E ﬀ ect of Various Nanoparticles (GaF 3 , ZnF 2 , Zn(BF 4 ) 2 and Ga 2 O 3 ) Additions on the Activity of CsF-RbF-AlF 3 Flux and Mechanical Behavior of Al / Steel Brazed Joints

: In this study, brazing AA6061 to Q235 steel using ﬂame brazing was performed with 70.9 wt.% CsF-0.5 wt.% RbF-28.6 wt.% AlF 3 ﬂuxes doped with GaF 3 , ZnF 2 , Zn(BF 4 ) 2 and Ga 2 O 3 nanoparticles, matched with a Zn-15Al ﬁller metal, and the spreadability of the ﬁller metal and the mechanical properties of brazed joints were investigated at the same time. The results showed suitable amounts of GaF 3 , ZnF 2 , Zn(BF 4 ) 2 and Ga 2 O 3 doped into the base ﬂux could strengthen the ﬁller metal in wetting and spreading on the surface of aluminum alloy and steel to di ﬀ erent degrees. The suitable ranges of GaF 3 , ZnF 2 , Zn(BF 4 ) 2 and Ga 2 O 3 , respectively, were 0.0075–0.01 wt.%, 0.0075–0.01 wt.%, 0.0075–0.01 wt.% and 0.009–0.01 wt.%, and the maximum spreading area was obtained via doping with GaF 3 . The shear strength of brazed joints reached the peak at 126 MPa when 0.075 wt.% GaF 3 was added. Comparative tests proved that the activity of the CsF-RbF-AlF 3 ﬂux doped with GaF 3 was the best. The reason was that the CsF-RbF-AlF 3 -GaF 3 ﬂux was competent in removing oxides of the base metal and decreasing the interfacial tension, in virtue of the activity of Ga 3 + as well as F − .


Introduction
The increase in joining dissimilar metals has been prompted by the increasing demand for light weights and fuel efficiency in modern industries in recent years [1]. The emerging trend towards lightweight, high performance and emissions reduction is leading to the increasing use of multi-material hybrid structures in electric vehicles [2]. The hybrid structure of aluminum and steel are superior to conventional materials because of the high mechanical properties of steel and the excellent corrosion resistance of aluminum [3]. Naturally, the combination of aluminum alloy and steel shows huge research value and future potential, and has become the subject of many investigators [4].
However, it remains an unconquerable problem to get stable brazed joints between aluminum/metals dissimilar to steel because of the formation of brittle and hard intermetallic compounds, like Fe x Al y [5]. The complexity and randomness of the brazing process makes the design and development of brazing materials more complicated and time-consuming than ordinary materials [6]. Therefore, it is absolutely imperative to control the formation of Fe-Al intermetallic compounds. EI-Sayed and Naka [7] found that the maximum bond shear strength of 127 MPa was obtained for an Al-steel joint brazed at 663 K with a 3 s ultrasound application time using a Zn-14Al alloy. It was reported that the shear strength of brazed joints reached the peak at 131 MPa when the Zn-15Al filler metal was added in the joining of lap joints of 6061 aluminum alloy to the 304 stainless steel via a flame brazing process, with the Zn-xAl filler metals matching the CsF-0.5 wt.% RbF-AlF 3 flux [8]. The interfacial layer in the weld made with the Zn-15Al filler metal was comprised of (FeAl 3 )Zn x and (Fe 2 Al 5 )Zn x [9].
Nanomaterials have been widely applied due to their unique properties. Joining technology at the nanometer scale has gradually developed with the popularity of nanomaterials, which have broad application prospects in the fields of electronics, aero-space, biology and health care [10]. Previous studies have indicated that [11] trace amounts of Ga 2 O 3 addition in the CsF-0.5 wt.% RbF-AlF 3 flux could obviously strengthen the Zn-2Al filler metal in its wetting and spreading on the surface of 5052 aluminum alloy and Q235 low-carbon steel. One previous study discussed how the addition of GaF 3 and Ga 2 O 3 nanoparticles influenced the wettability and spreadability of the CsF-AlF 3 flux under the same conditions [12].
The influence that adding heavy metal fluoride, ZnF 2 , SnF 2 , CdF 2 , PbF 2 and KBF 4 into the KF-AlF 3 flux has on the spreadability of brazing aluminium was investigated, and adding ZnF 2 could greatly improve the brazing area [13]. Therefore, the fourth component, ZnF 2 , has been considered for doping into CsF-0.5 wt.% RbF-AlF 3 flux in order to reduce the price in this paper. ZnCl 2 and SnCl 2 were added into the CsF-AlF 3 flux for connecting aluminium alloys [14]. The joints were bonded soundly when the mass fractions of ZnCl 2 and SnCl 2 are about 4%.
The effect of KBF 4 addition on the microstructure of the Mg-6Zn-1Si alloy has been investigated [15], and the morphology of the Mg 2 Si phase changed with the addition of 1.5 wt.% KBF 4 . To compare the activity of BF 4 − and its unified positive ions in reducing the variables, Zn(BF 4 ) 2 has been chosen.
In this paper, brazing AA6061 to Q235 steel using flame brazing has been performed with improved CsF-0.5 wt.% RbF-AlF 3 fluxes doped with a GaF 3 , ZnF 2 , Zn(BF 4 ) 2 and Ga 2 O 3 nanoparticles-matched Zn-15Al filler metal, and the spreadability of the filler metal and the mechanical properties of the brazed joints were investigated at the same time. XDR analysis was carried out and the reaction mechanism was analyzed. The results could be useful for brazing AA6061 to Q235 steel while choosing a suitable flux.

