This thesis describes the structural and preliminary electrical characterisation of various “shifted” hexagonal B-site deficient perovskites, of generic formula, A4B3O12. These ceramics are reported to possess promising microwave dielectric properties, which are a requirement for use in the ever evolving mobile telecommunications industry. The 12R crystal structures of the Ba3-xSrxLaNb3O12 series (x = 1 – 3) were refined at variable temperatures using NPD data in space group R-3 and some interesting behaviour was exhibited. All compositions were found to be antiphase tilted, the extent of which being dependent on the A-cation size. Upon heating, the tilt angle for all members of the series was found to decrease, although no tilt transitions from R-3 to R-3m were witnessed for all phases up to 900 °C, and a change in the magnitude of permittivity at 1 MHz was observed for all members except, BaSr2LaNb3O12. Additionally evidence from preliminary low temperature data suggest relaxor ferroelectric type behaviour. Ba2SrLaNb3O12 was found to possess the most promising dielectric properties although room exists for further improvement of the quality factor (Q value). To facilitate further tuning of the dielectric properties, specifically Q, the Ba3- xSrxLaNb3-yTayO12 series (x = 0 - 3, y = 1 - 2) was investigated, with the introduction of Ta5+ on the B-site. Some interesting ordering of A- and B-sites was found to occur, and furthermore, Nb5+ was found to display a preference to inhabit the distorted environment of B(1), adjacent to the “shifted” region, compared to the undistorted B(2) site, in the perovskite block. Unusual microwave dielectric properties were displayed with all compositions exhibiting small and positive temperature coefficient of resonant frequency (TCF) values and significantly small Q values, the reverse of what is expected, thus implying that the B-site has some influence on TCF values in this case. The interesting dielectric properties may be explained by several factors, including, the increased strength associated with the covalency of the Ta-O bond compared to that of Nb-O and the presence of some significant impurity phases, particularly in the intermediate compositions that are richer in Ta5+. The LCR data collected revealed similar results as found for the Ba3-xSrxLaNb3O12 series (x = 1 – 3), highlighting the limiting nature of the temperature range which did not, in fact reflect the true response of the compositions’ behaviour. All of the compositions are antiphase tilted at room temperature therefore the collection of variable temperature diffraction data is warranted. A range of novel Ta-based intergrowth compounds were synthesised and preliminarily characterised. No detailed structural or electrical data were collected for these compounds, however, the ability to form Ta-based intergrowths, in the first instance, has been established. It is postulated that by the formation of these intergrowth phases, the dielectric properties will be effectively tuned in comparison to those of the parent oxides that they are composed of, due to the regularisation of the octahedral layers that occurs upon the formation of such intergrowth structures. This often leads to improved TCF values, and by the introduction of Ta5+, can also give improved Q values. Ta5+ was substituted for Nb5+ in the untilted 5x6y (x = 1 – 3, y = 1) layered compounds, however, it was found that to facilitate this, the replacement of Ba2+ for Sr2+ was also required on the A-site. This double substitution was a requirement to obtain compositions closer to single phase Ta-rich phases compared to Nb-rich compositions.

