Thesis (MScEng (Process Engineering))--University of Stellenbosch, 2006. / This work focuses on the effects of hydrogen peroxide concentration on the catalytic activity and product selectivity in the liquid-phase hydroxylation of phenol over titanium-substituted zeolites Al-free Ti-Beta and TS-1 in water and methanol solvents. Hydroquinone is typically the desired product, and these solvents employed have previously been shown to be of importance in controlling the selectivity of this reaction. Different volumetric quantities of an aqueous 30 wt-% peroxide solution were added to either water or methanol solutions containing the catalyst and phenol substrate, and the reaction monitored by withdrawing samples over a period of 6-8 hours. For Al-free Ti-Beta catalysed reactions, the peroxide concentration affects the selectivity and activity differently in water and methanol solvents. Using methanol solvent, the selectivity to hydroquinone formation is dominant for all peroxide concentrations (p/o-ratio > 1), and favoured by higher initial peroxide concentrations (> 1.27 vol-%), where p/o-ratios of up to can be reached; in water solvent, increasing the peroxide concentration above this level results in almost unchanging selectivity (p/o-ratio of ca. 0.35). For lower peroxide concentrations in water, the p/o-ratio increases slightly, but never exceeds the statistical distribution of ca. 0.5. Using water as a solvent, higher phenol conversion is obtained as the initial peroxide concentration increases; in methanol the phenol conversion is largely independent of peroxide concentration. As expected for the smaller pore TS-1, higher hydroquinone selectivity is obtained in methanol than for Al-free Ti-Beta, which is consistent with shape-selectivity effects enhanced by the use of this protic solvent. Interestingly, with TS-1 the p/o-ratio is higher at lower phenol conversions, and specifically when the initial peroxide concentration is low (p/o-ratio exceeding 3 were obtained at low phenol conversion), and decreases to a near constant value at higher conversions regardless of the starting peroxide concentration. Thus, low peroxide concentrations favour hydroquinone formation when TS-1 is used as the catalyst. Comparing the performance of the two catalysts using methanol solvent, the phenol conversion on TS-1 is more significantly influenced by higher hydrogen peroxide concentrations than Al-free Ti-Beta. However, with higher initial concentrations the unselective phenol conversion to tars is more severe since the hydroquinone selectivity is not higher at these high peroxide concentrations. The increased tar formation, expressed as tar deposition on the catalyst or as the tar formation rate constant, confirms that the greater amount of free-peroxide present is mainly responsible for the non-selective conversion of phenol. Kinetic modelling of the reaction data with an overall second-order kinetic model gave a good fit in both solvents, and the phenol rate constant is independent of changing hydrogen peroxide concentration for the hydroxylation over Al-free Ti-Beta using water as the solvent (kPhenol = 1.93 x 10-9 dm3/mmol.m2.s). This constant value suggests that the model developed to represent the experimental data is accurate. For TS-1 in methanol solvent the rate constant is also independent of peroxide concentration (kPhenol = 1.36 x 10-8 dm3/mmol.m2.s). The effect of the method of peroxide addition was also investigated by adding discrete amounts over a period of 4.5 hours, and was seen to improve hydroquinone selectivity for reaction on both catalysts, and most significantly for Al-free Ti-Beta in methanol solvent. With TS-1, the mode of peroxide addition had little influence on phenol conversion, but the initial selectivity to hydroquinone was ca. 1.6 times higher than for an equivalent single-portion addition (at a similar phenol conversion). Discrete peroxide addition for hydroxylation in methanol over Al-free Ti-Beta gave greatly improved hydroquinone selectivities compared to the equivalent single-dose addition. Compared to TS-1, the initial selectivity was not as high (p/o-ratios of 0.86 and 1.40 respectively at 10 mol-% phenol conversion), but this can be explained on the basis of geometric limitations in the micropores of TS-1 favouring hydroquinone formation. The final selectivity, however, is marginally higher (using the same mode of peroxide addition, and at the same phenol conversion). Discrete peroxide addition has an additional benefit in that it also reduces the quantity of free-peroxide available for product over-oxidation, and consequently reduces the amount of tars formed. Thus, the interaction of the effects of peroxide concentration and the solvent composition and polarity on the product selectivity and degree of tar formation is important. Particularly with TS-1, lower peroxide concentrations in bulk methanol solvent are highly beneficial for hydroquinone formation, because of the implicit geometric constraints in the micropores, the lower water concentration, and the decreased tar formation associated with high methanol concentrations. This could have significant reactor design implications, as the results