Faculty of Engineering, Built Environment and Information Technology
School of Engineering
Department of Chemical Engineering
Selected Highlights from Research Findings
Chamotte Holdings recently started local production of synthetic hydrotalcite for PVC stabiliser applications. Synthetic hydrotalcites are also known as layered double hydroxides. They are synthetic anionic clays with a composition analogous to that of the mineral hydrotalcite. The key feature of these clays is that they have a unique layered structure that allows the intercalation of guest molecules and guest anions. Such intercalated materials belong to the class of nanostructured materials. Their properties are determined by the nature and activity of the guest molecules. Since the choice of guest/intercalant is virtually limitless, there is immense scope for tailoring these clays for a wide range of applications. The distinctive properties of these modified layered double hydroxides permit a wide range of uses, including polymer additives, precursors for catalysts and magnetic materials. Nanocomposites can be prepared by exfoliation in polymer matrices. Their generally non-toxic nature and membrane-like structure can be harnessed to protect, carry, deliver and controllably release active compounds such as pesticides, pharmaceuticals and even genes. They can be used as absorbents to remove contaminants from water. Hydrotalcite-like anionic clays already have existing applications. In medicine, they are utilised as antacids and antipeptins. In polymer technology, they function as halogen scavengers, flame retardants and PVC stabilisers. They are employed as catalysts and catalyst supports and their absorbent and ion exchange properties are of interest in waste water treatment. Surfactant intercalated layered double hydroxides are of interest for a variety of reasons. Hydrophobisation of the layered double hydroxides by ion exchange yields new types of thickening agents. It also facilitates sorption of nonionic organic compounds, for example, trichloroethylene and tetrachloroethylene
Contact person: Prof WW Focke.
A research project was undertaken that focuses on the synthesis and modification of hydrotalcites as functional additives in polymers. In these applications, surfactant intercalated clays assist in the dispersion of nanoparticles in the polymer matrix. Graduate students Edith Landman, Nontete Nhalpo and Lumbe Moyo have been studying the intercalation of anionic surfactants and long chain fatty acids in these materials. These researchers have developed novel and environmentally friendly processes for intercalating either long chain fatty acids or anionic surfactants using surfactant or acid assisted strategies. The stearic acid intercalation reaction is conducted at elevated temperatures with the layered double hydroxide (LDH) powder as a suspension in a stearic acid oil-in-water emulsion. The acidic fatty acid, for example, stearic acid, reacts with the basic carbonate anions. CO2 is released and the fatty acid is intercalated as a bilayer. High concentrations of anionic or nonionic surfactants, that is, sodium dodecylsulfate or Tween 60, aid the intercalation process by emulsifying the molten acid and dispersing the hydrotalcite particles. The method is convenient, economical and environmentally friendly. It employs the readily available carbonate form as starting reagent. Water is used as a medium rather than organic solvents. Low reaction temperatures suffice, that is, calcinations of the clay are superfluous and there is no need for working under a CO2-free atmosphere. Similarly, anionic surfactants are intercalated into the layered double hydroxides when used in combination with a short chain carboxylic acid. Best results were obtained using water-soluble organic acids, for example, acetic, butyric or hexanoic acid, to aid decarbonation of LDH-CO3. Intercalation proceeded at ambient temperatures with the precursor powder suspended in an aqueous dispersion of the anionic surfactant. The carboxylic acids are believed to assist intercalation by facilitating the elimination of the carbonate ions present in the anionic clay galleries. These processes are currently employed to synthesise novel nanostructured hydrotalcites for use in pharmaceutical and polymer applications
Contact person: Prof WW Focke.
