In this paper,by combining the characteristics of electrochemical reaction with that of the engineering aspects of the fluidized bed, the voltage fluctuation phenomena have been theoretically studied and a new mechanism has been put forward. The main point of the mechanism is that, for a given reaction system, the polarity state on the solid particles surface changes with the degree of fluidized bed expansion. For a certain fluidized bed electrochemical reactor there must be an optimal degree of bed expansion.at which the highest superficial current and space-time yield can be obtained. An experimental apparatus for studying the voltage fluctuations is presented with a side-by-side type reactor, silver particles and AgNO3 solution as the electrolyte. The voltage fluctuation values can be measured with a composite probe and an A/D converter as programmed by computer. The power spectra of voltage fluctuation can be calculated with Fast Fourier Transformation.

In this paper, the voltage fluctuation characteristics are experimentally studied for a silver deposition system. At the same time, the influences on voltage fluctuation of the degree of expansion of fluidized bed, of the voltage difference between collector and counter, and of the fluidization state are investigated. The conduction mechanism proposed in Part( I ) is corroborated experimentally.The experimental results show that there are two manners for the particle and electrolyte voltage fluctuations, one of high frequency but low amplitude and the other of low frequency but high amplitude. When slugging or channeling appears, the particle voltage and the electrolyte voltage fluctuate irregularly while the overvoltage has essentially no irregular fluctuation. The influence of the voltage difference between collector and counter on the voltage fluctuations is rather pronounced. When the voltage difference is lower than about 1V for the reactor used there are essentially not low frequency, high amplitude voltage fluctuations. When the voltage difference is greater than about 1V, the low frequency, high amplitude voltage fluctuations for both the particle and electrolyte become more and more remarkable.

In this paper, the voltage distributions and the probability distribution densities within the range of voltage fluctuations in a silver deposition system are investigated. The optimal degree of expansion of fluidized bed is obtained and some other important problems such as the influence of temperature on the silver deposition rate, the change of current efficiency with the concentration of Ag+ are also studied. The experimental results show that, under smooth fluidized state, all the particle voltage, electrolyte voltage and overvoltage follow the Gaussian probability distribution density within their fluctuation ranges, but when slugging or channeling appears such relationships no longer exist. The mean values of the voltage do not change in the x and z directions, but change in the y direction for the reactor used. Near the collector and the membrance the over-voltage is higher. The influence of temperature on the deposition rate of silver is rather small. The experimental results also show that the Ag+ in the electrolyte can be deposited rather completely, but when the electrolyte becomes very dilute the current efficiency decreases sharply.

The extents of axial gas mixing for turbulent fluidization, fast fluidization and dilute phase transport regimes in a circulating fluidized bed testing unit of 90mm inside diameter and 8m height were measured and characterized by dynamic response technique consisting of tracer pulsed input, dual probe detection and computer on-line analysis system. Experimental results indicated that the extents of axial gas mixing were related to the bed voidage mainly determined by both the gas velocity and the solids circulating rate. A one dimensional pseudo-homogeneous dispersion model was used with the least squares method in time domain to fit in with the experimental data. As a result, a generalized correlation of the gas dispersion coefficient is obtained:

The mechanism of interaction between gas and particles have been investigated with an one-dimensional gas-solids flow model for fast fluidized bed. It was found that the aggregation of particles (i. e. shielding effect) weakened the interaction between gas and solids, thus the drag coefficient of aggregated particles in fast fluidization was less than that of individual particle in an infinite field, e. g., CD/CDS<1.0. The value of CD/CDS not only decreased with decreasing bed voidage, but also is a function of gas-solids flow pattern, particle properties and bed diameter. The empirical correlation to predict the particle drag coefficient in fast fluidization regime is as follows:

The periodical operation behavior of a patented TPCIX column was investigated. The continuous ion exchange process was analysed through the concept of probability density and conversion distribution function. The numerical solution of a group of partial differential equations describing the operation characteristics of the pseudo steady state process was found with the leading moments method. The relation between moments of different orders was established by means of the Laguerre polynomial. The simulated results agreed with the experimental data.