obtained here suggest that the reaction should be terminated after approximately 30 minutes to maximise hydroquinone production (under the conditions evaluated in these experiments), even though the corresponding phenol conversions are low (ca. 10 mol-%). The higher hydroquinone selectivities reached at low phenol conversions for the discrete peroxide addition experiments also confirm this. Practically, to enhance the hydroquinone selectivity for reaction over TS-1, the initial phenol-peroxide molar ratio should be ca. 10, methanol should constitute not less than 90 vol-% of the reaction volume, and the peroxide should be added in discrete amounts. For reaction over Al-free Ti-Beta, methanol solvent also enhances the hydroquinone formation as expected. At low phenol conversions (ca. 10 mol-%) hydroquinone is still the preferred product, although in contrast to TS-1 the selectivity increases with phenol conversion, and is higher with higher initial peroxide concentrations. Under the best conditions evaluated here for optimal hydroquinone formation, the initial phenol-peroxide molar ratio should be ca. 2.5, with methanol making up at least 90 vol-% of the total volume. Discrete peroxide addition in methanol solvent for the Al-free Ti-Beta catalysed hydroxylation gives excellent improvements in hydroquinone selectivity (2.5 times higher than water solvent), and the addition in more discrete portions might further improve hydroquinone formation, and should therefore be examined.

Thesis (MScEng (Process Engineering))--University of Stellenbosch, 2006. / Filamentous fungi have been employed for production of heterologous proteins such as enzymes, antibiotics and vaccines due to their good secretion capacities and effective posttranslational modifications of these proteins. With an improvent in recombinant DNA technologies it has become possible to express many useful proteins in species such as the Aspergilli. However the submerged cultivation of filamentous fungi is complicated by the difficulties in mixing and oxygen and nutrient transfer in the highly viscous culture fluids that result. The purpose of the project was to investigate the potential of simultaneous control of morphology and production of enzymes in the dimorphic fungus, Mucor circinelloides, in order to overcome problems associated with the submerged cultivation of filamentous fungi. Dimorphic M. circinelloides, a zygomycete in the order Mucorales, occurs in a filamentous form or a yeast-like morphology in response to environmental conditions. Recently, advances were made in transformation of Mucor, and it has become possible to transform M. circinelloides to express heterologous proteins. The first example of a strong, regulated promoter from M. circinelloides being used for recombinant protein production was the expression of the glucose oxidase gene (from Aspergillus niger) under the control of the glyceradehyde-3-phosphate dehydrogenase (gpd1) promoter. Glucose oxidase (GOX) is an enzyme used to prevent oxidation of foods to extend shelf-life, to produce low-kilojoule beverages and to measure glucose levels in medical diagnostic applications. The scope of this project was to establish the conditions for yeast and filamentous growth of M. circinelloides in order to allow control of morphology, and to evaluate enzyme production under these conditions. Enzyme production of the GOX producing mutant strain, that was recently constructed, was compared to that of a wild type M.circinelloides strain. M. circinelloides was cultured in two-stage batch fermentations, firstly a yeast stage and then a filamentous stage. The yeast morphology was induced by anaerobic conditions while the filamentous morphology was achieved by exposure to air. The enzyme, biomass and metabolite production of the glucose-oxidase producing mutant strain and the wild type were monitored during the two-stage fermentations. GOX from the mutant and native amylase activity levels from the wild type were compared to each other and to other production systems for these enzymes. The morphology could be maintained in a yeast form under N2 with addition of ergosterol and Tween 80. The GOX activity levels in the culture fluid were comparable to some of the unoptimized GOX production systems in literature, but much lower than the optimized, recombinant GOX production systems that employ certain yeasts, or Aspergilli or Penicillium. The intracellular GOX levels were almost 6-fold higher than the extracellular levels which was unexpected as GOX is usually well-secreted. The morphological control improved the morphology for the initial yeast-stage of the fermentation but did not improve the morphology during the filamentous, enzyme-producing stage and it decreased the biomass yield and enzyme production by 50%. The constraint of Mucor to its yeast-like form did not improve the broth homogeneity or enzyme production and increased the time required for enzyme production. In this study M. circinelloides did not perform that well against other species already used to produce these enzymes. However, M. circinelloides could be used to produce enzymes from zygomycetes that systems such as A. niger do not produce well.

