Project topics and materials
Get Complete Project Material File(s) Now! »
Chapter 3 A Nuclear Magnetic Resonance study: Implications for coal formation in the Witbank Coalfield, South Africa
ABSTRACT
In this study, carbon-13 cross-polarization magic-angle-spinning solid-state nuclear magnetic resonance (13C CP-MAS SS NMR) was undertaken on a parent coal sample and its density-fractionated derivatives, an inertinite-rich and a vitrinite-rich sample, obtained from the Witbank Coalfield of South Africa (medium rank C bituminous coal). The formation of inertinite macerals has been extensively researched and the conclusions remain largely controversial. However, most research has been confined to northern hemisphere coals, which are typically dominated by vitrinite. South African coals are widely known for their high inertinite content. Earlier workers have ascribed the inertinite macerals, in their diversity, to aerial oxidation. However, this oxidation-only pathway fails to recognize that inertinite macerals can form through other processes such as charring of plant matter. Microscopically, charred matter possesses anatomical structures that closely resembles those observed in inertinite macerals, with lignin and/or cell walls largely preserved, perhaps mechanically fragmented or compressed in some instances. Based on the NMR structural parameters, the average aromatic cluster size in the inertinite-rich sample (petrographically dominated by fusinite, semifusinite of varying reflectance, and inertodetrinite) corresponds to monocyclic, 6-aromatic carbon rings. In the vitrinite-rich sample of the same coal, the cluster sizes are larger, corresponding to multi-ring aromatic hydrocarbons. The 6-carbon rings in the inertinite-rich sample are interpreted to correspond to guaiacol and syringol, the principal products of low-temperature (below 400 °C) lignin pyrolysis. Since charring depletes cellulose and moisture, fire-affected plant matter is interpreted to have a lower compaction potential than vitrinite-forming material, sustaining 6-aromatic carbon ring isolation in the former. Lignin-derived aromatic rings in compactable plant matter merge to form the multi-ring clusters present in the collotelinite- and collodetrinite-rich sample of the same coal.
Keywords: 13C CP-MAS SS NMR; Karoo Basin; Coal-formation; Charring; Monocyclic aromatic rings; Semifusinite; Fusinite.
Introduction
The formation of inertinite macerals in coal through peat fire and/or wildfire combustion has been the subject of controversial discussion for some time (Austen et al., 1966; Hunt and Smyth, 1989; Scott, 1989, 2002, 2010; Teichmüller, 1989; Diessel, 1992, 2010; Moore et al., 1996; Shearer and Moore 1996; Moore and Shearer, 1997; Guo and Bustin, 1998; International Committee for Coal and Organic Petrology (ICCP), 2001; Glasspool, 2003a, 2003b; Scott and Glasspool, 2007; Hower et al., 2009, 2011a, 2011b; O’Keefe and Hower, 2011; Richardson et al., 2012; O’Keefe et al.,2013). South African coals, specifically those of the Witbank-Highveld, are rich in inertinite, dominated by semifusinite and inertodetrinite, with variable proportions of fusinite. A low-reflecting semifusinite is recognized, termed reactive semifusinite due its coking and combustion properties, with reflectance values more comparable to that of vitrinite rather than other inertinite macerals within the same or adjacent microlithotypes of the same coal (Falcon and Ham, 1988; Hagelskamp and Snyman,1988; Snyman, 1989; Cadle et al., 1993; Glasspool, 2003a, 2003b; Van Niekerk et al., 2008, 2010; O’Keefe et al., 2013; Richards et al., 2013). The main point of contention between different workers relates to whether some inertinite macerals, including those from South African coals, owe their (partial or complete) inert character to prehistoric fires and/or aerial oxidation (e.g., Falcon, 1986; Hagelskamp and Snyman, 1988; Snyman, 1989; Cadle et al., 1993; ICCP, 2001; Glasspool, 2003a, 2003b). Various workers have argued for and against the formation of inertinite macerals through charring, with fusinite generally accepted as representing fossilized charcoal (Austen et al., 1966; Falcon, 1986; Hagelskamp and Snyman, 1988; Hunt and Smyth, 1989; Scott, 1989, 2002, 2010; Teichmüller, 1989; Jones and Chaloner, 1991; Diessel, 1992, 2010; Cadle et al., 1993; Moore et al., 1996; Shearer and Moore 1996; Moore and Shearer, 1997; Guo and Bustin, 1998; ICCP, 2001; Glasspool, 2003a, 2003b; Scott and Glasspool, 2007; Hower et al., 2009, 2011a, 2011b; O’Keefe and Hower, 2011; Richardson et al., 2012; O’Keefe et al., 2013).
