<p>Mass transfer in a porous solid, partially saturated with asingle solvent and multicomponent liquid mixtures, has beenexperimentally and theoretically studied. A porous materialcontaining single liquids and mixtures of organic solvents wasisothermally dried. Experiments were performed using a jacketedwind tunnel, through which a humidity andtemperature-controlled air stream flowed. The wetted porousmaterial was placed in a cylindrical vessel, whose top isexposed to the air stream until the material became dried to acertain extent. Drying experiments with the single solventswater, methanol, ethanol and 2-propanol, were performed atdifferent temperatures and transient liquid content profileswere determined. In isothermal drying experiments with liquidmixtures,the transient concentration profiles of thecomponents along the cylindrical sample as well as the totalliquid content were determined. The liquid mixtures examinedwere water-methanolethanol and isopropanol-methanol-ethanol.Two different temperatures and initial compositions were usedin the experiments. Mathematical models that describe nonsteadystate isothermal drying of a solid containing single liquidsand multicomponent liquid mixtures were developed. In the solidwetted with a single liquid, capillary movement of the liquidwas the main mechanism responsible for mass transfer. In thesolid containing liquid mixtures, interactive diffusion inliquid phase was superimposed to the capillary movement of theliquid mixture. In addition, interactive diffusion of thevapours in empty pores was considered. The parameters todescribe the retention properties of the solid and thecapillary movement of the liquid were determined by comparingtheoretical and experimental liquid content profiles obtainedduring drying of the solid wetted with single liquids. Tosimulate the transport of the liquid mixtures these parameterswhere weighed according to liquid composition. A fairly goodagreement between theoretical and experimental liquidcomposition profiles was obtained if axial dispersion isincluded in the model when the moisture consists of amixture.</p><p><b>Keywords:</b>Internal mass transfer, capillary flow,multicomponent, diffusion, solvent mixtures</p>

Mass transfer in a porous solid, partially saturated with asingle solvent and multicomponent liquid mixtures, has beenexperimentally and theoretically studied. A porous materialcontaining single liquids and mixtures of organic solvents wasisothermally dried. Experiments were performed using a jacketedwind tunnel, through which a humidity andtemperature-controlled air stream flowed. The wetted porousmaterial was placed in a cylindrical vessel, whose top isexposed to the air stream until the material became dried to acertain extent. Drying experiments with the single solventswater, methanol, ethanol and 2-propanol, were performed atdifferent temperatures and transient liquid content profileswere determined. In isothermal drying experiments with liquidmixtures,the transient concentration profiles of thecomponents along the cylindrical sample as well as the totalliquid content were determined. The liquid mixtures examinedwere water-methanolethanol and isopropanol-methanol-ethanol.Two different temperatures and initial compositions were usedin the experiments. Mathematical models that describe nonsteadystate isothermal drying of a solid containing single liquidsand multicomponent liquid mixtures were developed. In the solidwetted with a single liquid, capillary movement of the liquidwas the main mechanism responsible for mass transfer. In thesolid containing liquid mixtures, interactive diffusion inliquid phase was superimposed to the capillary movement of theliquid mixture. In addition, interactive diffusion of thevapours in empty pores was considered. The parameters todescribe the retention properties of the solid and thecapillary movement of the liquid were determined by comparingtheoretical and experimental liquid content profiles obtainedduring drying of the solid wetted with single liquids. Tosimulate the transport of the liquid mixtures these parameterswhere weighed according to liquid composition. A fairly goodagreement between theoretical and experimental liquidcomposition profiles was obtained if axial dispersion isincluded in the model when the moisture consists of amixture. Keywords:Internal mass transfer, capillary flow,multicomponent, diffusion, solvent mixtures