The objective of this work has been to study the binding of metal ions to complex ligands expressing two or more metal binding sites, in terms of the thermodynamics of the binding events, and to use this information to contribute to the understanding of both the metal binding processes and the properties of the metal/ligand complexes. Two ligands have formed the basis of the study, one a naturally occurring enzyme and the other a newly synthesised ligand with two distinct binding sites (to form supramolecular complexes). The principle technique used to determine thermodynamic parameters has been isothermal titration calorimetry (ITC). The first system studied involved the enzyme β-lactamase II from Bacillus cereus 569/H/9 (BcII). The binding of zinc (II), cobalt (II) and cadmium (II) was investigated in aqueous solution at pH’s from 5.2 to 7.2. By using three or more buffers at each pH it has been possible to determine not only the binding constants of the metal ions at these pH’s but also the molar enthalpies of binding (and hence molar entropies). In addition, the dissociation of hydrogen ions from the enzyme on metal binding was also observed. In the case of all three metal ions, one metal ion per enzyme appears to bind at low pH and two metal ions at high pH (within the pH 5.20-7.20 range). In the case of zinc and cobalt ions two metal ions seem to bind per mole of enzyme in a cooperative mechanism. In contrast two cadmium ions appear to bind to the active sites of the enzyme non-cooperatively. For all three metal ions, models have been developed to describe the two binding sites and the sources of H+ ions that dissociate on metal ion binding. The second system studied involved the supramolecular synthesised ligand bipy-aza-crown-4. This ligand has two metal binding sites, one at the 2,2-bipyridine and the other at the aza-crown-4. The objective was to compare the way copper (II) binds in acetonitrile solution with the way in which the copper (II) complex with this ligand is crystallised from methanol solution with progressively increased copper (II) concentration, as revealed by single crystal x-ray diffraction. Smaller ligands that represent the two individual complexation regions of the synthesised ligand were also used for comparison. Both ITC and solution spectroscopic measurements were used to compare the binding model that was proposed in methanol solution. It was demonstrated in acetonitrile solution that copper (II) binds to the 2,2-bipyridine site first in a 1:2 (Cu2+:L) ratio, then binds to the aza-crown site in a 1:1 (Cu2+:L) ratio . So it was demonstrated that the overall stoichiometric ratio of copper ion to bipy-aza-crown-4 in the fully formed complex is 3:2 (Cu2+: L). Possibly the main conclusion drawn is that the order in which the binding sites are occupied in acetonitrile solution is different to that in which they are occupied in methanol.

Despite the fact that in a general logistics context, value-adding services (VAS) have been acknowledged in various publications, there is a dearth of available empirical works that have particularly investigated the competitive capacity of these services in the formulation of port strategy. Hence, the development of models that are useful to understanding and assessing the capacity of value-adding services in port strategies is necessary to bridge this knowledge gap. In the context of this study, value-adding services are defined as customised and additional services which complement the core or main offerings of a firm. A critical realist approach to research was deployed. Questionnaires were used to survey port users, while in-depth interviews were carried out with port management of the multiple case study ports. The case study ports were Rotterdam (The Netherlands), PD Teesport (United Kingdom), Damietta (Egypt) and Apapa (Nigeria) ports. Ten key value-adding services were identified. The research showed that value-adding services may provide a useful input into the competitive strategy of attracting and retaining port users to a port. The services were found to offer opportunities for the diversification of business for port authorities and port operators. Original contributions of the study include that it brings to the maritime logistics body of knowledge an understanding of the importance of the different value-adding services, principally based on port users‟ assessment. In sequence, the most important value-adding services were transport service, warehousing, water supplies and technical support, whilst the least important services were canteen/catering and advertising support services. Hitherto, there is no published study that has provided insight into the importance port users attach to various value-adding services. In port strategy formulation by the deployment of value-adding services, there is a stage in which dynamism sets in and the uniqueness of the services tend to wear-off, requiring concerted rejuvenation to sustain the potential of value-adding services in competitive strategies. The main challenging issues to port management in the deployment of value-adding services in strategy were found to be concerned with legislation, availability of traffic (cargo and vessel), duration of contracts, adequate land and space availability. The study also developed models, one to stimulate theoretical understanding of port value-adding services, the other to facilitate informed decision-making on the viability of offering value-adding services in ports.

The development of an enzyme biosensor employing a novel functionalised polythiophene matrix is presented. The research upon conducting polymer platforms for biological immobilisation is extensive but by no means exhaustive and therefore this investigation contributes to the field of glucose detection with covalently immobilised glucose oxidase upon novel copolymers of N-succinimido thiophene-3- acetate/3-methylthiophene (STA-MT), trans-3-(3-thienyl) acetic acid/3- methylthiophene (TTA-MT) and N-succinimido trans-3-(3-thienyl) acetate/3- methylthiophene STTA-MT. Polymer characterisation was performed using electrochemical techniques, primarily cyclic voltammetry. The examination of various substituted conducting polymers from the polythiophene family was performed where 3-methylthiophene was selected for copolymerisation due to its low oxidation potential and redox behaviour. Copolymerisation of the novel monomers provided reversible and stable films and succeeded in generating a range of copolymer ratios characterised by cyclic voltammetry with the available binding sites calculated using SEM-EDX. Film morphology and dopant intercalation were investigated, providing supporting evidence for the successful copolymerisation of novel monomers with 3- methylthiophene. Optimisation of the biosensor format was investigated through analysis of film thickness, copolymer ratio and enzyme immobilisation time. A film thickness generated with 50 mC provided a stable film with good response to glucose over a wide range of immobilisation times and was employed for all other biosensor investigations. A 10% functional monomer content provided enhanced current signals upon glucose addition which supported the findings of Kuwahara et al where a 10% thiophene-3-acetic acid content within a thiophene-3-acetic acid/3-methylthiophene copolymer demonstrated enhanced biosensor behaviour. All studies were compared against the thiophene-3-acetic acid and 3-methylthiophene system generated under the optimised conditions of the novel copolymer films. The evaluation of the immobilised glucose oxidase as an amperometric biosensor within a mediated system employing p-benzoquinone was performed generating fast response times within 4 seconds, good stability and excellent current response to glucose between 1.4 and 1.9 mA/cm2 after the addition of 15 mM of glucose. All three novel biosensor systems exhibit good analytical performance demonstrating excellent repeatability typically over 90%, reproducibility of over 80% and shelf-life stability of 85-96% over a seven day analysis and between 50-78% after 1 month.

