L’énantiomérisme est un sujet majeur dans des domaines divers, en particulier celui de la pharmacie (principe actifs et métabolites). La commercialisation de médicaments sous forme racémique a longtemps été privilégiée au dépend d’éventuels effets secondaires. Depuis les années 90, le développement du nombre d’entités énantiomériquement pures est en hausse, notamment grâce au développement des techniques de séparation. La chromatographie chirale, à l’aide de support énantiosélectif, s’est très vite imposée pour la résolution rapide et directe de stéréoisomères. Cependant, la faible compréhension des mécanismes intervenant dans la reconnaissance chirale limite un choix rationnel reposant sur la structure des composés chiraux, conduisant à une procédure de développement de méthode de séparation chirale systématique, fastidieuse et coûteuse d’essais et erreurs. Ce projet de recherche se situe à l’interface des sciences analytiques et de la chémo-informatique. Il consiste en une amélioration de nos connaissances sur le mécanisme de reconnaissance chirale afin d’aborder le développement de méthode chirale de manière plus rigoureuse. / The subject of enantiomer separation is a major issue in various fields, particularly pharmaceuticals (metabolites and active principle). To prepare pharmaceutical formulations, the racemic form has long been favored at the expense of possible side effects. Since the 90s, the development of the number of enantiomerically pure entities is rising, particularly through the development of separation techniques. Chiral chromatography using enantioselective stationary phases is an excellent technique for fast and direct resolution of stereoisomers. However, limited understanding of the mechanism leads to tedious and expensive trial-and-error systematic chiral method development. No clear guideline for choosing a chromatographic system is available for a new chiral drug. In this project, we wish to achieve a better knowledge of enantioselective separation techniques in order to help in the choice of separation method that will be the most appropriate for any given chiral separation. This project is based on the rationalization of large amounts of experimental data with the help of modelling and chemometric techniques to unravel the enantioselective recognition mechanism.

Le développement de molécules chirales est devenu un enjeu majeur pour l’industrie pharmaceutique. Lors de la mise au point d’un nouveau principe actif, la chromatographie en phase supercritique (CPS) sur phases stationnaires énantiosélectives, est devenue une méthode incontournable pour l’analyse ou la purification d’énantiomères. Il n’en demeure pas moins que certains composés tels que les composés basiques restent délicats à analyser par cette technique et que l’ajout d’un additif basique dans la phase mobile est très souvent nécessaire. La quantité d’additif ajoutée dans la phase mobile est souvent choisie de manière arbitraire en se basant sur les spécifications des fournisseurs de phases stationnaires et sur les travaux publiés dans le domaine, sans se soucier de l’effet potentiel sur la résolution d’une modification de sa concentration. Ce travail de thèse a démontré l’importance de la concentration en additif dans la phase mobile lors de l’analyse de mélanges d’énantiomères par CPS, avec notamment des effets significatifs sur l’énantiosélectivité allant jusqu’à une inversion de leur ordre d’élution selon la concentration en additif. Dans un deuxième temps cette étude a consisté en une amélioration de notre compréhension des modes d’actions de l’additif, sur la phase stationnaire et sur la phase mobile permettant de poser des hypothèses sur les effets observés lors de la séparation d’énantiomères. La dernière partie est consacrée à la présentation de différents exemples, justifiant l’intérêt de moduler la concentration en additif lors de séparations d’énantiomères. / Nowadays, the development of chiral molecule is a strategic issue for the pharmaceutical industry. During the development of a new molecular entity, supercritical fluid chromatography using enantioselective stationary phases is essential for the analysis or the purification of racemates. Despite the advantages of SFC, the analysis of some compounds like basic compounds may still be a challenge. Their analysis most of the time, requires an additive in the mobile phase. In many cases additives concentration is chosen arbitrarily, based on the columns supplier guidelines and work already published, regardless of the impact of the additive concentration on the resolution. This work has shown the impact of the additives concentration for the analysis of enantiomers by SFC, in particular a significative impact on the enantioselectivity of chiral stationary phases, with for some of them a reversal elution order depending on the additive concentration. The second objective is to improve our knowledge on the mechanism of action of the additive both on the stationary phase and the mobile phase. The last part is dedicated to the presentation of a few examples highlighting the positive impact of the variation of the additives concentration for enantiomer separation.