Thesis (MScEng (Process Engineering))--University of Stellenbosch, 2001. / Current methods for wax fractionation result in products with large polydispersity, and due to the high temperatures required, thermal degradation of the wax is often incurred. The need for an alternative process thus exists. The purpose of this project is to investigate the technical viability of supercritical fluid processing as an alternative wax fractionation technology. The main aims of this project are to select a suitable supercritical solvent, to conduct binary phase equilibrium experiments, to determine if the process is technically viable and to investigate the ability of various equations of state to correlate the phase equilibrium data. Based on limited data from the literature, propane and a propane rich LPG (Liquefied Petroleum Gas) were selected as suitable solvents. Literature data for propane and high molecular weight alkanes is scares and incomplete, thus necessitating experimental measurements. A phase equilibrium cell was designed, constructed and commissioned. The cell was designed for pressures up to 500 bar and temperatures to 200 oC, and with the aid of an endoscope, the phase transitions were detected visually. The measurements correspond well to literature values from reliable research groups. Phase equilibrium data sets for propane with nC32, nC36, nC38, nC40, nC44, nC46, nC54 and nC60 as well as LP Gas with nC36 were measured. At temperatures just above the melting point of the alkanes, the phase transition pressures can be considered to be moderate, which will positively impact the economics of the process. The phase transition pressure increases with increasing carbon number, the relationship being found to be linear when the pressure is plotted as a function of carbon number at constant mass fractions and temperature. The increase in phase transition pressure with increasing carbon number indicates that the solvent will be able to selectively fractionate the wax. At higher temperatures the gradient of the line is larger and may thus lead to improved selectivity. The higher temperatures will also lead to better mass transfer. The linear relationship indicates that limited extrapolation to higher carbon numbers may be possible. However, this needs to be verified experimentally. The inability to measure the critical point and vapour pressure curves of the higher molecular weight normal alkanes, as well as the inability of cubic equations of state to predict liquid volumes and to capture the chain specific effects such as internal rotations, results in cubic equations of state requiring large interaction parameters to fit the data. The alternative, statistical mechanical equations of state, have difficulty in predicting the critical point of the solvent correctly and thus overpredicts the mixture critical point, yet require smaller interaction parameters to fit the data. Further work is required to improve the predictability of these non-cubic equations of state. This project has proven that wax fractionation by supercritical extraction with propane is technically feasible.

Thesis (MScEng (Process Engineering))--University of Stellenbosch, 2009. / Aromatic precursor compounds are derivatives that play an important role in biosystems and are useful in the production of fine chemicals. This work focuses on the catalytic synthesis of 2-methyl-1, 4-naphthoquinone and cresols (para- and ortho) using aqueous hydrogen peroxide as an oxidant in liquidphase oxidation of 2-methylnaphthalene and toluene over titanium-substituted zeolite TS-1 or Ti-MCM-41. Catalysts synthesised in this work were calcined at 550°C, extensively characterised using techniques such as X-ray Fluorescence for determining the catalyst chemical composition; BET for surface area, pore size and micropore volume; Powder X-ray diffraction for determining their crystallinity and phase purity and SEM was used