Effects of Adding Active Elements to Aluminum-Based Filler Alloys on the Bonding of 6061 Aluminum Alloy and Alumina

: In this study, AA6061/AA6061 and AA6061/alumina were directly brazed with Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE ﬁller alloys at 530 ◦ C for 10 min without the use of ﬂux. The addition of titanium and rare-earth elements into Al10.8Si10Cu alloy effectively improved the bonding shear strengths of AA6061/AA6061 and AA6061/alumina joints. The highest joint shear strengths were 61.1 and 19.2 MPa, respectively. The Al10.8Si10Cu ﬁller alloy without titanium and rare-earth elements could not wet on the alumina and caused failure of the AA6061/alumina joint. The shear strengths of the AA6061/AA6061 and AA6061/alumina joints both strongly depended on the active element addition. Due to the high chemical activity of the rare-earth elements, they formed AlLa between the Al10Si10Cu4Ti0.1RE ﬁller alloy and alumina. The addition of rare-earth elements into Al10Si10Cu4Ti ﬁller alloy resulted in signiﬁcant enhancement of the average bond strength of AA6061/alumina joints, from 8.0 to 14.8 MPa.


Introduction
The joining of alumina to aluminum alloy is widely used in electronic devices, chemical equipment and biomedicine [1][2][3]. In recent years, several bonding methods for joining metal with ceramic have been developed [4,5]. Among the variety of joining technologies, active brazing is an excellent method of joining metal with ceramic [6,7]. Ag-Cu-Ti alloys are adopted as highly reliable filler metals for ceramic/metal brazing [8][9][10]. However, the bonding temperature is above 850 • C, which is too high relative to the aluminum alloys. Therefore, it is necessary to design filler alloys with low melting points for joining aluminum alloy and ceramic. Active low-melting-point filler alloys such as Sn-, In-and Zn-based filler alloys with elements having high chemical activity, such as titanium, magnesium and rare-earth elements (RE), have been developed in the past few years [11][12][13][14][15][16]. Due to the low mechanical properties at high temperature, the Sn-and In-based filler metals are limited for high-temperature application. Zn-based filler metals with higher melting points can be used at temperatures below about 350 • C. However, zinc is prone to oxidation and causes problems with wettability and corrosion. In addition, Zn-based filler metals are not acceptable in many high-vacuum and high-temperature applications due to the high vapor pressure of zinc. The Al12Si eutectic alloy has been adopted as a reliable filler alloy for aluminum alloy brazing [17]. However, the brazing temperature is about 600 • C, which is too high relative to the solidus temperature of many aluminum alloys. Thus, it is necessary to design low-melting-point aluminum filler alloys. The general trend is that the addition of copper, zinc, magnesium and germanium into the eutectic Al12Si filler alloy lowers the brazing temperature [18][19][20]. In a series of investigations [18,19,21,22], the

Experimental
The Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys used in the study were prepared by melting 99.99% purity copper and titanium slugs, mixed rare-earth elements and the master alloy of Al-12 wt.% Si (supplied by Degussa AG, Hanau, Germany) in an electric arc furnace under a high-purity inert atmosphere. The rare-earth elements used in the study were mixed rare-earth elements. The chemical composition was 77.82 wt.% lanthanum, 16.84 wt.% praseodymium, 3.24 wt.% cerium and 2.10 wt.% neodymium. The chemical composition of the filler alloys used in the study is shown in Table 1. To ensure a homogeneous composition within the filler alloys, the melt was stirred by manual operation for 10 min. The filler alloys were solidified in a mold with an internal diameter of 20 mm under water cooling, and the cast ingots were rolled to a thickness of 300 µm. Prior to microstructural observations, the three filler alloys were metallographically ground and polished. The microstructures were observed with a field emission scanning electron microscope (FE-SEM, JEOL JSM-6700F). The chemical composition was analyzed by energy dispersive X-ray analysis (EDX). The existing phases of the filler alloys were identified with an X-ray diffractometer (Rigakue ATX-E) with Cu Kα radiation (λ = 1.54 Å) and a scanning rate of 1.8 • /min. The thermal properties of the Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu0.1RE filler alloys were determined by differential thermal analysis (DTA, TA Instruments SDT-2960) at a heating rate of 10 • C/min from ambient to 700 • C under a high-purity inert atmosphere. The aluminum alloy joined in this study was in the form of 2 mm thick AA6061 aluminum plates, with dimensions of 40 × 10 × 2 mm. The chemical composition of AA6061 is provided in Table 2. Bulk alumina ceramic with dimensions of 7 × 7 × 4 mm was fabricated by pressure casting. After 24 h of drying at 60 • C, the alumina was sintered at 1600 • C for 2 h. A lap joint configuration was used to evaluate the bonding strength. The geometry and dimensions of the AA6061/AA6061 and AA6061/alumina joints subjected to shear testing are demonstrated in Figure 1a,b, respectively.  