Energy Fuels 2010, 24, 6248–6256 Published on Web 11/30/2010
: DOI:10.1021/ef100780c
Factors Influencing Ultrafine Particulate Matter (PM0.1) Formation under Pulverized Coal Combustion and Oxyfiring Conditions Francesco Carbone,*,† Federico Beretta,† and Andrea D’Anna‡ †
Istituto di Ricerche sulla Combustione, Consiglio Nazionale delle Ricerche (CNR), Piazzale Vincenzo Tecchio 80, 80125 Napoli, Italy, and ‡Dipartimento di Ingegneria Chimica, Universit a degli Studi di Napoli “Federico II”, Piazzale Vincenzo Tecchio 80, 80125 Napoli, Italy Received June 21, 2010. Revised Manuscript Received October 30, 2010
This paper explores the early processes of coal ultrafine ash (D < 100 nm) formation under both conventional air-blown and oxyfiring conditions. An innovative flow reactor and a high-resolution differential mobility analysis technique have been coupled to measure the particle size distribution functions (PSDFs) in a size range extending down to 1 nm. Information on the formed fly ash chemical nature has been obtained by ultraviolet-visible (UV-vis) light absorption and scanning electron microscopy with energydispersive X-ray spectroscopy (SEM-EDXS). Five coals of different rank, covering a broad range of ash compositions, have been tested under three oxygen concentration levels. A multimodal behavior of coal ultrafine ash PSDFs has always been observed. The first mode at 1-5 nm has been attributed to carbonaceous particles based on the UV-vis light absorption measurements and the results obtained burning a carbon black powder under the same operating conditions. The volume fractions of larger mode particles have been correlated to coal components. SEM-EDXS analyses have mostly supported the correlation indications. Results suggest that the particle size modes derive from size-selective nucleation of refractory oxides and metal nanoparticles and their subsequent growth. The oxygen concentration influences the size of nucleating particles and the preferential vaporization of some compounds with respect to others through both char-burning temperatures and the local reducing properties of the gas environment. Nevertheless, an enhanced oxygen concentration promotes ultrafine particle formation.
Ash formation mechanisms in conventional coal combustion systems have been extensively studied.4,5 On the contrary, how oxycombustion conditions affect ash formation is still under investigation.6 This is particularly true for ultrafine ashes, whose formation is strongly affected by the temperature and gas-phase composition surrounding the coal particles. Indeed, both direct vaporization of volatile metals,7 which react in the gas phase and, subsequently, nucleate or condense on the surface of existing particles, and refractory species, after their reduction to sub-oxides,8 is strongly affected by local flame temperatures and oxygen concentrations. The process involving inorganic ash transformation is also coupled to carbonaceous particle formation because of the pyrolysis of tar and volatile organic compounds released during coal devolatilization and coal/char particle fragmentation.9 Experimental results performed in oxyfiring conditions have shown that the mean size of the fine particles became smaller.6 The ultrafine ash particles (D < 100 nm) represent just a small percentage of total ash mass at the exhaust of coal furnaces.10,11 Nevertheless, the ultrafine fraction is of great