to investigate the catalyst morphologies. The BET surface areas for Ti-MCM-41 showed a surface area of 1025 m2/g, and a 0.575 cm3/g micropore volume. However, zeolite TS-1 showed a BET surface area of 439 m2/g and a 0.174 cm3/g micropore volume. The initial experiments on 2-methylnaphthalene hydroxylation were performed using the normal batch method. After a series of batch runs, without any success as no products were generated as confirmed by GC, a second experimental tool was proposed. This technique made use of the reflux system at reaction conditions similar to that of the batch system. After performing several experimental runs and optimising the system to various reactor operating conditions and without any products formed, the thought of continuing using the reflux was put on hold. Due to this, a third procedure was brought into perspective. This process made use of PTFE lined Parr autoclave. The reactor operating conditions were changed in order to suit the specifications and requirements of the autoclave. This process yielded promising results and the formation of 2-MNQ was realised. There was a drawback when using an autoclave as only one data point was obtained, at the end of each run. Therefore, it was not possible to investigate reaction kinetics in terms of time. Addition of aqueous hydrogen peroxide (30 wt-%) solution in the feed was done in one lot at the beginning of each reaction in all oxidation reactions, to a reactor containing 2-methylnaphthalene and the catalyst in an appropriate solvent of choice (methanol, acetonitrile, 2-propanol, 1-propanol, 1-pentanol, and butanol), with sample withdrawal done over a period of 6 hours (excluding catalytic experiments done with a Parr autoclave as sampling was impossible). As expected, 2-methylnaphthalene oxidation reactions with medium pore zeolite TS-1 yielded no formation of 2-methyl-1, 4-naphthoquinone using various types of solvents, with a batch reactor, reflux system, or a Parr PTFE autoclave. This was attributed to the fact that 2-methylnaphthalene is a large compound and hinders diffusion into zeolite channels. With the use of an autoclave, Ti-MCM-41 catalysed reactions showed that the choice of a solvent and reaction temperature strongly affect 2- methylnaphthalene conversion and product selectivity. This was proven after comparing a series of different solvents (such as methanol, isopropanol, npropanol, isobutanol, n-pentanol and acetonitrile) at different temperatures. Only reactions using acetonitrile as a solvent showed 2-MNQ. Formation of 2- MNQ, indicating that acetonitrile is an appropriate choice of solvent for this system. The highest 2-methylnaphthalene conversion (92%) was achieved at 120 ˚C, with a relative product selectivity of 51.4 %. Temperature showed a major effect on 2-MN conversion as at lower reaction temperature 100˚C, the relative product selectivity (72%) seems to enhance; however, the drawback is the fact that lower 2-methylnaphthalene conversions (18%) are attained. Another important point to note is the fact that using an autoclave (with acetonitrile as a solvent), 2-methyl-1-naphthol was generated as a co-product. In conclusion, it has been shown that the hydroxylation of different aromatic compounds over zeolites conducted in this study generated interesting findings. In 2-MN hydroxylation over Ti-MCM-41 as a catalyst, only acetonitrile is an appropriate choice of solvent using an autoclave. In addition, zeolite TS-1 is not a suitable catalyst for 2-MN hydroxylation reactions. It is ideal to optimise an autoclave in order to investigate reaction kinetics and optimum selectivity. Toluene hydroxylation reactions yielded para and ortho-cresol as expected with either water or acetonitrile as a solvent. No meta-cresol was formed. The kinetic model fitted generated a good fit with water as a solvent or excess toluene, with acetonitrile as a solvent generating a reasonable fit.

