Dissimilar Infrared Brazing of CoCrFe(Mn)Ni Equiatomic High Entropy Alloys and 316 Stainless Steel

Dissimilar infrared brazing of CoCrFeMnNi/CoCrFeNi equiatomic high entropy alloys and 316 stainless steel using MBF601 and BNi-2 foils was evaluated. The wetting angles of the two fillers at 50 °C above their liquidus temperatures on the three substrates were below 40 degrees. The CoCrFeMnNi/316 SS joint had the highest shear strength of 361 MPa with BNi-2 filler brazing at 1020 °C for 180 s, and fractured at the CrB compound in the joint. The CoCrFeMnNi/MBF601/316 SS joint contained a CoCrFeMnNi-based matrix, phosphides and B-containing compounds. The CoCrFeNi/316 SS joint had the highest shear strength of 374 MPa when brazed with BNi-2 filler at 1020 °C for 600 s, and fractured at the CrB in the joint. The CoCrFeNi/MBF601/316 SS joint consisted of a (Fe,Ni)-rich matrix, phosphides and B/Cr/Fe/P compounds, and the highest shear strength of 324 MPa was achieved when it was brazed at 1080 °C for 600 s.

The joining of HEAs is crucial in the application of such new alloys for industrial applications. There are many studies on welding HEAs [35][36][37]. Good mechanical properties have been achieved in previous studies [38][39][40]. However, very limited researches are focused on brazing HEAs [41]. In comparison with welding, the major advantage of brazing is the dissimilar joining of two alloys with a huge difference in melting points [42,43]. In other words, brazing is usually applied for cases not suitable for welding.  The fractured surfaces, cross-sections of joints and shear test specimens were observed with a field emission scanning electron microscope (FESEM, NOVA NANO 450, FEI Corp., OR, USA) and the quantitative chemical compositions of the brazed specimens were examined with an electron probe microanalyzer (EPMA, JXA-8200, JEOL, Tokyo, Japan) equipped with a wavelength dispersive spectroscope (WDS). Shear tests were performed with a universal tensile test machine (AG-IS, Shimadzu Corp., Kyoto, Japan) with a constant compressive strain rate of 0.0167 mm/s. Three brazed specimens were shear tested to obtain the average shear strength for each brazing condition. Selected fracture surfaces were examined with a Rigaku TTRAX III monochromatized X-ray diffractometer (XRD) for structural analyses. Figure 2 shows the dynamic wetting angle results of BNi-2 and MBF601 filler alloys on 316 SS substrate. For the BNi-2 filler metal at 1020 °C, the wetting angle remained at 155 degrees for the first 20 s and then it fell quickly to 70 degrees at 100 s before decreasing slowly to 53 degrees at 300 s. The wetting angle of the BNi-2 filler metal at 1050 °C fell quickly to 68 degrees in the first 20 s and then decreased slowly to 40 degrees at 300 s. For the MBF601 filler metal at 1050 °C, the wetting angle was 166 degrees at the beginning and fell quickly to 30 degrees at 100 s, after which it decreased slowly to 14 degrees at 300 s. At 1080 °C, the wetting angle was 155 degrees at first, dropped fast to 34 degrees at 30 s and then fell continuously to 6 degrees at 300 s.  The fractured surfaces, cross-sections of joints and shear test specimens were observed with a field emission scanning electron microscope (FESEM, NOVA NANO 450, FEI Corp., OR, USA) and the quantitative chemical compositions of the brazed specimens were examined with an electron probe microanalyzer (EPMA, JXA-8200, JEOL, Tokyo, Japan) equipped with a wavelength dispersive spectroscope (WDS). Shear tests were performed with a universal tensile test machine (AG-IS, Shimadzu Corp., Kyoto, Japan) with a constant compressive strain rate of 0.0167 mm/s. Three brazed specimens were shear tested to obtain the average shear strength for each brazing condition. Selected fracture surfaces were examined with a Rigaku TTRAX III monochromatized X-ray diffractometer (XRD) for structural analyses. Figure 