Materials and Methods
The 6061 aluminum alloy and Q235 steel were used in this work as base metals. The compositions of the base metals were listed in Tables 1 and 2. Zn-15Al alloys were chosen as the filler metal. A CsF-0.5 wt.% RbF-AlF 3 flux was prepared by using commercial CsF-AlF 3 flux and an RbF of AR purity (Zhejiang Xinrui Welding Materials Co., Ltd., Shengzhou, China). CsF-RbF-AlF 3 -GaF 3 fluxes with different compositions of AR purity GaF 3 were confected and the range of GaF 3 was 0.0001-0.125 wt.%. Same amounts of ZnF 2 , and Zn(BF 4 ) 2 of AR purity were doped in the same way to obtain CsF-RbF-AlF 3 -ZnF 2 fluxes and CsF-RbF-AlF 3 -Zn(BF 4 ) 2 fluxes. Nano Ga 2 O 3 powder in the same range was doped into the CsF-RbF-AlF 3 flux to get the corresponding CsF-RbF-AlF 3 -Ga 2 O 3 fluxes. After mixing into a liquid, the fluxes were then dried to powder with the oven. The chemical compositions of these four nanoparticles doped into the fluxes has been listed in Table 3.   Zn-15Al alloys were extruded into a 2 mm diameter wire as an advanced preparation. SiC paper was used for mechanically polishing the specimens and filler metals. Before brazing, these materials were all degreased with acetone and cleaned by ethanol.
In order to prepare for the spreading test, the base metals were processed into plates of 40 mm × 40 mm × 3 mm. The spreading test was performed according to China's National Standard GB 11364-2008. The weights of the solder and flux used in the test should be 100 mg and 15 mg, respectively. The filler metals were placed in the center of the base metals covered with the fluxes prepared previously, as in Figure 1, and then put into the electrical resistance furnace (Zhejiang Xinrui Welding Materials Co., Ltd., Shengzhou, China). The heating temperature was uniformly set at 530 • C. The holding time of each test was set as 60 s. After spreading, the test boards were cleaned using ultrasonic wave waits and the spreading areas were calculated by the software Image-ro Plus (Image-Pro Plus Version 6.0). To be specific, we photographed the spreading boards with the graduated ruler and imported the image into Image-Pro Plus. By determining the actual scale in the picture and using chromatic aberration to circle the spreading outline, the spreading areas could be calculated. For the credibility of the results, the above tests of each group were repeated 5 times and the spreading areas of each group were averaged. At last, the residues of the fluxes were collected and the components in the residues were analyzed with a Brucker D8 XRD analyzer (Ningbo Institute of Materials Technology & Engineering, Ningbo, China).
Crystals 2020, 10, x FOR PEER REVIEW 3 of 10 Table 3. Chemical composition of nanoparticles doped into the fluxes (wt.%). In order to prepare for the spreading test, the base metals were processed into plates of 40 mm × 40 mm × 3 mm. The spreading test was performed according to China's National Standard GB 11364-2008. The weights of the solder and flux used in the test should be 100 mg and 15 mg, respectively. The filler metals were placed in the center of the base metals covered with the fluxes prepared previously, as in Figure 1, and then put into the electrical resistance furnace (Zhejiang Xinrui Welding Materials Co., Ltd., Shengzhou, China). The heating temperature was uniformly set at 530 °C . The holding time of each test was set as 60 s. After spreading, the test boards were cleaned using ultrasonic wave waits and the spreading areas were calculated by the software Image-ro Plus (Image-Pro Plus Version 6.0). To be specific, we photographed the spreading boards with the graduated ruler and imported the image into Image-Pro Plus. By determining the actual scale in the picture and using chromatic aberration to circle the spreading outline, the spreading areas could be calculated. For the credibility of the results, the above tests of each group were repeated 5 times and the spreading areas of each group were averaged. At last, the residues of the fluxes were collected and the components in the residues were analyzed with a Brucker D8 XRD analyzer (Ningbo Institute of Materials Technology & Engineering, Ningbo, China). For the shear performance tests, the supplied base metals for the brazed joint were processed into plates with the size of 60 mm × 25 mm × 3 mm. Figure 2 shows the schematic illustration of the brazed joint. The shear performance tests of the AA6061/Q235 brazed joints were carried out in strict accordance with China's National Standard GB 11363-2008. The equipment was an SAMS-CMT5105 universal tensile testing machine (Nanjing University of Aeronautics and Astronautics, Nanjing, China), and the loading rate in the tensile process was 3 mm/min. The test results for each group of samples were all averaged from 5 samples. The brazing temperature was detected and controlled at around 530 °C.