Tungstated Zirconia (WO3-ZrO2) is a technologically important catalyst; however, there is still a considerable lack of understanding for the presence of different WO3 surface species found in this catalyst which are responsible for its high activity. We report here the application of atomistic simulation techniques based on interatomic potentials to explore the nature of WO3 species on WO3-ZrO2 systems. At first modelling of the pure structure of t-ZrO2 was investigated and the reliability of computed structures was found to agree with earlier theoretical and experimental studies. An investigation of all low miller index pure surfaces of t-ZrO2 revealed the highest stability in the following order {101} > {001} > {111} > {110} > {100}. The adsorption of WO3 at partial monolayer coverages of 20 % and 50 % were then investigated onto the surfaces of t-ZrO2. Conclusions that can be drawn from this work are that the preference WO3 species for a particular surface depended on the type of the surface and temperature considered. The highest favourability to WO3 addition was detected on the {111} surface, where a 50 % monolayer coverage was found to be stable. The {100} and {110} surfaces were favourable to WO3 addition, although this favourability depended on the temperature and amounts of WO3 added. The surfaces {001} and {101} were not favourable to WO3 addition at low temperatures. The addition of WO3 resulted in stabilizing of surfaces which were otherwise unstable in t-ZrO2 such as the {111}, {110} and {100}. The study also confirms the formation of polymerized WO3 layers, dimers, monomers and Zr-WO3 linkages which were detected in earlier experiments. The nature of WO3 species were found to be surface specific, which also depended on the surface area and temperature. The knowledge obtained from this study could be used to design and optimize an efficient catalyst.

This project investigated the structures and physical characteristics of exopolysaccharides (EPSs) secreted by lactic acid bacteria. The structure of a novel exopolysaccharide (EPS) produced by Lactobacillus acidophilus 5e2 has been characterised. Analysis of the anomeric region of the H-NMR showed that the repeating oligosaccharide contained seven monosaccharides. GC-MS showed the structure to consist of D-glucose, D-galactose and D-N-acetyl-glucosamine in a molar ratio of 3:3:1. The linkage analysis showed that there were two terminal, three di-linked and two tri-linked monosaccharides, and in collaboration with data generated from a series of D-NMR experiments, an overall structure was determined. The weight-average molecular weight (Mw) of the EPS secreted by Lactobacillus acidophilus 5e2 when grown in skimmed milk was monitored during extended fermentation times. During the exponential growth phase, the increase in Mw closely followed the increase in yield of EPS. Under the fermentation conditions applied in this study, few if any new polysaccharide chains were formed during this growth phase despite a twenty five-fold increase in the cell count; almost the entire increase in yield can be accounted for by an increase in chain length. These results suggested that synthesis of new EPS chains is switched off during the exponential and stationary phase of fermentation. The increase in yield observed in this period is a consequence of the bacteria's ability to extend existing chains right up to the mid-stationary phase. These results raise questions about the factors that control EPS production and chain length. Depolymerisation techniques have been shown to reduce the Mw of the polysaccharide in a controlled manner. The H-NMR results have shown that the physical methods, constant pressure and ultrasonic disruption break the EPS randomly through the repeating oligosaccharide unit; polydispersity data suggests that the breakages were occurring midchain. A change to the peaks in the anomeric region of the H-NMR spectrum showed that depolymerisation, by acid hydrolysis, was chemically modifying the EPS structure. The approximate intrinsic viscosities of the EPS produced by Lactobacillus acidophilus 5e2 were determined to range between 0.6–2.0 dL g-1 for the Mw range of 1.59x105 – 4.78x105 g mol-1. A capillary zone electrophoresis method was developed to determine the monosaccharide composition of two EPS samples. The method successfully determined D-glucose and Dgalactose, but a peak for D-N-acetyl-glucosamine was not seen. The method was sensitive compared to current techniques, but not as low as using a HP-AEC-PAD. A novel method using LC-MS was developed for the linkage analysis of EPSs. Methylation, hydrolysis and reductive amination were used to derivatise the polysaccharide, and the fragmentation patterns were examined to determine the different linkage positions. Due to undesirable further fragmentation the method could not unequivocally differentiate between the different linkage positions, but the method was capable of resolving the monosaccharides residues with different linkage positions, at approximately the correct relative ratio.

