Zugl.: Tübingen, Univ., Diss., 2000.

Zugl.: Kassel, Univ., Habil.- Schr.

Thesis (doctoral)--Göteborg University, 2003. / Includes bibliographical references (p. 137-157).

Trotz hochmoderner Technologien und ausgefeilter therapeutischer und rekonstruktiver chirurgischer Heilungsmethoden beträgt die 5-Jahres Überlebensrate bei der Diagnose PECA der Mundhöhle im Durchschnitt auch im Jahre 2017 nur 55 % und die Heilungsmethoden haben sich in den letzten drei Jahrzehnten kaum verbessert. Umso wichtiger ist es deshalb, die Forschung voranzutreiben und ein aussagekräftiges Tumormodell zu etablieren, das bei der Entwicklung neuer Therapieansätze schnell und sicher gute Ergebnisse liefert. In dieser Studie soll mit Hilfe des Tissue Engineering (TE) ein in gesunder Mundschleimhaut (MSH) integriertes 3D-Tumormodell generiert werden, welches bestmöglich die Analyse pathologischer Mechanismen im Tumorzentrum, sowie im Randbereich von gesundem und erkranktem Gewebe, und auch die Analyse der Auswirkungen neuartiger Chemotherapeutika auf gesunde und maligne Zellen in direkter Nachbarschaft ermöglicht – ohne Tierversuche. In der Konsequenz könnte ein erheblicher Fortschritt mit höheren Erfolgsaussichten der Therapieansätze erzielt werden. Es wird ein Tumormodell generiert, in dem auf Basis eines gesunden MSH-Modells Tumorzellen eingebracht werden, um - genauso, wie die Tumorentstehung in vivo von statten gehen würde – Tumorentstehung und Tumorwachstum in der Umgebung von gesunder MSH analysieren zu können. Das Modell basiert dabei auf einer Matrix aus dezellularisierter, porciner, small-intestinal-submucosa (SIS/MUC), die mit primären Fibroblasten, primären Keratinozyten und Tumorzellen der Tumorzelllinie FaDu besiedelt wird. Eine Besonderheit der FaDu-Zellen ist die vorangegangene Transduktion mit dem Lentivirus RFP – um die eingewanderten Zellen von gesunden Zellen unterscheiden zu können. Der Vorgang der Transduktion war gelungen und es konnte eine Fluoreszenz der noch in Zellkulturschalen kultivierten Zellen erzielt werden. Allerdings waren die fluoreszierenden Zellen in den fixierten Schnitten nicht mehr nachweisbar. Zur Generierung eines Tumormodelles wurden auf Basis eines OMÄ drei unterschiedliche Applikationsformen zur Integration von Tumorzellen getestet. Die Integration von Tumorzellen fand in Form von Spots, Sphäroiden oder Tumorzellgemischen (prim. Keratinozyten/FaDu-Zellen) in zuvor kultivierte gesunde OMÄ statt. Dabei sollte das Applizieren von Spots oder Sphäroiden das Tumorzellwachstum auch in der Umgebung von gesundem Gewebe initiieren. Dies würde die Möglichkeit schaffen, auch in vitro, gesundes neben pathologischem Gewebe und den Übergang dazu genau analysieren zu können. Es sollen sowohl die optimale Konzentration der Tumorzellen, welche für die Entstehung von Tumoren nötig ist, als auch die geeignetste Applikationsmethode eruiert werden, um optimale Tumormodelle zuverlässig reproduzierbar ansetzen zu können. Die Modelle wurden histologisch und immunhistochemisch analysiert und die Ergebnisse mit ermittelten TEER-Werte in Korrelation gesetzt. In dieser Arbeit konnte mit der Applikation von Spots oder Sphäroiden kein suffizientes Tumorwachstum in Umgebung von gesunder MSH erzielt werden. Die Zellen lagen ohne Reaktion des angrenzenden Stratum corneums auf der zu stark ausgeprägten Hornschicht der OMÄ auf und es war keine Einwanderung in das darunterliegende Gewebe möglich. Allerdings ist es gelungen, durch Applikation eines Zellgemisches variierender Mischungsverhältnisse von primären Keratinozyten und Tumorzellen der Zelllinie FaDu ein 3D-Tumorwachstum unterschiedlicher Malignitätsstufen zu initiieren. Je kleiner das Mischungsverhältnis und je höher in der Konsequenz die Anzahl der FaDu-Tumorzellen, desto ausgeprägter waren die morphologischen Anzeichen einer Tumorbildung. Abhängig vom Mischungsverhältnis war dabei die Ausprägung des Tumors. Auch wenn dadurch keine Kombination von gesundem und pathologischem Gewebe in einem Modell mehr imitiert werden konnte, so konnten zumindest nach histologischen