The reaction equilibrium constant of Equation (4), calculated by HSC Chemistry 6.0, is K eq = 3.827·10 2 . This reaction also occurs in flux #5. Thus, the reaction corresponds to the endothermic peak shared by fluxes #4 and #5 at 460 °C. The heating curves of fluxes #4 and #5 move in the positive direction of the y-axis after the reaction occurs, indicating that the reaction absorbs a quantity of heat. In the range of 25 °C to 100 °C, KMgCl 3 (H 2 O) 6 was confirmed to exist in the MgCl 2 –KCl–H 2 O system by Cui et al. [ 50 ]. Since it is obtained by heating the mixture of MgCl 2 and KCl after absorbing water, the reaction might be as follows:
For more information, please visit Kingsun.
c,f,h–j show the XRD spectra of fluxes #3, #6, #8, #9, and #10, respectively. All these fluxes have no change in the physical phase before and after melting. The main components of flux #4 in d are KCl and MgCl 2 , and then its composition changed to KCl, MgF 2 , and KMgCl 3 (H 2 O) 6 after melting. The reaction for the formation of MgF 2 is shown in the following response equation [ 49 ]:
The calculated reaction equilibrium constant of Equation (2) is K eq = 9.132·10 −30 , and Equation (3) is confirmed in the NaF–Na 2 SO 4 system, and Na 3 FSO 4 exists in the NaF–Na 2 SO 4 system. The reaction corresponds to the small endothermic peak of flux #2 at 700 °C (as shown in b). The heating curve of flux #2 moves in the negative direction of the y-axis after the reaction occurs, indicating that the reaction releases a quantity of heat.
b shows the XRD spectrum of flux #2. The main components of flux #2 are Na 2 SO 4 and Na 2 CO 3 , the phases shown in the figure are NaF, Na 2 CO 3 , and Na 3 FSO 4 , NaF, and Na 3 FSO 4 are produced in flux #2 after the reaction. Their reactions are shown in the following response equations [ 48 ]:
The XRD spectra of the fluxes is shown in . a shows the XRD spectrum of flux #1. The main components of flux #1 are KCl and Na 2 CO 3 . After melting, no other phases were found in the XRD diffraction spectrum, indicating that the composition of the flux after melting is stable, and there is no reaction between the main components.
In this study, the melting point of the flux was defined as the onset point [ 45 ] of the first stable endothermic peak in the heating curve [ 46 ] (i.e., there is a corresponding thermal effect peak in the cooling curve). Thus, the melting points of fluxes #1, #3, #6, #8, and #10 are the onset points of the endothermic peaks. Several endothermic peaks were present in the heating curves of fluxes #4, #5, and #7 at 100–250 °C. According to , fluxes #4 and #5 contain MgCl 2 , and flux #7 contains K 2 CO 3 , the three compounds have extremely high water absorption. So, it is inferred that the endothermic peak is caused by sample dehydration. As a consequence, the melting points of fluxes #4 and #5 are the onset point of the endothermic peak that appears at 400–500 °C, as shown in b. For flux #2, a small endothermic peak appears at 400–500 °C. Due to the high temperature, and the fact that the corresponding thermal effect peak does not appear in the cooling curve, it is inferred that it is caused by an irreversible reaction. As a result, the melting point of flux #2 should be the onset point of the endothermic peak appearing at 600–700 °C, as shown in b. In the heating curve of flux #9, there is an endothermic peak at 300 °C, but it does not correspond to the cooling curve (shown in b). It is conjectured that sodium nitrate undergoes a second-order (order–disorder) phase transition at this temperature [ 47 ]. Therefore, the melting point of flux #9 should be the onset point of the endothermic peak appearing at 600–700 °C. All the melting points of the fluxes, as shown in d, will be further confirmed in the high-temperature contact angle experiments.
As shown in a, the heating curves of fluxes #3, #6, #7, #8, and #10 showed only a single endothermic peak. Fluxes #1, #2, #4, #5, and #9 showed multiple endothermic peaks in the heating curves. The presence of multiple endothermic peaks in the heating curve indicates that there are multiple thermal effects during sample heating, while the presence of a single endothermic peak indicates that there is only one thermal effect during sample heating [ 43 , 44 ].
shows the variation in contact angle between fluxes and alumina over the aluminum alloy refining temperature range (730–760 °C). At the refining temperature range, flux #4 has the largest contact angle with alumina. Fluxes #2, #7, and #10 all contain Na 2 CO 3 as the main component, so their contact angle is close, but flux #10 contains NaF. NaF can decrease the contact angle, which correlates with interfacial tensions among the aluminum melt, molten flux, and inclusions [ 18 ], accordingly, reducing the contact angle of flux #10 and alumina.