2 shows the dynamic wetting angle results of BNi-2 and MBF601 filler alloys on 316 SS substrate. For the BNi-2 filler metal at 1020 • C, the wetting angle remained at 155 degrees for the first 20 s and then it fell quickly to 70 degrees at 100 s before decreasing slowly to 53 degrees at 300 s. The wetting angle of the BNi-2 filler metal at 1050 • C fell quickly to 68 degrees in the first 20 s and then decreased slowly to 40 degrees at 300 s. For the MBF601 filler metal at 1050 • C, the wetting angle was 166 degrees at the beginning and fell quickly to 30 degrees at 100 s, after which it decreased slowly to 14 degrees at 300 s. At 1080 • C, the wetting angle was 155 degrees at first, dropped fast to 34 degrees at 30 s and then fell continuously to 6 degrees at 300 s. The fractured surfaces, cross-sections of joints and shear test specimens were observed with a field emission scanning electron microscope (FESEM, NOVA NANO 450, FEI Corp., OR, USA) and the quantitative chemical compositions of the brazed specimens were examined with an electron probe microanalyzer (EPMA, JXA-8200, JEOL, Tokyo, Japan) equipped with a wavelength dispersive spectroscope (WDS). Shear tests were performed with a universal tensile test machine (AG-IS, Shimadzu Corp., Kyoto, Japan) with a constant compressive strain rate of 0.0167 mm/s. Three brazed specimens were shear tested to obtain the average shear strength for each brazing condition. Selected fracture surfaces were examined with a Rigaku TTRAX III monochromatized X-ray diffractometer (XRD) for structural analyses. Figure 2 shows the dynamic wetting angle results of BNi-2 and MBF601 filler alloys on 316 SS substrate. For the BNi-2 filler metal at 1020 °C, the wetting angle remained at 155 degrees for the first 20 s and then it fell quickly to 70 degrees at 100 s before decreasing slowly to 53 degrees at 300 s. The wetting angle of the BNi-2 filler metal at 1050 °C fell quickly to 68 degrees in the first 20 s and then decreased slowly to 40 degrees at 300 s. For the MBF601 filler metal at 1050 °C, the wetting angle was 166 degrees at the beginning and fell quickly to 30 degrees at 100 s, after which it decreased slowly to 14 degrees at 300 s. At 1080 °C, the wetting angle was 155 degrees at first, dropped fast to 34 degrees at 30 s and then fell continuously to 6 degrees at 300 s.   Figure 3 shows the dynamic wetting angles of BNi-2 and MBF601 fillers on the CoCrFeNi substrate. The wetting angle of the BNi-2 filler metal at 1020 • C remained at 149 degrees for the first 30 s, dropped to 40 degrees at 100 s and fell to 28 degrees at 300 s. For the BNi-2 filler metal at 1050 • C, the wetting angle remained at 149 degrees for the first 10 s and dropped to 40 degrees at 50 s before reaching 26 degrees at 300 s. The wettability of the MBF601 filler metal at 1050 • C was not good-the wetting angle was 147 degrees for the first 10 s, after which it dropped to 110 degrees at 50 s before reaching 77 degrees at 300 s. The wetting angle at 1080 • C was 144 degrees at first, dropped to 28 degrees at 45 s and reached 19 degrees at 300 s. Crystals 2019, 9, x FOR PEER REVIEW 4 of 17 Figure 3 shows the dynamic wetting angles of BNi-2 and MBF601 fillers on the CoCrFeNi substrate. The wetting angle of the BNi-2 filler metal at 1020 °C remained at 149 degrees for the first 30 s, dropped to 40 degrees at 100 s and fell to 28 degrees at 300 s. For the BNi-2 filler metal at 1050 °C, the wetting angle remained at 149 degrees for the first 10 s and dropped to 40 degrees at 50 s before reaching 26 degrees at 300 s. The wettability of the MBF601 filler metal at 1050 °C was not good-the wetting angle was 147 degrees for the first 10 s, after which it dropped to 110 degrees at 50 s before reaching 77 degrees at 300 s. The wetting angle at 1080 °C was 144 degrees at first, dropped to 28 degrees at 45 s and reached 19 degrees at 300 s.