No description available.

No description available.

Addition of a small amount of polar solvent (i.e. modifier) which contains an ionic component (i.e. additive) to a CO2 mobile phase has shown major improvement in the elution of ionic analytes via packed column supercritical fluid chromatography (SFC). Firstly, we focused on the elution of sodium arylsulfonate analytes by using various ionic additives, such as lithium acetate, ammonium acetate, tetramethylammonium acetate, tetrabutylammonium acetate, and ammonium chloride. The analytes were successfully eluted with all additives with good peak shape under isocratic/isobaric/isothermal conditions. Three stationary phases with different degrees of deactivation were considered. They were conventional Cyanopropyl, Deltabond Cyanopropyl, and non-chemically bonded silica. The effect of additive concentration and additive functionality on retention was also investigated. Secondly, solid state NMR of the silica packing material before and after being flushed with supercritical CO2 modified by methanol containing the ionic additives was performed to gain some insight into the retention mechanism(s). A fraction of silanol protons were undetected after being treated with the mobile phase which suggested replacement by the cationic component of the additive. CaChe calculations were carried out on several of the additives in an attempt to explain why different ionic additives produce different effects on chromatographic retention. Modification of the stationary phase and ion pairing with the analyte are two possible retention mechanisms being considered. As ion-pair formation was considered to be one of the retention mechanisms, the use of sodium sulfonates as mobile phase additives to elute secondary and quaternary ammonium salts was then studied. Propranolol HCl, benzyltrimethylammonium chloride, and cetylpyridium chloride were chosen as the probe analytes. Sodium ethansulfonate, sodium 1-heptanesulfonate, and sodium 1-decanesulfonate were studied as mobile phase additives. The analytes were successfully eluted from Deltabond Cyano phase within 5 minutes, but were retained strongly without additive or with ammonium acetate as the additive. An Ethylpyridine column showed dramatic advantages on the elution of these ammonium analytes. No additive was required to elute these ionic compounds. Protonation of some fraction of the pyridine functional groups and the deactivation of active silanol sites were believed to be the major mechanisms responsible for this behavior. Lastly, we successfully eluted large peptides (up to 40 mers) containing a variety of acidic and basic residues in SFC. We used trifluoroacetic acid as additive in a CO2/methanol mobile phase to suppress deprotonation of peptide carboxylic acid groups and to protonate peptide amino groups. The Ethylpyridine column was used for the majority of this work. The relatively simple mobile phase was compatible with mass spectrometric (MS) detection. To our knowledge, this is the first report of the elution of peptides of this size with a simple, MS-compatible mobile phase. Fast analysis speed, the possibility of coupling multiple columns to achieve desired resolution, a normal-phase retention mechanism, and less use of organic solvents are the advantages of SFC approach for peptide separation. / Ph. D.

The need for sensitive and selective detectors in supercritical fluid chromatography (SFC) is particularly evident since SFC can be used to analyze classes of compounds that are not readily amenable to either gas chromatography (GC) or liquid chromatography (LC). These compounds include species that are nonvolatile or thermally labile and , in addition, contain no chromophore that can be used for spectrophoto detection. The objective of this research is therefore to interface selective chemilumninescent detectors with SFC in the sensitive detection of nitrogen- and/or sulfur containing compounds. The chemiluminecent nitrogen detector (CLND), a gas-phase detector which is specific for nitrogen-containing compounds, was first evaluated as a detector for use with capillary SFC. The potential use of the CLND for food flavor and petroleum samples was demonstrated. In addition to equimolar nitrogen response, the CLND showed good sensitivity and large linear dynamic range. Minimum detectable quantity (MDQ) was 60 pg of nitrogen with a linear range of over 3 orders of magnitude. Nitrogen to carbon selectivity of 105 was obtained. Capillary SFC with simultaneous flame ionization and chemiluminescent detection was also demonstrated. The second portion of the research investigated the CLND for packed column SFC with methanol modified CO2. The only modification made in the CLND for packed column SFC is the pyrolysis furnace. The CLND and UV were used to interface with SFC via a post-column split. Methanol-modified CO2 was also demonstrated to be compatible with the CLND even with a high mobile phase flow rate. The use of pressure and modifier programs appears to be feasible as is evidenced by the baseline studies which have been performed, as well as by the applications demonstrated. The last portion of the research focused on the evaluation of a new generation sulfur chemiluminescent detector (SCLD), which is also a gas-phase detector, with packed column SFC using both pure and methanol modified CO2. The minimum detectable quantities were determined to be 2.6 pg or 14 pg sulfur for mobile phase employing pure CO2 or 8% methanol modified CO2 respectively. The evaluation study also showed excellent selectivity and linearity, as well as day-to-day repeatability. The capabilities of the SFC-SCLD system for sulfur speciation and detection of thermally labile pesticides and polar sulfonamides, as well as petrochemical samples were presented. / Ph. D.