Polymers, or plastics as they are more commonly known, find increasing use as structural and functional materials. Owing to their combustible, organic nature, they pose a fire risk in some applications. Hydrated fillers, for example, aluminium trihydrate and magnesium hydroxide, have utility as flame retardants. They belong to the class of endothermic flame retardants. The candle model for polymer combustion assumes independent pyrolysis and flame zones. Heat transfer from the flame provides coupling between the gas and condensed phases. It drives the thermal degradation reactions that produce volatile fuel fragments that, in turn, feed the flame. This model suggests three possible strategies for flame-proofing combustible polymers: interference with the gas phase combustion and with the substrate pyrolysis reactions or their decoupling via a physical barrier to heat and mass transport. With endothermic flame retardants this may correspond to the cooling of the substrate and promotion of charring, dilution of the flammable gas with inert gases and, potentially, the formation of an ash-char barrier layer. Endothermic flame retardants contribute significantly to the first two flame retardancy mechanisms mentioned above. The endothermic decomposition reaction releases water vapour. The cooling of the polymer substrate inhibits the solid phase decomposition reactions. Simultaneously, the steam released dilutes the surrounding atmosphere with an inert gas. This reduces the rate of fuel production from the solid substrate and decreases the rate of combustion in the gas phase. The end result is that the flames are snuffed out. The use of these flame retardants is associated with some problems. In particular, on decomposition magnesium, hydroxide forms a high surface area, powdery residue that tends to catalyse oxidation of char residues leading to an afterglow effect. As a result, effective barrier formation is realised only when additional additives are incorporated. Secondly, high loadings are required to achieve adequate flame resistance. This impairs the fluidity of the corresponding melts and the mechanical properties of the solid compounds. The hydrophilic nature of the filler surfaces further compromises mechanical integrity. Surface coating modifications can compatibilise inorganic fillers with hydrophobic polymer matrices. Stearic acid is widely used for basic fillers such as calcium carbonate and magnesium hydroxide
Contact person: Prof WW Focke.
The Institute of Applied Materials conducted research in collaboration with the Institute of Polymer Research in Dresden, Germany, in an attempt to improve the performance of hydrated fillers in low density polyethylene (LDPE) and ethylene-vinyl acetate copolymer (EVA). These two polymers are widely used as cable sheathing compounds for cables used in underground mining that must be flame-retarded. In this work, the limiting oxygen index (LOI) was used as a measure of flammability. Cable specification often calls for an LOI value greater than 30. The research strategy had two aspects. First, the use of red phosphorus as afterglow suppressant and as synergist for the base inorganic flame retardant was considered. Secondly, the utility of blending the polyethylene or the EVA with other polymers carrying hydroxyl functionality was investigated. Hydroxyl groups are known to aid char formation by dehydration reactions with phosphorus compounds. The key findings of this research were that the combination of red phosphorus and hydroxyl-containing polymers had a beneficial effect with respect to LOI values in alumina trihydrate (ATH)-filled blends. An LOI value of 30 was achieved in ethylene-vinyl alcohol copolymers (EVOH) with as little as 32% ATH and 3% red phosphorus. These loadings are much lower than the >60% required when ATH is used on its own. The main advantage is that compounds containing less filler are easier to process and also feature better mechanical properties
Contact person: Prof WW Focke.
A dynamic distillation column model has been developed using the Modelica language. Modelica is an object-oriented, declarative, multi-domain modelling language for component-oriented modelling of complex systems, for example, systems containing mechanical, electrical, electronic, hydraulic, thermal, control, electric power or process-oriented subcomponents. The model utilises the object-oriented nature of Modelica by using generic models for the different distillation column parts. Together with the column model, a stream class has been developed that can use the external interface of Modelica to access thermodynamic information via Computer-aided Process Engineering (CAPE)-OPEN using Common Object Requesting Broker Architecture (CORBA), a standard defined by the Object Management Group (OMG) that enables software components written in multiple computer languages and running on multiple computers to work together. CAPE-OPEN standards are the uniform standards for interfacing process modelling software components developed specifically for the design and operation of chemical processes. They are based on universally recognised software technologies such as the Component Object Model (COM) and CORBA. Using CORBA enables thermodynamic properties to be calculated by any CAPE-OPEN-compliant property package. In this study, a Modelica library, allowing interface between Modelica and CAPE-OPEN, was developed. Its functionality was demonstrated using a model of a ten-plate distillation column simulated in OpenModelica on a Linux machine, with thermodynamic and property data from Honeywell Unisim on a Windows machine. The data interfacing was done over a network using CORBA. It was found that real-time operation is possible, but that network overhead makes up a significant fraction of the running time, posing problems for off-line simulation and optimisation
Contact person: Prof PL de Vaal.
