Investigation into char structure using Raman and petrographic techniques to assess combustion reactivity
No Thumbnail Available
Date
2011-04-12
Authors
Chabalala, Vongani
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Coal petrography, micro Raman spectroscopy (MRS) and thermogravimetric analysis
(TGA) were employed to obtain further inside into the evolution of char structure and its
reactivity during heat treatment in the temperature range of 300-1400°C on inertinite-rich
South African coals. The lack of publications particularly on South African coals, relating
Raman spectroscopy to coal petrography and thermogravimetric analyses when
investigating the evolution of char structure was a motivation for the study. A laboratory
scale Packed Bed Balance Reactor (PBBR) was used to prepare coal char samples at
various temperatures; that is 300, 600, 800 and 1000 °C. A drop tube furnace (DTF) was
used to prepare chars at 1400°C. Raman spectra of coal and chars were measured on the
first-order in the range 800-2000cm-1. Characteristic bonds for amorphous carbons, G
band (graphitic) and D band (disorder), were deconvoluted and curve fitted using the
OPUS software. Three bands for coal particles were determined; that is the G band at
~1590-1603cm-1, D1 band at ~1343-1355cm-1 and D3 band at ~1507-1557cm-1. Four
bands were determined for char particles; that is the G band at ~1590-1603cm-1, D1 band
at ~1343-1355cm-1, D3 band at ~1507-1557cm-1 and D4 band at ~1200-1232cm-1. All the
bands were fitted with a mixture of Lorentzian and Gaussian functions except the D3
band for which only a Gaussian function was used. It was found that sp2-sp3 bonding
(reactive sites/crystallites) occurred in dense chars (originating from inertinite particles)
at the initial heat treatment temperature, and these sp2-sp3 bondings are known to be
consumed later at high temperature. Earlier consumption of sp2-sp3 bonding was
observed in porous chars, since they were vitrinitic in origin and contained more reactive
sites. The D1 and G bandwidths showed a significant change with heat treatment, which
was consistent with structural modification due to high temperatures. Reflectance
measurements, that is: mean vitrinite reflectance (MVR) and mean total reflectance
(MTR), showed an increase with heat treatment temperature. MVR and MTR were
successfully correlated with Raman parameters (D1 and G bandwidth). MVR and MTR
also showed a good correlation with combustion reactivity measured by TGA. Char
morphology analyses were carried out petrographically by point counting for
quantification and qualification purposes. The char morphology data showed a significant increase in the amount of dense/solid chars as compared to the porous chars with an
increase in temperature which is in-line with expectations from the inertinite-rich parent
coal. Correlations between the D1 and G bandwidths and char morphology counts were
carried out. An inverse of D1 and G bandwidth showed good correlation with the
proportion of dense/solid and porous char. It was concluded from this study that the best
correlations between Raman spectroscopy and coal petrography was through reflectance
measurements, and the identified Raman D1 and G bandwidths. A good linear correlation
was also found between Raman D1 and G bandwidths and combustion reactivity. These
correlations confirm the strong connection between char structure and its reactivity and
illustrate the advantage of Raman spectroscopy in conjunction with coal petrography with
respect to other structural analyses. Therefore, the use of MRS and petrography on coal
chars enhances the understanding of char structural evolution on a molecular level and
may lead to enhanced understanding of pulverised fuel (pf) coal combustion.