3.1 Wettability of AgCuInTi20-20-4.5 on Al2O3 substrate
The wetting experiments of AgCuInTi20-20-4.5 on Al2O3 substrate were conducted at 700 ℃ for 5 min. As given in Fig. 3, the contact angle of AgCuInTi20-20-4.5 on Al2O3 substrate after brazing is 22 °, and the wettability parameters were listed in Table 2. It can be seen the AgCuInTi20-20-4.5 alloy has an appropriate spreading coefficient.
Table 2
Wettability of AgCuInTi20-20-4.5 on Al2O3
Specimen number | Spreading area /cm2 | Spreading coefficient/% |
1 | 1.42 | 96.5 |
2 | 1.43 | 96.3 |
3 | 1.40 | 96.1 |
3.3 Microstructure of AgCuInTi20-20-4.5 powder and Al2O3/Cu brazed joint
SEM micrograph of the AgCuInTi20-20-4.5 powder is shown in Fig. 4. It can be seen the average particle diameter of the spherical alloy particles is approximate 35 µm.
The Al2O3/AgCuInTi20-20-4.5/Cu brazed joint was fabricated at 700 ℃ for 10 min. The microstructure of the brazed joint is shown in Fig. 5. The interface microstructure has a great influence on the mechanical property of the joint. Obviously, there is no micro defects such as voids and cracks on the interface. The microstructures of the interlayer adjoined different substrates are obviously different. It can be seen from the Fig. 5 (a) that the boundary between Al2O3 and filler metal layer is more tortuous. According to the EDS result of the brazing joint shown in Fig. 5(b), the majority of Ti element remained in the middle of the filler, some Ti atoms distributed along the Al2O3/filler interface. This indicates that the filler metal has active reaction with the Al2O3 matrix. Additionally, large white blocks were observed in the bonding area. According to Refs. [16], the white blocks were Ag-based solid solution containing the In element. The area distribution of the In element was in good agreement with that of the Ag element.
To further elucidate the microstructures, the partial enlarged drawings of the areas marked in Fig. 5(a) are shown in Fig. 6(a) and Fig. 6(b). EDS point analysis of the brazed joint was done to determine the chemical composition and identification of individual phases (Table 3). According to previous research [17–20], active Ti element would react with the Al2O3 matrix and form TiO and TiO2 during brazing, as given in Eq. (1). In this study, TiO2 was found at the bonding interface (marked as A). The white block adjacent to Al2O3 (marked as B) was Cu3TiO4. According to the EDS analysis of the microdomain “C, D and F” shown in Table 3, the light gray phase in the middle of the filler is mainly composed of Cu, Ti, and In, it can be inferred that the light gray phase is Cu2InTi. The light gray Cu4In phase was found in the matrix in the middle of the brazing area (marked as E and H). At the same time, a large white silver-rich phase (marked as I) was observed, which is consistent with previous EDS scan results shown as Fig. 5 (b). Around the silver-rich phase, a small group of black particles (marked as J) were observed and the details were explained in Fig. 6(b). Based on the composition of position J, L and M, main phase in this region was Cu4Ti. During brazing, Ti atoms diffused to the Cu side and formed Cu-Ti phase with Cu atoms. Some short rod-shaped Ag (marked as N) were embedded into matrix. In addition, some small black particles were also observed inlaid on the silver-rich phase matrix (marked as K), which may be Cu2InTi combined with EDS analysis.
Ti + xAl2O3→Ti(O)3x+2xAl (1)
If x = 1/3, TiO is formed. If x = 2/3, TiO2 is formed.
Table 3
Chemical analyses at areas shown in Fig. 6
Symbol | Average chemical analyses, at.% | Possible phase |
O | Al | Ti | Cu | Ag | In |
A | 33.27 | 23.52 | 18.24 | 12.93 | 10.01 | 2.03 | TiO2 |
B | 31.81 | 14.15 | 20.45 | 14.28 | 16.19 | 3.12 | Cu3TiO4 |
C | 10.32 | 2.08 | 24.34 | 44.61 | 1.10 | 17.54 | Cu2InTi |
D | 12.45 | 2.10 | 23.34 | 44.27 | 1.95 | 15.88 | Cu2InTi |
E | 7.88 | 2.17 | 24.00 | 47.20 | 1.18 | 17.57 | Cu4In |
F | 12.58 | 2.14 | 22.15 | 48.02 | 0.83 | 14.28 | Cu2InTi |
G | 19.58 | 1.90 | 7.14 | 10.88 | 50.99 | 9.52 | Ag3In |
H | 5.35 | 1.05 | 23.01 | 51.83 | 1.96 | 16.8 | Cu4In |
I | 4.69 | 0.56 | 5.94 | 9.39 | 65.67 | 13.75 | Rich Ag |
J | 7.32 | 1.89 | 15.49 | 65.31 | 1.30 | 8.69 | Cu4Ti |
K | 18.60 | 1.45 | 8.16 | 22.34 | 42.30 | 7.15 | Ag + Cu2InTi |
L | 7.85 | 2.15 | 21.62 | 51.49 | 0.81 | 16.08 | Cu4Ti |
M | 8.91 | 1.87 | 18.58 | 64.88 | 1.25 | 4.51 | Cu4Ti |
N | 12.34 | 1.85 | 0.18 | 8.92 | 67.14 | 9.56 | Rich Ag |
3.4 Joining performance of the Al2O3/Cu brazed joints
In order to evaluate the joining performance of the Al2O3/Cu joints brazed with AgCuInTi20-20-4.5, shear tests of joints were conducted. The mean shear strength of the joints is 89 MPa. The fractured surfaces of joints were analyzed for more exact identification of the bond formation mechanism. Figure 7 shows the fractured surface at the interface of the joint. The fractured surface, on the side containing Al2O3, remained completely covered with filler metal. A ductile fracture occurred within the bonding area. XRD analysis was carried out on this fracture surface, as shown in Fig. 8. The results indicated that except the diffraction peaks of Al2O3, Ag, those of TiO2, Cu3TiO4, Cu4Ti, Cu4In and Ag3In intermetallic phases also appeared in the XRD pattern, which was correspond of the EDS analyses. The active Ti element diffused to the base materials on both sides under the driving force of the chemical potential difference, TiO2 and Cu4Ti formed. Some undiffused Ti atoms were still in the middle of the filler, which reacted with Cu and In in the AgCuIn eutectic to form the Cu2InTi phase [21].