Flow sheeting has assumed a central position in modern process engineering practice. Diagrams describing the flow of material and energy through a plant are the main mode of communicating process design and several packages exist that solve the associated equations. Steady state flow sheeting is accepted in industry to such an extent that it is unlikely that a chemical plant of any size is designed without the use of at least one such tool. Acceptance of dynamic simulation is lagging behind the steady state simulations, partly due to the computational requirements associated with such simulation and partly due to the increased workload of developing dynamic simulations. There is, however, increased interest in the development of dynamic plant simulations. In addition to dedicated process engineering tools like HYSIS, Aspen, ChemCAD and SimSci, several modelling languages have emerged in the last 15 years that aim to provide an environment for modelling dynamic processes. Ascend and gPROMS are two examples. On another front, modelling systems have been developed for multidiscipline simulation. Modelica was specifically designed as a standard for such simulation, incorporating aspects from other languages. Simulation of a distillation column was done before in Modelica. However, the modelling strategy followed here is more modular, and abstracts the thermodynamics to the streams, allowing easy interfacing with an external thermodynamics package
Contact person: Mr C Sandrock.
South Africa has substantial reserves of fluorine-containing minerals. Fluorine also happens to be a very versatile element that, in various compounds, has a variety of applications in such diverse areas as microprocessor technology, special materials, oxygen carriers and the nuclear industry. Research in this field is focused on local beneficiation of these deposits to enhance value-addition and in that way contributes to the development of infrastructure and increasing the value of exports. Despite the fact that the thermal decomposition of polytetrafluoroethylene (PTFE) has been extensively studied over the past six decades, some inconsistencies regarding the kinetic parameters, for example, the order of the reaction, remain. Representative kinetic data is essential for practical purposes, such as reactor design and scaling. In general, the literature data refers to homogeneous bulk heating, whereas the case of the nonhomogeneous heating of a single particle has not received attention. Data (reaction rate and pre-exponential factor) applicable to this latter case was experimentally determined from isothermal thermogravimetric analyses of the depolymerisation reaction of PTFE. The kinetic data obtained on coarse granules (800 to 1 000 micron) were reported. The rate law is consistent with a shrinking particle kinetic model, with chemical kinetics controlling phase-boundary movement. The mass loss rate is directly proportional to surface area. A rate law applicable to this case and useable for geometries of arbitrary shape was derived. It has long been known that the thermal decomposition of PTFE predominantly yields tetrafluoroethylene (TFE), hexafluoropropylene (HFP) and octafluorocyclobutane (OFCB). The product ratios can be tailored by manipulating the working temperature, the pressure, the residence time of the gaseous product stream in the hot zone and the quench rate. PTFE is non-melt-processible, hence one of the recycling methods is via pyrolysis of the solid waste and subsequent recovery of the monomer for re-use. For this process, solid waste is ground down to coarse granules rather than powder for economic and practical reasons. Knowledge of depolymerisation kinetics is essential for reactor design and industrial scaling
Contact person: Prof PL Crouse.
The least constrained mass transfer mathematical formulation for freshwater minimisation in multipurpose batch chemical processes with central reusable water storage is presented in this research. The mathematical formulation is an extension of the model developed by Prof Thokozani Majozi, in which four scenarios were considered with various limitations or constraints. In the scenario presented in this research, only the mass load is fixed, while both the quantity of water used in a particular operation and outlet concentration are allowed to vary. In essence, fixing the mass load is more representative of the practical case. A solution procedure for the resultant nonconvex mixed integer nonlinear programming (MINLP) model is also presented. The solution procedure first involves reformulating the MINLP into a relaxed linear model (MILP). The MILP is first solved, forming a feasible starting solution for the MINLP. The production of high value, specialty chemicals is generally performed in batch type processes. Typical industries in this category include pharmaceuticals, food and agrochemicals. Batch processing is suited to the production of such products since the products are characterised by identical recipes that allow for the sharing of equipment. One of the defining characteristics of batch processing is the discrete nature of the tasks. All operations and material flows are time dependent, which means that scheduling the operations becomes an important, but complex task. As with continuous processes, batch processes produce effluents. The volume of effluent from a batch process is much less than a similar continuous process. However, the toxicity of the effluent is much higher. Coupled with this is the mounting environmental pressure on industry to reduce effluent released in the environment. The least constrained mass transfer mathematical model for wastewater minimisation in batch plants has been presented together with an applicable solution procedure. Application of the mathematical model to the first illustrative example involving four water-using operations has shown 32% improvement in water use. The solution was obtained in a reasonably short central processing unit (CPU) time, that is, 20.9 CPU seconds, which could be ascribed to the structure of the overall model. The model was applied to a second illustrative example and an 84% reduction in the amount of wastewater was achieved. Furthermore, the solution to the exact model in the second illustrative example could only be determined once a linearised model was first solved and then used as an initial solution for the exact model. It is worth mentioning, however, that the overall mathematical model is a non-convex MINLP, which implies that global optimality cannot be guaranteed
Contact person: Prof T Majozi.