shows the contact angle of flux #10 and alumina at different temperatures. The flux did not start melting until 541 °C, and the contact angle with the alumina remained at 90°. At 553 °C the flux started to soften, which was close to the DSC test result (562.2 °C). The error values may be caused by different test conditions and equipment. When flux #10 was heated to 555 °C, the flux started to spread on the alumina ceramic plate, and the contact angle was reduced to 61.51°. Then, as the heating temperature increased, the contact angle between the flux and the alumina showed a trend of gradually decreasing, and the flux also gradually spread out on the surface of the alumina in the image, which also indicated that there was a good wettability between flux #10 and alumina. Then, aluminum alloy is heated to a refining temperature of 760 °C and the contact angle is continuously reduced, where the contact angle changes, as shown in , with the contact angle eventually reducing to 12.78°.
shows the contact angle of flux #2 and alumina at different temperatures. The temperature at which the sample begins to soften is 653 °C, which is close to the melting point (653.8 °C) measured experimentally, shown in d. The contact angle becomes progressively smaller with increasing temperature and reaches a minimum of 28.41° at the refining temperature (760 °C). and show the contact angles of flux #4 and flux #7, with alumina at different temperatures, respectively. The softening temperatures of the samples are both close to the DSC results. Flux #4 has a larger temperature range from the onset of softening to the liquid state, presumably due to the phase change at a high temperature, which increases the solid–liquid phase zone of the mixture and also results in a larger contact angle of the mixture with alumina at the refining temperature compared to the other fluxes.
Based on the data in and d, fluxes #2, #4, #7, and #10 were selected for high-temperature contact angle testing with alumina in this paper. According to the research of Utigard et al. [ 24 ], salts such as chloride and carbonate in the flux will not affect the wettability of the flux, which contributes to the flux grouping of this work. Since only one of the main components in the compositions of fluxes #1, #2, and #3 is different, they were grouped. Fluxes #4 and #5 were grouped because the chloride in the main components is different. Fluxes #6 and #7 were grouped because the main components are different. After all, the main components of flux #6 are chloride and carbonate and #7 is all carbonate. Fluxes #8, #9, and #10, for which the main components are all made up of three kinds of molten salt, were grouped.
The driving force for the diffusion of hydrogen at any given moment is the difference between the hydrogen concentration in the melt and the hydrogen concentration at the interface between the liquid and gas phases. The [H] in the melt diffuses through the boundary layer between the liquid and gas phases to the surface of the bubble, leading to an increase in hydrogen concentration at the interface, and the hydrogen is generated on the bubble surface and diffuses into the bubble rapidly. With the diffusion of hydrogen in the boundary layer, the hydrogen is formed at the bubble surface, and diffuses to the inside of the bubble rapidly, increasing the partial pressure of hydrogen inside the bubble. Until the interface reaches chemical equilibrium, and satisfies [PctH] = [PctH] e , the driving force for diffusion and the difference in partial pressure of hydrogen between the inside and outside of the bubble is zero, and the hydrogen that is formed at the interface can no longer diffuse into the bubble. Finally, the bubbles float up to the surface of the aluminum melt and burst. According to the studies of Utigard et al. [ 19 ], Na 2 CO 3 in the flux will form bubbles within the aluminum melt by the following reaction, and remove the hydrogen that is dissolved in the aluminum melt [ 57 ]:
It is assumed that the concentration of hydrogen in the aluminum melt at any given moment is [PctH], the equilibrium concentration of hydrogen on the bubble surface is [PctH] e , the activity coefficient of hydrogen is f H , the partial pressure of hydrogen inside the bubble is P H2 , and the equilibrium constant is K, that is temperature dependent only. The hydrogen in the aluminum melt reacts at the surface of the bubbles as follows [ 56 ]:
where V is the volume of the molten flux droplet (defined as constant), and R (t) is the time–dependent base radius of the droplet, which is in the shape of a spherical cap [ 18 ]. According to Equations (9), (12) and (13), through increasing the size of flux powders (i.e., the volume of molten flux droplets at the refining temperature), and adding NaF into the flux (i.e., decreasing θ, which correlates with interfacial tensions among the three phases), it is possible to achieve a greater driving force for spreading, and realize a wetting mechanism transition from adsorbing to engulfing, which will then respectively improve the efficiency of wetting and rate of floatation of a molten flux droplet. In consequence, as shown in and , flux #10 has a better spreading effect and a smaller contact angle, because the addition of NaF in flux #10 results in a reduction in surface tension between flux #10 and the inclusions. In addition, Fu et al. [ 9 ] indicated that the removal of oxide inclusions is important to eliminate pores, because part of the hydrogen will accumulate and adsorb around the inclusions, as shown in a. The hydrogen that accumulates around the inclusions, and is adsorbed by the inclusions, could be removed along with the removal of inclusions. The hydrogen could be adsorbed in the inclusions, which could also be eliminated through the inclusions’ removal. However, there is still some dispersive hydrogen within the aluminum melt, which is removed by the alternative scheme shown in c. When the flux is added to the aluminum melt, some components of the flux react, to form bubbles, and the hydrogen that is dissolved in the aluminum melt diffuses into the bubbles. Then, it is removed from the aluminum melt with the bubbles. That is the reason we present the degassing scheme in c.
