A new five-parameter corresponding states model has been developed for viscosity and proton spin-lattice relaxation rate applicable to a variety of hydrocarbons in the dense fluid phase region. This model is the key to the development of unprecedented, accurate correlations between the proton spin-lattice relaxation rate, R$\sb1$, and the viscosity, $\eta$, of pure hydrocarbons, with an observed root-mean-square deviations of less than 8.0%. For hydrocarbon mixtures, the notion of "effective" spin-lattice relaxation rate, introduced by Zega et al$\sp{1,2}$, is pursued. Then, simple mixing rules are successfully applied to extend the model to mixtures. The four-parameter corresponding states treatment, utilizing the shape factors as proposed by Leach and Leland$\sp3$, yields correlations of comparable accuracy, when confined to long-chain alkanes.

In vivo and ex vivo studies with whole vessels demonstrated that the endothelium comprises the major permeability barrier in the blood vessel wall. As a lining between the blood and the underlying tissue, the endothelium is exposed to biochemical components of the blood and wall shear stress. The endothelium responds to these biochemical and mechanical agonists both metabolically and morphologically. Using a model system consisting of human umbilical vein endothelial cells cultured on a permeable polycarbonate membrane, the permeability barrier function of endothelial monolayers was characterized quantitatively by determination of permeability coefficients. The effect of biochemical agonists and wall shear stress on permeability was assessed by the extent to which the agonists modulated the baseline permeability coefficient. Permeability increased nearly 10-fold over baseline values in the presence of 0.1 U/ml thrombin. Using altered forms of thrombin, it was shown that both proteolytic activity and binding to the endothelium are necessary for thrombin to increase permeability. Permeability was unaltered in the presence of 1 $\mu$M bradykinin. Permeability decreased to 25% of baseline during treatment with 6 nM iloprost, a stable functional analog of the vasodilator and anti-platelet aggregating agent prostacyclin. Prostacyclin often acts through the second messenger cAMP. The permeability also decreased in cells treated with combinations of cAMP analogs and phosphodiesterase inhibitors which act in concert to increase intracellular cAMP concentrations. Another anti-platelet aggregating agent and vasodilator, endothelial derived relaxing factor, acts by increasing intracellular cGMP concentrations. The cGMP analog dibutyryl cGMP also decreased endothelial permeability. Pretreatment of endothelial monolayers with iloprost, cAMP analog/phosphodiesterase inhibitor combinations or dibutyryl cGMP diminished the permeability increasing effect of thrombin. A parallel plate flow chamber was used to subject endothelial monolayers cultured on permeable membranes to steady, laminar flow under the condition of no volume flux across the membrane. After 1 hour of 7 dynes/cm$\sp2$ shear stress, permeability increased 7-fold in a partially reversible manner. When cells were pretreated with 6 nM iloprost, the effect of shear stress was diminished to a 2-fold increase which was completely reversible.

Escherichia coli has been widely used to produce high-value recombinant proteins for years. Although high recombinant protein productivity can be attained, perhaps the most important goal for such processes, that of achieving both high gene expression and high cell density simultaneously, is still challenging to both biochemists and biochemical engineers. A series of approaches to overcome this problem are evaluated in this study. First, a novel pH-inducible gene expression system, in which the expression of foreign gene products is directed by a pH down-shift, was chosen and characterized. This system was shown to have many attractive features, including high-level expression (40% of total cellular protein), fast response, and easy manipulation. It thus can serve as a proper model system for studying the fundamental mechanism of high-level gene expression in E. coli. Second, several factors limiting the culture performance were identified by systematically optimizing the culture conditions. Among those, acetate overproduction was shown to be critically involved. Various approaches on the basis of biochemical and genetic engineering techniques were successfully exploited to bypass such a cultivation bottleneck. Finally, the effects of various genetic elements, including the genes responsible for carbohydrate uptake and several stationary-phase genes, on recombinant protein production were investigated by genetic manipulation of the host strain. Several strategies were then developed to genetically construct more potent strains for recombinant protein production. The information is important not only for the modification of several structural models developed recently, but also for economic interest in terms of improving bioprocesses without further investment in equipment.