und immunhistochemischen Untersuchungen eindeutige pathologische, maligne Tumormodelle generiert werden. Die Tumormodelle zeigten durchgehend Zell- und Kernpleomorphismen, atypische und vermehrte suprabasale Mitosen, eine Störung der normalen Gewebearchitektur, die Ausbildung von Interzellularbrücken, Einzelzelldyskeratosen und Verhornungsknospen, sowie Stellen der Durchbrechung der Basalmembran und Invasion von Tumorzellen in die darunterliegende Lamina propria. All das sind eindeutige Kennzeichen malignen Wachstums Auch die Ergebnisse der TEER-Wert Messung stimmten mit den morphologischen Entwicklungen der Modelle überein. So stiegen die TEER-Werte der Kontrollmodelle konsequent an, was für eine deutliche Entwicklung von kontinuierlichem Gewebe spricht und im Gegensatz dazu fielen die TEER-Werte im zeitlichen Verlauf der Tumormodelle, bei denen die Basalmembran und somit die Kontinuität des Gewebes durchbrochen wurde rapide ab, bzw. lagen im konstant niedrigen Bereich. Der Erfolg der Etablierung dieses zuverlässig rekonstruierbaren 3D, in vitro generierten Tumormodells, das der in vivo Situation eines Plattenepithelkarzinoms sehr nahekommt, bietet der Wissenschaft eine sehr gute Möglichkeit, weitere Studien zum Tumorwachstum durchzuführen. Außerdem kann die Weiterentwicklung und Verbesserung vielversprechender, neuartiger chemotherapeutischer und radiologischer Therapieverfahren erheblich voran¬getrieben und dadurch die Heilungschancen mit geringeren Nebenwirkungen für den Patienten verbessert und eine erhöhte Lebensqualität erzielt werden. / Despite state-of-the-art technologies and sophisticated therapeutic and reconstructive surgical methods, the average 5-year survival rate of patients diagnosed with oral squamous cell carcinoma (OSCC) is still only 55% in 2017. Healing methods have barely improved over the last three decades. Therefore, it is important to establish a meaningful tumour model that delivers fast and reliable results in the development of new therapeutic approaches. In this study, Tissue Engineering is used to generate a three-dimensional tumour model integrated into healthy oral mucosa. This enables an ideal analysis of pathological mechanisms in the tumour center, as well as in the margins of healthy and diseased tissue. It also allows the analysis of the effects of novel chemotherapeutic agents on healthy and malignant cells in proximity - without animal testing. Consequently, a considerable progress could be achieved with a higher chance of success of therapeutic approaches. A tumour model, based on a healthy oral mucosa equivalent (OME), is generated in which tumour cells are integrated in order to be able to analyse tumourigenesis and tumour growth in the environment of healthy oral mucosa just as the tumour development would take place in vivo. For this primary fibroblasts, primary keratinocytes and tumour cells were cultured on a matrix of decellularized, porcine, small intestinal submucosa (SIS/MUC). For this FaDu cells were transduced with the lentivirus RFP to be able to distinguish the immigrated cells from healthy cells. The transduction was successful. It was possible to achieve a fluorescence of the cells still cultivated in cell culture dishes. However, the fluorescent cells could no longer be detected in the fixed tissue sections. For the tumour model three different forms of application of the tumour cells on the OMEs have been tested. The application of cell-spots, spheroids or cell mixtures of primary keratinocytes and FaDu tumour cells in previously cultivated OME. The application of spots or spheroids should initiate tumour cell growth even in the environment of healthy tissue. This would enable the in vitro analysation of the area of healthy and pathological tissue in one model. Therefore, the optimal concentration of tumour cells, which is necessary for the tumour development, and the most suitable application method are to be determined to be able to apply a suitable reproducible tumour model. The models were analysed histologically and immunohistochemically, and the