This work examines the use of supercritical fluid extraction (SFE)as a sample introduction technique for supercritical fluid chromatography (SFC) and Fourier transform infrared spectrometry (FT-IR). In order to study the effects of a supercritical mobile phase on the resulting IR spectra, carbon dioxide was compared to xenon as a supercritical fluid mobile phase for flow cell SFC/FT-IR. A packed capillary column (30 cm x 320 μm, Deltabond® Cyano, 5 μm particle) was employed for the chromatographic portion of the study. Various samples were tested, such as an ethoxylated alcohol mixture and a unique polarity mix. The roles of temperature and density on the IR spectra were also examined as each was independently varied for the analysis of five different probe compounds. Comparisons of spectra produced in supercritical CO₂ and xenon each matched well with the Nicolet vapor phase library. CO₂ spectra matched the vapor phase spectra equally as well as did the xenon spectra, despite the fact that CO₂ spectra have blanked regions where the mobile phase absorbs in the IR. Following this study, SFE was coupled directly to FT-IR to produce the novel method of on-line SFE/FT-IR. This technique was optimized for the analysis of n-tetracosane, yielding a detection limit of 74 ng. The method was later applied to the quantitative and qualitative analysis of fiber finishes from textile matrices with equivalent success. Finish was extracted directly from the fiber or textile surface and passed through the IR flow cell as an analyte "plug". In this way, the entire finish was quickly quantified, and if desired, qualitatively analyzed as well, without need of prior chromatographic separation. The method required no organic solvent and was proven to be fairly reproducible for four fiber finish types tested. Lastly, supercritica1 fluid extraction (SFE) was examined and optimized as a sample introduction technique for on-line SFE/SFC. Trapping and recovery of analytes in SFE/SFC were studied under varying conditions using a currently marketed system. The system was replumbed using a 12 cm x 100 μm fused silica capillary (100% methyl, d_f = 0.25 μm) for trapping and increased solute focusing. These changes nearly doubled the total analyte recovery (as based on FID peak areas) and lowered the overall system recovery RSDs from 30 % to 4 %. / Ph. D.

The introduction of supercritical CO₂ at packed column supercritical fluid chromatography (SFC) flow rates is shown to significantly affect the excitation characteristics of a helium microwave induced plasma (He MIP). In this work, the influence of CO₂ on specific atomic and ionic transitions of Cl, Br, I, P, and S is described. Also presented is the determined relationship between transition energy and degree of signal depression resulting from the introduction of CO₂ to the plasma. Attempts to enhance the emission signals of non-metals by introducing H₂ into a He MIP are discussed. The inadequacy of excitation temperature, ionization temperature, rotational temperature, and electron number density measurements to determine the effect of CO₂ on the excitation characteristics of a He MIP is also described. However, application of kinetic theory and a recently developed theory on charge transfer allows a reasonable series of mechanisms to be developed that describe the excitation processes of a He MIP to which supercritical CO₂ is added. The remainder of this work describes a direct interface between a supercritical fluid extraction (SFE) vessel and an Ar MIP for the purpose of element selective detection. The response of the plasma to the introduction of sample via SFE at a variety of extraction temperatures, pressures, and sample concentrations is presented. / Ph. D.