Flux
Filler Metal For the shear performance tests, the supplied base metals for the brazed joint were processed into plates with the size of 60 mm × 25 mm × 3 mm. Figure 2 shows the schematic illustration of the brazed joint. The shear performance tests of the AA6061/Q235 brazed joints were carried out in strict accordance with China's National Standard GB 11363-2008. The equipment was an SAMS-CMT5105 universal tensile testing machine (Nanjing University of Aeronautics and Astronautics, Nanjing, China), and the loading rate in the tensile process was 3 mm/min. The test results for each group of samples were all averaged from 5 samples. The brazing temperature was detected and controlled at around 530 • C.

The Spreadability and Wettability of Zn-15Al Filler Metal
The influence of GaF3, ZnF2, Zn(BF4)2 and Ga2O3 particles, doped into CsF-0.5 wt.% RbF-AlF3 fluxes, on the spreadability of the Zn-15Al filler metal was studied via spreading tests performed on the surface of the 6061 aluminum alloy and the Q235 steel. The results were as follows.
The relationships between the concentrations of GaF3, ZnF2, Zn(BF4)2 and Ga2O3 and the spreading areas on the 6061 aluminum alloy are shown in Figure 3a. While the content of GaF3 was 0.01 wt.%, the spreading area was maximized to 329 mm 2 , which was 65% larger than the area without any addition. With 0.0075 wt.%-doped ZnF2, the spreading area was maximized at 312 mm 2 , and showed a 56% increase compared with the results derived without adding ZnF2. While the content of Zn(BF4)2 was 0.0075 wt.%, the spreading area was maximized at 304 mm 2 , and was 52% larger than the area without any addition. With 0.009 wt.%-doped Ga2O3, the spreading area was maximized at 321 mm 2 and showed a 61% increase compared with the results derived without adding Ga2O3.

The Spreadability and Wettability of Zn-15Al Filler Metal
The influence of GaF 3 , ZnF 2 , Zn(BF 4 ) 2 and Ga 2 O 3 particles, doped into CsF-0.5 wt.% RbF-AlF 3 fluxes, on the spreadability of the Zn-15Al filler metal was studied via spreading tests performed on the surface of the 6061 aluminum alloy and the Q235 steel. The results were as follows.
The relationships between the concentrations of GaF 3 , ZnF 2 , Zn(BF 4 ) 2 and Ga 2 O 3 and the spreading areas on the 6061 aluminum alloy are shown in Figure 3a. While the content of GaF 3 was 0.01 wt.%, the spreading area was maximized to 329 mm 2 , which was 65% larger than the area without any addition. With 0.0075 wt.%-doped ZnF 2 , the spreading area was maximized at 312 mm 2 , and showed a 56% increase compared with the results derived without adding ZnF 2 . While the content of Zn(BF 4 ) 2 was 0.0075 wt.%, the spreading area was maximized at 304 mm 2 , and was 52% larger than the area without any addition. With 0.009 wt.%-doped Ga 2 O 3 , the spreading area was maximized at 321 mm 2 and showed a 61% increase compared with the results derived without adding Ga 2 O 3 .