Combining Equations (8) and (11), it can be concluded that ΔG < 0. The experiments show that the flux and the oxide inclusions have good wettability, from which it can be seen that the oxide inclusions can be spontaneously adsorbed by the flux at the refining temperature.
The process of the inclusions being adsorbed by the flux is shown in a. The process of inclusions migrating from alloy melt to flux was analyzed thermodynamically. Assuming that the surface area of the inclusions is S, the surface free energy G 1 of the inclusions before they are adsorbed by the flux can be expressed as:
When the fine inclusions that are floating in the melt are close to the flux interface, the inclusions migrate into the molten flux under the adsorption effect between the two phases, and the adsorption process is roughly divided into three stages, as shown in :
According to Majidi and Liu et al. [ 26 , 51 ], although the density of oxide inclusions (3.43–3.72 g/cm 3 ) are higher than that of molten aluminum (2.3–2.5 g/cm 3 ), they float in the melt because of their high area to volume ratio, and can increase their buoyancy by adsorbing hydrogen. Usually, because of the good wettability of the inclusions, and the lower density (1.9–2.1 g/cm 3 ) compared to the aluminum melt (2.3–2.5 g/cm 3 ) [ 19 , 52 ], fluxes are often used to remove the oxide inclusions by absorbing the inclusions, forming clusters, and then floating up to the surface of the molten aluminum together.
If you are looking for more details, kindly visit Aluminium Casting Flux.
shows the microstructures of the AA6111 alloy before and after flux refinement, and the pores and inclusions content before and after flux refinement.
a shows a metallographic photograph of a cast sample of untreated AA6111 aluminum alloy. Numerous pores and inclusions can be seen. The pores are nearly 100 µm in diameter and irregularly distributed. According to Wang’s [58] research, aluminum oxide inclusions within the aluminum melt often show an amorphous state, in the form of strings and films. After the melt cooling, the oxide formed stripe-like inclusions, as shown in a. In addition, the aluminum oxide cluster makes the ambient Al atoms disordered in the aluminum melt, and creates more vacancies to accumulate hydrogen, which causes the formation of pores in the aluminum [9]. Thus, inclusions in the form of strings and films are associated with concomitant pores, as shown in a–j.
b–d show the microstructure of AA6111 aluminum alloy refined by fluxes #1, #2, and #3, respectively. It can be seen that the pores in b are significantly smaller and more uniform in size than in a, and there are fewer inclusions than before the refinement. Only a few small inclusions remain, which are also more evenly distributed than before the flux refinement. Compared to the large area and size of the pores and inclusions in a, the pores and inclusions in c, after refining with flux #2 have been significantly removed, with the pores and inclusions being smaller and less concentrated, and more evenly distributed throughout the alloy. After melting, NaF formed in flux #2 reduced the interfacial tension between the flux and inclusions, and reduced the contact angle between flux #2 and the inclusions. In addition, flux #2 contains 51.8wt.%Na2CO3, which will react as in Equation (16), to form CO2 bubbles and remove the hydrogen. Thus, compared with fluxes #1 and #3, the result of refinement (as shown in ) was better. The pores and inclusions of the sample after refining were reduced to 1.35%. A comparison shows that the number and size of pores in d are slightly improved compared with a, but the improvement is limited and the reduction in inclusions is more pronounced than the pores.