Globally, plants started forming in the late Silurian, although initially sparsely distributed and with limited differentiation compared to the various organs of present-day plants; the plant structures were simple (Plumstead, 1961; Remy et al., 1994; Diessel, 2010). In the Permian, specifically during the formation of the upper coal seams of the Witbank Coalfield (South Africa), Glossopterid gymnosperms became abundant and were the major coal-forming plants (Falcon, 1986; Bamford, 2004; Ruckwied et al., 2014; Slater et al., 2015). Aside from the availability of vegetation, the main prerequisite for the formation of macerals through charring is the occurrence of fires within the coal-forming environments. The occurrence of wild-fires during coal-forming periods, as early as the Devonian and specifically in the Permian, is well-documented in the fossil record and is dependent, in part, on the availability of free oxygen in the atmosphere (Jones and Chaloner, 1991; Diessel, 2010; Glasspool and Scott, 2010; Scott, 2010; Slater et al., 2015). Similar processes continue to operate in modern peat-forming environments (Hope et al., 2005). Charring of plant matter before incorporation into peat would increase the rigidity of the coal-forming material so that it will be less affected by humification-related decomposition and confining pressures during diagenesis and subsequent coalification (Austen et al., 1966; Scott, 1989, and references therein; Jones and Chaloner, 1991; Diessel, 1992; Ascough et al., 2010). The charring process also imparts distinctive physical morphologies on fire-affected plant matter, and subsequently, on the macerals produced (Scott, 1989, 2010; Jones and Chaloner, 1991; Guo and Bustin, 1998).
The basic building units for the polycyclic aromatic clusters in coal, namely 6-carbon monocyclic rings, have been found to exist as aromatic alcohols in the lignin component of wood (Saiz-Jimenez and de Leeuw, 1986; Kirk and Farrell, 1987; Teichmüller, 1989; Haenel, 1992; Simoneit et al., 1993; Hatcher and Clifford, 1997, and references therein; Kawamoto, 2017). During the early stages of coal-formation, the cellulose component is preferentially eliminated through either enzymatic action or charring (Hatcher and Clifford, 1997; Spiker and Hatcher, 1987; Guo and Bustin, 1998; Czimczik et al., 2002; Yang et al., 2007). Coalification to higher coal ranks parallels an increase inaromaticity, which involves the condensation and merging of individual aromatic rings to form multi-ringed clusters (Solum et al., 1989; Haenel, 1992; Davidson, 2004; Suggate and Dickinson, 2004; Van Niekerk et al., 2008), physically manifested by cellular wall and bordered pit distortion (Hatcher and Clifford, 1997). The deformation is accompanied by major chemical and structural changes to the lignin macromolecule, with vitrinite becoming visually lustrous and more homogenous at the rank of bituminous coal (Hatcher and Clifford, 1997). Despite the structural changes observed at this coal rank, the chemical structure largely remains that of lignin although having been subjected to demethylation, dehydroxylation, and the expulsion of side chains. Low-temperature charring of lignin-rich wood produces phenolic moieties, thus preserving 6-carbon ring monocyclicity, with multi-ring aromatic clusters only forming at high temperatures (Saiz-Jimenez and de Leeuw, 1986; Simoneit et al., 1993; Kawamoto, 2017).
In this contribution, solid-state carbon-13 cross-polarization magic-angle-spinning solid-state nuclear magnetic resonance (13C CP-MAS SS NMR) evidence for the presence of wood-derived (lignin) charring products in an inertinite-rich coal is provided. A medium rank C bituminous coal was obtained from the No. 4 Seam Upper of the Witbank Coalfield, South Africa. Density fractionation was performed to prepare samples corresponding to an inertinite-rich and a vitrinite-rich sample of the same coal. The aromatic cluster sizes in each sample were determined using 13C CP-MAS SS NMR. These were subsequently used to infer the coalification pathways for the inertinite macerals specifically. Both the detailed maceral and microlithotype compositions are presented, with the latter used to understand the associations between the individual macerals. Published NMR parameters for a vitrinite-rich Waterberg and an inertinite-rich Highveld coal (Van Niekerk et al., 2008) are included for comparison. This study focusses primarily on the formation of fusinite and semifusinite, and given that inertodetrinite is a secondary maceral it is largely omitted from the discussions. Similar, woody botanical precursors are assumed for vitrinite and the specific inertinite macerals, fusinite and semifusinite (ICCP, 1998, 2001; O’Keefe et al., 2013), so that the differences observed in the NMR data derive mainly from the coalification process. The present paper forms part of a larger project considering the origin and geochemistry of inertinite-rich coals in the No. 4 Seam Upper of the Witbank Coalfield using advanced analytical techniques.