Introduction The increasing growth in the demand for electrical power is actually largely satisfied with coal-fired plants because coal is a cheaper and more abundant resource over other fossil fuels, such as oil and natural gas.1 The main drawback of large coal use for power generation is the production of greenhouse gases, primarily carbon dioxide, and pollutant emissions, such as nitrogen and sulfur oxides and particulate matter. Nextgeneration coal power plants require a carbon dioxide capture and sequestration system to be environmentally sustainable.2 Oxycombustion technologies are becoming widely used to allow for carbon dioxide capture in either new plants or retrofitting existing air-blown furnaces. Retrofitting requires oxygen-enriched concentrations in recirculated flue gas to satisfy the heat-transfer request.3 The change in the combustion atmosphere and the potential change in the local particle temperature between oxy- and air-fired combustion may have an effect on the ash formation mechanisms and, hence, the ash composition and quality. *To whom correspondence should be addressed. Telephone: þ39817682239. Fax: þ39815936936. E-mail: francesco.carbone@ irc.cnr.it. (1) International Energy Agency (IEA). World Energy Outlook 2008; IEA Publications: Paris, France, 2008. (2) Wall, T. F. Proc. Combust. Inst. 2007, 31, 31–47. (3) Buhre, B. J. P.; Elliott, L. K.; Sheng, C. D.; Gupta, R. P.; Wall, T. F. Prog. Energy Combust. Sci. 2005, 31, 283–307. (4) Flagan, R. C.; Friedlander, S. K. Recent Developments in Aerosol Science; John Wiley and Son: New York, 1978; pp 25-59. (5) McElroy, M. W.; Carr, R. C.; Ensor, D. S.; Markowski, G. R. Science 1982, 215, 13–19. r 2010 American Chemical Society
(6) Suriyawong, A.; Gamble, M.; Lee, M. H.; Axelbaum, R.; Biswas, P. Energy Fuels 2006, 20, 2357–2636. (7) Sarofim, A. F.; Howard, J. B.; Padia, A. S. Combust. Sci. Technol. 1977, 16, 187–204. (8) Quann, R. J.; Sarofim, A. F. Proc. Combust. Inst. 1982, 19, 1429– 1440. (9) Lighty, J. A. S.; Veranth, J. M.; Sarofim, A. F. J. Air Waste Manage. Assoc. 2000, 50, 1565–1618. (10) Linak, W. P.; Miller, C. A.; Seames, W. S.; Wendt, J. O. L.; Ishinomori, T.; Endo, Y.; Miyamae, S. Proc. Combust. Inst. 2002, 29, 441–447. (11) Zhang, L.; Ninomiya, Y. Fuel 2006, 85, 194–203.
6248
pubs.acs.org/EF
Energy Fuels 2010, 24, 6248–6256
: DOI:10.1021/ef100780c
Carbone et al. Table 2. Minor Elements in Coal as Received (ppm)
Table 1. Properties of the Used Coals
HHV (MJ/kg)
UTS
PRB
MTL
COL
IDN
29.3
21.1
18.3
28.7
30.3
Proximate Analysis (wt % on Coal as Received) ash 8.83 4.94 6.90 6.11 volatile 38.60 33.36 25.02 38.31 moisture 3.18 23.69 37.20 4.90 fixed carbon 49.39 38.01 30.89 50.68
7.40 40.20 3.11 49.29
Ultimate Analysis (wt % on Coal as Received) C 70.60 53.72 40.56 70.95 H 5.41 6.22 2.63 5.32 N 1.42 0.78 0.62 1.32 S 0.53 0.23 0.48 0.40 O (by difference) 13.21 34.11 11.62 11.00
71.68 5.62 1.38 0.53 10.28
SiO2 Al2O3 Fe2O3 MgO CaO Na2O K2O MnO SO3
Ash Composition (wt % on Ash) 60.89 30.46 26.79 14.52 14.78 13.57 5.09 5.20 6.16 1.39 5.17 10.02 6.11 22.19 24.73 1.41 1.94 0.24 0.57 0.35 0.29 0.02 0.01 0.07 2.33 8.83 14.13
47.05 23.94 5.99 1.43 3.29 2.03 1.40 0.04 na
Ti V Cr Co Ni Cu Zn As Se Mo Sr Zr Cd Sn Ba Hg Pb
44.47 22.28 9.67 1.41 1.98 1.14 2.31 0.04 na
UTS
PRB
MTL
COL
IDN
327 7 5.4 0.8 3.1 7.9 7 0.6 1.5 0.4 84.1 5.8 0.3 0.4 44 0.1 3
352 9.3 3 1.3 2.7 10.5 12.5 0.8 0.6 0.3 178.5 5.3 0 0.1 280 0.2 1.6
380 185.7 69.2 0 48.5 6.9 0.1 0.5 0.1 0 178.2 12.3 0.04 0 14.5 0.7 3
176 21 7 56.1 10 5.7 7.8 2.5 4 2.1 71.8 9.6 0.1 0.2 241 2.7 2.3
114 16 8 24.5 8.7 9.7 13.7 1.9 0.4 0.7 96 4.4 0 0.3 62.7 6.8 3.5