The Spreadability and Wettability of Zn-15Al Filler Metal
The influence of GaF3, ZnF2, Zn(BF4)2 and Ga2O3 particles, doped into CsF-0.5 wt.% RbF-AlF3 fluxes, on the spreadability of the Zn-15Al filler metal was studied via spreading tests performed on the surface of the 6061 aluminum alloy and the Q235 steel. The results were as follows.
The relationships between the concentrations of GaF3, ZnF2, Zn(BF4)2 and Ga2O3 and the spreading areas on the 6061 aluminum alloy are shown in Figure 3a. While the content of GaF3 was 0.01 wt.%, the spreading area was maximized to 329 mm 2 , which was 65% larger than the area without any addition. With 0.0075 wt.%-doped ZnF2, the spreading area was maximized at 312 mm 2 , and showed a 56% increase compared with the results derived without adding ZnF2. While the content of Zn(BF4)2 was 0.0075 wt.%, the spreading area was maximized at 304 mm 2 , and was 52% larger than the area without any addition. With 0.009 wt.%-doped Ga2O3, the spreading area was maximized at 321 mm 2 and showed a 61% increase compared with the results derived without adding Ga2O3.    Figure 3b showed that the spreadability of the Zn-15Al filler metal over Q235 steel was clearly improved with GaF 3 doped into the CsF-RbF-AlF 3 flux, compared to others. While the content of GaF 3 was 0.0075 wt.%, the spreading area was maximized at 189 mm 2 . With 0.01 wt.%-doped ZnF 2 , the spreading area was maximized at 155 mm 2 . While the content of Zn(BF 4 ) 2 was 0.01 wt.%, the spreading area was maximized at 149 mm 2 . With 0.01 wt.%-doped Ga 2 O 3 , the spreading area was maximized at 161 mm 2 . With the addition of GaF 3 , the spreading areas over Q235 steel underwent a 105% increase, compared with the results derived without adding GaF 3 (92 mm 2 ), and this substance showed clear advantages over ZnF 2 , Zn(BF 4 ) 2 and Ga 2 O 3 .
The pictures of the best spreadings of Zn-15Al alloys on AA6061 and Q235, with CsF-RbF-AlF 3 flux doped with different additions, are shown in Figure 4. Taking the test results from two kinds of base metal together, it could be concluded that the suitable ranges of GaF 3 , ZnF 2 , Zn(BF 4 ) 2      showed that the spreadability of the Zn-15Al filler metal over Q235 steel was clearly improved with GaF3 doped into the CsF-RbF-AlF3 flux, compared to others. While the content of GaF3 was 0.0075 wt. %, the spreading area was maximized at 189 mm 2 . With 0.01 wt. %-doped ZnF2, the spreading area was maximized at 155 mm 2 . While the content of Zn(BF4)2 was 0.01 wt. %, the spreading area was maximized at 149 mm 2 . With 0.01 wt. %-doped Ga2O3, the spreading area was maximized at 161 mm 2 . With the addition of GaF3, the spreading areas over Q235 steel underwent a 105% increase, compared with the results derived without adding GaF3 (92 mm 2 ), and this substance showed clear advantages over ZnF2, Zn(BF4)2 and Ga2O3.
The pictures of the best spreadings of Zn-15Al alloys on AA6061 and Q235, with CsF-RbF-AlF3 flux doped with different additions, are shown in Figure 4. Taking the test results from two kinds of base metal together, it could be concluded that the suitable ranges of GaF3, ZnF2, Zn(BF4)2 and Ga2O3 in the CsF-RbF-AlF3 flux respectively were 0.0075-0.01 wt. %, 0.0075-0.01 wt. %, 0.0075-0.01 wt. % and 0.009-0.01 wt. %, and the maximum spreading area of all the tests was obtained via doping with GaF3.