Materials and methods
SamplSamples and basic characterization
The coal sample (~50 kg) was received from an operating mine in the Witbank Coalfield (Seam 4 Upper), South Africa. In order to concentrate the vitrinite and inertinite components, the sample was crushed to –2 mm and the fines (–0.5 mm) removed. A quarter of the particles were retained to represent the “parent”, and the rest of the sample (~17 kg) was subjected to float-sink experiments (South African National Standards (SANS) 7936, 2010). The density fractionation was performed in low and high-density liquids (low-density, RD = 1.3; high-density, RD = 1.8) to create separate samples enriched in vitrinite and inertinite respectively. The liptinite content of Witbank coals is below 5 vol. % and should not influence the separation or subsequent analysis. Following density fractionation, a representative portion of each sample was mounted in epoxy resin and polished for petrographic analysis (SANS 7404-2, 2015).
Further representative samples were crushed (–212 μm) and split using a rotary splitter for the determination of proximate and ultimate/elemental analyses as well as calorific value (CV). These analyses were undertaken at a commercial laboratory (Bureau Veritas, Centurion, South Africa) following prescribed SANS standards (SANS 334, 1992; SANS 17247, 2006; SANS 1928, 2009; SANS 17246, 2011).
Petrographic analysis
A minimum of 500 unique points were recorded on each polished block to quantify and qualify the maceral components (SANS 7404-3, 2016) using a Zeiss AxioImager m2M reflected light microscope retrofitted with Hilgers Diskus Fossil components and software, at a total magnification of x500 using immersion oil. The terminology used follows that used by Taylor et al. (1998) and the International Committee for Coal and Organic Petrology (ICCP, 1998, 2001; Pickel et al., 2017). The mean random vitrinite reflectance (%RoVmr) was determined as per SANS standard 7404-5 (2016) on the parent sample.
The microlithotype composition of each sample was determined by a 500 point-count in 50 by 50 μm areas following the method described in Teichmüller (1989), Diessel (1992), as well as Hower and Wagner (2012). Depending on which of the three maceral groups or a combination of two or all three dominated the area under observation, the microlithotype group was classed as mono-, bi-, or trimaceral, respectively. In addition to vitrite and liptite, the inertite microlithotype was further subdivided into semifusite, fusite/secretite, and inertodetrite depending on the dominant inertinite maceral. When the area under observation also contained a proportion of mineral matter it was classed as a carbominerite. Areas comprising only mineral matter (more than 60% of the field of view) were classified as minerite. It is beneficial to undertake microlitholitype analyses in order to gain an understanding of the size of the maceral components and their associations.
Nuclear magnetic resonance
13C SS NMR has been widely used in the past to characterize the complex molecular structure of coal and other carbonaceous materials, as reported by Solum et al. (1989), Supaluknari et al. (1990), Edwards and Choate (1993), Hu et al. (2001), Suggate and Dickinson (2004), Van Niekerk et al. (2008), Erdenetsogt et al. (2010), De Goede et al. (2010) and Mao et al. (2010). The methodology developed by Solum et al. (1989) and Hu et al. (2001) has improved theunderstanding of the carbon skeletal structure of coal. They utilized cross-polarization, magic angle spinning (13C CP-MAS), variable contact time (VCT) and dipolar dephasing (CP-DD-MAS) techniques, which provided a direct measurement of the different carbon functionalities in coals. Ultimately, twelve carbon structural parameters can be determined from the combination of these experiments. The method relies on the integration of selected chemical shift ranges assigned to various carbon functional groups. Additionally, the structural parameters can be used to estimate the aromatic cluster size and extent of bridging between clusters in coal samples. This method has been widely employed by other researchers in the field, such as Supaluknari et al. (1990), Edwards and Choate (1993), Hu et al. (2001), Suggate and Dickinson (2004), Van Niekerk et al. (2008), De Goede et al. (2010) and Mao et al. (2010).
The NMR experiments discussed were mostly carried out on low magnetic field instruments (60-100MHz) and generally, broad resonances were observed for the aromatic and aliphatic carbon signals. Several research groups have used higher magnetic field instruments (200-300MHz) and have successfully demonstrated an increased insensitivity and spectral resolution (e.g., Supaluknari et al., 1990). However, the CP-MAS spectra obtained at the higher field with the sample spinning rate at 3-5 kHz were usually obscured by spinning sidebands, which made spectral interpretation difficult, as mentioned by Mao et al. (2010). Chemical-shift anisotropy (CSA) is known to manifest as spinning sidebands and their intensity correlates with increasing magnetic field strength. Therefore, the intensity of the downfield sideband is an important consideration in the total intensity of the total aromatic/anomeric and aliphatic carbon content at the higher field strengths. One way to address this problem is through using a sideband suppression pulse sequence such as Total Suppression of Sidebands (TOSS). Supaluknari et al. (1990) reported a combined CP-MAS/TOSS technique that they used to characterize 29 Australian coals. By using CP-MAS/TOSS together with CP-MAS/TOSS/DD, they could estimate the structural parameters of the different coals similarly to Solum et al. (1989). Thus, their findings indicated more accurate measurements of carbon fractions. Supaluknari et al. (1990) also investigated the dynamics of cross-polarization with the variable contact time experiment. Their findings suggest that there is an optimum contact time of ~1.5 ms which is suitable for a single contact time experiment in the absence of the more time-consuming variable contact time experiment. In the same way, while a full DD interrupt delay array analysis is useful for the ratio estimation of protonated and non-protonated carbons, the approach is tedious. Supaluknari et al. (1990) have indicated that for coal it is adequate to use only two spectra, one taken at an interrupted decoupling time constant, td = 0 (CP-MAS/TOSS) and one taken at td = 40μs (CP-MAS/TOSS/DD), as this delay provides the best suppression of protonated carbons.
