Mechanism And Kinetic Studies Of A Two-step Reaction Process For Metallurgical Silicon To Silane

Compared with the modified Siemens process to prepare polysilicon rod through thermal hydrogenation of trichlorosilane(Si HCl3, TCS) in a bell-jar reactor, silane(Si H4) pyrolysis to granular polysilicon in a fluidized bed reactor(FBR) has been attracting more and more attentions because of its much lower energy consumption and plant investment, so the Si H4/FBR process is nowdays considered to be the most promising technology for polysilicon production. Besides, a clean Si H4 preparation process, from metallurgical silicon(MG-Si) to Si H4, can be specifically realized via coupling the TCS resdistribution and co-hydrogention of silicon tetrachloride(Si Cl4, STC) with MG-Si. Therefore, this work is focused on `the mechanism and kinetic investigations on this clean Si H4 preparation process.The main contents and results in this work are as follows:1. Thermodynamic analysis reveal that the STC-Si-H2 process is equilibrium liminted and slightly exothermal. The STC equilibrium conversion decreases slightly with increasing temperature but increases remarkably with increasing pressure and H2/STC feed ratio. TCS is the major product with trace amounts of DCS and HCl as the thermodynamically feasible by-products in the STC-Si-H2 reaction system and the TCS equilibrium selectivity decreases slightly with the increase of temperature, pressure and H2/STC feed ratio, respectively.2. The STC-Si-H2 process was performed in a fixed bed reactor packed with silicon, and the effect of the silicon purity was firstly considered. Without catalyst addition, STC hydrogenation with very pure silicon is essentially kinetically inhibited below 600oC. The impurities in MG-Si exhibited some catalytic activity. Subsequently, various transition metal compounds were evaluated and Cu Cl exhibited the excellent catalytic activity. Cu3 Si was preliminarily determined as the active species through the XRD analysis. Cu Cl was employed in the rest of the experiments, and the optimal Cu Cl loading based on the Cu Cl/Si mass ratio is 10 wt.%. The effects of the operating parameters on the Cu Cl-catalyzed STC-Si-H2 process were examined at steady state.The STC conversion increases with increasing STC space time until the reaction system reaches equilibrium. The reaction rate increases remarkably with increasing temperature and STC partial pressure at the expense of the STC equilibrium conversion. Nevertheless, increasing the pressure could improve both the reaction rate and STC equilibrium conversion. Furthermore, the reaction rate is independent of the silicon particle size.3. The mechanism of the Cu Cl-catalyzed STC-Si-H2 process was further investigated. A novel reaction mechanism was proposed on the characterization results under different reation stages. This mechanism is remarkable in that the Cu-Si surface species acts as both a catalyst and a solid reactant and can be regenerated via Cu diffusion into the bulk silicon phase. We also formulated various kinetic models based on the L-H and E-R mechanisms. Model discrimination was performed for differential operation of the reactor. The initial STC consumption rate was observed to be first order with respect to pH2. An increase in pSTC initially increased the rate; however, at higher pSTC values, the rate was less sensitive to changes in pSTC. The kinetic model based on an E-R mechanism provided a satisfactory description of the experimental results. The selected kinetic model indicated that the surface reaction between the adsorbed STC and gas-phase H2 was the RDS. The parameters were estimated using a non-linear least-squares regression. Excellent agreement was observed between the model predictions and the experimental results, indicating that the kinetic model could be applied over a wide range of conditions.4. Thermodynamic results reveal that the three-step reactions involved in the TCS redistribution system are equilibrium limited. The first two steps are endothermic, while the third is exothermal, and their respective reaction enthalpy is low. Although the equilibrium constant increases as the redistribution proceeds, the formation of Si H4 from TCS is thermodynamically unfeasible. Based on the total Gibbs energy minimization results, STC and DCS are the only main products with TCS as the feedstock, and the Si H4 concetraction is about 3% and 35% with DCS and MCS as the feedstock, respectively.5. Various kinds of ion exchange resin were evaluated in a fixed bed reactor and the weak basic anion resin coupled with tertiary amine as the functional group exhibited excellent activity and stability for TCS redistribution. Degradation of the functional group at higher operating temperature is the main reason for deactivation. The effects of the operating parameters reveal that STC and DCS are the only main products with TCS as the feedstock, and the TCS consumption rate increases with temperature and its partial pressure. The redistribution of DCS and MCS proceeds simultaneously with DCS as the feedstock. The selectivity of Si H4 increases with temperature and DCS inlet partial pressure.6. Various kinetic models were formulated based on the L-H and E-R mechanisms. Model discrimination was performed for differential operation of the reactor. Finally, the kinetic model based on E-R mechanism provided a satisfactory description of the experimental results. The selected kinetic model indicated that the surface reaction between one adsorbed chlorosilane and another gas-phase chlorosilane was the rate-determing step(RDS). The parameters were estimated using a non-linear least-squares regression. Excellent agreement was observed between the model predictions and the experimental results, indicating that the kinetic models could be applied over a wide range of conditions. With the selected resin catalyst, a continues Si H4 production was realized by a simple reaction-reluxing operation, which proves the feasibility of reactive distillation process for commercial Si H4 production.