The CuCl-catalyzed hydrogenation of silicon tetrachloride (STC) in the presence of silicon was conducted in a fixed-bed reactor. The solid mass from different reaction stages was collected and characterized using X-ray diffraction (XRD) and scanning electron microscopy coupled with energy-dispersive X-ray spectroscopy (SEM-EDX). The characterization results were used to develop a novel reaction mechanism. This mechanism is remarkable in that the Cu–Si surface species acted as both a catalyst and a solid reactant and could be regenerated via Cu diffusion into the bulk silicon phase. Different kinetic models were formulated by employing the Langmuir–Hinshelwood (L–H) and Eley–Rideal (E–R) approaches. Model discrimination was used to select a kinetic model based on the E–R mechanism, in which the surface reaction between the adsorbed STC and gas-phase H2 over the Cu3Si active sites was assumed to be the rate-determining step (RDS). The kinetic parameters were estimated using nonlinear regression, and the model predictions were in good agreement with the experimental data.

Reactive distillation (RD) was applied for manufacturing silane from trichlorosilane (TCS). The previous studies resorted to a single RD column to attain a silane purity of 99% and to intermediate condensers to alleviate the cryogenic refrigeration load of the overhead condenser. In this paper, we present an extended scheme with the RD column assisted by two purification columns, which adopted a lowered silane purity at the RD column and achieved an ultra high purity of 99.9999% at the final column. Comparative simulation was conducted for three configurations of the RD column with one, two, and no intermediate condensers at different operating conditions for an optimal economy of operation. The cost dependency on the configuration was elucidated by the fluid flow rates and reaction efficiency in the RD column. The results showed that pursuing a silane purity over 99% at RD overhead was unadvisible due to the enormous cost for cryogenic refrigeration. Inserting two intermediate condensers was also unnecessary because it would interfere the reaction and eventually increase the overall operation cost. The optimal configuration was to insert a single intermediate condenser in between the reaction and rectifying sections, with an overhead silane content in the order of 82% at an overhead condensing temperature of about −40 °C and an intermediate condensing temperature in the range of 7–10 °C.

•HZSM-5/cordierite monolithic catalyst was prepared by washcoating method.•Monolithic catalysts produce 2.78–6.66 times larger reaction rate than the conventional ones.•90% paraffin and aromatics are reduced in the co-reaction of methanol and butene.•Calculated propylene yield is 81.6% in a six-stage monolithic reactor with SV = 5.27 gMeOH/gcat/h.Methanol to propylene (MTP) reaction system over HZSM-5 zeolites is characterized by obvious consecutive side reactions to paraffin and aromatics, which practically gives rise to a low propylene yield. In this work, HZSM-5/cordierite monolithic honeycomb catalyst was prepared by washcoating method and investigated in a lab fixed-bed reactor. Compared to the traditional extrudes, the monolithic honeycomb accelerated methanol conversion rate by 2.78 and 6.66 times when methanol was fed alone and co-fed with butene, respectively, and reduced the output of paraffin and aromatics by 90% when methanol was co-fed with butene. A two-dimensional mathematic model was then developed to simulate the HZSM-5/cordierite catalyst and validated by experimental data. On the basis of a commercial configuration of MTP reactor, a six-stage adiabatic fixed-bed reactor was simulated with ethylene and C4–C6 olefins recycled, the calculated space velocity and propylene selectivity for monolithic catalyst can be as high as 5.27 gMeOH/gcat/h and 81.6%, with significant enhancement achieved compared to the conventional catalyst with the results of 0.741 and 62.6%, respectively.

Chemical vapor deposition (CVD) at high temperature and pressure in a unique bell-jar reactor has been widely applied for high-pure polysilicon production using trichlorosilane (TCS) as a precursor. Silane is an alternative to TCS and used for ultrapure polysilicon. Nevertheless, silane is so reactive that results in significant homogeneous nucleated fines which lead to low yield and quality of crystalline silicon product. In this paper, we aimed to minimize the homogeneous nucleation by modulating flow pattern and temperature field in reactor. A novel bell-jar reactor with cooling jacket for each rod was modeled and simulated based on a three-dimensional CFD model. Contribution of homogeneous nucleation to rod growth on a basis of kinetic theory was included. The result showed that, in the novel reactor, the fresh feed was restricted in channels between cooling jacket and rod in which the favorable temperature and velocity profiles were achieved, and the gas phase near the rod was cooled so that the homogeneous nucleation was suppressed. In addition, the silane pyrolysis with simultaneous homogeneous nucleation and heterogeneous deposition was calculated and analyzed for various combinations of operating conditions.