In this paper, the multibead-rod molecular model is used to simulate the molecular structure of the dilute solution of sesbania gum. The flow of sesba-nia gum solution in a circular tube model with successive abrupt contractions and expansions in a ratio of four to one is studied by both numerical and experimental methods. The finite element method is used in numerical simulation. No special treatment is required in calculating the vortex flow. The calculated pressure drops and flow patterns by this method are in a good agreement with the experimental ones. In the experimental study, the pressure drop is found to be a quadratic function of flow rate for both the Newtonian and non-Newtonian fluids (glycerine solution and sesbania gum solution). The results show that such a molecular model can be used successfully to replace the constitutive equation.

A new generalized algorithm for chemical equilibrium calculation is presented. This method can effectively determine whether a pure solid phase exists and accurately evaluate the compositions of trace species with the extents of reaction and moles of components as the iterative variables. The efforts required for the inverse of Jacobian matrix are equivalent to the method where only the extents of reaction are iterative variables by a linear transformation. This approach begins with estimating the initial value with relaxation method and convergence can be accelerated by Newton-Raphson method. The singularity of Jacobian matrix can be overcome with the help ot damped Newton-Raphson method. The approach has the same robustness as Marquardt method used by Xiao et al. (1989). All the examples tested converge to the correct solutions from zero iterative variables. The carbonation of methane steam shift system and the chemical vapor deposition of Ti(C, N) are analy-sized. The computed results of NH3-CO2-SO2-H2S-CH4-C2H6 system are also presented.

A density dependent local composition version of M-H(81) equation of state(DDLC M-H(81)) tor mixtures has been developed to improve the application of M-H(81) equation of state to VLE calculation for systems containing polar and supercritical components. It was tested with 45 sets of low-pressure and high-pressure binary VLE data. When applied to low-pressure polar systems, the new model had the same accuracy for the vapor composition and better result for the bubble point pressure as compared with the common activity-coefficient models. When applied to high-pressure systems (including those containing supercritical components) the model showed significant improvement in data correlations as compared with the conventional quadratic mixing rule.

An apparatus which had an extremely small dead volume and could determine intrapellet gas effective diffusivities by chromatography rapidly and accurately at 250℃ was built. The statistical moment theory was introduced and equations describing the relations between the diffusivities and the experimental moment were derived. For a B110 catalyst pellet, (De)Ar-N2 and (De)He-N2 at 29℃ and 1.114 × 105Pa were 0.009344 and 0.02936 cm2/s respectively and its tortuosity factor was equal to 3.16. The results agree well with theoretical calculation. Both the experimental technique and the apparatus are reliable.

The equations relating the adsorption equilibrium constant and reactionrate constants on a single catalyst pellet to measureable properties of pulse-response experiments were derived by means of statistical moment theory. Under atmospheric pressure and at a temperature upto 340℃ the effective diffusivities and the adsorption equilibrium constants oi carbon monoxide and the rate constants for shift reaction CO + H2O = CO2 + H2 on a B110 catalyst pellet were measured by using ordinary thermal conductivity cell(TCC). It has been proved to be a new and valuable technique for determining the kinetic parameters of gas-solid heterogeneous catalytic reaction.

Solvent impregnated resins (SIR) as extractants evenly dispersed in a solid polymeric medium are used to separate spiramycin, a macrolide antibiotic. In this paper a particle-diffusion control mathematical model based on linear adsorption isotherm on the inner surface of SIR is developed and its analytical solution under the condition of finite fluid volume can be expressed as follows: where u(t) is referred to as equilibrium fraction of spiramycin, sn are the roots of the equation: snctg sn = 1 + sn2/(3ω), τ is dimensionless time, and ω is a characteristic constant depending on the properties of both SIR itself and the bulk solution of spiramycin. Freundlich adsorption isotherm is also used for the differential equation of diffusion in order to modify the linear type isotherm and its numerical solution is calculated by using finite difference scheme in steps of time and radius. Batch experimental results showed good agreement of the theoretical bulk spiramycin concentration history calculated from the analytical solution with the real values measured. The effective diffusion coefficients of spiramycin were estimated in the range of 3.7× 10-12-29.4 × 10-12 m/s for Amberlite XAD-4 resin impregnated with amyl acetate. The differences between analytical solutions with respect to linear isotherm and the relevant numerical solutions with respect to nonlinear isotherm were within the range of experimental errors indicating that the assumption of linear isotherm is a good approach to the practical process.