results were correlated with determined TEER values. In this work, the application of spots or spheroids did not achieve tumour growth in the environment of healthy oral mucosa. The cells were not responsive to the adjacent stratum corneum on the highly pronounced horn layer of the OME and no migration into the underlying tissue was possible. However, by applying a cell mixture of varying mixing ratios of primary keratinocytes and tumour cells of the FaDu cell line, it has been possible to initiate 3D tumour growth of different malignant stages. The smaller the mixing ratio and the higher the number of FaDu tumour cells, the more pronounced have been the morphological signs of tumour formation. Even if it was no longer possible to mimic a combination of healthy and pathological tissue in a model, clear pathological, malignant tumour models could be generated at least after histological and immunohistochemical investigations. The tumour models consistently showed cellular- and nuclearpleomorphisms, atypical and increased suprabasal mitoses, disruption of normal tissue architecture, the formation of intercellular bridges, single cell dyskeratosis and cornification buds, as well as sites of disruption of the basement membrane and invasion of tumour cells into the underlying lamina propria. All these are clear signs of malignant growth. The results of the TEER value measurement were also consistent with the morphological developments of the models. Thus, the TEER values of the control models rose consistently, which indicates a significant development of continuous tissue. In contrast, the TEER values over the course of time of the tumour models, in which the basal membrane and thus the continuity of the tissue was broken, fell rapidly or were in a constantly low range. The success of the establishment of this reliably reconstructable 3D, in vitro generated tumour model, which is very close to the in vivo situation of a squamous cell carcinoma, offers the science a very good opportunity to carry out further studies on tumour growth. In addition, the further development and improvement of promising, novel chemotherapeutic and radiological therapy methods can be considerably advanced, thereby improving the chances of recovery with fewer side effects for the patient and achieving an increased quality of life.

Darmstadt, Techn. Universiẗat, Diss., 2007. / Dateien im PDF-Format.

Die Vergleichbarkeit von Messergebnissen ist für eine weitere Datenverarbeitung und den darauf aufbauenden Auswertungen unumgänglich. Hier machen jedoch eine Vielzahl von Messverfahren und -methoden sowie fehlende Angaben zu den Messbedingungen und der Messgenauigkeit einen Vergleich unterschiedlicher Datensätze nahezu unmöglich. Dabei kann bereits durch die Angabe einiger wesentlicher Analyse- und Prozessparameter eine projektübergreifende Auswertung erfolgen und ein erheblicher Wissensgewinn generiert werden. Besonders dann, wenn kostenintensive Messkampagnen nur im begrenzten Umfang durchführbar sind, verbessern sich durch weitere Messdaten aus anderen Studien die Sicherheit und damit die Aussagekraft der Ergebnisse bei gleichzeitig geringem Mehraufwand. Die folgenden Ausführungen zu verschiedenen Messgeräten und -methoden zur Bestimmung von Permanentgaskonzentrationen sowie von Teergehalten im Produktgas von Holzvergasungsanlagen erheben nicht den Anspruch auf Vollständigkeit. Dennoch soll die exemplarische Darstellung allgemeine Hinweise für den Umgang mit Messdaten geben und die Nutzer von Messdaten für mögliche Probleme sensibilisieren, um Fehler bei der Auswertung zu minimieren. Damit ergänzt das Handbuch „Produktgasmessmethoden – Biomassevergasung“ neben den Methodensammlungen „Biogas“ und „Feinstaub“ die Arbeiten rund um die Methodenharmonisierung im Förderprogramm „Energetische Biomassenutzung“. Großer Dank für die unermüdliche Unterstützung gilt allen Projektpartnern, den Stadtwerken Rosenheim, der Burkhardt GmbH, der Spanner Re² GmbH und dem Projektteam der Programmbegleitung des Förderprogramms „Energetische Biomassenutzung“.