The Mechanical Properties of Brazed Joints
The variation of brazed joint shear strength was shown in Figure 5. It could be seen that the addition of GaF3 has a significant impact on the brazed joint shear strength of AA6061/Q235. While the content of GaF3 was 0.075 wt. %, the shear strength reached a maximum at 126 MPa, which was 110% higher than that achieved without GaF3 addition (64.5 MPa). While the content of ZnF2 was 0.01 wt. %, the maximum shear strength was 111 MPa. While the content of Zn(BF4)2 was 0.01 wt. %, the shear strength reached the top at 105 MPa. While the content of Ga2O3 was 0.01 wt. %, the maximum shear strength was 116 MPa.
It could be seen that, only considering the shear strength of the AA6061/Q235 brazed joints, the addition range should be controlled as 0.0075-0.01 wt. %, and this appropriate addition range coincides with the results of the spreading test. Among the four doped ingredients selected, GaF3 showed the best performance as regards the mechanical properties of brazed joints. The second-best performance was achieved with Ga2O3. At the same time, Zn(BF4)2 was inferior to others.
The typical fracture modes of the 6061/Q235 brazed joints were shown in Figure 6. The fracture of the joint occurred mainly at the interface layer of the brazing joint and the Q235 steel. It

The Mechanical Properties of Brazed Joints
The variation of brazed joint shear strength was shown in Figure 5. It could be seen that the addition of GaF 3 has a significant impact on the brazed joint shear strength of AA6061/Q235. While the content of GaF 3 was 0.075 wt.%, the shear strength reached a maximum at 126 MPa, which was 110% higher than that achieved without GaF 3 addition (64.5 MPa). While the content of ZnF 2 was 0.01 wt.%, the maximum shear strength was 111 MPa. While the content of Zn(BF 4 ) 2 was 0.01 wt.%, the shear strength reached the top at 105 MPa. While the content of Ga 2 O 3 was 0.01 wt.%, the maximum shear strength was 116 MPa.  It could be seen that, only considering the shear strength of the AA6061/Q235 brazed joints, the addition range should be controlled as 0.0075-0.01 wt.%, and this appropriate addition range coincides with the results of the spreading test. Among the four doped ingredients selected, GaF 3 showed the best performance as regards the mechanical properties of brazed joints. The second-best performance was achieved with Ga 2 O 3 . At the same time, Zn(BF 4 ) 2 was inferior to others.
The typical fracture modes of the 6061/Q235 brazed joints were shown in Figure 6. The fracture of the joint occurred mainly at the interface layer of the brazing joint and the Q235 steel. It demonstrated that the interface layer between the filler metal and the Q235 steel was the weakest area of the whole joint, and that when there were layers of brittle compounds at the interface, the brazed joint would crack at the layers of brittle compounds first.

Effect of Zn(BF4)2
The XRD analysis results for the residues of the fluxes on Q235 low-carbon steel are shown in Figure 7. In inndividually evaluating the spreadability of the Zn-15Al filler metal, and the mechanical 40 μm

Effect of Zn(BF 4 ) 2
The XRD analysis results for the residues of the fluxes on Q235 low-carbon steel are shown in Figure 7. In inndividually evaluating the spreadability of the Zn-15Al filler metal, and the mechanical properties of the brazed joints with CsF-RbF-AlF 3 -Zn(BF 4 ) 2 over the 6061 aluminum and Q235 steel, it was clear that the addition of Zn(BF 4 ) 2 significantly improved the activity of the CsF-RbF-AlF 3 flux, with a 52% increase in spreading area and a 92% increase in shear strength.
Zn(BF 4 ) 2 → ZnF 2 + 2BF 3 (1) Zn(BF4)2 could produce BF3 as shown Equation (1). BF3 could react with oxides such as FeO, Fe2O, NiO, Cr2O3 and ZnO, thus playing the role of removing the oxide film and promoting the wetting, spreading and flowing of the molten filler metal in the brazed part [16]. However, BF3 could not react with oxides such as Al2O3 and MgO on the surface of t he aluminum alloy, thus limiting the activity of the CsF-RbF-AlF3-Zn(BF4)2 flux.

Effect of ZnF2
It was found in the study [13] that adding ZnF2 to the flux could produce a mass transfer effect, and improve the activity of the flux. This was because the ZnF2 was reduced by the base metal, and molten Zn has great solubility in aluminum. It was also observed that adding ZnF2 could reduce the initial temperature of the flux, which was also related to the reduction of the precipitation of molten Zn. Zn(BF 4 ) 2 could produce BF 3 as shown Equation (1). BF 3 could react with oxides such as FeO, Fe 2 O, NiO, Cr 2 O 3 and ZnO, thus playing the role of removing the oxide film and promoting the wetting, spreading and flowing of the molten filler metal in the brazed part [16]. However, BF 3 could not react with oxides such as Al 2 O 3 and MgO on the surface of t he aluminum alloy, thus limiting the activity of the CsF-RbF-AlF 3 -Zn(BF 4 ) 2 flux.