Prior to joining, the bonding surfaces of AA6061 and the surfaces of the filler alloys were ground with 1200-grit silicon carbide paper and then cleaned in acetone. Copper and aluminum have high vapor pressure at the bonding temperature in order to prevent the evaporation of the elements during the bonding process. Therefore, the argon partial pressure is introduced into a vacuum furnace during the bonding process. The brazing process was conducted with a heating rate of 10 °C/min in a vacuum furnace under a high-purity Ar gas atmosphere at 530 °C for 10 min. For the metallographic study of the bonding interfaces, the joined specimens were cross-sectioned. In order to avoid damage to the joints caused by metallographic production, after the completion of the bonding, a set of test specimens was hot-embedded, and then was cut with a precision low-speed diamond saw for further grinding and polishing with SiC paper and diamond paste, respectively. The microstructures of the bonding interfaces were characterized with a field emission scanning electron microscope (FE-SEM) and energy dispersive X-ray (EDX). The microhardness of the filler alloys and the joint interfaces was measured with a micro-sclerometer (Shimadzu HMV-2), using a 20 gf indenting load and a dwell time of 10 s. The shear strengths of the joints were determined with a tensile testing machine (Hung-Ta HT-2102) at room temperature. To ensure the accuracy of the results, five measurements of joint strength were performed for each brazing condition and the average was calculated. The fractographs of the brazed joints after the shear tests were characterized with a FE-SEM coupled to an EDX.

Results and Discussion
Micrographs of the as-solidified Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys are shown in Figure 2a Al:Cu = 65.1:34.9 (at.%) respectively, which corresponded to the Al2Cu phase. According to the results of EDX analysis, the Al-Si binary phase diagram [24] and the Al-Si-Cu ternary phase diagram [25], Spot 2, Spot 4 and Spot 11 in Figure 2a-c respectively, show a gray particle containing a high amount of Si, which can be interpreted as Si-rich solid solution particles. Spot 3, Spot 6 and Spot 8 in Figure 2a-c respectively, can be inferred as the α-Al solid solution phase. The chemical compositions of the phases (gray) at Spot 7 in Figure 2b and Spot 9 in Figure 2c were identified by EDX as Al:Si:Ti = 14.8:55.0:30.2 (at.%) and Al:Si:Ti = 14.6:55.7:29.7 (at.%) respectively, which corresponded to the Al5Si12Ti7 intermetallic compound [26]. The EDX analysis results of the filler alloys are listed in Table  3. Prior to joining, the bonding surfaces of AA6061 and the surfaces of the filler alloys were ground with 1200-grit silicon carbide paper and then cleaned in acetone. Copper and aluminum have high vapor pressure at the bonding temperature in order to prevent the evaporation of the elements during the bonding process. Therefore, the argon partial pressure is introduced into a vacuum furnace during the bonding process. The brazing process was conducted with a heating rate of 10 • C/min in a vacuum furnace under a high-purity Ar gas atmosphere at 530 • C for 10 min. For the metallographic study of the bonding interfaces, the joined specimens were cross-sectioned. In order to avoid damage to the joints caused by metallographic production, after the completion of the bonding, a set of test specimens was hot-embedded, and then was cut with a precision low-speed diamond saw for further grinding and polishing with SiC paper and diamond paste, respectively. The microstructures of the bonding interfaces were characterized with a field emission scanning electron microscope (FE-SEM) and energy dispersive X-ray (EDX). The microhardness of the filler alloys and the joint interfaces was measured with a microsclerometer (Shimadzu HMV-2), using a 20 gf indenting load and a dwell time of 10 s. The shear strengths of the joints were determined with a tensile testing machine (Hung-Ta HT-2102) at room temperature. To ensure the accuracy of the results, five measurements of joint strength were performed for each brazing condition and the average was calculated. The fractographs of the brazed joints after the shear tests were characterized with a FE-SEM coupled to an EDX.