Thesis (MScEng (Process Engineering))--University of Stellenbosch, 1997. / This study focuses on the preparation and characterization of palladium-silver-nickel composite membranes. Electroless plating was used to deposit thin metal films on aluminazirconia membrane supports. Palladium conversion, in the electroless palladium plating process, was optimized with the aim of minimizing expensive palladium losses. The effect of deposition order on alloy composition and heat treatment on structural characteristics of the composite membrane was investigated. The inorganic support membranes were thoroughly cleaned and pretreated prior to plating to catalyze the surface. Factorial designs were used to obtain the maximum palladium conversion. Tetra amine palladium nitrate gave better solution stability and resulted in higher conversions than tetra amine palladium chloride. Buffer pH values of 9 to 11 caused little variation in palladium conversion. Moving outside this range resulted in a sharp decline in palladium conversion. At a pH of 9 to 11 the stabilizer is in the correct ionic form (EDTA3 and EDTA4,) to best stabilize the palladium ions, and hydrazine acts as a proper reducing agent. Significant interactions existed between the EDTA concentration (stabilizer) and hydrazine concentration (reducing agent) and between EDTA and temperature. The EDTA concentration was the most sensitive variable. A 27.5 g 10% tetra amine palladium nitrate solution per liter plating solution was used. Conversions exceeding 80% were obtained after three hours plating with 20 ml plating solution at the following conditions: temperatures from 71 to 77 DC, 40-60% molar excess hydrazine, EDTA:Pd-salt molar ratios between 30:1 and 40: 1 and buffer pH = 11. Silver plating rates for two hours plating of up to 2 mg/cm2 were obtained using a dilute silver nitrate solution with hydrazine as reducing agent. Electroless nickel plating was performed in a low temperature bath (40 DC) with nickel sulphate as source of metal ions and sodium hypophosphite as reducing agent. Metal films were fully characterized before and after heat treatment for 5 hours in a hydrogen atmosphere at 650 dc. Scanning electron microscopy (SEM) was used to analyze the surface structure. X-ray diffraction (XRD) patterns were taken to examine alloying and detect changes in the crystal structure .after heating. Energy dispersive X-ray maps (EDS) were used to visualize the diffusion process and particle induced X-ray emission (PIXE) was used to construct concentration profiles across the thickness of the metal films. Palladium deposits were dense, but columnlike, with a purity of 99.75%. Silver deposits were non-homogeneous, in other words it did not cover the entire substrate. The purity of the silver films was 99.5%. The nickel films were about 97% pure, very dense and defect free. When silver was deposited on palladium, the alloy penetrated more than 3 microns into the support and the palladium and silver concentrations varied across the thickness of the film after heating. By depositing palladium on silver, there was very little penetration into the support membrane pores (about 1 micron) and the palladium to silver ratio remained constant across the thickness of the film after heating. Silver-palladium-nickel alloy films call be prepared by first depositing silver, then palladium and finally nickel. During heat treatment, a counter diffusion process took place and the smaller nickel atoms penetrated into pores and defects that might be present in the palladium-silver solid solution. By adding more than 3% nickel, dense defect free films can be prepared, which is much thinner than conventional palladium-silver films. This method makes it possible to reduce the film thickness of dense, non-porous films to less than 5 microns, reducing fabrication cost and increasing the hydrogen flux through the film. Dense, non-porous palladium-silver films are usually in the range of 10-15 microns.

Thesis (MScEng (Process Engineering))--University of Stellenbosch, 2007. / The synthesis of mixed metal oxides, specifically the need and ability to successfully and accurately control the particle size, their stability and the reactivity of these nanoparticles is required, so as to allow the attachment of catalyst nanoparticle to the surface of a substrate or to other particles without leading to coalescence of the catalyst particle and hence to loss of their size induced properties. However, the synthesis of mixed metal oxides is a complex problem. Though various methods of preparing these types of oxides have been reported and applied, such methods they rarely produced pure forms and have often been recorded as having been contaminated with other phases. Often the particle sizes are too large in the micrometer range, and the size distribution is overly wide. Moreover, even if particles of nanometer size are formed, they tend to aggregate or agglomerate. In the current research, microemulsions were used to synthesize the nanoparticles. Such microemulsion consists of water droplets encapsulated by surfactant molecules in a pool of oil, comprising: water in oil (w/o) or reverse micelles. Reverse micelles in the nanometer size range are thermodynamically stable and optically transparent in the solution. They are believed to be highly dynamic structures whose components rearrange themselves over time and space through interaction or collision, coalescing and redispersing. However, the advantage of this method over using the standard method is that the particle size can largely be controlled, and a narrow size distribution obtained. The aim of the research was to investigate the feasibility of using the reverse micelle technique for the synthesis of mixed metal oxides - specifically α-bismuth molybdate (α- Bi2Mo3O12) with a controlled and desirable particle size and a narrow size distribution...