In a study by Mao et al. (2010), advanced solid-state NMR has been employed in the characterization of coal. The techniques used in the study included quantitative direct polarization/magic angle spinning (DP-MAS) at a high MAS rate of 14kHz, CP-MAS/TOSS, CP-MAS/TOSS/DD, 13C CSA filtering and two-dimensional (2D) 1H-13C heteronuclear correlation NMR (HETCOR) experiments. Direct polarization combined with spectral editing techniques, allowed Mao et al. (2010) to quantify eleven different types of functional groups with confidence. Unfortunately, neither the CP-MAS/TOSS/DD or DP-MAS/TOSS/DD pulse sequence are currently commercially available for Agilent users, even though their required CP-MAS/TOSS or DP-MAS/TOSS counterparts are. The experiments, while probably slightly more quantitative because of the direct polarization techniques, each took several days to run. Delay times of up to 30s were reported and the number of scans for different experiments varied from 4000 to over 70 000 scans. While the CP-MAS experiment can only be regarded as a semi-quantitative technique, it is practically more feasible than DP-MAS to evaluate differences between coal samples where the sample’s cross-polarization dynamics are similar. CP-MAS is much more cost effective and, more importantly, their required DP-MAS/TOSS/DD and CP-MAS/TOSS/DD versions needed for the dipolar dephasing experiments are not available commercially. Thus, CP-MAS in combination with a CP-DD experiment at 12 kHz MAS still holds excellent value, at high field, for the relative comparison of coal samples. We, like others, dealt with the sidebands by subtracting the downfield sideband signal area from the corresponding upfield area in the calculations. The use of CP with DD NMR experiments is still popular for coal analysis (Malumbazo et al., 2011; Hattingh et al., 2013; Okolo et al., 2015).
Declaration
Abstract
Notes on the structure of the thesis.
Research outputs
Acknowledgements
Dedication
List of Tables
List of Figures
Chapter 1: Introduction, hypothesis, and scope
1.1. Introduction
1.2. Coal use and geological setting
1.3. Macerals, with a focus on inertinite
1.4. Brief overview of previous work on the origin of inertinite in South African coals
1.5. Hypothesis
1.6. Aim and objectives
1.7. Scope
1.8. Contribution to science
1.9. References
Chapter 2: Characterization of coal using Electron Spin Resonance: Implications for the formation of inertinite macerals in the Witbank Coalfield, South Africa
2.1. Introduction
2.2. Materials and methods
2.2.1. Sample, preparation, and basic characterization
2.2.2. Electron spin resonance analysis
2.3. Results
2.3.1. Basic characterization and aromaticity inferred from H/C atomic ratio
2.3.2. Electron spin resonance: Spectral amplitudes, line-shapes, and line-widths
2.4. Discussion
2.4.1. Radical properties for vitrinite-rich and inertinite-rich Witbank coals
2.4.2. On the formation of the inertinite macerals in the Witbank Coalfield
2.5. Conclusion
2.6. References
Chapter 3: A Nuclear Magnetic Resonance study: Implications for coal formation in the Witbank Coalfield, South Africa
3.1. Introduction
3.2. Materials and methods
3.3. Results
3.4. Discussion
3.5. Conclusion
3.6. References
Chapter 4: Using δ15N and δ13C and nitrogen functionalities to support a fire-origin for certain inertinite macerals in a No. 4 Seam Upper Witbank coal, South Africa
4.1. Introduction
4.2. Methodology
4.3. Results
4.4. Discussion
4.5. Conclusion
References
Chapter 5:Comparative study of a vitrinite-rich and an inertinite-rich Witbank coal using pyrolysis-gas chromatography, South Africa
5.1. Introduction
5.2. Samples and analytical methods
5.3. Results and discussion
5.4. Conclusion
5.5. References
Chapter 6: Summary and conclusions
6.1. References
Chapter 7: References
GET THE COMPLETE PROJECT