Results and Discussion
Micrographs of the as-solidified Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys are shown in Figure 2a respectively, which corresponded to the Al 2 Cu phase. According to the results of EDX analysis, the Al-Si binary phase diagram [24] and the Al-Si-Cu ternary phase diagram [25], Spot 2, Spot 4 and Spot 11 in Figure 2a-c respectively, show a gray particle containing a high amount of Si, which can be interpreted as Si-rich solid solution particles. Spot 3, Spot 6 and Spot 8 in Figure 2a-c respectively, can be inferred as the α-Al solid solution phase. The chemical compositions of the phases (gray) at Spot 7 in Figure 2b and Spot 9 in Figure 2c were identified by EDX as Al:Si:Ti = 14.8:55.0:30.2 (at.%) and Al:Si:Ti = 14.6:55.7:29.7 (at.%) respectively, which corresponded to the Al 5 Si 12 Ti 7 intermetallic compound [26]. The EDX analysis results of the filler alloys are listed in Table 3. Appl. Sci. 2021, 11, x FOR PEER REVIEW 4 of 16   The X-ray diffraction patterns for the Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys are presented in Figure 3. The X-ray diffraction was carried out in the presence of the Al 2 Cu intermetallic compound and the Si-rich and Al-rich alloy of the as-solidified filler metals. Both X-ray diffraction analysis and SEM-EDX observation confirmed the presence of the Al 5 Si 12 Ti 7 intermetallic compound in the Al10Si10Cu4Ti and Al10Si10Cu0.1RE filler alloys.
The X-ray diffraction patterns for the Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys are presented in Figure 3. The X-ray diffraction was carried out in the presence of the Al2Cu intermetallic compound and the Si-rich and Al-rich alloy of the as-solidified filler metals. Both X-ray diffraction analysis and SEM-EDX observation confirmed the presence of the Al5Si12Ti7 intermetallic compound in the Al10Si10Cu4Ti and Al10Si10Cu0.1RE filler alloys.  Figure 4 shows the DTA curves of the Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler metals. A prior study [19] indicated that the liquidus and solidus temperatures of the Al12Si alloy were 586.1 and 591.7 °C, respectively. The addition of 10 wt.% copper into the Al12Si filler alloy formed the Al10.8Si10Cu alloy and lowered the solidus and liquidus temperatures to 522.25 and 569.99 °C, respectively [19]. The first exothermic peak of the Al10.8Si10Cu filler metal at 527.72 °C was attributed to the reaction of θ(Al2Cu) + Si + α(Al)→L. The second exothermic peak at 569.99 °C was related to melting of the Al-Si eutectic. When 4 wt.% Ti was added into the Al-Si-Cu filler alloys, the solidus and liquidus temperatures were reduced, and the Al-Si eutectic melting reaction  Figure 4 shows the DTA curves of the Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler metals. A prior study [19] indicated that the liquidus and solidus temperatures of the Al12Si alloy were 586.1 and 591.7 • C, respectively. The addition of 10 wt.% copper into the Al12Si filler alloy formed the Al10.8Si10Cu alloy and lowered the solidus and liquidus temperatures to 522.25 and 569.99 • C, respectively [19]. The first exothermic peak of the Al10.8Si10Cu filler metal at 527.72 • C was attributed to the reaction of θ(Al 2 Cu) + Si + α(Al)→L. The second exothermic peak at 569.99 • C was related to melting of the Al-Si eutectic. When 4 wt.% Ti was added into the Al-Si-Cu filler alloys, the solidus and liquidus temperatures were reduced, and the Al-Si eutectic melting reaction was inhibited. As a result, a low-melting-point alloy reaction was initiated, and only one exothermic peak appeared for the Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler metals.