Pulverized Coal Combustion Reactor. The reactor consists of a fuel-lean flat laminar premixed flame sustained by ethanol vapor and operated at atmospheric pressure. Pulverized coal particles, monodisperse in size, are homogeneously added in the flame, so that their oxidation occurs in the post-flame gases. The flame is stabilized on a burner consisting of two coaxial stainless-steel tubes and a flat plate positioned 90 mm downstream of the burner mouth. A silicon carbide honeycomb [1 in., 300 cells/in.2 (CPSI), CTI s.a.] covers the top of the inner tube (18 mm inner diameter) to laminarize the flow and stabilize the premixed flame. The tube wall temperature is kept constant at 65 °C. Sheath argon (9.0 L/min) is flowed through the ring (24 mm inner diameter and 34 mm outer diameter) between the tubes to prevent surrounding air entrainment and minimize flame flickering. The burner is fed with 2.33 L/min [standard temperature and pressure (STP)] oxidant stream and 0.30 cm3/min liquid ethanol, in the form of 80 μm monodisperse droplets, in which pulverized coal is suspended (about 1.5 wt % of coal). Gas flows are controlled by rotameters, while droplets are produced by a vibrating orifice aerosol generator (VOAG, model 3450, TSI, Inc.), whose syringe pump is immersed into an ultrasonic thermostatic bath at 40 °C to ensure the coal suspension stability. Coal is milled in a Planetary Mono Mill (Pulverisette 6, Fritsch) prior to being suspended in ethanol and intensively sonicated, to prevent the VOAG orifice from clogging. Monodisperse coal agglomerates (D > > > 6ln > < X π Ni 16 7 = Di 7 3 7 pffiffiffiffiffiffi exp - 6 ¼ ½MD - D0 > 6 24 lnðσi Þ 5 > > > i lnðσ i Þ 2π > > : ;
Table 3. Gas Composition and Temperatures oxidant
air
Species O2 CO2 H2O N2
5.0 9.5 14.0 71.5
HAB (mm) 5 25 50
1730 1590 1290
O2/CO2
O2
O2/N2
Exhaust Gas Composition (vol %, (0.5) 48.0 76.5 48.0 38.0 9.5 9.5 14.0 14.0 14.0 0.0 0.0 28.5 Gas Temperature (K, (30) 1600 1640 1430 1450 1140 1150
1670 1490 1190
Adiabatic Flame Temperature (K, (50) ∼2050 ∼2900 ∼3350 ∼2950
oxygen and carbon dioxide (2.4:1.1 ratio). A mixture of oxygen and nitrogen in the same ratio is also used for the intermediate oxygen concentration. The gas composition and temperature in which coal combustion occurs are summarized in Table 3. Gas concentrations are calculated by mass balance, while excess oxygen is measured at the flame exhaust. Temperatures are measured along the reactor axis using a 250 μm Pt/Pt-13% Rh thermocouple (type R, Omega Engineering) and properly corrected for radiative losses.17 Pure ethanol without coal is fed during temperature measurements to avoid particle deposition on the thermocouple junction. In all investigated cases, the temperature similarly decreases at increasing HAB and is slightly higher using air as the oxidant because of its lower specific heat over the other used gas mixtures.16 Char burning temperatures in each condition are also estimated calculating the adiabatic flame temperature18 and are also reported in Table 3. Such temperatures reasonably agree with the char surface temperature calculated using the model described by Mitchell and Madsen19 in the steady-state approximation. Measurements Techniques and Data Analysis. PSDFs are measured online using a horizontal rapid dilution probe (8 mm inner diameter and 9 mm outer diameter) delivering the aerosol to a TapCon 3/150 differential mobility analyzer (DMA), equipped with a diffusion charger (Am-241 bipolar) to neutralize the particles and a Faraday cup electrometer detector. The aerosol is drawn through a pinhole, drilled on the probe wall, because of the effect of a slight underpressure into the tube where particle-free nitrogen steadily flows (29.5 