Die Durchfallerkrankung Giardiasis wird durch den Protisten Giardia duodenalis ausgelöst. Die Infektion erfolgt fäkal-oral, meist über kontaminiertes Trinkwasser. Der Parasit kolonisiert den oberen Bereich des Dünndarms und heftt sich an das Epithel, wodurch es die Krankheitsbeschwerden auslöst. Allerdings sind Details über die Mechanismen der Pathogenese unbekannt. Dazu kommt, dass der Ausgang einer Infektion fallspezifisch starken Schwankungen unterworfen ist, von selbst-limitierend bis chronisch und asymptomatischer Kolonisierung bis hin zur schweren Enteritis. Ein möglicher Pathomechanismus ist der Wegfall der Barrierefunktion des Dünndarmepithels, z.B. durch Beeinträchtigung von tight junctions oder Zelltod. In dieser Arbeit wurden Effekte von G. duodenalis auf in vitro Modellsysteme des humanen Dünndarmepithels untersucht. Dazu wurden hauptsächlich Daten über die Barrierefunktion sowohl von der weit verbreiteten Caco-2 Zelllinie, als auch über ein neu etabliertes humanes Dünndarmorganoidsystem, erhoben. Es konnte gezeigt werden, dass mehrere - mitunter in der Literatur als hochvirulent beschriebene - G. duodenalis Isolate zu keinerlei Beeinträchtigung der Barrierefunktion oder irgendeiner anderen untersuchten potenziellen Schädigung an zwei unterschiedlichen Caco-2 Zelllinien unter diversen Infektions- und Kulturbedingungen führte. Jedoch andererseits das neu entwickelte Dünndarmorganoidsystem mit pseudo-luminalem Medium TYI S 33 reproduzierbar die Zerstörung des Epithelmodells mit Zellverlust, Zelltod (apoptotisch und nicht-apoptotisch), Störung der tight junctions (Abbau und Dislokation von Claudinen und ZO-1) und den Verlust von Mikrovilli innerhalb ein bis zwei Tage nach Parasiteninfektion zeigen konnte. Zudem wurde das Auftauchen von ClCa-1-Signalen unter andauerndem Infektionsstress beobachtet, was die Differenzierung bzw. Metaplasie zu Becherzellen nahelegt, jedoch keine Wirtsreaktion auf die Gewebszerstörung zu sein scheint. / The protozoan parasite Giardia duodenalis is the etiological agent for the intestinal diarrheal disease giardiasis. Infections are acquired via the fecal-oral route, mostly via uptake of cysts from contaminated drinking water. The colonization of the hosts’ duodenum and upper jejunum and the attachment of Giardia trophozoites onto the epithelium is the cause of a variety of gastrointestinal complaints but the exact pathomechanisms are unknown. Furthermore, the outcome of Giardia infections varies greatly between individuals, ranging from self-limiting to chronic, and asymptomatic to severe enteritis. One proposed mechanism for the pathogenesis is the breakdown of intestinal barrier function, e.g. by tight junction impairment or induction of cell death. In this work, effects of G. duodenalis on in vitro models of the human small intestinal epithelium were investigated by studying mainly barrier-related properties and changes of widely used Caco-2 cells as well as newly established human small intestinal organoid-derived monolayers (ODMs). It could be shown that several isolates of G. duodenalis, some described as highly virulent, fail to induce barrier dysfunction or any other investigated pathological effect on two Caco-2 cell lines under various infection and culturing conditions. On the other side, by developing a new organoid-based model system and the use of luminal mock medium TYI-S-33, considerable epithelial disruption (including loss of cells), cell death (apoptosis and non-apoptotic), tight junction impairment (degradation and dislocation of claudins and ZO-1), and microvilli depletion reproducibly induced by G. duodenalis trophozoites between one and two days after infection could be observed. Moreover, emergence of ClCa-1 positive cells with ongoing parasite infections suggest epithelial differentiation or metaplasia towards goblet cells, which is furthermore not associated to tissue damage.

Blood-brain barrier (BBB) is a unique membranous barrier, which segregates brain from the circulating blood. It works as a physical and metabolic barrier between the central nervous system (CNS) and periphery. In mammals, endothelial cells were shown to be of BBB and are characterized by the tight junctions along with efflux system which are responsible for the restriction of movement of molecules within the cells. Efflux system consists of multidrug resistance proteins such as P-glycoprotein (P-gp). P-gp removes substances out back from the brain to the blood before they reach to the brain. So the barrier is impermeable to many compounds such as amino acids, ions, small peptides and proteins, making it the most challenging factor for the development of new drugs for targeting CNS. Curcumin is a bioactive compound that has a number of health promoting benefits such as anti-inflammatory, anticancer, anti-oxidant agent; as well as a role in neurodegenerative diseases, but low oral bioavailability is the major limiting factor. Low water solubility and rapid metabolism are the two important factors responsible for poor bioavailability of curcumin. Galaxolide is a musk compound and previously known for the bioaccumulation of toxic components in the aquatic animals by interference with the activity of multidrug/multixenobiotic resistance efflux transporters (MDR/MXR). The bioavailability of curcumin can be enhanced when administered with galaxolide. This study was carried out to investigate the effect of galaxolide on the permeation of curcumin through the epithelial cell monolayers. MDCKII-MDR1 cell monolayer is used an in vitro blood-brain barrier model while Caco-2 monolayer is used as an in vitro intestinal model, which also expresses the P-glycoprotein. The curcumin and galaxolide were separately solubilised in the DMSO and used in combination to perform permeation study, to determine the effect of galaxolide on curcumin permeation through epithelial cell monolayers. The galaxolide shows an efflux protein inhibition activity and this activity was used to enhance permeation of curcumin through the Caco-2 monolayer. In summary, galaxolide is a novel permeation enhancer molecule, which can be used for the improvement of drug delivery of other bioactive compounds in future.

