基于CFD的流態(tài)化多晶硅CVD過程研究
本文選題:多晶硅 + CVD; 參考:《青島科技大學》2015年碩士論文
【摘要】:本文基于CFD分析手段,針對流化床反應(yīng)器內(nèi)復(fù)雜的多晶硅CVD氣-固反應(yīng)過程編寫Mfix代碼,擴充了硅氫化合物熱力學數(shù)據(jù)庫,建立了耦合化學反應(yīng)、熱量、動量、質(zhì)量傳遞的數(shù)值模型,對流化床內(nèi)的反應(yīng)過程和溫度、壓力、氣速、顆粒分布等情況進行了分析,并對無定形硅粉成核機理進行了探索性研究。具體工作如下:1、根據(jù)量子化學計算結(jié)果,確定了102個硅氫化合物(Si1-Si10)的分子結(jié)構(gòu)式;采用基團因子貢獻法確定了各硅氫化合物的標準生成焓、標準摩爾熵以及七個溫度(300K、400K、500K、600K、800K、1000K、1500K)下的比熱容;利用Matlab軟件編程計算了104種硅氫化合物的熱力學物性多項式系數(shù)a1-a7、a15,借助化學分子量計算器計算了上述所有物質(zhì)的分子量;將NIST數(shù)據(jù)庫、商業(yè)軟件(Fluent、Chemkin、Mfix等)、實驗和模擬方面的文獻數(shù)據(jù)與本文計算結(jié)果進行了對比,校正了H2、SiH4等Mfix已有2種物性,擴充了SiH2、Si2H6等Mfix缺乏的102種硅氫化合物的熱力學數(shù)據(jù),建立了硅氫化合物熱力學數(shù)據(jù)庫。該數(shù)據(jù)庫可準確計算各物質(zhì)的比熱、熵、焓等熱力學物性,為本文數(shù)值模擬反應(yīng)源項的計算和無定形硅粉成核機理的研究打下了基礎(chǔ)。2、確定了多晶硅CVD表面化學反應(yīng)、氣相反應(yīng)共222個反應(yīng)的動力學參數(shù);通過在控制方程中加入反應(yīng)速率源項、化學反應(yīng)熱、輻射傳熱、相間傳熱系數(shù)、顆粒碰撞耗散項等實現(xiàn)了化學反應(yīng)和質(zhì)量、動量、熱量傳遞過程的耦合;采用卡迪爾網(wǎng)格劃分方法劃分網(wǎng)格,提高了計算速度和準確性;采用二階迎風格式對控制微分方程進行離散,SIMPLE算法求解壓力耦合方程,進而完成對流態(tài)化多晶硅CVD過程的數(shù)值模擬計算。3、采用Caussat和Hsu的實驗條件模擬流態(tài)化多晶硅CVD過程,多晶硅CVD速率誤差分別為0.8%-30%、19%-28%,表明本文建立的數(shù)學模型能確切描述流態(tài)化多晶硅的CVD過程。綜合多晶硅CVD反應(yīng)速率、抑制無定形硅粉形成兩方面因素,選擇氫氣為流化氣體,基于CFD分析了多晶硅CVD速率隨硅烷進口濃度、進口操作氣速、操作溫度、操作壓力變化情況,確定硅烷與氫氣最佳操作進氣比為3/17(即硅烷進口濃度15%),最佳操作氣速為最小流化速度的5.5倍,最佳操作溫度為963.15 K,最佳操作壓力為0.2 MPa。4、對反應(yīng)器出口處氣相質(zhì)量、速度、溫度分布,反應(yīng)器內(nèi)軸向、徑向氣相分布,反應(yīng)器軸向固含率分布進行了分析,對無定形硅粉成核機理進行了初步探索,結(jié)果表明:反應(yīng)器內(nèi)以甲硅烷和硅烯的非均相熱裂解反應(yīng)為主,氣相反應(yīng)為輔;氣相質(zhì)量分布與速度分布、溫度分布直接相關(guān),軸向速度對氣相分布的影響高于徑向速度分布;無定形硅粉在反應(yīng)器稀相區(qū)質(zhì)量分率較大,低溫利于抑制無定形硅粉的形成。本文建立的熱力學數(shù)據(jù)庫可準確表達各硅氫化合物在不同壓力、溫度下的比熱、熵、焓等熱力學物性,為本文計算結(jié)果的可靠性奠定了基礎(chǔ);本文建立的綜合氣固流體力學、動力學且耦合化學反應(yīng)的數(shù)學模型能確切描述流態(tài)化多晶硅CVD過程,給出不同工藝條件對反應(yīng)過程的影響,實時分析流化床內(nèi)溫度、氣速、反應(yīng)速率、硅氫化合物質(zhì)量等分布情況。本文研究結(jié)果可對多晶硅CVD流化床反應(yīng)器內(nèi)構(gòu)件的設(shè)計提供指導,對高純度多晶硅實際生產(chǎn)工藝的設(shè)計提供理論支持。
[Abstract]:In this paper, based on CFD analysis, the Mfix code is written for the complex polysilicon CVD gas solid reaction process in a fluidized bed reactor, and the thermodynamic database of the silicon hydrogen compound is expanded. The numerical model of the coupling chemical reaction, heat, momentum, mass transfer, the reaction process in the fluidized bed and the temperature, pressure, gas velocity, particle distribution and so on are established. The analysis is carried out and the nucleation mechanism of amorphous silicon powder is explored. The specific work is as follows: 1, according to the results of quantum chemical calculation, the molecular structure of 102 silicon hydrogen compounds (Si1-Si10) is determined, and the standard formation enthalpy, standard molar entropy and seven temperatures of each silicon hydrogen compound are determined by the group factor contribution method (the group factor contribution method). The specific heat capacity of 300K, 400K, 500K, 600K, 800K, 1000K, 1500K) was programmed by Matlab software to calculate the thermodynamic property polynomial coefficient A1-A7, A15, and the molecular weight of all the above substances was calculated with the chemical molecular weight calculator. Compared with the results of this paper, 2 kinds of physical properties of H2, SiH4 and other Mfix have been corrected. The thermodynamic data of 102 kinds of silicon hydrogen compounds, such as SiH2 and Si2H6, are expanded, and the thermodynamic database of the silicon and hydrogen compounds is established. This database can accurately calculate the thermodynamic properties of the specific heat, entropy and enthalpy of each substance, which is the value of this paper. The calculation of the simulated reaction source term and the study of the mechanism of amorphous silicon powder nucleation have laid the foundation.2, determined the surface chemical reaction of the polysilicon CVD, and the kinetic parameters of 222 reactions in the gas phase reaction; by adding the reaction rate source term in the control equation, the chemical reaction heat, the radiation heat transfer, the interphase heat transfer coefficient, the particle collision