L/min).20 Three probes with 0.3, 0.9, and 1.5 mm sampling pinhole diameters were used to perform dilution ratios ranging from 1103 to 30.15,16 The diluted aerosol temperature does not exceed 45 °C, while the cooling rate is on the order of 105-106 K s-1. Dilution is required to control coagulation during sampling and to reduce the particle concentration within the electrometer detection range. Condensation in the probe has been evaluated; it does not significantly influence the measured PSDFs.14,15 The DMA was operated in three modalities, selected changing the maximum voltage applied to the electrostatic classifier and the sheath air flow rate. The nominal mobility diameter (MD) ranges from 0.6 to 28 nm and from 2.1 to 100 nm in the high-flux (50 L/min), low- and high-voltage modes, respectively. The range is further enlarged (2.9-151 nm) operating the DMA in the low-flux (25 L/min), high-voltage mode. Both the burner and probes are electrically grounded, whereas background aerosol measurements were performed in coal-free (only ethanol droplets) flames.16 All measurements are performed by positioning the sampling pinhole on axis 50 mm downstream of the burner. The distance
ð1Þ Total and modal volume fractions (Fv = Vparticles/Vaerosol) are calculated by PSDF integrations. The aerosol generated in the low oxygen condition is sampled (0.8 L/min) on axis 50 mm above the burner using a vertical tube probe (8 mm inner diameter) and bubbled for 1 h into 15 cm3 of pure water. The bubbler is ice-cooled to condense combustion water. Ultraviolet-visible (UV-vis) spectra of water samples, previously filtered with 500 nm fiber glass filter, are measured on an Agilent 8453 diode array spectrophotometer using a standard 1 cm path-length cuvette. Particles generated from the UTS coal combustion in all investigated conditions have also been thermophoretically collected on square pure gold foils (3 mm side and 0.05 mm thick, Sigma-Aldrich, Inc.) inserted parallel to the gas streamline at HAB of 50 mm. The samples have been examined by a Philips XL30 scanning electron microscope (SEM) with a LaB6 filament equipped with an energy-dispersive X-ray spectroscopy (EDXS) DX-4i microanalysis device to determine particle elemental composition. A double-acting pneumatic actuator has been used to realize a quick insertion and a constant sampling time to ensure the sampling of particles smaller than 30 nm.14 A total of 1000 substrate insertions in flame have been required to collect enough matter for EDXS analyses.
Results and Discussion Both number and volume PSDFs of the ultrafine ashes formed during MTL coal combustion at a low oxygen concentration are plotted in Figure 1. This illustrative PSDF clearly shows the multimodal nature of ultrafine coal ashes and the huge number concentration (on the order of 1011 cm-3) of particles smaller than 30 nm. Data fitting has been performed adding five log-normal distributions, and it is also plotted in Figure 1. The modes of the PSDF are more clearly identified in the volume PSDF, showing concentrations that range from a hundredth parts per billion (ppb) to a unit ppb. A reasonable fitting of experimental data could be performed using only four log-normal modes.16 The use of five modes is preferred because of the possibility to fit all collected data with
(17) Rolando, A.; D’Alessio, A.; D’Anna, A.; Allouis, C.; Beretta, F.; Minutolo, P. Combust. Sci. Technol. 2004, 176, 945–958. (18) Bejarano, P. A.; Levendis, Y. A. Combust. Flame 2008, 153, 270–287. (19) Mitchell, R. E.; Madsen, O. H. Proc. Combust. Inst. 1986, 21, 173–181.