Thesis (MScEng (Process Engineering))--University of Stellenbosch, 2005. / The thesis reports the investigation of various distillery processes and wastewater streams. The aim was to evaluate the processes and thereafter design interventions for improved wastewater treatment at the respective distilleries. An integrated environmental approach was adopted based on the principle that prevention of pollution is the preferred option and end-of-pipe treatment the least favoured option. As such, feed material to the processes was studied to determine whether some of the components that are not required in the distillation process could in fact be removed prior to entering the system. The results indicate that organic constituents such as phenol and tartaric acid could be removed using physico-chemical and biological treatment methods. The treatment of effluent was studied using an Upflow Anaerobic Sludge Blanket (UASB) set-up to determine the reduction in Chemical Oxygen Demand (COD) in the wastewater. Thereafter the UASB treated effluent was exposed to aeration for further treatment. Summary of conclusions • Pretreatment of wine feed material with calcium hydroxide is effective in removal of 98% tartaric acid, 30% COD and a total phenol content of 57%. • Bio-augmentation results showed that the soil inoculum was the most effective treatment method with reductions of 61% COD at a temperature of 30°C, tartaric acid removal of 98% at the same temperature and 25% reduction in total phenol at 26°C. • UASB was effective with soil inoculum and removed approximately 90% of COD although operational problems were experienced and hindered the operation of the plant. • Aeration of UASB effluent further reduced the COD by a further 60% with a total COD reduction of 96% after both UASB and aeration treatment. • Effective reduction of total phosphorus by 70% and the total phenol content by 80% was achieved by UASB treatment followed by aeration.

Thesis (MScEng (Process Engineering))--Stellenbosch University, 2008. / With the harvesting of fish and other aquatic organisms from natural waters having reached its upper limit, aquaculture is vital in providing for the ever increasing demand for fishery products (Boyd, 1999). Not surprisingly, aquaculture has seen considerable growth over the last decade or more. With the rising importance of aquaculture, there is an increased emphasis on cost and reducing of waste for environmental reasons. Therefore, attempts to automate or increase efficiency of feeding are constantly being explored. On an aquaculture unit approximately 60% of all costs are for feed; therefore high quality feeding management is essential for all fish farmers. The rainbow trout farm at Jonkershoek Aquaculture Research farm near Stellenbosch currently have a feeding management system which makes use of traditional hand feeding. Handfeeding is not considered optimal, as the feed intake or pellet loss is not closely monitored resulting in higher operating costs. Automation of aquaculture systems will allow the industry to produce closer to markets, improve environmental control, reduce catastrophic losses, minimize environmental regulation by reducing effluents, reduce production costs and improve product quality. The history of automated control in aquaculture has been brief; most of the systems have been custom-designed, personal computer systems. A very popular approach for an automated feeding system is to monitor waste pellets beneath the feeding zone of the fish, with a feedback loop that can switch off the feeder if this waste exceeds a predetermined threshold. Other approaches use hydroacoustics to monitor waste pellets or demand feeders have also been implemented. These approaches however are not considered optimal as automatic feeders do not necessarily ensure optimal feed intake. Social dominance using demand feeders does not allow even feeding distribution among all sizes of fish. In this project it was investigated whether an automated feeding system can be developed based on fish feeding behaviour. After facing problems with poor visibility at the Jonkershoek Aquaculture farm near Stellenbosch, video data were acquired from the Two Oceans Aquarium in Cape Town. Since it was a feasibility study, the focus was rather to investigate whether a predictive model could be generated for fish feeding behaviour in a more ideal environment which can form a foundation for further research. The well-established multivariate methods of principal components

Thesis (MScEng (Process Engineering))--University of Stellenbosch, 2007. / Adequate provision of oxygen to aerobic bioprocesses is essential for the optimisation of process kinetics. In bioprocesses in which the feedstock is an alkane, the supply of sufficient oxygen is of particular concern because the alkane molecular structure is deficient in oxygen. As a result, the oxygen demand has to be met solely by transfer of oxygen to the culture, necessitating a proportionately higher requirement for oxygen transfer. Maximisation of the rate of oxygen transfer is therefore of key importance in optimising the potential for alkane bioconversion, with respect to both operation and scale up. Nevertheless, the oxygen transfer rate (OTR), and its dependence on the overall volumetric mass transfer coefficient (KLa) in alkane-aqueous dispersions is not yet well understood. In view of the importance of an adequate OTR in the optimisation of alkane bioconversion, this study has focused on the identification and elucidation of the factors which underpin the behaviour of KLa in an alkane-aqueous dispersion. KLa behaviour was quantified in terms of the pressures imposed by turbulence