The ditopic ligand, L1, contains both a bipyridine domain and a tetra-azacrown binding domain. Introduction of one equivalent of Cu2+ ions results in co-ordination of the ion by all four of the available N-donor atoms in the azacrown unit. However, introduction of further Cu2+ ions results in the formation of a variety of species depending on the ratio of Cu2+ ions to ditopic ligands available. The bipyridine site is capable of behaving as a bidentate chelator as co-ordination at the bipyridine unit by a further Cu2+ ion allosterically alters the behavior of the azacrown domain. The co-ordination of a Cu2+ ion at the bipyridine unit causes the unit to adopt a planar conformation, preventing all four available N-donor atoms in the tetra-azacrown from co-ordinating an encapsulated copper ion. Therefore coordination of the bipyridine domain changes the tetra-azacrown unit from a tetradentate to a tridentate N-donor unit. A diamino-functionalised cryptate, L2, was synthesized, which is capable of reacting irreversibly with butanal in aqueous media, when in the presence of an excess of metal ions. Co-ordination of the cryptate unit is required in order for the cyclisation to occur, excess metal ions act as Lewis acids in order to promote the reaction, and also co-ordinate the cryptate ion. Co-ordination of the cryptate ion forces the bipyridine unit into a more planar position, allowing the amine groups to move into closer proximity to each other. Subsequent reaction with butanal induces the ligand to form a cyclised bis-aminal complex, as the amino groups react readily with the aldehyde to form a seven-membered bis-aminal species. The cyclised ligand was found to display metal dependent luminescent properties, with zinc ions producing the most intense increase in emissive properties by a significant margin as well as increasing the luminescent lifetime. A potentially hexadentate ligand, L3, was synthesized, which upon co-ordination with dicationic metal ions forms dinuclear double stranded helicate species. The ligand partitions into two separate tridentate binding domains, consisting of two N-donor atoms from a pyridyl-triazole unit and an O-donor atom from a carbonyl oxygen group on the coumarin unit. The ligand contains two coumarin fluorophores, and the luminescent properties of the dinuclear helicate complexes were investigated. It was found that upon co-ordination of Zn2+ the emissive properties were enhanced in comparison with the parent ligand, however introduction of Co2+, Cu2+, Cd2+ and Hg2+ ions induced fluorescent quenching of varying degrees.

