oxides were Cr2O3, TiO2, Al2O3, and spinel (mainly composed of NiCr2O4), as in Figure 8c. Besides, a Fe-Cr sigma phase with a tetragonal structure (P42/mmm, 136) was detected in AlCoCrFeNiTi0.5 alloy oxidized shown in Figure 8b. A similar phenomenon was described by Wang et al., in which the transformation occurred from BCC to sigma phase at 650 °C [38,39].

Cross-Sectional Morphology of Oxide Scales
The cross-sectional backscattered electron images and the elemental composition distribution of the Ti6Al4V and AlCoCrFeNiTi0.5 HEA oxidized at 700 °C and 800 °C are present in Figure 9. The oxide scales of the Ti6Al4V had a thickness of 25.54 ± 1.85 µm and they were loose, porous, and some cracks can be observed in Figure 9a. Its main composition was TiO2, as the content of Ti was ~25-28 atom% and ~47-60 atom% of O (as in Figure 9b). EDS mapping analysis of the cross-sectional HEA oxidized at 700 °C and 800 °C was performed to reveal the oxide scales better, and the results are For the AlCoCrFeNiTi 0.5 alloy oxidized at 700 • C, Al 2 O 3, together with B2 and A2 phases were detected, as shown in Figure 8b. When oxidized at 800 • C, the oxides were Cr 2 O 3 , TiO 2 , Al 2 O 3 , and spinel (mainly composed of NiCr 2 O 4 ), as in Figure 8c. Besides, a Fe-Cr sigma phase with a tetragonal structure (P4 2 /mmm, 136) was detected in AlCoCrFeNiTi 0.5 alloy oxidized shown in Figure 8b. A similar phenomenon was described by Wang et al., in which the transformation occurred from BCC to sigma phase at 650 • C [38,39].

Cross-Sectional Morphology of Oxide Scales
The cross-sectional backscattered electron images and the elemental composition distribution of the Ti6Al4V and AlCoCrFeNiTi 0.5 HEA oxidized at 700 • C and 800 • C are present in Figure 9. The oxide scales of the Ti6Al4V had a thickness of 25.54 ± 1.85 µm and they were loose, porous, and some cracks can be observed in Figure 9a. Its main composition was TiO 2 , as the content of Ti was~25-28 atom% and~47-60 atom% of O (as in Figure 9b). EDS mapping analysis of the cross-sectional HEA oxidized at 700 • C and 800 • C was performed to reveal the oxide scales better, and the results are shown in

Discussion on the Oxidation Behavior
The oxidation mechanism of Ti6Al4V and AlCoCrFeNiTi0.5 HEA can be obtained from the above analysis. The predominant oxide in Ti6Al4V is TiO2, with a small amount of Al2O3 at 700 °C. Since TiO2 is brittle and loose in the oxide film, it is vulnerable to detach from the substrate when it comes to high-temperature oxidation. For the AlCoCrFeNiTi0.5 HEA, it is slightly oxidized at 700 °C, as a thin Al2O3 oxide layer is observed, and the Fe-Cr sigma phase occurs due to the phase transformation. The Al, Cr, and Ti are selectively oxidized and diffuses and enriches into the oxide layer at 800 °C. The TiO2 is distributed in the outermost oxide layer, and continuous protective Cr2O3 and Al2O3 scales are located beneath the TiO2 layer. The behavior of the AlCoCrFeNiTi0.5 HEA is similar to Group II in Ni-Cr-Al alloy systems [25,41]. The oxide map can be explained that the

Discussion on the Oxidation Behavior
The oxidation mechanism of Ti6Al4V and AlCoCrFeNiTi0.5 HEA can be obtained from the above analysis. The predominant oxide in Ti6Al4V is TiO2, with a small amount of Al2O3 at 700 °C. Since TiO2 is brittle and loose in the oxide film, it is vulnerable to detach from the substrate when it comes to high-temperature oxidation. For the AlCoCrFeNiTi0.5 HEA, it is slightly oxidized at 700 °C, as a thin Al2O3 oxide layer is observed, and the Fe-Cr sigma phase occurs due to the phase transformation. The Al, Cr, and Ti are selectively oxidized and diffuses and enriches into the oxide layer at 800 °C. The TiO2 is distributed in the outermost oxide layer, and continuous protective Cr2O3 and Al2O3 scales are located beneath the TiO2 layer. The behavior of the AlCoCrFeNiTi0.5 HEA is similar to Group II in Ni-Cr-Al alloy systems [25,41]. The oxide map can be explained that the