Coal pyrolysis and combustion have been systematically investigated at high temperatures where external and intraparticle transfer limitations become important. A thermogravimetric reactor equipped with in-situ video imaging capabilities provided the reaction rate measurements while its video microscopy system simultaneously allowed observation of the pyrolyzing or combusting coal particles. Video microscopy permitted direct observation of several transient phenomena occurring during combustion (particle ignition, macropore opening, particle fragmentation) or pyrolysis (particle swelling and bubbling) and these phenomena have been correlated with the combustion or devolatilization rate measurements. Particle ignition causes a sharp increase in the char combustion rates. The probability of particle ignition increases with increasing particle size, increasing porosity, increasing oxygen concentration and decreasing gas flow rate. Macropore opening also enhances char reactivity. During pyrolysis, the most vigorous bubbling of the particles occurred when the devolatilization rate was at its maximum. Pyrolysis conditions also affect char ignition behavior. Increasing pyrolysis heating rates result in chars with more open macropore structure and higher reactivity in the diffusion-limited regime. However, heating rates do not affect reactivity in the kinetic control regime. Chars pyrolyzed in 5% oxygen are more swollen and more porous than chars produced in pure nitrogen and are also more reactive in the diffusion-limited regime. Finally, increasing soak times and heat treatment temperatures deteriorate char reactivity in all regimes. A cellular automaton algorithm was developed to simulate combustion of chars with complex macropore structures in the diffusion-limited regime. This model accounts for diffusional limitations by assuming a finite penetration length of gas inside the porous solid and by treating the closed pores as inaccessible to the reactants. Computational grids were generated to model the structure of chars prepared at three different heating rates. Simulation results suggest that char reactivity depends strongly on macroporosity and macropore specific surface area. In agreement with our experimental reactivity measurements, the simulations show significant reactivity differences of the studied chars, even under isothermal conditions. The simulations do not detect significant particle fragmentation at conversions as high as 81%. Small fragments were produced though, at all conversions and their number reached a maximum at about 95% conversion.

A mathematical model is developed to predict oxygen transport from large capillaries. The model includes diffusion and convection in the plasma and rbc and the reaction in the rbcs. It also includes the radial distribution of red cells and flow velocities of red cells and plasma. The calculated oxygen saturations are in good agreement with the results from experiments in which artificial rubber capillaries are perfused with red blood cell suspensions. It is found that in the large capillaries most of the resistance to Oxygen transport lies in the plasma. The Nusselt number for mass transfer is determined as a function of various parameters. The fluxes from large capillaries under various conditions can be easily predicted from these Nusselt numbers. The resistance to oxygen transport is found to be greater with rbc suspensions than with equivalent Hb solutions. A mathematical model is developed for small capillaries. The effects of various parameters on oxygen transport are studied using this model. The physiological significance of these effects is discussed. The Nusselt number is calculated as a function of various parameters. It is found that a significant fraction of the transport resistance lies outside the tissue. The model is applied to certain physiological situations and is found to predict the observed behavior. The experimental methodology developed by Boland et al (13) is validated with the help of a well established mathematical model for hemoglobin solutions. The experimental system is characterized with the help of the model. The effects of physiological factors on oxygen transport from hemoglobin solutions is studied.

A microvisual flow cell was used to observe the effects of fluid viscosity ratio and of pore size distribution on the mechanisms of steady cocurrent two-phase flow in porous media. The transition in flow mechanisms was found to have the same dependence on capillary number, N$\sb{\rm ca}$, for all viscosity ratios provided N$\sb{\rm ca}$ was defined in terms of the interstitial velocity of the wetting phase. A simple theoretical model of ganglia flow through an idealized pore constriction was developed. Although velocity varied greatly during ganglion passage, the average volumetric flow rate of the nonwetting phase agreed well with relative permeability theory. A series of displacement experiments were performed with several oil-water-alcohol systems where diffusion occurred between oil and water phases. The amount of spontaneous emulsification observed was found to be greatest in systems where both Marangoni flow and local super saturation due to diffusion were expected.

The selective hydrogenation of 1,3-butadiene in the presence of n-butenes over supported palladium catalysts, an industrially important reaction, was investigated at atmospheric pressure between 308K and 393K. The study included measuring the reaction kinetics both in the presence and absence of CO, examining of the differences between palladium supported on six different carriers, testing for the effects of hydrogen spillover, and investigating the catalyst's aging behavior. Isotopic tracers (single $\sp{13}$C labeled 1-butene) provided direct information about the reaction network. The kinetics of butadiene hydrogenation were measured on a 0.02% Pd/alumina catalyst. Reaction orders were one in hydrogen and zero in butadiene. Addition of CO decreased the rate of butadiene hydrogenation; the reaction rate was proportional to the negative one-half power of the CO pressure in the range of 25 to 145 ppm CO. CO affects the hydrogenation rate by reducing the surface concentration of hydrogen. Activation energies for butadiene hydrogenation were 13.3 + 0.8 kcal/mole in the absence of CO and increased to 15.5 + 0.5 kcal/mole with the addition of 40 ppm CO. Six supports were tested to determine what effect the support might have on the catalytic properties of the supported palladium catalyst. The six materials tested were: activated carbon, alumina, barium carbonate, barium sulfate, silica gel, and titania. Alumina, silica, and titania all displayed approximately the same behavior; barium carbonate and sulfate supported catalysts exhibited significantly higher production of butane than those supported on refractory oxides. High palladium loading (all catalysts were 5 weight percent palladium) may have obscured real differences between alumina, silica, and titania. The presence of support activity induced by spiltover hydrogen was also investigated. Blending the supported Pd catalyst with additional support containing no Pd usually caused some increase in both hydrogenation and isomerization activity, but the results were not sufficiently reproducible to provide conclusive evidence for hydrogen spillover effects. The variability in the results, probably caused by the low reduction temperatures used. Through the use of $\sp{13}$C labeled 1-butene, n-butane was observed to come from the hydrogenation of both 1-butene and 2-butene during the selective hydrogenation of 1,3-butadiene in the presence on n-butenes.