The chloride concentration of a hazardous leachate of approximately 22 000 mg/l is too high for the direct discharge requirement of less than 1 000 mg/l into the municipal sewerage system. Electrodialysis (ED) was selected among various other alternative technologies for the desalination/concentration of the leachate. Lime pre-treatment of the leachate was superior to caustic soda treatment for the removal of scale-forming chemicals (Ca, Ba, Sr, Fe, Mn). Membrane fouling took place during ED treatment. However, membrane fouling should be restored with polarity reversal and mechanical cleaning of the membranes. Batch ED tests have shown that ED performance remained more or less constant during eight desalination/concentration runs. The electrical energy consumption for ion transport and brine volume varied between 9.6 and 11.4 kWh/m3 feed and 17 and 35% of the treated feed, respectively. A very good quality water could be produced with ED treatment of the leachate. The electrical conductivity of the leachate could be reduced from 5 490 to 139 mS/m (97.5% removal). Chloride was reduced from 21 957 to 345 mg/l (98.4% removal). Arsenic was reduced from 8.8 to 1.37 mg/l. The capital cost for a 100 kl/d ED plant is estimated at US$ 0.6 million. Operational costs (ion transport) are estimated at US$ 0.38/m3 leachate. The chloride concentration of a hazardous leachate of approximately 22 000 mg/l is too high for direct discharge into the municipal sewerage system. The chloride concentration should be reduced to less than 1 000 mg/l to make it suitable for discharge. ED, reverse osmosis (RO) and evaporation are technologies that can be considered for the desalination of the leachate. However, the osmotic pressure of the leachate should be too high to consider conventional RO for the treatment of the leachate. Evaporation technologies are very expensive for this type of application. ED, on the other hand, is not dependent on the osmotic pressure of the leachate and can be used to reduce high saline concentrations in effluents to low concentration levels. Membrane fouling, however, may adversely affect the process. Nonetheless, it was decided to evaluate ED for the desalination of the leachate. The objectives of this investigation were therefore to evaluate ED for the desalination/concentration of the leachate and to determine the preliminary economics of the process. Excellent removals of calcium, barium, strontium, iron and manganese were obtained with lime treatment of the leachate. Sludge volume will comprise approximately 10% of the treated leachate. More membrane fouling was experienced at higher current densities (100 to 120 mA/cm2) than at lower current densities (20 to 80 mA/cm2). It appears to be possible to control membrane fouling with polarity reversal and mechanical cleaning of the membranes
Contact person: Prof JJ Schoeman.
In this groundbreaking study, the researchers investigated the use of the capillary-gate mechanism to explain the multiplicity behaviour in trickle bed reactors. They reviewed the experimental trends in literature on the basis of a limiting cases framework and evaluated the three-dimensional pore-scale liquid distribution using computed tomography (CT) data. This led to the identification of 20 phenomenological trends that characterise hydrodynamic multiplicity, including hydrodynamic flow hysteresis, as well as the effects of pre-wetting. The CT study yielded additional experimental insight into the role of capillary pressure and ultimately led to the proposal of a capillary-gate mechanism based on contact angle hysteresis as the root cause of multiplicity. The mechanism was incorporated into a simple pore network model and it was shown that the qualitative performance of the model corresponds closely to the majority of phenomenological trends
Contact person: Prof W Nicol.