Discussion on the Oxidation Behavior
The oxidation mechanism of Ti6Al4V and AlCoCrFeNiTi 0.5 HEA can be obtained from the above analysis. The predominant oxide in Ti6Al4V is TiO 2, with a small amount of Al 2 O 3 at 700 • C. Since TiO 2 is brittle and loose in the oxide film, it is vulnerable to detach from the substrate when it comes to high-temperature oxidation. For the AlCoCrFeNiTi 0.5 HEA, it is slightly oxidized at 700 • C, as a thin Al 2 O 3 oxide layer is observed, and the Fe-Cr sigma phase occurs due to the phase transformation. The Al, Cr, and Ti are selectively oxidized and diffuses and enriches into the oxide layer at 800 • C. The TiO 2 is distributed in the outermost oxide layer, and continuous protective Cr 2 O 3 and Al 2 O 3 scales are located beneath the TiO 2 layer. The behavior of the AlCoCrFeNiTi 0.5 HEA is similar to Group II in Ni-Cr-Al alloy systems [25,41]. The oxide map can be explained that the concentrations of Al and Cr facilitate the external Cr 2 O 3 with an internal subscale Al 2 O 3 scale. In a review of thermodynamic data, the standard Gibbs free energy of Al 2 O 3 (−891 KJ/mol at 800 • C) formation is more negative than other possible oxides (i.e., Cr 2 O 3 : −569 KJ/mol) in the HEA [42], the growth of Al 2 O 3 should be favorable during the initial stage of oxidation.
It is important and interesting to compare the oxidation rates from our work with other HEAs and conventional alloys. Table 4 collects the parabolic constants K p measured for different HEAs (CrMnFeCoNi , FeCoNiCrAl and Al 0.5 CoCrFeNiTi 0.5 ) and alumina-forming austenitic (AFA) stainless steel. It is worth noting that the examined HEAs are comparable to those similar types of HEAs. The AlCoCrFeNiTi 0.5 HEA is believed to own good oxidation properties due to the sluggish diffusion effect and formation of Al 2 O 3 and Cr 2 O 3 oxides. However, its oxidation resistance is shown to fall short of FeCoNiCrAl, Al 1.5 CoCrFeNiTi 0.5 and AFA steel, i.e., 2-3 order of differences in K p values. According to Butler et al. [24,25], increased Al content enhanced the continuity of the Al 2 O 3 scale, leading to improved oxidation resistance. The formed alumina could exhibit a protective effect at low Ti content. With the addition of Ti, it negatively affected the oxidation behavior of the aluminum-containing HEAs. As described in Erdogan's work [43], a good barrier against oxidation did not form in the Ti-rich CoCrFeNiAl 0.5 Ti due to the fast-growing oxides. Based on these facts, the laser processed AlCoCrFeNiTi 0.5 HEA has great potential under the high-temperature application. A great effort should be put on (1) the investigation of the sequence of oxide formation at the early stage, and (2) the improvement of the oxidation resistance by alloying addition, e.g., Al, Si.

Conclusions
The AlCoCrFeNiTi 0.5 high-entropy alloy (HEA) coating was fabricated by laser metal deposition (LMD) on a Ti6Al4V substrate. The microstructure and isothermal oxidation behavior at 700 • C and 800 • C in air atmosphere were investigated, and the underlying mechanisms were discussed. The main phase constitutions in as-deposited HEA were the Fe, Cr-rich A2 and Al, Ni, and Ti-enriched B2 phases. The isothermal oxidation testing demonstrated that the HEA coatings could effectively improve the oxidation resistance of the Ti6Al4V substrate. The oxidation kinetics of the HEA and Ti6Al4V met the parabolic rate law, while the weight gain and parabolic rate constant of the HEA were lower than Ti6Al4V, implying a better oxidation resistance. The scales of the Ti6Al4V were mainly composed of TiO 2 at 700 • C, and it suffered from spalling at 800 • C. The AlCoCrFeNiTi 0.5 HEA was slightly oxidized at 700 • C as a few oxides were formed. At 800 • C, the formation of continuous Al 2 O 3 , Cr 2 O 3 scales could be ascribed to their good oxidation resistance of the HEA. This work provides an approach to enhance the oxidation resistance of Ti6Al4V alloy and accelerate the broad adoption of HEAs in high-temperature applications.