The rate constants for the reactions of a variety of nucleophiles reacting with substituted benzyl chlorides in liquid ammonia (LNH3) have been determined. To fully interpret the associated linear free-energy relationships, the ionisation constants of phenols ions in liquid ammonia were obtained using UV spectra. These equilibrium constants are the product of those for ion-pair formation and dissociation to the free ions, which can be separated by evaluating the effect of added ammonium ions. There is a linear relationship between the pKa of phenols in liquid ammonia and those in water of slope 1.68. Aminium ions exist in their unprotonated free base form in liquid ammonia and their ionisation constants could not be determined by NMR. The rates of solvolysis of substituted benzyl chlorides in liquid ammonia at 25 oC show a Hammett ρ of zero, having little or no dependence upon ring substituents, which is in stark contrast with the hydrolysis rates of substituted benzyl halides in water, which vary 107 fold. The rate of substitution of benzyl chloride by substituted phenoxide ions is first order in the concentration of the nucleophile indicative of a SN2 process, and the dependence of the rate constants on the pKa of the phenol in liquid ammonia generates a Brønsted βnuc = 0.40. Contrary to the solvolysis reaction, the reaction of phenoxide ion with 4-substituted benzyl chlorides gives a Hammett ρ = 1.1, excluding the 4-methoxy derivative, which shows the normal positive deviation. The second order rate constants for the substitution of benzyl chlorides by neutral and anionic amines show a single Brønsted βnuc = 0.21 (based on the aqueous pKa of amine), but their dependence on the substituent in substituted benzyl chlorides varies with a Hammett ρ of 0 for neutral amines, similar to that seen for solvolysis, whereas that for amine anions is 0.93, similar to that seen for phenoxide ion. The rates of aromatic nucleophilic substitution reactions in liquid ammonia are much faster than those in protic solvents indicating that liquid ammonia behaves like a typical dipolar aprotic solvent in its solvent effects on organic reactions. Nitrofluorobenzenes (NFBs) readily undergo solvolysis in liquid ammonia and 2-NFB is about 30 times more reactive than the 4-substituted isomer. Oxygen nucleophiles, such as alkoxide and phenoxide ions, readily displace fluorine of 4-NFB in liquid ammonia to give the corresponding substitution product with little or no competing solvolysis product. Using the pKa of the substituted phenols in liquid ammonia, the Brønsted βnuc for the reaction of 4-NFB with para-substituted phenoxides is 0.91, indicative of the removal of most of the negative charge on the oxygen anion and complete bond formation in the transition state and therefore suggests that the decomposition of the Meisenheimer σ-intermediate is rate limiting. The aminolysis of 4-NFB occurs without general base catalysis by the amine and the second order rate constants generate a Brønsted βnuc of 0.36 using either the pKa of aminium ion in acetonitrile or in water, which is also interpreted in terms of rate limiting breakdown of Meisenheimer σ-intermediate. Nitrobenzene and diazene are formed as unusual products from the reaction between sodium azide and 4-NFB which may be due to the initially formed 4-nitroazidobenzene decomposing to give a nitrene intermediate, which may dimerise and be trapped by ammonia to give the unstable hydrazine which then yields nitrobenzene. We have developed a method for the amination of aryl halides in liquid ammonia using copper (I) catalysis which enables direct synthesis of a number of primary amines with excellent yields. This method does not require strong base and ligands as additives and the amination in liquid ammonia has exclusive selectivity for the formation of primary amines, even under relative higher temperature. The amount of catalyst required for the reaction is relatively lower than that generally used, and the convenience of products separation with liquid ammonia as reaction medium indicate its potential industrial application. The preliminary mechanistic investigation indicates that the rate of the amination is first order dependence on the concentration of copper (I) catalyst, and the formation of triamminecopper (I)-aryl ring intermediate is probably the rate limiting step in liquid ammonia. Due to strong coordination of solvent molecules to the copper (I) ion, the kinetics of the reaction are generally insensitive to the addition of other conventional ligands in liquid ammonia. The copper (I) catalysed 1,3-Huisgen cycloaddition reaction of azide and alkynes (CuIAAC) in liquid ammonia requires less catalyst than those in conventionally used solvents. The excellent yield, exclusive selectivity, and most importantly, the ease of separation of the product indicate the potential advantages of using liquid ammonia as the solvent for this reaction. The preliminary mechanistic investigation suggests that CuIAAC reaction in liquid ammonia is a stepwise process with the initial formation of copper (I)-acetylide ion complex, followed by its combination with copper (I) coordinated azide.

