In recent years, photovoltaic power generation using silicon solar cells has been increasing significantly. Currently, the solar-grade silicon (SOG-Si) with 6-7N (99.9999~9%) level of purity, for photovoltaic power generation is manufactured using Siemens process, which consumes drastic amount of energy & time, and uses “silica ore” as raw material that might be possibly depleted soon. There are number of alternative approaches that have been developed for solar-grade silicon such as the metallurgical process, zinc reduction process or hydrogenation process. However, these processes still require substantial energy input for the fusion or gasification and reduction of metallurgical-grade silicon (MG-Si). Therefore, there is a need to develop an alternate production process for SOG-Si that would stabilize the problems including the cost factor. Our research group focuses on the development on new preparation process of SOG-Si from diatomaceous earth as that is one of the promising candidate with abundant raw material of silica reserves: combination of wet chemical process and channel reactor process.

The wet chemical process includes dissolution/precipitation treatment of diatomaceous earth with NaOH + TMAH mixture solvent with pH adjustment, followed by acid leaching ^[1-2]. The dissolution/precipitation treatment preliminarily purifies silica by utilizing difference in pH dependences of solubility of Si and other impurities. In the acid leaching, HCl aqueous solution subsequently removes impurities from gel-state silica. These processes are profitable to remove large amount of impurities, which were however found to be hard to eliminate B that affects significantly the electrical properties of SOG-Si.

Series of our studies formerly found the followings: (i) liquid-liquid extraction process using 2,2,4-trimethyl-1,3-pentanediol (TMPD) as an extractant in organic phase (toluene) works very well to extract the B (boric acid in solution) from aqueous solution, (ii) flow type reactors are more suitable and profitable than batch type reactors for practical uses as it provides a long contact period ^[3-4], and (iii) the liquid-liquid interfacial reaction rate of boric acid and TMPD is sufficiently high to set the process diffusion-limited. Based on these findings, this study focuses on the design and testing of the practical channel rector in a larger scale (135x100x10 mm), which is fabricated using 3D printing technology. The channel was designed to have 2 characteristic structures: a throat structure that shows gradually diverging part after converging (converging/diverging structure) near Y shaped inlets and obstacles structure on downstream, as shown in Fig.1. The silica solution (aqueous phase) was prepared and injected into one of the inlet and TMPD (organic phase) solution into the other inlet. Vortices were generated around diverging part, leading to efficient mixing of each phase to produce smaller droplets of one phase dispersed into the other, which was retained by obstacles part efficiently. This continuous droplet production enhances mass transfer of boric acid to the interface to react with TMPD. 5 times repeating application found to be 99.65% removal of B at the flow rate 350 ml/min.

From these results, this kind of converging/diverging structure and presence of obstacles will enhance the mixing efficiency between aqueous and organic phases, which in turn produces maximum extraction efficiency. These kinds of channel reactors are profitable for silica purification at industries.

References:

[1] M. Bessho, Y. Fukunaka, H. Kusuda and T. Nishiyama: Energy & Fuels, 23, 4160–4165 (2009).

[2] T. Homma, N. Matsuo, X. Yang, K. Yasuda, Y. Fukunaka and T. Nohira: Electrochimica Acta 179 (2015) 512–518.

[3] N. Matsuo, Y. Matsui, Y. Fukunaka and T. Homma: Journal of The Electrochemical Society, 161 (5) E93-E96 (2014).

[4] N. Matsuo, Y. Matsui, Y. Fukunaka, and T. Homma: ECS Transactions, 50 (48) 103-108 (2013).