With the rapid development of photovoltaic industry in recent years, improved Siemens method has been widely used as the main method to produce polysilicon. One of the notable characteristics of this method is the large amount of hydrogen consumption. Therefore, it is necessary to use the method of hydrogen recovery to reduce the amount of hydrogen added from the outside, and the use of activated carbon adsorption purification to recover hydrogen is a common method. Although activated carbon is widely used in industry, the design of adsorption device is still based on production experience or a variety of tests, and the design results have a large deviation, mainly because the surplus is too large, resulting in waste of investment. In the past, it was difficult to accurately design the activated carbon adsorption device for two main reasons: first, people did not master the adsorption performance of activated carbon under various conditions, or did not master the rule of its change. Up to now, with the continuous efforts of countless researchers, dozens of adsorption equilibrium equations were summarized, but they were not a simple relationship with temperature, pressure or composition. Adsorption process was controlled by a variety of related process variables and basic equilibrium and kinetic relations. Another important fact was that currently there was no good model to accurately simulate the adsorption and desorption process of activated carbon. Firstly, the internal pore of the adsorbent, the shape of the adsorbent particles, the adsorbent in the adsorbent internal heap, the gap and pore of the adsorbent bed were not regular, and the same equation was not applicable to describe the law of mass transfer and heat transfer in all nodes. Secondly, no suitable software could directly simulate the adsorption and desorption process of activated carbon. Based on the industrial production demand and existing problems, this paper carried out in-depth research, and put forward effective solutions. Firstly, the author studied some representative adsorption equilibrium equations, and selected the most suitable adsorption equilibrium equation as the basis, which was improved to describe the adsorption process and desorption process of activated carbon. Secondly, the author simplified the adsorbent bed model according to the production experience and fitted some parameters of adsorption equilibrium equation according to the known experimental data. Finally, the author chose Fluent software as the platform of simulation calculation, and applied UDF method to load the adsorption equilibrium equation and heat absorption into Fluent software. Based on the above analysis and research, this paper discussed the using activated carbon adsorption purification of hydrogen in the process of polysilicon production methods and models, introduced the selection of physical adsorption, studied the choice of multicomponent Langmuir adsorption balance equations to describe the mass transfer in the process of adsorbate on the adsorption of hydrogen, and calculated according to the mass transfer and energy transfer. In order to simulate the adsorption process more accurately, Fluent software was used to simulate the adsorption process of hydrogen adsorption and purification in polysilicon production process. In the simulation process, the adsorption bed was simplified in this paper, and the porous region model and UDF method were used to simulate the adsorption process of hydrogen chloride and chlorosilane by coconut shell activated carbon adsorbent. In the process of compiling UDF, the author referred to some relevant materials, sorted out and summarized the relevant contents, and published the correct UDF usage method in this paper, hoping to be helpful to the scientific and technological workers in related work. After the model was established, Fluent software was used to simulate the adsorption process of activated carbon, and the mass distribution maps of hydrogen chloride and chlorosilane in the adsorbent bed at different operation stages were obtained, analyzed and summarized. The simulation results showed that the activated carbon had good adsorption effects on hydrogen chloride and chlorosilane, and the purity of hydrogen reached 99.9999% after adsorption. Simulation results showed that the method used in this paper, the simulation of the adsorbent bed height derived from the calculation results and the actual adsorber bed height, simulation of the distribution of components and actual production equipment running results, which proved that this article chose the adsorption balance equations and appropriate simplification of adsorbent bed model was feasible. In this paper, for the first time the Langmuir equation and the Fluent software were used to simulate the combination of activated carbon for chlorosilane and the adsorption process of hydrogen chloride successfully, and the simulation results coincided with the operation of the actual device testing data, showing that using this method could design new adsorber device accurately, which would be helpful to save investment and reduce the operation cost. At the same time, the old device could also be reformed and optimized, which was conducive to improve production capacity and product quality. Therefore, the method described in this paper provided an advanced design means for the design and optimization of the adsorption process in polysilicon production, which had good promotion value and was an important development direction in the field of adsorption research, and made it possible to use computers to study the adsorption process from theoretical research to practical application. © Editorial Office of Chinese Journal of Rare Metals. All right reserved.