Wetting Behavior
There were rapid/slow changes in the wetting angles of Figure 2; Figure 3. The initial wetting of the braze melt is strongly related to rapid solid/liquid interfacial reaction at the early stage of the wetting test. The wetting angle was stabilized by increasing the test time because the mass transport of elements is confined by diffusion. Much slower changes in wetting angles were observed in Figures  2 and 3. The BNi-2 filler metal wet the CoCrFeNi substrate well at both 1020 °C and 1050 °C, but its wettability was poor on 316 SS substrate. The MBF601 filler metal wet 316 SS well at 1050 °C and 1080 °C and wet CoCrFeNi well at 1080 °C, but the wettability was poor at 1050 °C. In a previous study [41], the wettability of the BNi-2 and MBF601 filler metals on the CoCrFeMnNi HEA was investigated, and the wetting angles of the BNi-2 filler metal at 1020 °C and 1050 °C were found to be 80 and 18 degrees at 300 s, respectively. For the MBF601 filler metal on the CoCrFeMnNi substrate, the wetting angles at 1050 °C and 1080 °C were 86 and 8 degrees at 300 s, respectively. Both the BNi-2 and MBF601 filler metals are nickel-based filler metals, which are applicable to brazing Fe/Ni/Cobased alloys [44]. Good wettability is also achieved by the interfacial reaction between the filler metal and substrate. This explains why the wettability was good on these three substrates at temperatures 50 °C above their liquidus temperatures.  Table 2 show BEIs (backscattered electron images) and the EPMA quantitative chemical analyses of joints infrared brazed at 1020 °C for 180 s and 600 s and at 1050 °C for 600 s. Both Ni-based filler alloys contain no carbon. The carbon content of 316 stainless steel is as low as 0.08 wt%. The solubility of carbon in Ni and γ-Fe is 0.6 wt% and 2.1 wt%, respectively [49]. There is no carbide formation in the experimental observation. In Figure 4a, the substrate marked A is the CoCrFeMnNi HEA and the substrate marked F is 316 SS. The joint brazed at 1020 °C for 180 s was dominated by a Ni-rich matrix, marked C, with continuous CrB, marked D, in the center of the brazed There were rapid/slow changes in the wetting angles of Figures 2 and 3. The initial wetting of the braze melt is strongly related to rapid solid/liquid interfacial reaction at the early stage of the wetting test. The wetting angle was stabilized by increasing the test time because the mass transport of elements is confined by diffusion. Much slower changes in wetting angles were observed in Figures 2 and 3.

CoCrFeMnNi/BNi-2/316 Infrared Joints
The BNi-2 filler metal wet the CoCrFeNi substrate well at both 1020 • C and 1050 • C, but its wettability was poor on 316 SS substrate. The MBF601 filler metal wet 316 SS well at 1050 • C and 1080 • C and wet CoCrFeNi well at 1080 • C, but the wettability was poor at 1050 • C. In a previous study [41], the wettability of the BNi-2 and MBF601 filler metals on the CoCrFeMnNi HEA was investigated, and the wetting angles of the BNi-2 filler metal at 1020 • C and 1050 • C were found to be 80 and 18 degrees at 300 s, respectively. For the MBF601 filler metal on the CoCrFeMnNi substrate, the wetting angles at 1050 • C and 1080 • C were 86 and 8 degrees at 300 s, respectively. Both the BNi-2 and MBF601 filler metals are nickel-based filler metals, which are applicable to brazing Fe/Ni/Co-based alloys [44]. Good wettability is also achieved by the interfacial reaction between the filler metal and substrate. This explains why the wettability was good on these three substrates at temperatures 50 • C above their liquidus temperatures. Figure 4 and Table 2 show BEIs (backscattered electron images) and the EPMA quantitative chemical analyses of joints infrared brazed at 1020 • C for 180 s and 600 s and at 1050 • C for 600 s. Both Ni-based filler alloys contain no carbon. The carbon content of 316 stainless steel is as low as 0.08 wt%. The solubility of carbon in Ni and γ-Fe is 0.6 wt% and 2.1 wt%, respectively [49]. There is no carbide formation in the experimental observation. In Figure 4a, the substrate marked A is the CoCrFeMnNi HEA and the substrate marked F is 316 SS. The joint brazed at 1020 • C for 180 s was dominated by a Ni-rich matrix, marked C, with continuous CrB, marked D, in the center of the brazed zone. There were two boride layers composed of fine borides at the interface between the braze and the two substrates. A few coarsened borides penetrated into the grain boundaries of both substrates. The boride layer on the CoCrFeMnNi side, marked B in Figure 4a, was thicker than that on the 316 SS side, marked E in Figure 4a. When the brazing time was increased from 180 s to 600 s, the amount of CrB in the center of the brazed zone diminished, but the CrB coarsened. However, the boride layers on both sides thickened. Increasing the brazing temperature from 1020 • C to 1050 • C caused the width of the Ni-rich matrix to thicken, and most of the CrB in the brazed zone disappeared. Boron atoms diffused into both substrates, and both boride layers became much thicker. Additionally, the borides growing along the grain boundaries penetrated much more deeply into the substrates. zone. There were two boride layers composed of fine borides at the interface between the braze and the two substrates. A few coarsened borides penetrated into the grain boundaries of both substrates. The boride layer on the CoCrFeMnNi side, marked B in Figure 4a, was thicker than that on the 316 SS side, marked E in Figure 4a. When the brazing time was increased from 180 s to 600 s, the amount of CrB in the center of the brazed zone diminished, but the CrB coarsened. However, the boride layers on both sides thickened. Increasing the brazing temperature from 1020 °C to 1050 °C caused the width of the Ni-rich matrix to thicken, and most of the CrB in the brazed zone disappeared. Boron atoms diffused into both substrates, and both boride layers became much thicker. Additionally, the borides growing along the grain boundaries penetrated much more deeply into the substrates.     Figure 5 and Table 3 show the BEIs and the quantitative chemical analyses of CoCrFeMnNi/MBF601/316 joints infrared brazed at 1050 • C for 180 s and 600 s and at 1080 • C for 600 s. As shown in Figure 5a, the brazed joint was composed of a CoCrFeMnNi-based matrix, marked I, and some phosphides, marked J, and there was a row of B/Co/Cr/Fe/Mn/Ni/P compounds, marked H, between the CoCrFeMnNi substrate and the brazed zone. Also, some phosphides, marked G, penetrated into the grain boundaries of the CoCrFeMnNi substrate. When the brazing time at 1050 • C was increased from 180 s to 600 s, the phosphides in the brazed joint diminished in number and size. When the brazing temperature was changed from 1050 • C to 1080 • C, the phosphides in the brazed zone almost disappeared and the B/Co/Cr/Fe/Mn/Ni/P compounds became much larger.