Appl. Sci. 2021, 11, x FOR PEER REVIEW 6 of 16 was inhibited. As a result, a low-melting-point alloy reaction was initiated, and only one exothermic peak appeared for the Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler metals. Figure 5a-c respectively present the microstructures under BSE mode of the AA6061 joints with Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys after brazing at 530 °C for 10 min. Some Si-rich phase particles and a large number of Al2Cu precipitates were found at grain boundaries in the Al10.8Si10Cu filler alloy. The Al2Cu particles were smaller in the titanium-containing Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys than in the Al10.8Si10Cu filler alloy. Only a very few tiny AlLa intermetallic compound particles with a size of 2-10 μm have been found in the Al10Si10Cu4Ti0.1RE filler alloy [27,28]. The addition amount of mixed rare-earth elements is very small, only 0.1 wt.%. Thus, except for the AlLa phase, another trace rare-earth element phase cannot be observed by SEM. In the study of Chen et al. [28], the slender grains of the AlLa intermetallic compounds with a size of about 5-15 μm can be found in the A356 aluminum alloy with 0.3% La and 0.2% Yb mixed rare-earth elements. Moreover, Samuel et al. [27] studied the metallography of rare-earth intermetallic compounds in the rare-earth element-containing Al-Cu-Mg and Al-Si-Cu-Mg alloys. The strip grains of aluminum-rare-earth intermetallic compounds with a size of 15-20 μm can be determined by SEM and EDX. According to the composition analysis of EDX and the studies of rare-earth elements containing aluminum alloys by Chen et al. [28] and Samuel et al. [27], it is believed that there are AlLa intermetallic compounds in the Al10Si10Cu4Ti0.1RE filler alloy. Figure 5a-c respectively present the microstructures under BSE mode of the AA6061 joints with Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys after brazing at 530 • C for 10 min. Some Si-rich phase particles and a large number of Al 2 Cu precipitates were found at grain boundaries in the Al10.8Si10Cu filler alloy. The Al 2 Cu particles were smaller in the titanium-containing Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys than in the Al10.8Si10Cu filler alloy. Only a very few tiny AlLa intermetallic compound particles with a size of 2-10 µm have been found in the Al10Si10Cu4Ti0.1RE filler alloy [27,28]. The addition amount of mixed rare-earth elements is very small, only 0.1 wt.%. Thus, except for the AlLa phase, another trace rare-earth element phase cannot be observed by SEM. In the study of Chen et al. [28], the slender grains of the AlLa intermetallic compounds with a size of about 5-15 µm can be found in the A356 aluminum alloy with 0.3% La and 0.2% Yb mixed rare-earth elements. Moreover, Samuel et al. [27] studied the metallography of rare-earth intermetallic compounds in the rare-earth element-containing Al-Cu-Mg and Al-Si-Cu-Mg alloys. The strip grains of aluminum-rare-earth intermetallic compounds with a size of 15-20 µm can be determined by SEM and EDX. According to the composition analysis of EDX and the studies of rare-earth elements containing aluminum alloys by Chen et al. [28] and Samuel et al. [27], it is believed that there are AlLa intermetallic compounds in the Al10Si10Cu4Ti0.1RE filler alloy. Appl. Sci. 2021, 11, x FOR PEER REVIEW 7 of 16 The average joint shear strengths of the AA6061/AA6061 joints bonded with Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys were determined to be 39.8, 47.8 and 57.3 MPa. The highest joint shear strengths were 46.54, 51.5 and 61.1 MPa, respectively. The shear strengths of the joints increased significantly with the addition of active elements, as shown in Figure 6. Aluminum alloys are prone to form a dense oxide shown in Figure 6. Aluminum alloys are prone to form a dense oxide layer on the surface that hinders the reaction of aluminum alloy with the filler alloy. In order to obtain a good joint, a highly corrosive specially formulated flux is necessary to be used. Niu et al. [20] used Al-Si-Ge, Al-Si-Zn and Al-Si-Ge-Zn filler metal with AlF 3 -KF-KCl-CsF flux to join 6061 aluminum alloys at higher joining temperature (570 • C), and the joining strengths reached 87.5, 103.6 and 138.2 MPa, respectively. The flux can cause environmental pollution, and the residual flux may also cause corrosion of the joint. In the study, the flux-less joining strength with Al10Si10Cu4Ti0.1RE filler alloys (57.3 MPa) is higher than the NaOH surface-treated adhesive bonding (21.8 MPa) [29], the adhesive bonding with mechanically pretreated rough surface (34.4 MPa) [30] and gas metal arc welding (53.3 MPa) [31].