Effect of ZnF 2
It was found in the study [13] that adding ZnF 2 to the flux could produce a mass transfer effect, and improve the activity of the flux. This was because the ZnF 2 was reduced by the base metal, and molten Zn has great solubility in aluminum. It was also observed that adding ZnF 2 could reduce the initial temperature of the flux, which was also related to the reduction of the precipitation of molten Zn. When heated, ZnF 2 would react as shown Equation (2). The Fgenerated by the reaction would react with the oxide on the surface of the base metals, as shown in Equations (3) and (4), respectively, thus enhancing the membrane removal effect of CsF-RbF-AlF 3 flux. The Zn atoms precipitated from the reaction would react with the aluminum atoms on the surface of the aluminum alloy and the Fe atoms on the surface of steel, and the wetting and spreading would be promoted on the surface of the base metals [17].
The presence of ZnSiO 3 , ZnFe 2 O 4 , AlPO 4 and FePO 4 in the residue on the one hand indicated that the ZnF 2 in the flux did participate in the reaction, and the Zn atoms from the reduced precipitation participated in the reaction and played a role in improving the activity. On the other hand, it indicated that the enrichment phenomenon of P occurred, which resulted in the formation of compounds such as AlPO 4 and FePO 4 , which also played an important role in the removal of the oxide film from the steel surface.

Effect of Ga 2 O 3
The "skin effect" of high-frequency currents is a well-known natural phenomenon in physics, especially in electromagnetism. However, the "skin effect" of some oxides or halides in brazing has rarely been reported. It was found that [18] adding a very small amount of Ga 2 O 3 to the CsF-RbF-AlF 3 flux could increase the brazed area by about 50~90%. Observing the surface of the spreading filler metal via its spectrum showed that much Ga 2 O 3 was enriched at the spreading surface and at the edge of the spreading area. Further, the relative amount of Ga 2 O 3 was higher here than in the middle part, from which it could be concluded that the chemical reaction mechanism resulted from the skin effect of the Ga 2 O 3 , which enabled the Ga 2 O 3 to flow fast and then promote the CsF-RbF-AlF 3 flux flow, so as to improve the spreadablity of the filler metal.
Extremely small amounts of Ga 2 O 3 were enriched to participate in the reaction, and played a nonnegligible role. The appearance of MgGa 2 O 4 indicated that the reaction between the flux and the MgO, which was more stable than Al 2 O 3 , was also liable to react as in Equation (5). However, RbF did not appear as a compound phase in the residue, which indicated that the compound of Rb was diffused-distributed, and acted as catalysis and welding aid. CsF always appeared in the form of Cs 11 O 3 , which indicated that CsF did undergo chemical reactions with Al 2 O 3 and MgO, and replaced the O atoms in Al 2 O 3 and MgO. This result explains why the spreading area increased with the increased addition of Ga 2 O 3 , shown in Figure 3.

Effect of GaF 3
From the above it was demonstrated that the CsF-RbF-AlF 3 flux doped with GaF 3 obtained the maximum spreading area, both on 6061 aluminum alloy and Q235 steel, and reached the highest shear strength of 126MPa. The CsF-RbF-AlF 3 flux doped with GaF 3 showed remarkable superiority to that doped with ZnF 2 , Zn(BF 4 ) 2 and Ga 2 O 3 .
The mechanism of GaF 3 could be summarized as a "synergistic effect" on the oxide removal of steel. When GaF 3 was doped, the flux could not only remove the oxide films on the surface of the aluminum alloy, but could also remove those on the surface of the steel at the same time, thus making it superior to brazed aluminum-steel heterogeneous materials.
4ZnO + 2GaF 3 → ZnGa 2 O 4 + 3ZnF 2 The ionic compound GaF 3 consists of cations Ga 3+ and anions F − , and makes it such that Ga 3+ and F − ions can be more easily dissociated from GaF 3 than from Ga 2 O 3 , as in Equation (6), while been