A DC arc plasma jet apparatus was established to prepare diamond film with good quality. High deposition rate of 65 μm/h was achieved on non-polished molybdenum substrate, and diamond films can be deposited in an area as large as 10-20mm in diameter. The diamond films were analyzed by Raman spectrum and SEM. Raman results showed only a peak at 1334cm-1, which fitted well with that of natural diamond. SEM results showed that diamond crystal grew well and the film was dense and uniform. Several forms of diamond crystal in our research were presented. Finally, a preliminary discussion on the formation of diamond film was made.

The kinetics of oxidative dehydrogenation of butene to butadiene over B-02 ferrite catalyst has been studied in an internal recirculation gradientless reactor. The experiments were conducted over a wide range of water to olefin ratio, and a large variation in the partial pressures of oxygen, butene, and butadiene. The kinetic models developed were based on a consecutive parallel reaction scheme: butene-butadiene carbon dioxide and were supported by the results obtained. The change in temperature had less effect on the ratio r8/r2. When the partial pressure of butadiene was high, the formation of carbon dioxide from butadiene was much greater than that from butene. A model with best fit was obtained as follows: The model obtained comprises both typical redox expressions and power rate expressions and is shown to be compatible with Mars-Van Krevelen redox mechanism. The model fitted the experimental data at high water/olefin ratio as well as those at very low water/olefin ratio. The effect of butadiene on the reaction rate was investigated through introducing additional butadiene to the fresh butene feed. It was observed that butadiene showed slight inhibition to r1, but hindered the combustion rate of butene to a much greater extent. It was also observed that the selectivity declined much at a high temperature. Low water/olefin ratio greatly enhanced carbon dioxide formation.

The impulse-response experiment was conducted to measure the mass transfer and adsorption coefficients and the rate constant of furfural hydrogena-tion reaction over catalyst BHK-1. The results showed that adsorption equilibrium constant, adsorption rate constant and surface reaction rate constants Ka, ka and Ki and interpartide diffusivity De did not change with the flow rate of the carrier and that adsorption heat, adsorption energy and surface reaction energy were 15.7, 29.4 and 43.8kJ/mol respectively.

In this paper, simulation calculation of liquid mixing in downcomer is carried out by using eddy diffusion model and mixing pool model. The mixing quantity among pools and the eddy diffusivity De expressing the mixing degree are obtained. The results indicate that mixing in downcomer is actually incomplete. It is a correction to the assumption of complete mixing in downcomer which is generally made in simulation calculations. The plate efficiency at various degrees of mixing in downcomer is calculated by the use of two-dimensional eddy diffusion model.

The catalytic activity of iron-molybdenum catalyst in ammoxidation of methanol to HCN in boiling bed microreactor has been investigated with XRD, IR, Mossbauer Spectrum, ESR and XPS. The studies of XRD and IR indicated that iron and molybdenum ions were present as Fe2(MoO4)3, MoOs and α-Fe2O3 in the catalyst, and that the change of the relative content of Fe2(MoO4)3 and catalyticac activity showed corresponding correlation in a peak shape with Fe/Mo ratio. Mossbauer Spectrum, ESR and XPS analysis showed that iron and molybdenum would be present as Fe3+ and Mo6+ in Fe2(MoO4)3,Fe2O3·xMoO3 and α-Fe2O3·Fe2 (MoO4)3 was in coordination of an octahedron, while Fe2O3·xMoO3(x<3) showed im perfect coordination.The former could be a basic active component, the latter was very likely the surface active center. All the three components synergistically performed the function of selective oxidation.

By treating salt as one kind of solvent, eighteen sets of vapor pressure data of salt-containing systems have been correlated with Wilson, NRTL and modified UNIQUAC models. The maximum likelihood method based on maximum liklihood principle was used in correlation. Satisfactory results were obtained for all the three models. The optimum parameters thus obtained were used to predict five sets of VLE data for multicomponent salt-containing systems reported in literature. The predicted results show that the precision with NRTL model is a little better than with the other models; and is comparable with "pseudo binary approach".