(20) Kasper, M.; Siegmann, K.; Sattler, K. J. Aerosol Sci. 1997, 28, 1569–1578. (21) Fernadez de la Mora, J. J.; De Juan, L. L.; Liedtke, K.; Schmidt-Ott, A. J. Aerosol Sci. 2003, 34, 79–98.
6250
Energy Fuels 2010, 24, 6248–6256
: DOI:10.1021/ef100780c
Carbone et al.
Air-Blown Coal Combustion. The yields on a coal mass basis of ultrafine ashes from the combustion of the five coals at a low oxygen concentration are plotted in Figure 2 as a function of the MD. The yields have been obtained by dividing the PSDFs by the coal concentration in the reacting mixture (Mcoal/Vaerosol), calculated by performing the mass balance of the fed coal.16 The number yields range from 106 to 1012 particles/μg of coal. The volume yield range from some tenths up to a few cubic millimeters per gram of coal, implying that ultrafine particles, assumed to have unit density, are from ∼1% (UTS) to ∼3% (IDN) of total ash mass. All of the PSDFs show that the largest fraction of PM0.1 is due to two classes of particles: particles smaller than 10 nm and particles larger than 30 nm. Each of the two classes is composed of two peaks, whose relative contribution depends upon the coal type. Mode I, centered between 1.5 and 2.5 nm, surprisingly accounts for a considerable fraction, from 15% (MTL and PRB) up to 50% (COL and IDN) of ultrafine particles. Particles smaller than 10 nm are at least 35% (UTS) because of mode II located at 5 nm that is prevalent for MTL (45%) and PRB (36%) coals. Mode III, centered at 18 nm, accounts for matter less than 10% in the intermediate-size range. Mode IV, accounting for 5-12% of PM0.1, is recognized at 45 nm, while the remaining part of PM0.1 belongs to the last mode that sweeps beyond the measurement range. Figure 2 also shows the unimodal PSDF centered at about 2 nm, resulting from CB oxidation at the same low oxygen concentration. Oxidation of CB in the flame reactor leaves about 0.01% of particles not oxidized at 50 mm above the burner after 60 ms in the reactor. These particles have sizes in the 1-5 nm range and can derive from fragmentation of CB, as observed for soot oxidation in a flow reactor22 or from a surface oxidative detachment. The particles formed during CB oxidation have the same sizes as and a slightly higher volume fraction of mode I particles from coal combustion than shown by the calculated volume fraction reported in Table 4 and evidenced by the parameters to fit the CB PSDF reported in Table 2A of the Supporting Information. Results from CB oxidation indicate that the carbonaceous matter directly generated during char oxidation can
almost the same modal median diameters and widths, so that each mode can be assumed to result from a characteristic formation pathway. The fitting parameters to reproduce the experimental data are reported in Table 1A of the Supporting Information, whereas the calculated PM0.1 and modal volume fractions are listed in Table 4. Only particles smaller than 100 nm have been considered in calculating the volume fraction of the last mode that is then labeled as mode V0.1.
Figure 1. Number (top) and volume (bottom) weighted PSDFs of ultrafine particles generated from MTL coal combustion in a 5.0% oxygen concentration. The solid line shows data fitting performed using five log-normal modes (dashed lines).