The objective of the work studied here was to relate the structure of the conventional Layered Double Hydroxide (LDH), hydrotalcite and several transition metal doped hydrotalcites to their function once calcined into catalysts for use in the production of biodiesel. Attention was paid to the preparation of the LDHs. Three preparative methods were investigated, using sodium hydroxide and carbonate, using ammonia, and using urea as precipitating agents. The properties of the resultant LDHs and those of the mixed oxides produced on calcination were shown to be relatively independent of the synthesis method. The importance of ensuring that sodium salts were removed from the catalyst precursors before use, when using the first synthetic method, was shown. Evidence was provided which showed the significant effect on activity of the calcined catalysts when sodium was present. The importance of thorough washing of the LDH precipitates was clearly demonstrated. The calcination process was also studied and materials were subjected to two methods of calcination, “feedback-controlled” microwave heating and conventional heating in a furnace. This method of controlled microwave calcination may offer promise in the production of optimised mixed metal oxide and other catalysts. A copper-substituted hydrotalcite was subjected to calcination under feedback-controlled microwave heating, in which microwave power is continuously modulated to generate a defined sample temperature programme or constant sample temperature. The results showed that microwave calcination resulted in enhanced crystallinity of the resultant oxides and spinel phase formed at high temperature. In addition, an additional phase, Cu2MgO3, was detected following microwave calcination, which was not formed at any temperature (up to 1000 oC) under conventional heating. The concentrations and strengths of surface basic sites were significantly higher for materials calcined using microwaves than using conventional heating. Catalytic activities in the base-catalysed transesterification of tributyrin with methanol were also higher. Microwave calcination under feedback-control, while allowing control of material bulk temperature during calcination and preventing major 6 temperature excursions, may allow quite large but highly localised temperature variation, for instance as water is released during dehydroxylation, which are beneficial in developing surface defects and surface basicity. Other LDHs were studied incorporating the transition metals, cobalt, nickel and iron. All three showed some activity but basicity appeared to be enhanced by doping hydrotalcite with copper (II) and cobalt (II) particularly. Whether this is because of the presence of these ions on the catalyst surface and their behaviour as Lewis bases, or whether their presence leads to surface defect sites that show electron donating abilities, is not clear. The effect of microwave calcination on these other substituted LDHs seems to be variable and not always as pronounced as it is with the copper containing LDH. It seems likely that the extent to which microwaves are effective depends on their capacity to couple with the metals in the structures and the fact that this varies between metals perhaps explains why the different LDHs show different behaviours. Although not the most active material, perhaps the most interesting material formed upon calcination was the mixed oxides of Mg5.82Al1.12Fe0.88(CO3)(OH)16.4H2O, possessing both acidic and basic sites. All other LDHs studied possessed basic sites only. This could be very useful for the production of biodiesel from waste oil containing free fatty acids (FFAs) which require the presence of an acid catalyst for pre-esterification of these free acids.

The ZSM-5 zeolites with varying silica to alumina ratios and with both hydrogen and sodium counter ions (Z1000H:SiO2/Al2O3 = 1000, Z900Na:SiO2/Al2O3 = 900, Z25H:SiO2/Al2O3 = 25 and Z25Na:SiO2/Al2O3 = 25) and γ-alumina have been selected as catalysts. The first investigation was initiated to study the mechanisms of catalytic ozonation on zeolites and alumina. The formation of reactive oxygen species (ROS) such as hydroxyl radicals (oOH), hydrogen peroxide (H2O2) and superoxide ion radical (oO2-) have been investigated by using coumarin (COU), amplex red and 4-chloro-7-nitrobenzo-2-oxa-1,3-dizole (NBD-Cl) as probes respectively. The effects of hydroxyl radical scavenger and phosphates have also been studied to investigate the mechanism. The results show that alumina catalyses radical pathways involving ROS, showing its highest activity at pH close to the point of zero charge. However, zeolites do not promote the formation of ROS. The presence of phosphates and t-butanol (TBA) significantly reduces the formation of ROS in the case of alumina. However, in the case of zeolites TBA and phosphates do not have a significant effect on ROS formation. The second investigation involved the study of the efficiency of catalysts to remove organic contaminants. The ibuprofen and volatile organic chemicals (VOCs) such as cumene, 1,2-dichlorobenzene and 1,2,4-trichlorobenzene have been selected as target pollutants. The results show that within the family of zeolites, silica to alumina ratio is important for the adsorption of pollutants and for catalytic efficiency of zeolites. Therefore, Z1000H and Z900Na have been found to effectively catalyse the removal of VOCS and Z25H and Z25Na were the better catalysts for the removal of ibuprofen in its ionized form. The alumina was found to be ineffective for the removal of VOCs. However, alumina effectively removes ibuprofen. This is because of high adsorption of ibuprofen on alumina. Therefore, it is hypothesized that zeolites operate through a simple mechanism involving the direct reaction of adsorbed species on their surfaces; their activity depends upon their silica to alumina ratios and is insensitive to the nature of counter ions. The alumina operates through a radical mechanism involving the formation of ROS. Furthermore, the adsorption of pollutants plays an important role in the catalytic ozonation process.