A stirred microcarrier bioreactor contains a variety of local fluid dynamic environments, including isotropic turbulence, boundary layers, and vortices trailing from the impeller blades, all of which are superimposed on a larger-scale rotation and figure eight circulation. Analysis of the interaction of a cell-covered microcarrier with this complex environment suggest there are three potentially important mechanisms of cell damage. These are interaction with turbulent fluid eddies of about the size of the microcarrier bead, collisions with other beads, and collisions against the impeller. These mechanisms may be characterized by the eddy/bead size ratio, turbulent collision severity, and impeller collision severity respectively. The severities are defined as the kinetic energy of that type of collision times the frequency at which it occurs. Measurement of growth and death rates of bovine embryonic kidney cells showed that both a modified eddy/bead size ratio and turbulent collision severity based on a shear collision mechanism fit the results for moderate to high levels of agitation. Both parameters must be calculated on the basis of power dissipation around the impeller, rather than throughout the entire reactor. Experiments with decreased bead size and increased viscosity of the medium gave improved growth rates in agreement with these parameters' predictions. At low levels of agitation the net growth rate of cells decreased. This was attributed to the formation of cellular bridges joining two or more beads together in a stable clump. The lack of branched clumps and the observation of artifacts of bridges indicate that these bridges are frequently destroyed. The increased death of cells may be explained by bridge breaking and by the clump's having a larger equivalent diameter which is predicted to be more damaging by both parameters. Modeling of the growth of contact-inhibited cells on a sphere predicts a continually decreasing apparent growth rate. Use of an exponential model for early growth gives approximately the correct growth rate. This model also predicts the observed beneficial effect of increasing the inoculum cell density.

Transient two dimensional mass and energy balance models of a monolith reactor are formulated and solved using a combination of analytical and numerical algorithms. The reactor passage has three distinct phases: the void phase where the reactant gases flow, the porous catalytic layer where the reactants diffuse and react, and the nonporous solid substrate layer. The concentration and temperature profiles are computed using a Fourier series solution for the fluid equations and a finite difference method of lines algorithm for the solid equations. Inter- and intra-phase heat and mass transport are quantified in terms of effectiveness factors and Sherwood and Nusselt numbers. A parametric sensitivity analysis for the automobile exhaust converter is carried out for two passage geometries: cylinder and parallel plate. These two reactors are compared under similar conditions to evaluate relative converter performance in terms of light off time and final conversion and effect of radiation.

The isochoric PVT behavior of two binary fluid mixtures has been studied over extensive ranges of temperature and pressure. Emphasis in both experimental investigations was on definition of single-phase compressed liquid behavior, as indicated by isochoric curvatures and isochoric inflection loci, although PVT behavior in the virial region was also determined with comparable accuracy. In the first study, a mixture of 4.65 mole % hydrogen and 95.35 mole % methane was studied at temperatures from 140 to 273.15 K, densities from 2.2 to 23.6 moles/liter, and pressures to 700 bar. Densities of the nine isochores were determined by coupling to a reference isotherm at 273.15 K, where isothermal Burnett experiments were conducted to determine the dependence of density on pressure. Results of this hydrogen-methane study showed that the mixture isochoric inflection locus is "open". This is surprising, as it qualitatively resembles the inflection locus of hydrogen more than that of methane, even though hydrogen is the minority constituent of the mixture. This has significant implications for prediction of hydrogen-methane mixture properties by corresponding states and equation of state methods. Conventional mixing rules cannot predict the quantitative and qualitative dominance of hydrogen at such low concentrations. The second isochoric study consisted of measurements on a mixture of 20.19 mole % methane and 79.81 mole % methanol. Thirteen isochores were measured at temperatures from 350 to 550 K, densities from 1.7 to 21.8 moles/liter, and pressures to 700 bar. A reference isotherm at 510 K was used to determine isochore densities. A gravimetric technique was used to characterize this isotherm. Methanol decomposition at the high temperatures of this study was modeled as a pseudohomogeneous zeroth-order reaction. Each isochoric data point was density-corrected using the model parameters. Results of the methane-methanol investigation show the mixture isochoric inflection locus is "open", like pure methanol. The exact shape of the locus could not be determined because points on the locus were outside the operating range of the apparatus and/or methanol's stability region. Strong positive curvature near the bubble point line at high densities indicates the "open" nature of the locus.