The oligomerisation of olefins allows the petrochemical industry flexibility in the production of gasoline and diesel from shorter chain olefins. This involves the use of a solid acid catalyst. In industry, mainly ZSM-5 and solid phosphoric acid (SPA) are used. Industrially, the oligomerisation of shorter chain olefins is more established in the production of petrol. The oligomerisation of heavier olefins, such as 1-hexene, can, however, also be used for the production of diesel and lubricating oils. In general, acid catalysed olefin oligomerisation may involve many different reaction steps, among which are isomerisation (for C4 and longer olefins), hydrogen exchange, dimerisation and trimerisation, and cracking. In previous studies on 1-hexene oligomerisation, the reaction selectivity towards the formation of dimers, trimers, heavier products and cracked products was investigated as a function of solvent type or catalyst type. Batch kinetic experiments at 150°C, 200°C and 250°C showed that the reaction can be modelled with a three-step sequential reaction scheme. Firstly, this involves linear isomerisation of 1-hexene, followed by skeletal isomerisation, and finally dimerisation and cracking. The first and last steps in the sequence are modelled as reversible reactions. When first-order reaction kinetics was assumed for each of the reactions, the model gave a very good representation of the experimental data. In order to test the validity of the series pathway hypothesis, the reaction was repeated with a skeletal hexene isomer – 2.3-dimethyl-2-butene (DMB) – as reactant. Although the rate and equilibrium constants for the third reaction step, as obtained from the 1-hexene conversion data, gave a good prediction of the DMB conversion at 200°C and 250°C, it failed to predict the reaction rate at 150°C. This suggests that a different reaction pathway – where linear hexene isomers are directly converted to dimer products – becomes more significant at lower temperatures. The same equilibrium conversions of both 1-hexene and DMB were observed at all three temperatures investigated, suggesting that the equilibrium conversion is independent of the type of hexene isomer in the reaction mixture
Contact person: Prof W Nicol.
Breakthroughs were achieved in the fields of environmental microbiology and advanced oxidation processes. One such breakthrough is the recent identification of 25 microbial culture species and serotypes capable of reducing Cr (VI) at a rate 8 to 10 times higher than previously observed in cultures isolated from other parts of the world. The fastest acting of these species in less than 110 hours completely reduced Cr(VI) in solutions with initial Cr(VI) concentrations as high as 300 mg/l. While working with sulphate-reducing bacteria from the South African gold mines, the research group discovered species that are capable of selectively extracting strontium (Sr(II)), one of the fission products in the thermonuclear reaction, from mixed waste. Strontium was utilised in this study to simulate the removal of metallic elements. Sr itself poses a serious environmental hazard. Sr(II) is mistakenly taken up and incorporated into the bone material of animals since living cells cannot distinguish it from the essential mineral calcium (Ca(II)). In this study, Sr removal was determined to be predominantly a surface reaction, which was determined by the fit to the common adsorption isotherms. However, the surface reactions and subsequent uptake into the cell were metabolically linked
Contact person: Prof EMN Chirwa.