A fundamental understanding of the process of devolatilization requires extensive knowledge of not only the intrinsic properties of the parent coal and its subsequent formed products (tars, gases and chars), but also its characteristic reaction rate behaviour. Devolatilization behaviour has been extensively addressed in literature with the use of powdered coal samples, which normally do not adhere to particle size constraints of coal conversion processes utilizing lump coal. The aim of this investigation was therefore to assess the devolatilization behaviour (with respect to product yield and -quality; and reaction rate modelling) of four typical South African coals (UMZ, INY, G#5 and TSH) confined to the large particle regime. All four coals were found to be bituminous in rank, with vitrinite contents ranging between 24.4 vol.% and 69.2 vol.% (mineral matter free basis). Two were inertinite-rich coals (UMZ and INY) and the other two were vitrinite-rich coals (G#5 and TSH). From thermoplasticity measurements it was evident that only coal TSH displayed extensive thermoplastic behaviour, while a comparison between molecular properties confirmed the higher abundance of poly-condensed aromatic structures (aromaticity of 81%) present in this coal. Product evolution was evaluated under atmospheric conditions in a self-constructed, large particle, fixed-bed reactor, on two particle sizes (5 mm and 20 mm) at two isothermal reactor temperatures (450°C and 750°C) using a combination of both GC and MS techniques for gas species measurement, while standard gravimetric methods were used to quantify tar- and char yield respectively. Elucidation of tar- and char structural features involved the use of both conventional- and advanced analytical techniques. From the results it could be concluded that temperature was the dominating factor controlling product yield- and quality, with significant increases in both volatile- and gas yield observed for an increase in temperature. Tar yields ranged between 3.6 wt.% and 10.1 wt.% and increased in the order UMZ < INY < TSH < G#5, with higher tar yields obtained for coal G#5, being ascribed to larger abundances of vitrinite and liptinite present in this coal. For coal TSH, lower tar yields could mainly be attributed to the higher aromaticity and extensive swelling nature of this coal. Evolved gases were found to be mainly composed of H2, CH4, CO and CO2, low molecular weight olefins and paraffins; and some C4 homologues. Advanced analytical techniques (NMR, SEC, GC-MS, XRD, etc.) revealed the progressive increase of the aromatic nature of both tars and chars with increasing temperature; as well as subsequent differences in tar composition between the different parent coals. In all cases, an increase in devolatilization temperature led to the evolution of larger amounts of aromatic compounds such as alkyl-naphthalenes and PAHs, while significant decreases in the amount of aliphatics and mixed compounds could be observed. From 13C NMR, HRTEM and XRD carbon crystallite results it was clear that an increase in temperature led to the formation of progressively larger, more aromatic and structurally orientated polycondensed carbon structures. Reaction rate studies involved the use of non-isothermal (5-40 K/min) and isothermal (350- 900°C) thermogravimetry of both powdered (-200 μm) and large particle samples (20 mm) in order to assess intrinsic kinetics and large particle rate behaviour, respectively. Evaluation of the intrinsic kinetic parameters of each coal involved the numerical regression of non-isothermal rate data in MATLAB® 7.1.1 according to a pseudo-component modelling philosophy. Modelling results indicated that the intrinsic devolatilization behaviour of each coal could be adequately described by using a total number of eight pseudo-components, while reported activation energies were found to range between 22.3 kJ/mol and 244.3 kJ/mol. Description of the rate of large particle devolatilization involved the evaluation of a novel, comprehensive rate model accounting for derived kinetics, heat and mass transport effects, as well as physical changes due to particle swelling/shrinkage. Evaluation of the proposed model with the aid of the COMSOL Multiphysics 4.3 simulation software provided a suitable fit to the experimental data of all four coals, while simulation studies highlighted the relevant importance of not only the effect of particle size, but also the importance of including terms affecting for heat losses due to particle swelling/shrinkage, transport of volatile products through the porous char structure, heat of reaction and heat of vaporization of water. / Thesis (PhD (Chemical Engineering))--North-West University, Potchefstroom Campus, 2013

