Quantitative Structure-Retention Relationships for Solid-Phase Extraction on Modified Silica

Quye, Anita (1989) Quantitative Structure-Retention Relationships for Solid-Phase Extraction on Modified Silica. PhD thesis, University of Glasgow.

Abstract

The work presented in this thesis was conducted in the field of analytical chemistry and was intended to improve the process of methodology development in the related fields of analytical toxicology and therapeutic drug monitoring. Most pharmaceutical and toxicological analyses reguire a sample clean-up step before quantitative detection of a drug by chromatography or spectroscopy. Solid phase extraction (SPE) techniques have recently assumed considerable importance for this purpose and many of these procedures utilise chemically modified silica as adsorbents, in which organic substituents have been introduced on to the silica surface. A diverse range of polar and non-polar substituents has been used to provide adsorbents for straight-phase, reversed-phase and ion exchange systems. Bonded silica is used as a stationary phase in high performance liquid chromatography (HPLC) and although much research has been directed towards understanding the solute retention process in HPLC, as yet no attempt has been made to apply this theory to SPE. In this study, the physical and chemical parameters determining the retention of analytes on SPE sorbents were examined to elucidate the underlying mechanisms of the SPE process on different sorbents. A mathematical model of the process could then be constructed to predict the retention of novel analytes based solely on their physical and chemical characteristics. These predictive rules are known as Quantitative Structure-Retention Relationships (QSRR). The substances of interest were B-adrenoreceptor antagonist drugs related to propranolol. In the first three chapters of the thesis, background material is given concerning silica stationary phases in liquid chromatography (LC), their synthesis and properties, the role of the mobile phase in (LC) systems and the interactions which take place between analytes and the mobile and stationary phases. Thereafter, QSRR are reviewed and parameters used to characterise analytes are introduced: these fall into two categories - dispersive parameters (e.g. molecular volume) and inductive parameters (e.g. dipole moment). Finally, particular problems arising in the SPE of basic substances are reviewed. From the physicochemical parameters selected by multiple linear regression analysis, the following conclusions were drawn: (i) The partition coefficient, log P, and the ionisation-corrected partition coefficient, log D, were dominant in the regression equations derived for the substituted benzene solutes. An inductive parameter, the number of hydrogen-bond donor groups, was also significant. With the n-alkyl bonded silica sorbents, solute retention increased with log P or log D. This was a reflection of the hydrophobic contribution to retention by the non-polar sorbent ligands. The hydrogen-bond donor term reflected the ability of the mobile phase components, methanol and water, to decrease retention by hydrogen-bonding to solutes i.e. log k' decreased as the hydrogen-bonding term increases. (ii) Increased hydrophobic retention by octadecylsilica enhanced the hydrogen-bond donor contribution to retention of substituted benzene solutes, and the coefficient of this term was greater than for ethyl- and octyl-silica. (iii) The volume of an acidic benzene solute was another retention-determining parameter on octadecylsilica. Such solutes resided in the most mobile part of the stationary phase where they could be enveloped by the flexible bonded chains if their size was appropriate. No volume parameter was observed in the regression equations for the shorter n-alkyl chains. (iv) Phenylsilica was shape-selective towards substituted benzene solutes as indicated by two additional terms, volume and connectivity. (v) Phenylsilica appeared to undergo a phase transition at 40% methanol in water which increased the hydrophobic surface area of the bonded ligands. (vi) Cyclohexylsilica could not be modelled successfully with substituted benzene solutes. (vii) The B-blocker test compounds were retained indefinitely on octyl-, phenyl- and cyclohexyl-silica through silanol interactions, unless tri-n-butylamine was added to suppress retention. Pretreating the sorbent with either plasma protein solution or fresh plasma also masked silanol behaviour. Use of tri-n-butylamine as well as pretreatment with the biological matrix aided fast elution of the solutes. (viii) The three selected bonded phases for B-blocker probes were all shape-selective, although octylsilica was not a suitable sorbent for modelling the retention behaviour of these solutes as the correlation between log k' and the selected physicochemical parameters was poor. Excellent correlations were achieved with the cyclic sorbents. The retention prediction equations derived for phenyl- and cyclohexyl-silica with the 3-blocker compounds could be used to either predict log k' for a particular chromographic system, or more useful for SPE, a suitable eluent or sorbent could be selected by setting log k' at 1.8 for retention of a solute, and at 1.6 for elution, thereby allowing prediction of suitable systems for method development. (Abstract shortened by ProQuest.).

• Item Type: Thesis (PhD)
• Qualification Level: Doctoral
• Keywords: Analytical chemistry, Forensic anthropology, Toxicology
• Date of Award: 1989
• Depositing User: Enlighten Team
• Unique ID: glathesis:1989-77832
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 14 Jan 2020 11:53
• Last Modified: 14 Jan 2020 11:53
• URI: https://theses.gla.ac.uk/id/eprint/77832