Table 4. Modal Volume Fractions of the Fitted PSDFs mode
UTS
PRB
I II III IV V0.1 PM0.1
5.37 10-11 4.53 10-11 2.48 10-11 3.46 10-11 1.21 10-10 2.79 10-10
6.44 10-11 1.36 10-10 9.92 10-12 4.45 10-11 1.17 10-10 3.72 10-10
I II III IV V0.1 PM0.1
7.15 10-11 2.72 10-10 4.96 10-10 2.97 10-10 8.00 10-11 1.22 10-9
I II III IV V0.1 PM0.1
5.37 10-11 2.72 10-10 7.94 10-10 8.24 10-10 9.81 10-11 2.04 10-9
MTL
IDN
CB
5.0% Oxygen Concentration 5.98 10-11 2.59 10-10 1.70 10-10 4.53 10-11 6.95 10-12 8.93 10-12 4.94 10-11 3.13 10-11 8.82 10-11 1.25 10-10 3.74 10-10 4.69 10-10
3.04 10-10 5.66 10-11 2.38 10-11 3.46 10-11 1.94 10-10 6.12 10-10
3.36 10-10
5.32 10-11 3.26 10-10 5.95 10-10 1.65 10-10 1.10 10-10 1.25 10-9
48.0% Oxygen Concentration 7.15 10-11 2.17 10-10 4.90 10-10 4.90 10-10 2.98 10-10 3.47 10-10 1.15 10-10 7.41 10-10 -10 1.19 10 5.66 10-10 1.09 10-9 2.36 10-9
2.17 10-10 7.07 10-10 6.95 10-10 2.14 10-9 4.55 10-10 4.22 10-9
1.22 10-10 2.51 10-11
7.15 10-11 3.10 10-10 2.28 10-10 9.88 10-10 1.21 10-10 1.72 10-9
76.5% Oxygen Concentration 7.87 10-11 1.02 10-10 4.90 10-10 9.79 10-11 1.29 10-09 2.98 10-10 -10 2.64 10 1.32 10-9 1.36 10-10 6.04 10-10 2.26 10-9 2.42 10-9
1.02 10-10 2.18 10-10 4.96 10-10 6.59 10-10 1.10 10-9 2.57 10-9
8.58 10-11 1.13 10-11
6251
COL
3.36 10-10
1.47 10-10
9.72 10-11
Energy Fuels 2010, 24, 6248–6256
: DOI:10.1021/ef100780c
Carbone et al.
Figure 3. UV-vis absorbance of particles collected in water by bubbling the aerosol generated by burning MTL coal at a low (5.0%) oxygen concentration and surviving the filtering procedure. (Inset) Percentage of mode I particles on PM0.1 plotted versus coal fixed carbon and carbon contents.
Figure 3 also reports, in the inset, the percentage of mode I particles on PM0.1 versus fixed carbon and carbon contents of the three UTS, PRB, and MTL coals investigated here. The graph shows a positive correlation of the mode I contribution to ultrafine particles with respect to the two variables, indicating that carbonaceous matter with sizes smaller than 5 nm is formed during air-blown pulverized coal combustion. The contribution of carbonaceous matter is even more evident considering that only mode I particles have been generated from CB oxidation. Coal Oxyfiring. The PSDFs of PM0.1 formed by burning the five coals in oxygen-enriched environments are plotted in Figure 4 as a function of the particle MD. Two extremely high oxygen concentrations of 48.0 and 76.5% are investigated to amplify the effects of oxyfiring conditions on ash formation. The number yields are of the same order of magnitude as that found in the air-blown case even if their decrease with increasing sizes is more gradual. Ultrafine particle volume yields range from a few up to several cubic millimeters per gram of coal in both conditions, so that it is from 3 (MTL-48% O2) to 7 times (UTS-76% O2) higher than a conventional air-blown coal combustion. Particles are rather uniformly distributed in the ultrafine size range, and each mode of the PSDFs gives a significant contribution to the particle volume fraction. Five modes have also been identified in the investigated oxyfiring conditions. The modal relative importance depends upon the coal type and oxygen concentration. Mode I is again centered at 1.5-2 nm and accounts for a percentage of PM0.1 ranging from 4.3% (PRB) to 9.2% (COL) and from 2.6% (UTS) to 4.2% (COL) at oxygen concentrations of 48.0 and 76.5%, respectively. The yield of such small particles is only slightly smaller than that measured in the airblown combustion case. Mode II particles are slightly larger (7 nm) than those measured in the air-blown combustion case. The larger size probably depends upon different ash nucleation locations in the reactor involving longer residence times before sampling.16 As for air blown combustion, particles smaller than 10 nm totally account for a significant percentage of PM0.1 also in oxyfiring conditions. In more detail, they are from 21.9% (IND) to 51.3% (MTL) and from 8.3% (COL) to 25.2% (MTL) of PM0.1 for oxygen concentrations of 48.0 and 76.5%, respectively.