•Amorphous silicon nano-powders are obtained by silane vapor decomposition.•Silicon powders transform to crystal ones prepared at 700 °C.•Higher sintering temperature and vacuum benefits crystallite size growth.•Higher temperature and longer duration strengthen powder aggregation.The effect of sintering treatment on the amorphous silicon powders from monosilane pyrolysis was investigated to enhance polysilicon yield in the FBR-CVD process. The crystallite size, hydrogen bond structure and morphology of the powders were characterized by X-ray Diffraction (XRD), Scanning Electron Microscopy (SEM), Diffuse Reflection Fourier Transform Infrared Spectroscopy (DR-FT-IR) and Zeta Potential Analyzer (ZPA). The results showed that the higher sintering temperature and lower pressure were more favorable to the growth of silicon crystallites and the liberation of hydrogen from silicon hydrides. The crystallite size increased significantly with a critical low FWHM when the sintering temperature was at 750 °C, and the hydrogen releasing from polysilanes and OSiH took place remarkably with a relative flat IR spectrum when the vacuum condition was implemented. Moreover, the particle aggregation was strengthened with broad particle size distribution as long sintering duration or high temperature applied, and the fusion occurred as the temperature was high enough but still with good crystallinity.

The Eulerian–Eulerian model with kinetic theory of granular flow was applied to evaluate the gas–solid flow behavior in a large-scale polydisperse fluidized system for polysilicon chemical vapor deposition growth. The uniform particle system was well validated on the basis of the empirical correlations and experimental data after parametric analysis on drag force, particle–particle collision, and wall boundary condition, and furthermore the fluidization quality in the polydisperse system by the validated computational fluid dynamics model was basically the same as that in the uniform one. The results showed that the energy minimization multiscale-based drag model had remarkable advantages in the prediction of the “core-annular” flow structure and mesoscale flow characteristics in a large-diameter fluidized bed, as compared to the Gidaspow model. Moreover, effects of particle aggregation and the fluidization ratio on solid distribution in the polydisperse system were also investigated, revealing an accumulation of small particles at the near-wall and upper regions and an increase of bubbles at high fluidization ratio, respectively.

Hydrogenation of silicon tetrachloride (STC) in the presence of silicon was investigated both thermodynamically and experimentally. The thermodynamic results revealed that the reaction is thermodynamically constrained and the STC equilibrium conversion is nearly temperature-independent but increases with increasing pressure and H2/STC feed ratio. Trichlorosilane (TCS) was calculated to be the dominant product, with trace amounts of dichlorosilane (DCS) and HCl as feasible byproducts. Experimental tests were conducted in a fixed-bed reactor loaded with silicon. The results indicated that the reaction is kinetically restricted in the absence of a catalyst. Adding CuCl to the silicon mass bed was found to result in a dramatic increase in the STC conversion, and Cu3Si was determined to be the active species. In all of the tests runs, TCS was determined to be the dominant product, and the STC reaction rate was found to increase markedly as the temperature and STC partial pressure increased but not to depend on the silicon particle size.

Methanol conversion to olefins (MTO) was investigated in an isotherm fixed-bed reactor at methanol partial pressures ranging from 5 to 50 kPa, water/methanol ratios from 0 to 9 and temperatures from 380 °C to 460 °C over H-ZSM-5 catalyst with a Si/Al ratio of 200. Product distribution was affected by operation conditions remarkably only at low methanol conversions (< 80%). When methanol conversion approaches 100%, product distribution is hardly affected by operation conditions. In addition, it was always found that propylene selectivity is enhanced at the expense of hexenes at high methanol conversions. The dependence of product distribution on methanol conversion derives from the change of dominant reaction pathway from olefin methylation and cracking to oligomerization and cracking. Product distribution at low methanol conversions may be determined by the relative rate between olefin methylation and cracking, while at complete methanol conversion product distribution is determined by oligomerization and cracking of C4=–C7= likely.Highlights► Reaction conditions affect product distribution in MTO evidently at low conversions. ► The main reaction pathway in MTO changes with increasing methanol conversion. ► C3= selectivity is increased at higher methanol conversions through cracking of C6=. ► C2= selectivity is increased only at high space timevia over-cracking of olefins.