CoCrFeMnNi/BNi-2/316 Infrared Joints
Crystals 2019, 9, x FOR PEER REVIEW 6 of 17 Figure 5 and Table 3 show the BEIs and the quantitative chemical analyses of CoCrFeMnNi/MBF601/316 joints infrared brazed at 1050 °C for 180 s and 600 s and at 1080 °C for 600 s. As shown in Figure 5a, the brazed joint was composed of a CoCrFeMnNi-based matrix, marked I, and some phosphides, marked J, and there was a row of B/Co/Cr/Fe/Mn/Ni/P compounds, marked H, between the CoCrFeMnNi substrate and the brazed zone. Also, some phosphides, marked G, penetrated into the grain boundaries of the CoCrFeMnNi substrate. When the brazing time at 1050 °C was increased from 180 s to 600 s, the phosphides in the brazed joint diminished in number and size. When the brazing temperature was changed from 1050 °C to 1080 °C, the phosphides in the brazed zone almost disappeared and the B/Co/Cr/Fe/Mn/Ni/P compounds became much larger.     Figure 6 and Table 4 show the BEIs and quantitative chemical analyses of CoCrFeNi/BNi-2/316 joints infrared brazed at 1020 • C for 180 and 600 s and at 1050 • C for 600 s. In Figure 6a, the areas marked K and P are the CoCrFeNi and 316 SS substrates, respectively. The CoCrFeNi/BNi-2/316 joint infrared brazed at 1020 • C for 180 s was composed of a Ni-rich matrix and tiny CrB in the brazed zone, and there were interfacial boride layers between the braze and the two substrates. Some borides penetrated along the grain boundaries of the two substrates. When the brazing time was increased from 180 s to 600 s at 1020 • C, the amount of CrB in the brazed zone was reduced and the size coarsened, and the boride layer thickened. Increasing the brazing temperature from 1020 • C to 1050 • C reduced the amount of central CrB in the brazed zone due to the dissolution of the CrB at higher temperatures.  Figure 6 and Table 4 show the BEIs and quantitative chemical analyses of CoCrFeNi/BNi-2/316 joints infrared brazed at 1020 °C for 180 and 600 s and at 1050 °C for 600 s. In Figure 6a, the areas marked K and P are the CoCrFeNi and 316 SS substrates, respectively. The CoCrFeNi/BNi-2/316 joint infrared brazed at 1020 °C for 180 s was composed of a Ni-rich matrix and tiny CrB in the brazed zone, and there were interfacial boride layers between the braze and the two substrates. Some borides penetrated along the grain boundaries of the two substrates. When the brazing time was increased from 180 s to 600 s at 1020 °C, the amount of CrB in the brazed zone was reduced and the size coarsened, and the boride layer thickened. Increasing the brazing temperature from 1020 °C to 1050 °C reduced the amount of central CrB in the brazed zone due to the dissolution of the CrB at higher temperatures.     Figure 7 and Table 5 show the BEIs and quantitative chemical analyses of CoCrFeNi/MBF601/316 SS joints infrared brazed at 1050 • C for 180, 300, and 600 s and at 1080 • C for 600 s. The joint was composed of a (Fe,Ni)-rich matrix, marked Q, phosphides, marked R, and a few B/Cr/Fe/P compounds, marked S. Because the P cannot be alloyed into transition metals as high as 31.4 at%, the R phase in Figure 7a and Table 5 was categorized as a phosphide. Since BNi-2 contains no P, only boride was observed in the BNi-2 brazed joint. In contrast, the phosphide was observed in the MBF601 brazed joint because it was alloyed with high P content. Both are brittle compounds. Some phosphides penetrated into the grain boundaries of the CoCrFeNi and 316 SS substrates. The microstructure of the phosphides and the (Fe,Ni)-rich matrix appeared to be eutectic. However, no related phase diagrams are available. When the brazing time or temperature was increased, the thickness of the brazed zone decreased. It resulted from the dissolution of these compounds into both CoCrFeNi and 316 SS substrates. Figure 8 presents EPMA mappings to reveal the distributions of alloy elements in the brazed zone illustrated in Figure 7a. Elemental distributions along phases in Figure 8 are consistent with those in Figure 7a and Table 5.  