Humic substances (HS) are described as a complex mixture of organic molecules formed by incomplete decomposition of plant, animal and microbial matter. They are found in soil, water and air and have many environmental roles, e.g. water retention and metal ion binding in soil. Despite their importance, the molecular composition of HS is poorly understood. This is mostly because of an inability to separate individual molecules from these complex mixtures and then characterise them by standard analytical methods such as NMR and MS. In order improve the understanding of these important mixtures I have studied them using a high-resolution analytical method, Fourier transform ion-cyclotron resonance mass spectrometry (FTICR MS). Initial efforts focussed on testing the, fast, automated data analysis of the large data sets produced. Two pieces of software were compared and the reliability of the formulae assigned by these was critically evaluated. This confident formula assignment was then applied to study the consequences of different ionisation and instrumental parameters on the mass spectra obtained. The use of laser desorption/ionisation (LDI) without the need to employ a matrix required in matrix assisted laser desorption/ionisation (MALDI) was explored. A comparison of LDI and electrospray ionisation (ESI) FTICR MS of natural organic matter samples showed that these methods ionise complementary sets of compounds. The LDI ionised compounds were characterised as aromatics or condensed aromatics and compounds belonging to lower oxygen classes (maximum number at O8), while ESI ionised higher oxygen classes (maximum number at O16) with a vast majority of compounds classified as aliphatic based on their modified aromaticity index. MALDI and LDI spectra produced very similar data with over 90% matching formulas implying that fragmentation is not caused by LDI, as taught previously. My work showed that to maximize the coverage by FTICR MS of the molecular space occupied by these complex mixtures, multiple ionization methods must be used. As a particularly convenient and readily deployable ionization technique, LDI should be included in standard analytical protocols for FTICR MS analysis of NOM. I have explored different parameters and experimental settings to obtain a fuller coverage of the molecular space of NOM, this showed that different experimental conditions enhance peak intensities in different m/z regions of the FTICR MS spectra and that information can be obtained outside of the narrow 200-700 m/z window. To gain chemical and structural information about humic substances beyond what is currently known, experiments aimed to label HS using different isotopes and at specific sites were developed and tested. Two methylation reactions were of particular interest. A methylation that selectively targeted carboxylic acid groups and incorporated deuterium in the form of CD3 groups. An international standard, Suwannee River fulvic acid, was methylated and analysed by high-resolution mass spectrometry in order to gain information on the number and distribution carboxylic acid groups. This proved challenging due to the reactivity of the unknown molecules being difficult to determine in advance. Additionally, the peak separation being reduced to as low as 1.5 mDa pushed the instrument resolution and assignment confidence to their limits. The second methylation method explored used 13CH3I, a nonselective agent reacting with any labile proton, particularly attaching 13CH3 groups to carboxylic, phenolic and alcoholic OH groups. I prepared a methylated sample of fulvic acid from a Red Moss raised bog (Balerno, near Edinburgh) ready for analyses by high field NMR. This investigation yielded structures of a number of phenolic compounds for the first time by NMR.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:743844 |
| Date | January 2018 |
| Creators | Blackburn, John William Teasdale |
| Contributors | Uhrin, Dusan ; Graham, Margaret |
| Publisher | University of Edinburgh |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Source | http://hdl.handle.net/1842/31149 |
Page generated in 0.0014 seconds