A properties model was developed for use in commercial process simulators to model pyrolysis of lignocellulosic biomass. The component list was chosen to enable process simulations based on a recently published lumped pyrolysis kinetics model. Since many of the compounds involved in pyrolysis are not found in simulator databanks, estimation based on available literature data was used to establish missing parameters. Standard solid enthalpy of formation, solid heat capacity, and solid density estimates calculated from the limited experimental data available were prepared for six biomass constituents and nine intermediate and end-products of their pyrolysis. Ideal gas enthalpy of formation and heat capacity, critical property, and vapor pressure estimates were prepared for another four pyrolysis end-products and one biomass component. The estimates were all validated against the closest available experimental data in the literature. The addition of these new components and properties allows thermodynamically rigorous simulation of lumped biomass pyrolysis reactions with accurate energy balances. Enthalpies of reaction calculated from the properties model were compared with reported reaction enthalpies for the same lumped biomass pyrolysis reactions and found to be in general agreement.

No abstract

No abstract

Metal-organic frameworks (MOFs) are a class of crystalline and porous adsorbents, with wide-ranging applications in gas separations, membrane materials as well as sensors. Commonly used batch synthesis techniques for MOF production are limited by low productivity, high operating costs, and slow crystallization timescales, severely impeding the large-scale manufacturing of these materials. However, batch synthesis is a useful and easy technique to screen multiple reaction parameters to find an optimal chemistry. Therefore, in this study, we have used the batch process and screened a multidimensional reaction space consisting of 45 sample variations based on the crystallinity, yield and instantaneous precipitation, which could lead to tube clogging under flow conditions. We have found one optimized reaction chemistry, that could be used in flow conditions, which in this study is a novel millifluidic droplet-based reactor for the continuous synthesis of HKUST-1 crystals. The biphasic flow in the millifluidic reactor consisted of droplets of the reactant solution, dispersed in a continuous phase of silicone oil. We investigate the differences in the quality and quantity of HKUST-1 synthesized via the continuous and batch techniques. Moreover, we have demonstrated that the HKUST-1 samples prepared via the continuous synthesis in a droplet based millifluidic reactor, at an ultra-low residence time exhibit excellent physical properties comparable to that obtained for the samples prepared by the traditional batch process. A clean, easy-to-install, and reusable millifluidic reactor presented in this work may pave the path for an economically viable, large-scale synthesis of HKUST-1.

Adsorption-based separation techniques are significantly energy efficient in comparison to the conventional thermal separation techniques such as distillation. Despite the extensive research and development activities undertaken for mixed gas adsorption, the use of adsorption techniques for the separation of multicomponent liquid mixtures is still limited. This is due to the lack of accurate adsorption thermodynamic models, which form the scientific foundation of process simulation of such systems, making the translation to the industrial scale challenging. In this work, we have rigorously computed the surface excess of adsorption for six binary liquid mixtures on silica gel at 303 K using the frameworks of the adsorbed solution theory and the generalized Langmuir isotherm model. The six binary liquid mixtures studied in this work were formed by the pair-wise combinations of four components: benzene, 1,2-dichloroethane, cyclohexane, and n-heptane. We have based our calculations by considering simultaneous equilibria of three phases: saturated binary vapor phase, binary liquid phase, and the adsorbed phase. The composition of the corresponding saturated vapor phase was determined by correlating the experimental vapor-liquid equilibria data using the Non-Random Two-Liquid activity coefficient model. The activity coefficients of the adsorbed phase were calculated using the adsorption Non-Random Two-Liquid activity coefficient model. Devoid of simplifying assumptions, our methodology for computing the surface excess of binary liquid adsorption should be applicable for the adsorption from a wide variety of liquid mixtures.