Figure 7 and Table 5 show the BEIs and quantitative chemical analyses of CoCrFeNi/MBF601/316 SS joints infrared brazed at 1050 °C for 180, 300, and 600 s and at 1080 °C for 600 s. The joint was composed of a (Fe,Ni)-rich matrix, marked Q, phosphides, marked R, and a few B/Cr/Fe/P compounds, marked S. Because the P cannot be alloyed into transition metals as high as 31.4 at%, the R phase in Figure 7a and Table 5 was categorized as a phosphide. Since BNi-2 contains no P, only boride was observed in the BNi-2 brazed joint. In contrast, the phosphide was observed in the MBF601 brazed joint because it was alloyed with high P content. Both are brittle compounds. Some phosphides penetrated into the grain boundaries of the CoCrFeNi and 316 SS substrates. The microstructure of the phosphides and the (Fe,Ni)-rich matrix appeared to be eutectic. However, no related phase diagrams are available. When the brazing time or temperature was increased, the thickness of the brazed zone decreased. It resulted from the dissolution of these compounds into both CoCrFeNi and 316 SS substrates. Figure 8 presents EPMA mappings to reveal the distributions of alloy elements in the brazed zone illustrated in Figure 7a. Elemental distributions along phases in Figure 8 are consistent with those in Figure 7a and Table 5.      Table 6 shows that the average shear strengths and thicknesses of CoCrFeMnNi/BNi-2/316 SS joints brazed at 1020 °C for 180 and 600 s and at 1050 °C for 600 s were 361, 294 and 328 MPa, respectively. The joint brazed at 1020 °C for 600 s had the lowest shear strength. Figure 9 shows the BEI cross-sections and secondary electron image (SEI) fractographs of joints infrared brazed at 1020 °C for 600 s and at 1050 °C for 600 s, respectively. As can be seen in Figures 9a and 4b, the fracture was located at the central continuous CrB in the brazed zone. The coarse central CrB in the joint resulted in decreased shear strength. The cleavage fracture surface is shown in Figure 9b. Figure 9c displays the XRD analysis of the fractured surface after the shear test. The fractured surfaces were primarily comprised of a FCC Ni-rich matrix and a CrB intermetallic compound for the BNi-2 brazed joint. When the brazing temperature was increased to 1050 °C, a fracture occurred in the coarsened  Table 6 shows that the average shear strengths and thicknesses of CoCrFeMnNi/BNi-2/316 SS joints brazed at 1020 • C for 180 and 600 s and at 1050 • C for 600 s were 361, 294 and 328 MPa, respectively. The joint brazed at 1020 • C for 600 s had the lowest shear strength. Figure 9 shows the BEI cross-sections and secondary electron image (SEI) fractographs of joints infrared brazed at 1020 • C for 600 s and at 1050 • C for 600 s, respectively. As can be seen in Figures 4b and 9a, the fracture was located at the central continuous CrB in the brazed zone. The coarse central CrB in the joint resulted in decreased shear strength. The cleavage fracture surface is shown in Figure 9b. Figure 9c displays the XRD analysis of the fractured surface after the shear test. The fractured surfaces were primarily comprised of a FCC Ni-rich matrix and a CrB intermetallic compound for the BNi-2 brazed joint. When the brazing temperature was increased to 1050 • C, a fracture occurred in the coarsened interfacial boride layer on the CoCrFeMnNi side, as shown in Figure 9d, due to the dissolution of the central CrB into both substrates. Its fractograph indicated a brittle fracture, as shown in Figure 9e. interfacial boride layer on the CoCrFeMnNi side, as shown in Figure 9d, due to the dissolution of the central CrB into both substrates. Its fractograph indicated a brittle fracture, as shown in Figure 9e.   In a previous study of a similar infrared-brazed CoCrFeMnNi/BNi-2/CoCrFeMnNi joint, large CrB formed after brazing at 1020 °C for 180 s, and the joint had the lowest shear strength of 193 MPa [41]. The major alloying elements in 316 SS are Fe, Ni and Cr. These elements are also included in HEAs. The dissolution/reaction of these elements in 316 SS with filler metal are similar to HEAs. In the CoCrFeMnNi/BNi-2/316 SS dissimilar joint infrared brazed at 1020 °C for 180 s, there were no large CrB in the joint, and it had the highest shear strength of 361 MPa. The CrB became larger when the brazing time increased to 600 s. The fracture was usually located at the central large and continuous CrB borides, which caused the decrease in shear strength. It is worth mentioning that good wettability is a principal precondition to form a high-quality brazing joint. On the other hand, the formation and growth of interfacial compounds, e.g., CrB, was responsible for the fracture origin in the joints.