In a parallel activity supported by the Water Research Commission, the research group investigated the photolytic and photocatalytic degradation of harmful organic compounds in water. Photolytic and photocatalytic processes are emerging as a cleaner solution for treating pollutants, such as refractory organics, pharmaceutical compounds and endocrine-disrupting compounds (EDCs) emanating from industrial, medical and cosmetic applications. Another troublesome class of pollutants is emerging from halogenated forms of algal metabolites in eutrophic water sources. These compounds, when chlorinated, form carcinogenic halogenated products such as trihalomethanes (THMs), haloacetic acid (HAA) and chlorophenols (CPs). Photocatalysis accelerates the photochemical transformation of substrates in the presence of a semiconducting material such as TiO2. The catalyst may accelerate the photoreaction by interaction with a substrate in its ground or excited state and/or with a primary photoproduct, depending on the mechanism of photoreaction. Sn energy source is needed during photocatalysis, usually ultraviolet light, to create reactive sites called electron band pairs in semiconductor particles. The potential difference between an electron band pair (that is, the -ve valence and +ve conduction band) forms a natural band gap for the given material. The band gap refers to energy difference between the valence and conduction bands. The band gap is also seen as a void region that extends from the top of the electron-filled valence band to the vacant conduction band and is dependent on the nature of the semiconductor material. In photocatalysis, adsorption of energy (hv) equal to or more than its band gap produces electron excitation in the catalyst and electrons gain sufficient energy to change levels from the valence to the conduction band. On the other hand, an electron vacancy or hole (h+) is created in the valence band. Excited-state conduction-band electrons and valence-band holes can recombine and dissipate the input energy as heat, get trapped in metastable surface states, or react with electron donors and electron acceptors adsorbed on the semiconductor surface or in the surrounding electrical double layer of the charged particles. The research succeeded in reducing the residence time required for the complete degradation of single ring aromatic compounds to less than two minutes under aerated conditions in a photocatalytic reactor. These results indicate a high potential for substituting slower acting biological processes with photocatalytic and photolytic systems to degrade refractory and hard-to-degrade organic compounds using a cheap source of energy such as sunlight
Contact person: Prof EMN Chirwa.
The German Federal Ministry for Education and Research and the NRF funded a scientific water management project, which deals with a technological unit called the communal water house (CWH). The aim was to improve the water situation, especially in rural communities. Kouga Local Municipality was the beneficiary of the project in South Africa. The CWH is a facility that aims to provide water of definite quality and temperature for household use, shower and laundry purposes in rural settlements. It consists of two or more separate rooms inside the building for laundry and shower activities, and tap water outlets outside the building. Water sources are tap water, rainwater or well water. The water for laundry and shower purposes is recycled inside the building and treated by a service water treatment station. This treatment station consists of three technological units that work in a controlled manner: a biological treatment unit, an ultraviolet light disinfection unit and membrane filtration modules. The quality of the treated water is in accordance with the European Union bathing water regulations. The water is heated by a solar heating system. Water is pumped by solar pumps. The researchers found that CWH improves water use efficiency and that twice as many people can be served by the same amount of water at the same level of services, highest sanitary standards and reduced costs. CWH is therefore very well suited for rural communities that lack water supply and energy to improve the living conditions of their people. It can be used as a full, sustainable water management solution or as an intermediate solution. Moreover, it can be used in tourist facilities, such as camping grounds, sport grounds and other camps. Sanitation for CWH users is provided by urine diversion toilet systems
Contact person: Prof EMN Chirwa.
The behaviour of cohesionless soils is known to be influenced by the method of reconstitution. It is generally accepted in the literature that different reconstitution methods produce samples of varying fabric and thus samples of varying behaviour. Very little evidence has been presented to validate this statement. The main aim of this research project was, firstly, to observe the fabric of in situ and reconstituted gold tailing samples and, secondly, to investigate the difference in behaviour between these samples at the same state. The study focused on testing in situ and reconstituted gold tailing samples obtained from three positions on a tailing dam: pond, middle beach and upper beach. Laboratory reconstitution methods included moist tamping and slurry deposition. Fabric analysis involved the use of scanning electron microscope (SEM) images to classify the observed differences in the fabric of the undisturbed and reconstituted gold tailing samples. A particle interaction model based on the observed fabric was postulated to explain the differences or similarities in behaviour. The scope of behaviour investigated included sedimentation, collapse and swell, consolidation and compressibility, creep, stiffness and shear behaviour. The fabric analysis indicates that differences in the fabric of undisturbed and reconstituted gold tailing samples are visible. Moist tamping produces an aggregated fabric, while slurry deposition yields a homogeneous fabric similar to that of the undisturbed samples. Comparison of behaviour indicates that neither moist tamping nor slurry deposition can fully replicate the behaviour of the undisturbed sample. Consolidation and compression is a function of the fabric, while friction angle is independent of the fabric. Available shear strength and liquefaction potential is also affected by the preparation method and the resulting fabric
Contact person: Prof PL de Vaal.
|