and alkane fluid properties, through their influence on the Sauter mean diameter (D32), gas hold up, gas-liquid interface rigidity and gas-liquid interfacial area per unit volume. These properties were correlated with KLa over a wide range of agitation rates and alkane concentrations in alkane-aqueous dispersions. Experiments were conducted in a 5 litre aerated and agitated bioreactor at agitation rates of 600, 800, 1000 and 1200 rpm and alkane (n-C10-C13 cut) concentrations of 0, 2.5, 5, 10, and 20% (v/v). KLa determination was executed using both the gassing out and pressure step methods. The accuracy and reliability of these methods were compared under the full range of agitation rates and alkane concentrations. The pressure step method was conclusively shown to be superior provided that probe response was taken into account, and was therefore used in the correlations. The interfacial areas corresponding to the KLa values were calculated from the combined effects of D32 and gas hold up. D32 was determined from the measurement of the dispersed air bubble diameters by means of a photographic technique and image analysis. Image analysis was performed by a program that was developed in Matlab® using image acquisition and image processing techniques. This program used these techniques to extract information of the gas bubbles in the image. The gas hold up was determined using the dispersion height technique. The behaviour of KLa was shown to be dependent on both agitation and alkane concentration. Increasing agitation from 600 to 1200 rpm increased KLa for each of the alkane concentrations. The influence of agitation on the interfacial area was evaluated over the same range of agitation rates and the relationship between the corresponding KLa values and interfacial areas assessed. Increasing agitation rate similarly enhanced the interfacial area available for transfer for each of the alkane concentrations, resulting in the concomitant increase in KLa. This increase in interfacial area was related directly to a shear-induced decease in D32 and indirectly to an increased gas holdup as a result of the lower rise velocity of the smaller bubbles. In addition to the agitation, the presence of alkane markedly influenced KLa behaviour, but in different ways, depending on the alkane concentration. Alkane concentration between 2.5 and 5% (v/v) reduced D32 at constant agitation of 800, 1000 and 1200 rpm, a likely consequence of decreased surface tension and retarded coalescence conferred by the alkane. The smaller D32 and the consequential enhanced gas hold up served to amplify KLa through increased interfacial area. However, as alkane concentration was increased above 5% (v/v), the gas hold up decreased despite a continued decrease in D32, resulting in a corresponding decrease in both the interfacial area and KLa. This suggests that at the higher alkane concentrations, the influence of viscosity predominated, exerting multiple negative influences on the interfacial area and oxygen transfer coefficient. The trends were however, not observed at the low agitation of 600 rpm, where turbulence was significantly reduced and KLa was repressed for all alkane concentrations. The pressures imposed by turbulence and alkane properties on the interfacial area defined locales of KLa behaviour and three distinct KLa behavioural trends were identified, depending on the agitation rate and alkane concentration. Regime 1 was constrained between 2.5 and 5% (v/v) for agitation rates of 800 rpm and above. Here KLa enhancement was directly associated with increased interfacial area which was the major factor defining KLa in this regime. Regime 2 was constrained by alkane concentrations higher than 5% (v/v) for agitation rates of 800 rpm and above. In this regime, the KLa depression was observed with increasing alkane concentration suggesting a predominant influence of viscosity which would be likely to exert multiple negative influences on KLa, through both the interfacial area and KL. The interfacial area in this regime decreased mainly due to the negative effect of viscosity on gas holdup. Regime 3, characterised by a decline in KLa irrespective of the alkane concentration, occurred at agitation rates smaller than 800 rpm. It is likely that at low agitation rates, the contribution of turbulence was insufficient to exert a positive influence on the interfacial area In this regime, the interfacial decreased through the combined negative effect of increased D32 and decreased gas holdup. The resultant variation in OTR depended directly on the relative magnitudes of the KLa and oxygen solubility and indirectly on the process conditions which defined these magnitudes. Under conditions of enhanced KLa, OTR benefited from the combined increases in KLa and oxygen solubility. However, under conditions of KLa depression, the elevated oxygen solubility did not invariably outweigh the influence of KLa depression on OTR. Consequently, despite the considerably increased solubility of oxygen in alkane-based bioprocesses a potential decrease in OTR through depressed KLa underlines the critical importance of the quantification of this parameter in alkane-aqueous dispersions and the necessity for a definition of the locales of optimal KLa. Through the identification of the parameters which underpin the behaviour of KLa in alkane-aqueous dispersions and the quantification of the effect of process conditions on these parameters, a fundamental understanding of the KLa and OTR in alkane-aqueous dispersions has been developed. This provides a knowledge base for the prediction of optimal KLa in these systems and has wide application across all alkane-based bioprocesses.