Sci. 2021, 11, x FOR PEER REVIEW 8 of 1 layer on the surface that hinders the reaction of aluminum alloy with the filler alloy. I order to obtain a good joint, a highly corrosive specially formulated flux is necessary t be used. Niu et al. [20] used Al-Si-Ge, Al-Si-Zn and Al-Si-Ge-Zn filler metal with AlF3-KF KCl-CsF flux to join 6061 aluminum alloys at higher joining temperature (570 °C ), and th joining strengths reached 87.5, 103.6 and 138.2 MPa, respectively. The flux can cause en vironmental pollution, and the residual flux may also cause corrosion of the joint. In th study, the flux-less joining strength with Al10Si10Cu4Ti0.1RE filler alloys (57.3 MPa) i higher than the NaOH surface-treated adhesive bonding (21.8 MPa) [29], the adhesiv bonding with mechanically pretreated rough surface (34.4 MPa) [30] and gas metal ar welding (53.3 MPa) [31].  Figure 7 presents the microhardness profiles measured across the brazing interfacia region after bonding at 530 °C for 10 min. Due to the large number of intermetallic com pounds and Si-rich particles in the filler alloys, the hardness values of the filler alloys wer higher than that of the AA6061 substrate. The hardness values of the Al10Si10Cu4Ti an Al10Si10Cu4Ti0.1RE filler alloys were higher than that of the Al10.8Si10Cu filler alloy du to Al5Si12Ti7 precipitation hardening and Al2Cu grain refinement. Large increments of th hardness values of the Al10Si10Cu4Ti0.1RE filler alloy resulted when the trace rare-eart elements were added.  Figure 7 presents the microhardness profiles measured across the brazing interfacial region after bonding at 530 • C for 10 min. Due to the large number of intermetallic compounds and Si-rich particles in the filler alloys, the hardness values of the filler alloys were higher than that of the AA6061 substrate. The hardness values of the Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys were higher than that of the Al10.8Si10Cu filler alloy due to Al 5 Si 12 Ti 7 precipitation hardening and Al 2 Cu grain refinement. Large increments of the hardness values of the Al10Si10Cu4Ti0.1RE filler alloy resulted when the trace rare-earth elements were added.   Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys, respectively. the fractographs of the joints indicated that the fractured surfaces were covered with alloys with intermetallic compounds and Si-rich solid solution particles, suggestin the fractures of the joints occurred mainly in the interior of the filler alloys. Mainly granular fracture occurred at the joint with Al-10.8Si-10Cu filler metal due to the c Al2Cu intermetallic compounds at the grain boundaries. In the joints brazed Al10.8Si10Cu4Ti and Al10.8Si10Cu4Ti0.1RE filler alloys, mixed intergranular and granular fractures were found at the brazing interfaces.  8a-c show the fractography of the AA6061/AA6061 joints brazed with Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys, respectively. All of the fractographs of the joints indicated that the fractured surfaces were covered with filler alloys with intermetallic compounds and Si-rich solid solution particles, suggesting that the fractures of the joints occurred mainly in the interior of the filler alloys. Mainly intergranular fracture occurred at the joint with Al-10.8Si-10Cu filler metal due to the coarse Al 2 Cu intermetallic compounds at the grain boundaries. In the joints brazed with Al10.8Si10Cu4Ti and Al10.8Si10Cu4Ti0.1RE filler alloys, mixed intergranular and transgranular fractures were found at the brazing interfaces.