Shear Strength and Failure Analyses
The shear strengths of CoCrFeMnNi/MBF601/316 SS infrared joints infrared brazed at 1050 °C for 180 and 600 s and at 1080 °C for 600 s were 216, 276 and 248 MPa, respectively. For the joint brazed at 1050 °C, the shear strength increased because of better wettability. The BEI cross-sections of a CoCrFeMnNi/MBF601/316 SS joint infrared brazed at 1080 °C for 600 s are shown in Figures 10a,b. In all the MBF601 joints, cracks propagated along the phosphides in the grain boundaries of the CoCrFeMnNi substrate and along the B/Co/Cr/Fe/Mn/Ni/P compounds between the CoCrFeMnNi substrate. The granular fracture surface of the CoCrFeMnNi substrate is shown in Figure 10c. In the joint brazed at 1080 °C for 600 s, a few solidification shrinkage voids were visible in the SEI fractograph, as shown in Figure 10d. The shrinkage voids in brazing HEA have been observed in a previous study [41]. The presence of solidification shrinkage voids caused the decrease in shear strength. In a previous study of a similar infrared-brazed CoCrFeMnNi/BNi-2/CoCrFeMnNi joint, large CrB formed after brazing at 1020 • C for 180 s, and the joint had the lowest shear strength of 193 MPa [41]. The major alloying elements in 316 SS are Fe, Ni and Cr. These elements are also included in HEAs. The dissolution/reaction of these elements in 316 SS with filler metal are similar to HEAs. In the CoCrFeMnNi/BNi-2/316 SS dissimilar joint infrared brazed at 1020 • C for 180 s, there were no large CrB in the joint, and it had the highest shear strength of 361 MPa. The CrB became larger when the brazing time increased to 600 s. The fracture was usually located at the central large and continuous CrB borides, which caused the decrease in shear strength. It is worth mentioning that good wettability is a principal precondition to form a high-quality brazing joint. On the other hand, the formation and growth of interfacial compounds, e.g., CrB, was responsible for the fracture origin in the joints.
The shear strengths of CoCrFeMnNi/MBF601/316 SS infrared joints infrared brazed at 1050 • C for 180 and 600 s and at 1080 • C for 600 s were 216, 276 and 248 MPa, respectively. For the joint brazed at 1050 • C, the shear strength increased because of better wettability. The BEI cross-sections of a CoCrFeMnNi/MBF601/316 SS joint infrared brazed at 1080 • C for 600 s are shown in Figure 10a,b. In all the MBF601 joints, cracks propagated along the phosphides in the grain boundaries of the CoCrFeMnNi substrate and along the B/Co/Cr/Fe/Mn/Ni/P compounds between the CoCrFeMnNi substrate. The granular fracture surface of the CoCrFeMnNi substrate is shown in Figure 10c. In the joint brazed at 1080 • C for 600 s, a few solidification shrinkage voids were visible in the SEI fractograph, as shown in Figure 10d. The shrinkage voids in brazing HEA have been observed in a previous study [41]. The presence of solidification shrinkage voids caused the decrease in shear strength. In a previous study [41], similar infrared brazed CoCrFeMnNi/MBF601/CoCrFeMnNi joints fractured at the phosphides that formed along the grain boundaries of the CoCrFeMnNi substrate, and some solidification shrinkage voids were observed in the joint brazed at 1080 °C for 600 s. A similar result was found in this study. The high brazing temperature caused the fast chemical reaction of phosphorus and the substrate. The consumption of P from the braze melt results in increasing the liquidus temperature of the braze melt. Isothermal solidification of the residual braze melt proceeds and forms solidification shrinkage voids, as illustrated in Figure 10d. Table 6 shows the average shear strengths of CoCrFeNi/BNi-2/316 SS infrared joints infrared brazed at 1020 °C for 180 and 600 s and at 1050 °C for 600 s, which were 349, 374 and 329 MPa, respectively. The fracture was located in the central CrB of the brazed zone in the specimens brazed at 1020 °C for 180 s, as shown in Figures 6a and 11a. There was no continuous line of CrB in the joint brazed at 1050 °C for 600 s, as displayed in Figure 6c, so the crack was located in the coarsened interfacial boride layer on the 316 SS side, as shown in Figure 11c. Cleavage fractographs are presented in Figures 11b,d. In a previous study [41], similar infrared brazed CoCrFeMnNi/MBF601/CoCrFeMnNi joints fractured at the phosphides that formed along the grain boundaries of the CoCrFeMnNi substrate, and some solidification shrinkage voids were observed in the joint brazed at 1080 • C for 600 s. A similar result was found in this study. The high brazing temperature caused the fast chemical reaction of phosphorus and the substrate. The consumption of P from the braze melt results in increasing the liquidus temperature of the braze melt. Isothermal solidification of the residual braze melt proceeds and forms solidification shrinkage voids, as illustrated in Figure 10d. Table 6 shows the average shear strengths of CoCrFeNi/BNi-2/316 SS infrared joints infrared brazed at 1020 • C for 180 and 600 s and at 1050 • C for 600 s, which were 349, 374 and 329 MPa, respectively. The fracture was located in the central CrB of the brazed zone in the specimens brazed at 1020 • C for 180 s, as shown in Figures 6a and 11a. There was no continuous line of CrB in the joint brazed at 1050 • C for 600 s, as displayed in Figure 6c, so the crack was located in the coarsened interfacial boride layer on the 316 SS side, as shown in Figure 11c. Cleavage fractographs are presented in Figure 11b The shear strengths of CoCrFeNi/MBF601/316 SS joints infrared brazed at 1050 °C for 180 and 600 s and at 1080 °C for 600 s were 293, 284 and 324 MPa, respectively. Figure 12 shows the BEI crosssection and SEI fractograph of the CoCrFeNi/MBF601/316 SS joint brazed at 1080 °C for 600 s. All the CoCrFeNi/MBF601/316 joints fractured at the phosphides in the brazed zone. Figure 12a shows the phosphides penetrating into the grain boundaries of the CoCrFeNi substrate and forming cracks, which are marked by arrows. Figure 12b presents the brittle fractograph of the CoCrFeNi substrate after the shear test. Figure 12c shows the XRD analysis of the fractured surface after the shear test. Only the FCC Ni-rich matrix can be identified from its fractograph. There are a few unidentified peaks in Figure 12c because of the existence of phosphides and B /Cr/Fe/P compounds in the joint. The shear strengths of CoCrFeNi/MBF601/316 SS joints infrared brazed at 1050 • C for 180 and 600 s and at 1080 • C for 600 s were 293, 284 and 324 MPa, respectively. Figure 12 shows the BEI cross-section and SEI fractograph of the CoCrFeNi/MBF601/316 SS joint brazed at 1080 • C for 600 s. All the CoCrFeNi/MBF601/316 joints fractured at the phosphides in the brazed zone. Figure 12a shows the phosphides penetrating into the grain boundaries of the CoCrFeNi substrate and forming cracks, which are marked by arrows. Figure 12b presents the brittle fractograph of the CoCrFeNi substrate after the shear test. Figure 12c shows the XRD analysis of the fractured surface after the shear test. Only the FCC Ni-rich matrix can be identified from its fractograph. There are a few unidentified peaks in Figure 12c because of the existence of phosphides and B /Cr/Fe/P compounds in the joint.
section and SEI fractograph of the CoCrFeNi/MBF601/316 SS joint brazed at 1080 °C for 600 s. All the CoCrFeNi/MBF601/316 joints fractured at the phosphides in the brazed zone. Figure 12a shows the phosphides penetrating into the grain boundaries of the CoCrFeNi substrate and forming cracks, which are marked by arrows. Figure 12b presents the brittle fractograph of the CoCrFeNi substrate after the shear test. Figure 12c shows the XRD analysis of the fractured surface after the shear test. Only the FCC Ni-rich matrix can be identified from its fractograph. There are a few unidentified peaks in Figure 12c because of the existence of phosphides and B /Cr/Fe/P compounds in the joint. In Table 6, the thickness of the brazed zone increased as the brazing temperature/time for the BNi-2 braze alloy increased. However, for the MBF601 filler, the thickness of the brazed zone decreased as the brazing temperature/time increased. For the BNi-2 filler, the dissolution of central CrB into the substrates resulted in the formation of an interfacial boride layer. There were no interfacial boride layers for the MBF601 filler brazed joint. Central phosphides and B/Co/Cr/Fe/Mn/Ni/P compounds were dissolved into both substrates, and the thickness of the brazed zone decreased.