Figure 2. PSDFs of PM0.1 formed from combustion at a low (5.0%) oxygen concentration of five coals: (0) UTS, (Ο) PRB, (4) MTL, (/) COL, and (þ) IDN. The fitted PSDFs (dotted lines) are also plotted, while the solid-line fits result from CB (2) oxidation.
contribute to the PM0.1 from an air-blown pulverized coal combustion. The UV-vis absorption spectrum of MTL coal ashes, which are collected by bubbling the exhaust aerosol in water and survive the filtering procedure, is plotted in Figure 3. Almost the same spectra are measured also for the UTS and PRB coal ashes. The absorption spectrum shows a strong signal in the UV range, which decreases approaching a low and noisy asymptotic value for a wavelength larger than 350 nm. A shoulder, which is stronger for the UTS coal ashes, is also evident around 215 nm, while three absorption bands can be identified after spectra magnification at 257, 264, and 271 nm, respectively. The UV absorption spectra are similar to those of the hydrophilic organic carbon nanoparticles with a size below 5 nm detected in rich combustion of hydrocarbons.23 The shoulder and the non-zero asymptotic value could be due to small amounts of elemental carbon (soot) collected in water. The attribution of the small bands to specific compounds has not yet been attempted. Some other compounds captured in water (such as sulfates) could also contribute to the observed UV absorption. The similarity of the spectra observed for the three different coals, having different ash compositions, supports the idea that the UV absorption mainly depends upon nanometric carbonaceous compounds formed during coal combustion. (22) Lighty, J. A. S.; Romano, V.; Sarofim, A. F. Combustion Generated Fine Carbonaceous Particles; KIT Scientific Publishing: Karlsruhe, Germany, 2009; pp 523-536. (23) Sgro, L. A.; Basile, G.; Barone, A. C.; D’Anna, A.; Minutolo, P.; Borghese, A.; D’Alessio, A. Chemosphere 2003, 51, 1079–1090.
6252
Energy Fuels 2010, 24, 6248–6256
: DOI:10.1021/ef100780c
Carbone et al.
Figure 4. PSDFs of PM0.1 formed from combustion of five coals: (0) UTS, (O) PRB, (4) MTL, (/) COL, and (þ) IDN at oxygen concentrations of 48.0% (left) and 76.5% (right). The fitted PSDF (dotted lines) are also plotted, while the solid-line fits result from CB (]) oxidation.
The yields of particles belonging to the third and fourth modes, still centered at 18 and 45 nm, respectively, are largely promoted by oxygen-enriched environments. Appreciable differences have also been found among the five coals and the two enhanced oxygen concentrations. Mode III represents a fraction ranging from 14.7% (COL) to 47.6% (PRB) and from 12.3% (COL) to 57.13% (MTL) for the intermediate and high oxygen concentrations, respectively. Results obtained at the intermediate oxygen level using the O2/N2 mixture have not been reported, but they only slightly differ from that obtained using the O2/CO2 mixture because of a larger contribution of mode IV. This enhancement of mode IV is much lower than that observed for an increased oxygen concentration. Mode V still sweeps the dimensional range under investigation, but it seems to be not largely affected by the oxygen concentration. In oxygen concentrations of 48 and 76.5%, CB oxidation produces the bimodal PSDFs, also reported in Figure 4. The parameters to reproduce such experimental data are reported in Table 2A of the Supporting Information, whereas the calculated volume fractions are shown in Table 4. A small fraction (