The use of modifiers in supercritical fluid chromatography and extraction has become quite common due to the inability of pure carbon dioxide alone to solvate many of the compounds of interest. The effects of modifiers in supercritical fluid chromatography have been more thoroughly studied than the effects of modifier in supercritical fluid extraction. The effects of modifier on trapping efficiencies for off-line supercritical fluid extraction have been evaluated in this work. Sorbent and solid phase traps were investigated with pure carbon dioxide in order to determine the effect of stationary phase identity, pretreatment, and rinse solvent on the recoveries of a test mixture of compounds of varying vapor pressure and molecular weight. The solid phase traps, which were polyethylene frits, performed as well as the sorbent traps in most cases, and significantly better than the sorbent traps in many cases. The ability to cool these traps to -20°C allowed for efficient trapping of volatile compounds without the benefit of sorptive interactions. Sorbent and solid phase traps were then studied with the addition of 1%, 2%, 4%, and 8% methanol to the mobile phase. The sorbent trap explored consisted of 40 µm ODS packing material, while the solid phase trap consisted of 100 yum stainless steel beads. In this work trap temperatures ranged from 5-80°C. It was found that trap temperature, modifier concentration, and trap type influenced recoveries of the test mixture components. Applications of these solid phase and sorbent traps explored were the extraction of polychlorinated biphenyls from river sediment and the extraction of the active components from a drug formulation. The separation of some compounds of pharmaceutical interest was also explored, where the addition of modifier, and in some cases an additive, was required to elute compounds from the chromatographic column. / Ph. D.

Supercritical fluid (SF) technologies are being investigated extensively by the food industry for a variety of applications. Carbon dioxide in the supercritical state is of particular interest to the food industry due to its extremely low toxicity in comparison with organic solvents. Three applications of chromatography or extraction employing SFs in the analysis of food components and natural products were investigated. These applications related to carbohydrate, lipid, and coumarin derivative analyses. The peracetylated nitrogen derivatives of carbohydrates were analyzed by supercritical fluid chromatography (SFC) with flame ionization detection (FID), Fourier-transform infrared (FT-IR) detection, and mass spectrometry (MS) detection. Although reports in the literature indicated that only one derivatized component was detected under GC conditions, the SFC analysis of peracetylated aldonitrile derivatives of monosaccharides resulted in detection of multiple reaction products. The identification of the peracetylated nitrile and acyclic oxime was accomplished using both SFC/FT-IR and SFC/MS. The spectroscopic data indicated that an additional reaction product was a peracetylated cyclic oxime. The data was not conclusive enough to determine if the structure contained a pyranose or furanose ring. Changes in chemical composition of vegetable oils due to processing were observed employing SFC/FT-IR. Refined soybean oil and soybean oil that was hydrogenated with a NI or a Ni-S catalyst were analyzed. SFC/FT-IR provided an opportunity to determine the extent of unsaturation and isomerization in a single analysis. The C-H deformation of trans R₁HC=CHR₂ groups was observed at 972 cm⁻¹. The double bond in the cis configuration was evident by absorptions above 3000 cm⁻¹l. On-line FT-IR spectra of triacylglycerols and free fatty acids readily revealed conversion to trans isomers in those samples that were hydrogenated with the Ni-S. Quantitation was not possible due to the incomplete resolution of the cis and trans isomers present in the hydrogenated samples. Apparent threshold densities were determined for a series of coumarin derivatives using a supercritical fluid chromatograph with a flame ionization detector. The extraction cell was a LC stainless steel precolumn. Milligram quantities of the model compounds were extracted. Functionality, extraction temperature, and matrix affected the measured threshold densities of the compounds. The addition of two hydroxyl groups onto the coumarin structure caused the compound to be unextractable at 60°C and densities up to 0.90 g/mL. For those compounds that were completely extracted, coumarin, 7-methoxycoumarin, 3-phenylcoumarin, and psoralen, the apparent threshold density increased with increasing melting point. The melting or decomposition points of those compounds that were partially extracted ranged between 200 and 240°C. For those compounds that were not extracted, under the constraints of the experiment, the melting or decomposition points were greater than 240°C. A higher extraction temperature resulted in a lower threshold density. Corn starch was found to be a noninteractive matrix. / Ph. D.