The analysis of mixtures of all stable C1 - C4 hydrocarbons has been studied by the gas-liquid (GLC), gas-solid (GSC), and gas-liquid-solid (GLSC) chromatographic techniques. A complete separation of all readily available compounds in the C1 - C4 range has been achieved by GLSC which has, so far, proved to be the only technique capable of solving this analytical problem in a practicable fashion. New packings for GSC and GLSC columns have been developed; in particular, a novel technique of dilution of active column packing with inert solid has provided excellent and rapid separation of the hydrocarbons and offers prospects of wider application. The dramatic effect of exposure of solid adsorbent to volatile liquids has been revealed by chromatographic analysis. The origins of this phenomenon have been sought and a variety of different physico-chemical techniques has been employed. The study of this phenomenon has provided evidence of an adsorptive mechanism not previously well documented. The effect revealed here may account for many of the reports in the literature of chromatographic irreproducibility of solid adsorbents. Further, the results indicate that such irreproducibility may be overcome. In consequence, it now seems likely that GSC and GLSC may find wider application than in the past.

Chapter one introduces chromatography and mass spectrometry used to conduct experiments in this thesis. An explanation in the importance and ways of implementing the interfacing of high-pressure liquid chromatography and supercritical fluid chromatography with mass spectrometry is briefly summarized. The feasibility of high-pressure liquid chromatography and supercritical fluid chromatography mass spectrometry for the analysis of β-blockers is shown in chapter two. Methods were developed for the separation and analysis of various β-blockers using SFC and HPLC coupled to MS for specificity. The expected advantages of using SFC in relation to HPLC for analysis were not met, resulting in favour of the faster HPLC method. Nevertheless, SFC is complementary to HPLC as it offers advantages in rapidity of method development, reduce cost of purchasing and disposal of solvents. Chapter three demonstrates the investigation into the fading products of Azo dyes using LC-MS. The separation and analysis of standard azo dyes using LC API MS was implemented to observe the behaviour of dyes under these set conditions. The importance in characterisation of impurities at low levels is critical from a customer safety and FDA regulations aspect. Chapter four demonstrates the analysis of two unknown degradation products of methyl-5-aminolevulinate as being related to the degradation of methyl-5-aminolevulinate using high-pressure liquid chromatography mass spectrometry. Finally Chapter five describes the investigation of high-pressure liquid chromatography and supercritical fluid chromatography mass spectrometry in the characterisation of an unknown penicillin G impurity present in batch samples. The impurity in question co-eluted with the main bulk drug and could only be analysed using ion-pair agents.

The first section of the thesis examines the problem of saccharide separation by H.P.L.C. with special reference to the speed as well as the ease of separation. Previous work is extensively reviewed and several types of common systems are studied with a view to elucidating the underlying problems of sugar separation. Inherent system inefficiency is identified as being responsible for the poor results obtained by earlier workers and confirmed here. An optimisation programme was developed to identify the best conditions for the best systems examined. Although the separations achieved are not comparable with those attainable with other sample types, they represent a substantial improvement so far as sugar analyses are concerned. The second section deals with developments in mixed-solvent theory for liquid chromatography. The theories developed over the past two decades are critically reviewed and tested with a variety of ternary systems over the complete range of binary solvent composition. It was found that the retention behaviour of the majority of systems could not be adequately explained by the theories so far developed. A comparison of the models showed the Langmuir/diachoric models which relate inverse corrected retention volume of the solute to volume fraction of the mobile phase to be the most suitable and further modifications to this model are made to encompass all of the systems studied. A general, semi-empirical equation has been found, that is capable of describing any of the six separate types-of inverse retention-volume fraction plot identified in this work. This is the first occasion on which the range of types of plot occurring has been established and also, the first, on which a satisfactory general equation has been identified.