This paper presents a reactive distillation column for the catalytic disproportionation of trichlorosilane to silane which includes three consecutive reversible reactions with a thermodynamic conversion to silane as low as 0.2% and is of no practical significance using the conventional reactors. This reaction system is however characterized by a large distinction in the boiling points of the components, which makes the reactive distillation extremely favored. Nevertheless, the normal reactive distillation column possesses the shortage of high refrigeration requirement as the standard boiling point of the overhead product silane is −112 °C. A novel reactive distillation column with intercondensers placed within has been simulated using Aspen Plus package for a feasible alternative to alleviate this drawback. The calculated results show that a reduction in the refrigeration load of more than 97% can be obtained when one intermediate condenser is inserted between the rectifying and reaction zones, and that another normal intermediate condenser equipped within the reaction section can further decrease the condensation requirement of the first intermediate condenser by 50%. Effects of the configuration parameters and operational conditions have also been investigated for optimal design and operation of the proposed reactor.

Dissociation of methyl nitrite is the first step during CO catalytic coupling to dimethyl oxalate followed by hydrogenation to ethyl glycol in a typical coal to liquid process. In this work, the first-principle calculations based on density functional theory were performed to explore the reaction mechanism for the non-catalytic dissociation of methyl nitrite in the gas phase and the catalytic dissociation of methyl nitrite on Pd(111) surface since palladium supported on alpha-alumina is the most effective catalyst for the coupling. For the non-catalytic case, the calculated results show that the CH3O–NO bond will break with a bond energy of 1.91 eV, and the produced CH3O radicals easily decompose to formaldehyde, while the further dissociation of formaldehyde in the gas phase is difficult due to the strong C–H bond. On the other hand, the catalytic dissociation of methyl nitrite on Pd(111) to the adsorbed CH3O and NO takes place with a small energy barrier of 0.03 eV. The calculated activation energies along the proposed reaction pathways indicate that (i) at low coverage, a successive dehydrogenation of the adsorbed CH3O to CO and H is favored while (ii) at high coverage, hydrogenation of CH3O to methanol and carbonylation of CH3O to methyl formate are more preferred. On the basis of the proposed reaction mechanism, two meaningful ways are proposed to suppress the dissociation of methyl nitrate during the CO catalytic coupling to dimethyl oxalate.Download full-size image

•CFD–PBM coupled model represents the silicon-CVD growth in a FBR successfully.•Scavenging dominated growth mechanism matches the experimental data well.•Disilane is the main silicon hydride in silane homogenous pyrolysis.•Defluidization appears during the silicon-CVD process.A Eulerian–Eulerian two-fluid model coupled with population balance equations was applied to simulate the evolution of silicon particle growth by chemical vapor deposition of silane pyrolysis in a three-dimensional slugging fluidized bed reactor using FLUENT. The simulation of the particle growth considering surface deposition, cluster scavenging, aggregation and wall deposition was carried out after the verification of flow and heat transfer characteristics based on the well-accepted correlations. The results showed that the scavenging effect was responsible for the particle growth, and the growth rate agreed well with the experimental data by Tejero-Ezpeleta et al. (2004) when the scavenging factor was set to 0.1 under the condition of 923 K and atmospheric pressure. Moreover, the formation of light silicon hydrides by silane homogeneous pyrolysis in the dilute phase was also investigated in the form of CHMEKIN mechanism, which showed that disilane turned to be the main silicon hydride and the silane conversion was underestimated by 12.5%. Finally, the effects of operating conditions on the growth rate were studied in detail with the observation of defluidization phenomenon during the evolution of particle growth.

The reaction pathway for propene formation in methanol to propene (MTP) process over a high silica H-ZSM-5 catalyst has been investigated in a fixed bed reactor by comparing the experimental results from three kinds of feeding: alkene only, methanol only and mixed alkene and methanol. The results show that alkene methylation with methanol is dominant for the case of methanol and individual C3–C6 alkenes co-feeding, C2= is almost un-reactive. C7= cracks to propene and butene immediately whether co-fed with methanol or not, and C6= cracks to propene readily when reacted alone. Oligomerization occurs but is suppressed by the co-fed methanol for light alkenes of C2–C5. Methylation-cracking has been verified as the main reaction mechanism of a typical MTP process in which recycling of C2= and C4=–C6= to the reactor inlet is required. Based on the relative reactivities of alkenes towards methylation and inter-conversion, a reaction scheme has been presented including a cycle composed of a consecutive methylation from C4= through C5= to C6= and further to C7=, the β-scission of hexene and heptene for propene, and the α-scission of hexene for ethene as well.Highlights► Methylation-cracking is the dominant reaction pathway in a typical MTP process. ► The rank of reactivities of C2-C6 alkenes towards methylation is C5=>C4=>C6=>C3=>C2=. ► Propene is formed mainly from cracking of C6=and C7=via methylation of C4=∼C6=. ► Ethene is formed mainly from C6=cracking.