e,f shows the metallograph of a cast sample after refining the melt with flux #4 and flux #5, respectively. It can be seen that the pores and inclusions in e are significantly reduced compared to the sample without any flux refining (shown in a), which shows the pores and inclusions are significantly smaller in size, averaging around 20 µm and 5 µm, respectively. The sample treated by flux #5 (shown in f) is similar to the result using flux #3 (shown in d), with the pores appearing more concentrated in the flux #5-treated sample, but slightly less numerous than in d. Although the contact angle between flux #4 and aluminum oxide is large, about 80° (shown in ), fluxes #4 and #5 contain MgF2. This can react in the molten aluminum as follows [24]:
MgF2+Al=AlF3+Mg
(17)
AlF3+Na/Ca=NaF/CaF2+Al
(18)
This compensates for the poor wettability of flux #4 with oxide at the refining temperature, and improves the refining effect of flux #4. In addition, flux #4 contains MgCl2 and KCl, a combination that has good cover properties [19] and prevents the aluminum melt from absorbing hydrogen. After refining, the pores and inclusions content of the sample is reduced to 1.35% (shown in ).
g,h shows the metallograph of a cast sample after refining by flux #6 and flux #7, respectively. In g, the inclusions were significantly reduced after being refined by flux #6, and the size and number of pores were also significantly smaller and less concentrated than those in the sample without any flux refinement. It can be seen that the pores and inclusions are significantly reduced and the size is reduced to around 10–20 µm, compared to a. Within the refining temperature range (730–760 °C), the contact angle between flux #7 and aluminum oxide is about 40°, which indicates that the flux can absorb inclusions well in the melt. The composition of flux #7 contains 46.5 wt.% Na2CO3 and 40.5 wt.% K2CO3, and the composition of flux #6 contains 46.3 wt.% K2CO3. According to Utigard et al. [19,24], K2CO3 and Na2CO3 can form CO2 bubbles in the aluminum melt and thus remove hydrogen. The pores and inclusions content of the sample after refining is reduced to 1.48% (as shown in ).
i–k show the microstructure of the AA6111 aluminum alloy refined by fluxes #8, #9, and #10, respectively. The effect of the refinement using flux #8 is similar to that using flux #7, although flux #8 is better at purifying inclusions. It can be seen that the pores and inclusions in j are more significantly reduced and the size is smaller than those before the refinement by flux #9. As can be seen in k, the number of pores and inclusions in the sample is reduced and the size is not as large as in a, and the size of the pores and inclusions is only about 5 µm, or smaller, and evenly distributed. In the group of fluxes including #8, #9, and #10, the best refining performance is from flux #10 (as shown in ). At the refining temperature (760 °C), the contact angle between flux #10 and aluminum oxide is 12.78°. Its good wettability with inclusions is due to the composition containing 11.0 wt.%NaF. In addition, 29.5 wt.%NaCl is present in flux #10, which makes it easy to separate the flux from the aluminum melt. The Na2CO3 (46.5 wt.%Na2CO3) in the flux can react (Equation (16)) to form CO2 bubbles and remove the hydrogen, as shown in the scheme in c. AA6111 alloy treated with flux #10 has the highest effect of refinement. After refining, the pores and inclusions content of the sample is reduced to 1.11% (as shown in ), and the inclusions and pores removal rate reaches 62.5%.
l shows the microstructure of the alloy refined by the STJ–A3 flux. It can be seen that the pores have reduced significantly in the sample, and only retained some fine and slender inclusions. According to the results in , the refining effect of the STJ–A3 flux is better than all the rest of the fluxes, and only slightly worse than flux #10.
From the results of this work, it can be concluded that the refining efficiency of different fluxes is highly dependent on their chemical composition, melting temperature, and wettability with aluminum oxide, which agrees with the conclusions of Gyarmati et al. [15]. The ten fluxes in this work all have significantly lower melting temperatures than the refining temperature, which is a prerequisite to enable the fluxes to have a refining effect in the melt. For the molten salt-based fluxes (fluxes #1–#7), fluxes #2 and #7 show better refining effects, because they contain two exothermic components. This is because the refining effect is determined by the coupling of the degassing and inclusions removing effects. The Na2CO3 and K2CO3 in fluxes #2 and #7 can react in the melt to form bubbles, removing some dispersive hydrogen dissolved in the aluminum melt. Thus, if the flux is used to refine alloys with a high hydrogen content, the content of components among which the exothermic reaction occurs should be increased accordingly. After comparing the refining effects of fluxes #1, #3, #4, and #5, it was found that fluxes #3 and #5, which contained NaCl, were not as effective as fluxes #1 and #4, which contained KCl. For the three molten salt-based fluxes, #8, #9, and #10, the refining effect of flux #10, which contained NaF, was better than that of flux #8, which contained KCl. Thus, if the flux is used to refine alloys with a high inclusions content, attention should be paid to the contact angle between the flux and the oxide [19,24], and molten salt, that improves the wettability of the flux and inclusions, could be added as appropriate.
Originally Posted by
Are you interested in learning more about High-Purity Metal Chromium for Sale? Contact us today to secure an expert consultation!