Conclusion
The dissimilar infrared brazing of the CoCrFeMnNi/CoCrFeNi HEA and 316 SS using BNi-2 and MBF601 filler metals was investigated. The target of this study was to optimize the conditions in the dissimilar brazing of HEAs and 316 stainless steel. The wettability, microstructure evolution and shear strength results are summarized below: 1. The wetting angles of BNi-2 and MBF601 fillers at 50 °C above their liquidus temperatures on the CoCrFeMnNi/CoCrFeNi HEA and 316 SS substrates were below 40 degrees. Both fillers showed acceptable wetting ability on the CoCrFeMnNi/CoCrFeNi HEAs and 316 SS for brazing.
2. The CoCrFeMnNi/BNi-2/316 SS joint brazed at 1020 °C was dominated by a Ni-rich matrix with continuous central CrB in the brazed zone. Increasing the brazing temperature caused the central CrB to dissolve into the substrates and form an interfacial boride layer, primarily along the grain boundaries of the substrate. The fracture location was changed from continuous central CrB in the brazed zone into the interfacial boride layer. Similar experimental results were observed in the dissimilar brazing of CoCrFeNi and 316 SS using the BNi-2 filler metal.
3. The CoCrFeMnNi/MBF601/316 SS joint was composed of a CoCrFeMnNi matrix, central phosphides and B/Co/Cr/Fe/Mn/Ni/P compounds, and a few phosphides penetrated into the grain In Table 6, the thickness of the brazed zone increased as the brazing temperature/time for the BNi-2 braze alloy increased. However, for the MBF601 filler, the thickness of the brazed zone decreased as the brazing temperature/time increased. For the BNi-2 filler, the dissolution of central CrB into the substrates resulted in the formation of an interfacial boride layer. There were no interfacial boride layers for the MBF601 filler brazed joint. Central phosphides and B/Co/Cr/Fe/Mn/Ni/P compounds were dissolved into both substrates, and the thickness of the brazed zone decreased.

Conclusions
The dissimilar infrared brazing of the CoCrFeMnNi/CoCrFeNi HEA and 316 SS using BNi-2 and MBF601 filler metals was investigated. The target of this study was to optimize the conditions in the dissimilar brazing of HEAs and 316 stainless steel. The wettability, microstructure evolution and shear strength results are summarized below: 1. The wetting angles of BNi-2 and MBF601 fillers at 50 • C above their liquidus temperatures on the CoCrFeMnNi/CoCrFeNi HEA and 316 SS substrates were below 40 degrees. Both fillers showed acceptable wetting ability on the CoCrFeMnNi/CoCrFeNi HEAs and 316 SS for brazing.
2. The CoCrFeMnNi/BNi-2/316 SS joint brazed at 1020 • C was dominated by a Ni-rich matrix with continuous central CrB in the brazed zone. Increasing the brazing temperature caused the central CrB to dissolve into the substrates and form an interfacial boride layer, primarily along the grain boundaries of the substrate. The fracture location was changed from continuous central CrB in the brazed zone into the interfacial boride layer. Similar experimental results were observed in the dissimilar brazing of CoCrFeNi and 316 SS using the BNi-2 filler metal.
3. The CoCrFeMnNi/MBF601/316 SS joint was composed of a CoCrFeMnNi matrix, central phosphides and B/Co/Cr/Fe/Mn/Ni/P compounds, and a few phosphides penetrated into the grain boundaries of the CoCrFeMnNi base metal. When the brazing time and temperature were increased, the central phosphides in the brazed zone almost disappeared and the B/Co/Cr/Fe/Mn/Ni/P compounds became larger. After the shear test, cracks were located at the B/Co/Cr/Fe/Mn/Ni/P compounds and the phosphides penetrating along the grain boundaries of the CoCrFeMnNi base metal.
4. The CoCrFeNi/MBF601/316 SS joint was composed of a (Fe,Ni)-rich matrix and a phosphide eutectic structure with a few B/Cr/Fe/P compounds. Some of the phosphides penetrated into the grain boundaries of the CoCrFeNi and the 316 SS substrates.