A fundamental understanding of the process of devolatilization requires extensive knowledge of not only the intrinsic properties of the parent coal and its subsequent formed products (tars, gases and chars), but also its characteristic reaction rate behaviour. Devolatilization behaviour has been extensively addressed in literature with the use of powdered coal samples, which normally do not adhere to particle size constraints of coal conversion processes utilizing lump coal. The aim of this investigation was therefore to assess the devolatilization behaviour (with respect to product yield and -quality; and reaction rate modelling) of four typical South African coals (UMZ, INY, G#5 and TSH) confined to the large particle regime. All four coals were found to be bituminous in rank, with vitrinite contents ranging between 24.4 vol.% and 69.2 vol.% (mineral matter free basis). Two were inertinite-rich coals (UMZ and INY) and the other two were vitrinite-rich coals (G#5 and TSH). From thermoplasticity measurements it was evident that only coal TSH displayed extensive thermoplastic behaviour, while a comparison between molecular properties confirmed the higher abundance of poly-condensed aromatic structures (aromaticity of 81%) present in this coal. Product evolution was evaluated under atmospheric conditions in a self-constructed, large particle, fixed-bed reactor, on two particle sizes (5 mm and 20 mm) at two isothermal reactor temperatures (450°C and 750°C) using a combination of both GC and MS techniques for gas species measurement, while standard gravimetric methods were used to quantify tar- and char yield respectively. Elucidation of tar- and char structural features involved the use of both conventional- and advanced analytical techniques. From the results it could be concluded that temperature was the dominating factor controlling product yield- and quality, with significant increases in both volatile- and gas yield observed for an increase in temperature. Tar yields ranged between 3.6 wt.% and 10.1 wt.% and increased in the order UMZ < INY < TSH < G#5, with higher tar yields obtained for coal G#5, being ascribed to larger abundances of vitrinite and liptinite present in this coal. For coal TSH, lower tar yields could mainly be attributed to the higher aromaticity and extensive swelling nature of this coal. Evolved gases were found to be mainly composed of H2, CH4, CO and CO2, low molecular weight olefins and paraffins; and some C4 homologues. Advanced analytical techniques (NMR, SEC, GC-MS, XRD, etc.) revealed the progressive increase of the aromatic nature of both tars and chars with increasing temperature; as well as subsequent differences in tar composition between the different parent coals. In all cases, an increase in devolatilization temperature led to the evolution of larger amounts of aromatic compounds such as alkyl-naphthalenes and PAHs, while significant decreases in the amount of aliphatics and mixed compounds could be observed. From 13C NMR, HRTEM and XRD carbon crystallite results it was clear that an increase in temperature led to the formation of progressively larger, more aromatic and structurally orientated polycondensed carbon structures. Reaction rate studies involved the use of non-isothermal (5-40 K/min) and isothermal (350- 900°C) thermogravimetry of both powdered (-200 μm) and large particle samples (20 mm) in order to assess intrinsic kinetics and large particle rate behaviour, respectively. Evaluation of the intrinsic kinetic parameters of each coal involved the numerical regression of non-isothermal rate data in MATLAB® 7.1.1 according to a pseudo-component modelling philosophy. Modelling results indicated that the intrinsic devolatilization behaviour of each coal could be adequately described by using a total number of eight pseudo-components, while reported activation energies were found to range between 22.3 kJ/mol and 244.3 kJ/mol. Description of the rate of large particle devolatilization involved the evaluation of a novel, comprehensive rate model accounting for derived kinetics, heat and mass transport effects, as well as physical changes due to particle swelling/shrinkage. Evaluation of the proposed model with the aid of the COMSOL Multiphysics 4.3 simulation software provided a suitable fit to the experimental data of all four coals, while simulation studies highlighted the relevant importance of not only the effect of particle size, but also the importance of including terms affecting for heat losses due to particle swelling/shrinkage, transport of volatile products through the porous char structure, heat of reaction and heat of vaporization of water. / Thesis (PhD (Chemical Engineering))--North-West University, Potchefstroom Campus, 2013