The studies described in this thesis have explored the scope and limitations of using thermospray LC-MS as a routine analytical technique for providing the quantitative and qualitative data necessary to support the drug development process. Repeller assisted discharge ionisation, also known as plasmaspray, has also been investigated as an adjunct to thermospray for structural analysis of both Phase 1 and Phase 2 drug metabolites. There are four main subject areas encompassed by these studies namely drug metabolism, pharmacokinetics, high performance liquid chromatography and combined high performance liquid chromatography-mass spectrometry. Optimisation of the numerous system parameters is a prerequisite for all analyses and the extent to which the mobile phase composition, vaporiser temperature and repeller voltage affect the sensitivity of detection has been investigated using model compounds. The absolute responses obtained were compound dependent but all of the compounds tested demonstrated best sensitivity when the proportion of methanol in the mobile phase was high, generally greater than 60%. The repeller voltage was also found to have considerable influence on the intensity of response for all compounds, whilst the concentration of ammonium acetate in the mobile phase and the vaporiser temperature were less critical. These findings have led to the adoption of a modified approach to LC-MS in which the chromatography is performed at reduced flow rates, typically 0.5 ml min<SUP>-1</SUP>, and the column effluent is then enriched with methanol containing ammonium acetate to give a total flow compatible with optimum performance of the thermospray interface. This procedure has been found to be suitable for quantitative and qualitative analyses. The adoption of post-column addition of methanol has enabled the use of lower operating temperatures on the thermospray interface which has in turn led to a reduction in the extent to which labile compounds undergo thermal decomposition. This has proven to be a simple yet successful approach, illustrated by the identification of two unusual conjugates, a sulphamate and a carbamoyl-glucuronide, of GR50360 and a glutathione conjugate of GR55721. The analysis of these metabolites had proven problematic prior to the use of the post-column addition of methanol.

The advantages of packed column supercritical fluid chromatography (SFC) over LC in terms of speed of analysis, increased efficiency and speed of method development have been shown to be of particular use for chiral separations of β-Blockers, where UV detection was employed. Interfaces that enabled SFC-MS to be effected were initially reviewed to enable the development of an interface to couple an HP G1205 A supercritical fluid chromatograph to the APCI source of the Finnigan TSQ 700 and LCQ ion trap. Effectively, SFC-MS has been effected with a pre-expansion split and a post-expansion interface thus allowing a portion of the eluent to enter the ionisation region of the mass spectrometer from the supercritical fluid chromatograph. For the work presented, approximately twenty five per cent of the eluent entered the ionisation source. Later, SFC was coupled with the ES ionisation source of the LCQ ion trap. The interfaces developed offer enhanced specificity of identification, with improved detection limits over UV and allowed a comparison of effecting SFC-API-MS with the two mass spectrometers. Being able to facilitate SFC-ES-MS and SFC-APCI-MS with an LCQ ion trap allowed a comparison of the ionisation techniques for the analytes of interest. The differences in effecting SFC-API-MS-MS with the LCQ ion trap and the SQ 700 were illustrated. Anti cancer agents are of great interest to the pharmaceutical industry. Two of synthetically produced groups of lignans were analysed using SFC-APCI-MS<SUP>n</SUP> and SFC-ES-MS<SUP>n</SUP>. Acceptable resolution of the enantiomers and low detection limits are shown, thus showing SFC-API-MS to be a potential tool to help solve chiral analysis problems within industry. Investigations in the use of a Shandon hypercarb column to effect packed column SFC-API-MS have also been explored. Packed column SFC-API-MS<SUP>n</SUP> is shown for benzodiazepines, sulphonamides and carbamate pesticides. Fast analysis times, acceptable resolution and low detection limits for the three classes of compounds are shown.

Chromatographic retention due to adsorption at the liquid/gas and liquid/solid interfaces is possible in gas-liquid chromatographic (GLC) systems, especially with polar solutes from a non-polar stationary phase. The variation of retention volumes with sample size at 60<SUP>o</SUP>C on several series of columns, using different support materials, was studied for di iso-propyl ether (DIPE), a polar solute, on squalane, a non-polar stationary phase. Isolation of the total adsorption contributions to retention was then performed using a semi-empirical curve fitting procedure devised by Mathiasson and Jonsson. The adsorption retentions at infinite dilution on a fully wetted porous silica support (Porasil F) was extrapolated to zero loading where the gas/liquid interfacial area (A<SUB>x</SUB>) approaches the value of the support surface area (A<SUB>s</SUB>). A<SUB>s</SUB> was measured using the BET nitrogen adsorption method, and the sum of the support/liquid and vapour/liquid adsorption coefficients was obtained. The solid/liquid interfacial adsorption coefficient, K<SUB>s</SUB>, was estimated from adsorption liquid chromatographic parameters, and was found to be small in comparison with the gas/liquid interfacial adsorption coefficient, K<SUB>x</SUB>. The variation with liquid loading of the adsorption contribution to retention on two silanised supports (Chromosorb P-AW DMCS and HMDS treated Porasil F) was studied. The gas liquid interfacial area at high loading was found to be small in comparison with that of untreated Porasil F, and comparable with values obtained by other workers. At squalane loadings where a wetting transition is postulated on the silanised support, the magnitude of A<SUB>x</SUB> approaches that of the unsilanised support, and helps confirm the model for the wetting transition proposed by Conder and coworkers.