dissipation term and so on. The coupling of the chemical reaction and the mass, momentum and heat transfer process is used. The computational speed and accuracy are improved by using the Kadeer grid division method. The two order upwind scheme is used to discretize the control differential equations. The SIMPLE algorithm is used to solve the pressure coupling equation, and then the numerical simulation of the CVD process of the convective state polysilicon is completed. .3, using the experimental conditions of Caussat and Hsu to simulate the flow of fluidization polysilicon CVD process, the CVD rate error of polysilicon is 0.8%-30%, 19%-28% respectively. It shows that the mathematical model established in this paper can describe the CVD process of the fluidization polysilicon. It combines the CVD reaction rate of polysilicon, inhibits the formation of amorphous silicon powder in two aspects, and chooses hydrogen as the fluidization Gas, based on the CFD analysis, the CVD rate of polysilicon with the concentration of silane inlet, inlet operating gas speed, operating temperature, and operation pressure change, the optimum operation intake ratio of silane and hydrogen is 3/17 (i. e. 15% of silane inlet concentration), the optimum operating gas velocity is 5.5 times of the minimum flow rate, the optimum operating temperature is 963.15 K, and the optimum operating pressure is 0. .2 MPa.4, the gas phase mass, velocity, temperature distribution at the outlet of the reactor, axial, radial gas distribution and axial solid content distribution in the reactor were analyzed. The nucleation mechanism of amorphous silicon powder was preliminarily explored. The results showed that the non homogeneous thermal cracking reaction of methylene silane and Silene was the main reaction in the reactor and the gas phase reaction was supplemented in the reactor. The gas phase mass distribution is directly related to the velocity distribution and the temperature distribution. The effect of the axial velocity on the gas phase distribution is higher than the radial velocity distribution; the mass fraction of the amorphous silicon powder in the dilute phase region of the reactor is larger, and the low temperature is beneficial to the formation of the amorphous silicon powder. The thermodynamic database established in this paper can accurately express the different silicon hydrogen compounds in different types. The thermodynamic properties such as pressure, specific heat, entropy and enthalpy at temperature have laid the foundation for the reliability of the results of this paper. The mathematical model of integrated gas-solid fluid mechanics, dynamics and coupling chemical reaction can describe the CVD process of fluidization polysilicon accurately, and the effect of different process conditions on the reaction process is given, and the real-time analysis of fluidization is given. The distribution of temperature in bed, gas velocity, reaction rate, and the quality of silicon and hydrogen compounds. The results of this paper can provide guidance for the design of the internal components of the polysilicon CVD fluidized bed reactor, and provide theoretical support for the design of the actual production process of high purity polysilicon.
【學位授予單位】:青島科技大學
【學位級別】:碩士
【學位授予年份】:2015
【分類號】:TQ127.2
【相似文獻】
相關(guān)期刊論文 前10條
1 席德清 ,余保發(fā) ,周淑芬;對Ⅰ皮磨最佳操作指標的探討[J];鄭州糧食學院學報;1984年03期
2 王靜康,廖立,張遠謀;間歇結(jié)晶過程最佳操作時間表安排的研究[J];化工學報;1990年04期
3 蔡平;金涌;俞芷青;;催化裂化再生器最佳操作氣速的選擇[J];石油煉制與化工;1991年06期
4 張懷信;胱氨酸制備中的最佳操作曲線研究[J];氨基酸雜志;1983年04期
5 趙宗模;;旋液器的物料平衡[J];廣東化工;1983年03期
6 裴定榮;;空分提氬(六上)[J];冶金動力;1985年04期
7 ;丙烯腈催化水合制丙烯酰胺反應(yīng)可逆性的研究[J];齊魯石油化工;1981年05期
8 ;南鋼15t轉(zhuǎn)爐LD-NGF高壓復(fù)吹技術(shù)的開發(fā)與應(yīng)用[J];北京科技大學學報;1992年04期
9 ;酒精廠蒸餾和發(fā)酵車間的最佳操作[J];甘蔗糖業(yè);1982年01期
10 何潤霞,白雪,秦海莉,楊瑞芳;CO中變反應(yīng)器最佳操作溫度分析[J];內(nèi)蒙古科技與經(jīng)濟;1999年S2期
相關(guān)重要報紙文章 前2條
1 葛梅偉;首鋼評出90項最佳操作法[N];中國冶金報;2007年
2 首鋼;美麗焊花點亮閃光人生[N];首都建設(shè)報;2010年
相關(guān)碩士學位論文 前1條
1 宗芳;基于CFD的流態(tài)化多晶硅CVD過程研究[D];青島科技大學;2015年
,本文編號:2039588
本文鏈接:http://sikaile.net/kejilunwen/huaxuehuagong/2039588.html