Thesis (MScEng (Process Engineering))--University of Stellenbosch, 2005. / Although selective oxidation catalysts are widely used and extensively studied for their industrial and academic value, their complex mechanisms are, to a large extent, still unclear. The field of so-called allylic (amm)oxidations reactions was chosen for further investigation, in particular the simplistic selective oxidation of propene to acrolein over an α-Bi2Mo3O12 catalyst. One of the most important approaches in selective oxidation is to try to correlate the physicochemical properties of catalysts with their catalytic performance (activity and selectivity). The most interesting, and seemingly most widely invoked parameter, is lattice oxygen mobility. The problem, however, is the difficulty encountered in measuring oxygen mobility. It is hypothesised that the depth of oxygen utilisation and lattice oxygen mobility of bismuth molybdate during the partial oxidation of propene to acrolein may be determined by measuring the rate of acrolein formation and lattice oxygen usage over a range of discrete particle sizes that could be synthesised using reverse micelle technology. Catalyst Preparation A preliminary investigation into the reverse micelle technique showed that discrete nanosized particles could be synthesised, but that there was no size control over the outcome and that, in most cases there were some degree of particle agglomeration. It was also found that nanorod formation occurred due to adsorbtion of surfactant. More in-depth investigation had to be done in order to achieve particle size control and the liberation of the calcined α-Bi2Mo3O12 catalyst particles required for kinetic experiments. Simple precipitation methods, the catalyst calcination step, and the formation and stability of reverse micelles were investigated. A simple precipitation method to prepare α-Bi2Mo3O12, suitable to be integrated into the reverse micelle technique was found by buffering the mixture of bismuth nitrate and ammonium molybdate solutions with an excess of molybdate. This prevented the pH from decreasing below a critical value of 1.3 (at which β-Bi2Mo2O9 forms as an impurity). The excess molybdenum caused the formation of MoO3 in the calcined product, which was selectively and successfully removed using a warm ammonium wash followed by a water rinse and a recalcination step. XRD of a temperature range calcination shows that the calcination starts at temperatures as low as 200°C and almost complete calcination of the catalyst at 280°C. DSC analyses show a 47.15 J/g crystal formation peak only at 351°C. The Mo18O56(H2O)8 4- anion or its double, Mo36O112(H2O)16 8-, is responsible for the formation of α-Bi2Mo3O12 in the precipitation calcination reaction. Reverse micelles were investigated using a Malvern Zetasizer and showed a complex dynamic system in which the reverse micelle sizes and size distributions change over time as a function of surfactant and aqueous concentrations, the salt used and aqueous phase salinity. Although much was accomplished in this study, more investigations into the constituent steps of the reverse micelle technique are needed to develop a method to synthesise the range of discrete catalyst particle sizes required for kinetic studies. Kinetic Studies For the purpose of kinetic experiments a metal reactor was found to be superior to that of a glass reactor. The reactor rig was adequate for these kinetic studies but do not meet the requirements for detailed reaction order experiments. The analysing apparatus could not measure CO2 formation accurately and it had to be calculated using a carbon balance. Only the model proposed by Keulks and Krenzke [1980a] was able to describe the kinetic result, but the model parameter describing the oxidative state of the catalyst surface could not be calculated due to the lack compatibility between published data. Values were awarded to this parameter so to give an Arrhenius plot which corresponded to published data. The parameter describing the oxidative state vs. temperature took on a function that was consistant with the reasoning of Keulks and Krenzke [1980a]. Comprehensive preliminary kinetic studies are needed, both in catalyst reduction and reoxidation, in order to determine the reaction conditions, explore more advanced kinetic models and investigate model parameters that are theoretically and/or empirically obtainable and quantifiable.