Capillary Electrochromatography (CEC) combines the specificity of high-pressure liquid chromatography (HPLC) with the efficacy of capillary electrophoresis (CE). CEC utilises electrically driven (electroosmotic) flow to drive the mobile phase through a narrow i.d. (typically 50-100 mm) fused silica capillary packed with stationary phase material. Separation of analytes occurs by partitioning between the stationary and mobile phases and by differences in electrophoretic mobilities (for charged molecules). Using electroosmotic flow rather than pressure-driven flow leads to greater separation efficiencies and less band broadening, due to the uniform flow velocity profile produced, as opposed to the parabolic flow profile found with LC. Furthermore, CEC is an ideal chromatographic method to couple with mass spectrometric detection, due to their compatible flow rates. Mass spectrometric detection allows for sensitive analysis and provides substantial structural information. This study commenced with work to understand the theory and practice of offline CEC coupled to UV-Vis detection, using mixtures of corticosteroids as test molecules. Different types of column packings and buffer types were investigated to find optimum operating conditions for CEC. Following this, research into CEC/MS was carried out, firstly using a CEC/MS probe already available, before a CEC inlet, microscopy-MS and nanospray-MS interfaces were designed and constructed. It was important when designing these CEC/MS interfaces to ensure that chromatographic integrity was maintained, to preserve peak efficiency into the mass spectrometer. These CEC/MS devices were evaluated on several mass spectrometers to investigate the limits of detection possible. In addition, offline nanospray, CEC/ESI-MS and automated CEC/MS were carried out to offer a comparison. A number of applications for the new technology were then investigated. Sulfonamide veterinary products were analysed, both as standards in water and then extracted from spiked milk samples. Pharmaceutical products and carbamate pesticides were also tested, the latter extracted from spiked river water samples. The limited sample volumes obtained from a combinatorial library, an ideal application for CEC/MS, were also analysed, along with some biological compounds.

The theory of liquid adsorption chromatography is reviewed. Preliminary experiments involving the eluting of aromatic alcohols with mixtures of petroleum spirit (bp 30-40°C) and ethanol from silica were carried out and a new approach to the theory is developed. This is based on a Langmuir competitive adsorption model and leads to an equation linearly relating inverse net retention volume of the solute to volume fraction composition of the mobile phase. The theory was tested with experiments on systems in which solvent composition was varied over the entire volume fraction range. Results are included also for elution of a series of ketones by mixtures of hydrocarbons and tetrachloromethane from a reverse phase silica. Good agreement was obtained between the partition coefficients calculated from the results, predicted by the model and values deriving from direct partition experiments based upon gas chromatographic analyses. Further modifications to the theoretical model were then introduced to cover the characteristic behaviour of all the systems studied.

In this thesis, the initial use of API-MS in CCC was demonstrated with the analysis of polar herbicides by CCC/ESI-MS including MS/MS fragmentation by collision-induced dissociation (CID). This study highlighted the traditional problems commonly associated with CCC/MS methodologies such as high backpressure and CCC coil failure. Six benzodiazepines were analysed by CCC/ESI-MS performed on two countercurrent chromatographs, demonstrating the advantages of small coil volumes in rapid CCC analysis. It was determined that for these compounds, modification of the mobile phase stream by addition of a dilute organic acid greatly increased signal responses under CCC/ESI-MS. Comparative studies were performed on LC/ESI-MS and compound structural information was successfully collected from both techniques. The development of a new solvent system consisting of butyronitrile:acetonitrile:water was demonstrated for the analysis of hydrophilic compounds by CCC and CCC/MS, as an alternative to the n-butanol:water based solvent systems. The successful employment of this system was shown in both reverse and normal phase CCC modes, in the study of stimulants and barbiturates by CCC/MS and LC/MS. Both ESI and APCI ionisation methods were employed in these studies, and this thesis presents the first known applications of APCI in CCC. The newly developed butyronitrile:acetonitrile:water solvent system was then further employed in the study of zwitterionic sulfobetaine surfactants by CCC/ESI-MS. The application of a small coil volume CCC instrument and the butyronitrile:acetonitrile:water based solvent systems was able to provide convenient analysis times of less than one hour for very hydrophilic substances, with the addition of mass spectrometry as a highly selective detection method.

No description available.