Separation and Recovery of Refined Si from Al–Si Melt by Modified Czochralski Method

Separation of refined silicon from Al–Si melt is still a puzzle for the solvent refining process, resulting in considerable waste of acid and silicon powder. A novel modified Czochralski method within the Al–Si alloy is proposed. After the modified Czochralski process, a large amount of refined Si particles was enriched around the seed crystalline Si and separated from the Al–Si melt. As for the Al–28%Si with the pulling rate of 0.001 mm/min, the recovery of refined Si in the pulled-up alloy (PUA) sample is 21.5%, an improvement of 22% compared with the theoretical value, which is much larger 1.99 times than that in the remained alloy (RA) sample. The content of impurities in the PUA is much less than that in the RA sample, which indicates that the modified Czochralski method is effective to improve the removal fraction of impurities. The apparent segregation coefficients of boron (B) and phosphorus (P) in the PUA and RA samples were evaluated. These results demonstrate that the modified Czochralski method for the alloy system is an effective way to enrich and separate refined silicon from the Al–Si melt, which provide a potential and clean production of solar grade silicon (SoG-Si) for the future industrial application.

Effect of Zirconium on Boron Removal in Electromagnetic Solidification Refinement of Silicon from Si-Al Melt

This work investigates the removal of B in Si by the addition of Zr in the electromagnetic solidification refinement of silicon-aluminum melts. As Zr has a strong affinity for B and can form the thermodynamically stable compound of ZrB2, the B content of lower-grade Si is expected to be effectively removed by adding a small amount of Zr to the Si-55 at% Al melt. The results show that Zr is strongly responsible for the decrease in B content of refined Si. The removal fraction of B significantly increased from 60.2% to 97.3% by adding a small amount of Zr (0 to 3500 ppmw). In addition, the removal fraction of Zr from Si was found to be as high as 98.6%; however, its residual content in the refined Si was significantly larger than its solid solubility in Si, possibly due to the non-equilibrium solidification occurring during the refining process.

Si Purity Control and Separation from Solidification of Al–30Si Melt under Pulse Electromagnetic Field

Purification of metallurgical-grade silicon (MG–Si) by a combination of Al-Si solidification refining and electromagnetic oscillating separation and acid leaching collection has been studied. The primary Si crystals and Al-Si alloy in hypereutectic Al-30%Si melt were separated during solidification under the pulse electromagnetic field (PEF). The results show that the Si content in Si-rich layer increases with increasing discharging frequency. The typical metallic impurities (Fe, Ti, and Ca) have removal fraction higher than 99.5%. The removal fractions of the impurities B and P which are more difficult to remove are over 90% and 85%, respectively.

Valorization of agricultural waste into activated carbons and its adsorption characteristics for heavy metals

Activated carbon derived from pine cones waste was prepared by carbonization at 450°C, activated by different activating agents: ZnCl 2, H 2 SO 4 and NaOH, and then pyrolyzed at 600°C. Adsorption of Cr VI and other heavy metals (Mn II, Fe II, Cu II) on activated carbons was investigated to evaluate the adsorption properties. Special attention was paid to the effects of carbon surface functionalities that were analyzed by FT-IR and zeta potential study. Moreover, XRD study of activated carbon was also carried out. Results had shown that activated carbon by NaOH was the best adsorbent for removal of chromium VI from wastewater. The solid-solution interaction was determined by analyzing the adsorption isotherms at room temperature at different pH. When pH is above 4, the removal fraction of Cr (VI) ions decreased with the increase of pH. The removal fraction of Cr (VI) ions decreased below pH 4. The preferable removal of Cu (II) over Mn(II) and Fe (II) could be due to its lower affinity to solvent.Pseudo-second order equation provided the better correlation for the adsorption kinetics data. Equilibrium isotherms were determined by Fowler-Guggenheim model.

Reductive Removal of Phosphorus in Silicon Using CaO-CaF2 Slag

In order to verify an alternative metallurgical process of phosphorus removal for solar grade silicon (SOG-Si), slag treatment of metallurgical grade silicon (MG-Si) was conducted followed by acid leaching in the present study. MG-Si containing certain amount of phosphorus and calcium was equilibrated at 1723 and 1773 K with several compositions of the CaO-CaF2 slags and phosphorus in molten silicon was confirmed to be removed into slag phase also by reducing reaction as a form of phosphide ion, P3-, in addition to the phosphate ion, PO43-. These contents were separately determined by a wet chemical analysis method developed by ourselves. Although the distribution ratio of phosphorus could not exceed the highest reported values of 3, subsequent leaching brought about considerably high fraction of P removal. The removal fraction of 95.6% was attained when 5 g of silicon was treated with 10 g of the slag at 1773 K followed by the acid leaching, which would be much higher than that expected by the ordinary oxidizing slag treatment. Although the possibility of reducing dephosphorization by slag treatment was clarified, more effective condition should be pursued by changing slag composition, calcium content of silicon, temperature, etc.

TOLERANCE OF SCALE-FREE NETWORKS UNDER DEGREE SEGMENT PROTECTION AND REMOVAL

We study the tolerance of scale-free networks (following a power-law distribution P(k) = c⋅kα) under degree segment protection and removal. We use percolation theory to examine analytically and numerically the critical node removal fraction pc required for the disintegration of the network as well as the critical node protection fraction ppc necessary to immunize the network against the disintegration. We show that when degree segment protection is prior to degree segment removal and 2 ≤ α ≤3, scale-free networks are quite robust due to the extremely low value of ppc. Meanwhile, if we protect a degree segment with a fixed fraction of nodes, the threshold pc has a generally downward trend as the degree sum of the segment decreases, but it is not strictly monotonic.