Cross-sectional SEM micrographs under BSE mode of the AA6061/alumina joints brazed with Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys at 530 • C for 10 min are presented in Figure 9a-c, respectively. Satisfactory bonding interfaces formed in the AA6061 and alumina joints. No obvious defects were found at the joint interfaces. A few Si-rich and Al 2 Cu intermetallic compound particles were found at the interface between the Al10.8Si10Cu filler alloy and alumina, as shown in Figure 9a. Large grains were observed in the Al10Si10Cu4Ti filler metal, as were Al 2 Cu and Al 5 Si 12 Ti 7 intermetallic compounds on the grain boundaries, as shown in Figure 9b. Compared with the Al10Si10Cu4Ti filler alloy, the Al10Si10Cu4Ti0.1RE filler alloy had a smaller grain structure, and rare-earth elements accumulated at the interface of the Al10Si10Cu4Ti0.1RE filler alloy and alumina. According to the EDX results, the composition of La segregates was 53.4 at.% of Al and 46.6 at.% of La, which corresponded to the AlLa intermetallic compound [27,28]. the fractographs of the joints indicated that the fractured surfaces were covered with filler alloys with intermetallic compounds and Si-rich solid solution particles, suggesting that the fractures of the joints occurred mainly in the interior of the filler alloys. Mainly intergranular fracture occurred at the joint with Al-10.8Si-10Cu filler metal due to the coarse Al2Cu intermetallic compounds at the grain boundaries. In the joints brazed with Al10.8Si10Cu4Ti and Al10.8Si10Cu4Ti0.1RE filler alloys, mixed intergranular and transgranular fractures were found at the brazing interfaces. Cross-sectional SEM micrographs under BSE mode of the AA6061/alumina joints brazed with Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys at 530 °C for 10 min are presented in Figure 9a-c, respectively. Satisfactory bonding interfaces formed in the AA6061 and alumina joints. No obvious defects were found at the joint interfaces. A few Si-rich and Al2Cu intermetallic compound particles were found at the interface between the Al10.8Si10Cu filler alloy and alumina, as shown in Figure 9a. Large grains were observed in the Al10Si10Cu4Ti filler metal, as were Al2Cu and Al5Si12Ti7 in-  The joint shear strengths of the AA6061/alumina joints with Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys are shown in Figure 10. The joint with the Al10.8Si10Cu filler alloy separated at the beginning of the shear test, so the joint shear strength was not obtained. Although, the Al10.8Si10Cu filler alloy can be spread on the alumina under a small pressure during the joining process, and there is no obvious The joint shear strengths of the AA6061/alumina joints with Al10.8Si10Cu, Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys are shown in Figure 10. The joint with the Al10.8Si10Cu filler alloy separated at the beginning of the shear test, so the joint shear strength was not obtained. Although, the Al10.8Si10Cu filler alloy can be spread on the alumina under a small pressure during the joining process, and there is no obvious defect in the interface between Al10.8Si10Cu filler alloy and alumina. Since there is no metallurgical reaction, that results in almost no bonding strength between Al10.8Si10Cu filler alloy and alumina. The highest shear strengths of the AA6061/alumina joints with Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys were 11.7 and 19.2 MPa, respectively. The shear strength of the AA6061/alumina joint with the Al10Si10Cu4Ti0.1RE filler alloy was significantly higher than that of the joint with the Al10Si10Cu4Ti filler alloy. The average shear strengths of the AA6061/alumina joint with Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys were 8.0 and 14.8 MPa, respectively. The role of titanium in Al-Si-Cu filler is the same as the titanium-containing filler metals, such as tin-based [13][14][15], silver-based [9,10] and indium-based [11] filler alloys, which can promote the wetting of the filler on the ceramic surface and the interfacial reaction between the filler and the ceramic. The shear strength of the AA6061/alumina joint with the Al10Si10Cu4Ti0.1RE filler alloy was significantly higher than that of the joint with the Al10Si10Cu4Ti filler alloy. The average shear strengths of the AA6061/alumina joint with Al10Si10Cu4Ti and Al10Si10Cu4Ti0.1RE filler alloys were 8.0 and 14.8 MPa, respectively. The role of titanium in Al-Si-Cu filler is the same as the titanium-containing filler metals, such as tin-based [13][14][15], silver-based [9,10] and indium-based [11] filler alloys, which can promote the wetting of the filler on the ceramic surface and the interfacial reaction between the filler and the ceramic. The fractography of joints brazed with Al10.8Si10Cu filler alloy after shear tests is shown in Figure 11. All the filler alloy was on the fracture surface of the AA6061 side of the joint brazed with Al10.8Si10Cu filler metal after the shear test. Many particles of Al2Cu intermetallic compounds and Si-rich phase were found on the fracture surface of the AA6061 side, and no filler alloy was observed on the alumina surface side. These results indicated that the Al10.8Si10Cu filler alloy could not form a wetting reaction on the alumina, which led to the joint having almost no strength. Fractography of the AA6061/alumina joint with the Al10Si10Cu4Ti filler alloy revealed that the brazed joint fractured along the filler metal/alumina interface, as shown in Figure 12a,b. After the shear test, most of the Al10Si10Cu4Ti filler alloy was on the AA6061 surface, and very little filler metal was observed on the alumina surface. Figure 13a,b show the fractography of an AA6061/alumina joint with Al10Si10Cu4Ti0.1RE filler alloy after shear test. Both of the fractured surfaces of the AA6061 and alumina were covered with Al10Si10Cu4Ti0.1RE filler alloy. Since more residual attached filler alloy was observed on the alumina surface of the AA6061/alumina joint brazed with Al10Si10Cu4Ti0.1RE filler alloy than on the joint brazed with Al10Si10Cu4Ti alloy, high AA6061/alumina joint strength was achieved by brazing with the Al10Si10Cu4Ti0.1RE filler alloy at 530 °C for 10 min. The fractography of joints brazed with Al10.8Si10Cu filler alloy after shear tests is shown in Figure 11. All the filler alloy was on the fracture surface of the AA6061 side of the joint brazed with Al10.8Si10Cu filler metal after the shear test. Many particles of Al 2 Cu intermetallic compounds and Si-rich phase were found on the fracture surface of the AA6061 side, and no filler alloy was observed on the alumina surface side. These results indicated that the Al10.8Si10Cu filler alloy could not form a wetting reaction on the alumina, which led to the joint having almost no strength. Fractography of the AA6061/alumina joint with the Al10Si10Cu4Ti filler alloy revealed that the brazed joint fractured along the filler metal/alumina interface, as shown in Figure 12a,b. After the shear test, most of the Al10Si10Cu4Ti filler alloy was on the AA6061 surface, and very little filler metal was observed on the alumina surface. Figure 13a,b show the fractography of an AA6061/alumina joint with Al10Si10Cu4Ti0.1RE filler alloy after shear test. Both of the fractured surfaces of the AA6061 and alumina were covered with Al10Si10Cu4Ti0.1RE filler alloy. Since more residual attached filler alloy was observed on the alumina surface of the AA6061/alumina joint brazed with Al10Si10Cu4Ti0.1RE filler alloy than on the joint brazed with Al10Si10Cu4Ti alloy, high AA6061/alumina joint strength was achieved by brazing with the Al10Si10Cu4Ti0.1RE filler alloy at 530 • C for 10 min. Appl. Sci. 2021, 11, x FOR PEER REVIEW 13 of 16 Figure 11. Fractography of the aluminum alloy/alumina joint bonded with the Al10.8Si10Cu filler alloy: (a) AA6061 side and (b) alumina side.