Biofilms of the bovine rumen bacterium Actinobacillus succinogenes have demonstrated their exceptional capabilities as biocatalysts for high productivity, titre and yield production of succinic acid (SA). Succinic acid is set to become a significant building block chemical in the biobased economy. Although substantial progress has been made towards understanding the productive aspect of this microorganism with regard to its metabolic limits and performance on unrefined biorefinery stream substrates, more research is still required to address other challenges. One aspect is to understand how the biofilm biocatalyst is affected by bioreactor conditions, which would help in developing stable and highly active biofilms. For this reason the aim of this thesis was (i) to characterise how the accumulation of acid metabolites in continuous operation impacts A. succinogenes biofilms with respect to biofilm development, biofilm structure and cell activity within the biofilm, (ii) to show how shear conditions in the fermenter can be used to manipulate the biofilm structure and viable cell content of biofilms, leading to improved cell-based succinic acid productivities, and lastly (iii) to investigate the internal mass transfer effects on biofilm performance, further showing the role played by differences in shear and acid accumulation conditions in this respect. The first part of the study addressed the interaction between the biofilm and the accumulation of metabolites produced. The results showed that biofilms of A. succinogenes develop rapidly and with high activity when cultivated under low product accumulation (LPA) conditions (< 10 g L-1 SA). High product accumulation (HPA) conditions considerably slowed down biofilm development, and increased cell mortality. Under HPA conditions some cells exhibited severe elongation while maintaining a cross-sectional diameter like the rod/cocci-shaped cells predominantly found in LPA conditions. The elongated cells formed in HPA conditions were found to be more viable and thus more resistant than the clusters of rod-shaped or cocci-shaped cells. The global microscopic structure of the HPA biofilms also differed significantly from that of the LPA biofilms. Although both exhibited shedding after 4 days of growth, the LPA biofilms were more homogenous (less patchy), thicker and had high viability throughout the biofilm depth. In the second part of the study, two custom-designed bioreactors were used to evaluate the effect of shear on the biofilms. The first bioreactor allowed for in situ removal of small biofilm samples used for microscopic imaging. The second bioreactor allowed for complete removal of all biofilm and was used to analyse biofilm composition and productivity. Results clearly indicated that high shear biofilm cultivation in LPA conditions has beneficial morphological, viability and cell-based productivity characteristics. The smooth, low-porosity biofilms obtained under high shear and LPA conditions had an average cell viability of 79% (over a 3-day cultivation period) compared with the low shear value of 57%, also developed under LPA conditions. The EPS content of the high shear biofilm was 58% compared with 7% of the low shear equivalent. The cell-based (EPS excluded) succinic acid productivity for the high shear biofilm was 2.4 g g-1DCW h-1 compared with the 0.8 g g-1DCW h-1 for the low shear biofilm. This threefold increase in productivity obtained from the second bioreactor corresponded to the cell viability differences obtained from the first bioreactor. Clear evidence was provided for shear-induced shaping of the biofilm which resulted in improved volumetric glucose turnover attributes within the biofilm matrix. The last section of the study investigated internal mass transfer effects in biofilm fermentations of Actinobacillus succinogenes by performing batch fermentations using attached and resuspended biofilms as biocatalysts. In the latter, the biofilms were resuspended after initial development to simulate mass transfer-free fermentations. Intrinsic kinetics for succinic acid production obtained from resuspended fermentations predicted faster production rates than for the attached biofilm runs (biofilm thicknesses in the range of 120–200 µm), indicating internal mass transfer limitations. A developed biofilm reaction diffusion model gave good prediction of attached biofilm batch operation results by accounting for internal mass transfer in the biofilm. Biofilm effectiveness factors ranged from 75% to 97% for all batches at the inception of batch conditions, but increased with the progression of batch operation due to the increased succinic acid titres which inhibited the production rates. Analysis of pseudo-steady-state continuous fermentation data from the literature, as well as from the second part of the study, using the model developed, showed that active biofilm thickness and effectiveness factors were dependent on the shear conditions and succinic acid titres in the biofilm reactors. A simplified algorithm was developed to estimate the pseudo-steady-state glucose penetration and biofilm effectiveness of A. succinogenes biofilms without the requirement to solve the overall mass transfer model. The results clearly showed that internal mass transfer needs to be considered in biofilm fermentations involving A. succinogenes as high biomass concentrations may not always equate to increased productivities if mass transfer effects dominate. / Thesis (PhD)--University of Pretoria, 2020. / NRF / Chemical Engineering / PhD / Unrestricted