ASH FREE COAL (AFC) FROM LOW-GRADE CANADIAN COAL BY SOLVENT EXTRACTION Moshfiqur Rahman*, Arunkumar Samanta, Arno de Klerk and Rajender Gupta Department of Chemical and Materials Engineering, University of Alberta, Edmonton, AB, T6G 2V4, Canada Introduction The low cost and availability of low-grade coals (lignites and subbituminous coals) make them attractive as feed materials for fuel, chemical and energy production. However, these coals have several limitations too, such as lower energy content due to their lower carbon content, higher mineral matter and moisture contents, as well as lower calorific values. Some of these drawbacks can be overcome by producing ash free coal (AFC). AFC is coal that has been demineralized. It is a preferred feed for some applications, such as catalytic gasification and direct combustion in gas turbines so as to get the effect of combined cycle. Since AFC is demineralized it causes less erosion and corrosion of turbine blades and less fouling due to ash deposition. Solvent extraction of coal using model and industrial solvents can produce AFC (<0.1% ash). 1-5 It not only removes ash forming coal minerals, but also heavy metals, almost all inorganic sulfur and it reduces the alkali content to less than 0.5 μg∙g -1 . Different regimes can be considered for the solvent extraction of coal: 6 physical dissolution (<150 °C), extractive disintegration (150-400 °C) and liquefaction (>400 °C). In this work we consider solvent extraction in the industrially interesting temperature range 360-400 °C, which is nominally inside the extractive disintegration regime. There are two key design decisions that affect the cost of solvent extraction at such conditions: solvent-to-coal ratio and hydrogenation of the solvent and/or coal. We evaluated the AFC yield loss from solvent extraction when extraction is not combined with hydrogen addition. Additionally we evaluated potential benefits of AFC compared to raw coal beyond that of demineralization. Of specific interest was properties affecting combustion, such as H:C ratio and particle size. Experimental Materials. Subbituminous coal (SBC) and lignite coal (LC) from Western Canada were used. The ultimate and proximate analyses are shown in Table 1. Model solvents were obtained commercially: 1-methylnaphthalene (1-MN) and quinoline (QN). Heavy Aromatic Based Solvent (HABS) with hydrotreatment and without were supplied by industry. Table 1. Ultimate and Proximate Analyses of Coals Description SBC LC Ultimate analysis (wt% daf) carbon 72.0 68.2 hydrogen 4.6 5.0 nitrogen 1.1 1.2 sulfur 0.9 0.93 oxygen 21.4 24.6 Proximate analysis (wt%) moisture 2.8 6.9 ash 10.6 19.2 volatile matter 34.3 31.7 fixed carbon 52.3 42.2 Equipment and procedure. Solvent extraction using different solvents were carried out under N 2 atmosphere in a 500 mL autoclave (batch reactor) at 360400 °C and 1 MPa. For each experiment 10 g of dry coal (particle size of <150 μm) and 100 ml of solvent were put in the reactor and purged it with N 2 . The reactor was heated in an oven to the desired temperature while stirring the slurry continuously. After the end of reaction time of 1 h, the heater was turned off. The reactor was allowed to cool down to about 100 °C. Hot filtration was carried out using a 0.1 μm filter. The residue was washed several times with hexane and finally with acetone. The filtrate was then poured into a large volume of hexane solvent (reaction solvent to hexane ratio of about 1:40) to precipitate the AFC. The precipitated AFC was filtered, washed and dried under vacuum. The solvents were recovered by rotatory evaporation. Analyses and calculations. The raw coal, AFC and residue were characterized using ultimate and proximate analyses, particle size distribution (PSD) using Mastersizer from Malvern instruments, thermogravimetric analysis (TGA) using a SDT Q600 TGA-DSC, infrared analysis using an ABB FTIR MB3000 and solid state 13 C nuclear magnetic resonance (NMR) spectrometry using a Bruker Avance 300 NMR spectrometer. The extraction yields were calculated based on solvent-free dried residue as indicated: daf] [wt% 100 (daf) coal Feed (daf) Residue (daf) coal Feed Yield Extraction Results and Discussion Yield of AFC. The extraction yields that were obtained with different solvents are listed in Table 2. The hydrogenated HABS resulted in the highest extraction yield. This highlighted the importance of hydrogen-donor properties, which are more important at high temperature than solvent properties. The HABS is a better solvent (contains a mixture of aromatic, aliphatic and polar compounds), yet, the hydrogenated HABS resulted in a better extraction yield. The non-polar 1-MN performed poorly, since it did not have any hydrogen-donor ability and was not a particularly good solvent either. Previous studies showed that the extraction of subbituminous coals with polar solvents such as N-methyl-2-pyrrolidone (NMP) and crude methylnaphthalene oil (CMNO) gave higher yields than that with non-polar solvents like 1-MN. 1-3 It was therefore surprising that 1-MN mixtures with QN performed worse than 1-MN. The lower observed yield may be due to acid-base type reactions between the QN and oxygenates such as carboxylic acids and phenols to form heavier products. Table 2. Solvent Extraction Yields at 400 °C Solvent Extraction yield (wt%) SBC LC 1-MN 30 30 90 % 1-MN + 10% QN 18 - 80 % 1-MN + 20% QN 22 - HABS - 42 hydrogenated HABS 62 67 Infrared analysis (IR). The infrared spectra of raw coal, residue and AFC were compared. It was observed that the raw and residue coal showed a broad signal in the range of 500-600 and 1000-1100 cm -1 , indicating the presence of mineral mater. 7 These absorptions were absent from the AFC in the coal, indicating the absence of mineral matter. Solid state 13 C NMR analysis. The 13 C NMR of raw coal, AFC and residue coal using 1-MN are shown in Figure 1. It is observed that raw and residue coal show two very broad peaks in the aromatic and aliphatic regions due to the presence of numerous aromatic and Prep. Pap.-Am. Chem. Soc., Div. Fuel Chem. 2011, 56 (2), 308 Proceedings Published 2011 by the American Chemical Society