Chemical Characterization of Unburned Carbon in ... - ACS Publications

Mar 20, 2015 - The ash in the reactor after the TPD was then subjected, without exposure to laboratory air, to the TPO experiment, in which the sample...
0 downloads 0 Views 1MB Size
Article pubs.acs.org/est

Chemical Characterization of Unburned Carbon in Coal Fly Ashes by Use of TPD/TPO and LRS Methods Naoto Tsubouchi,*,† Yasuo Ohtsuka,‡ Hiroyuki Hashimoto,‡ Tetsuo Yamada,§ and Harumi Hashimoto§ †

Center for Advanced Research of Energy and Materials, Hokkaido University, Kita 13 Nishi 5, Kita-ku, Sapporo 060-8628, Japan Institute of Multidisciplinary Research for Advanced Materials, Tohoku University, Katahira 2-1-1, Aoba-ku, Sendai 980-8577, Japan § Department of Applied and Environmental Chemistry, Kitami Institute of Technology, Koencho 165, Kitami 090-8507, Japan ‡

ABSTRACT: Functional forms of the unburned carbon present in six kinds of coal fly ashes have been examined mainly by the temperature-programmed desorption (TPD)/temperature-programmed oxidation (TPO) and laser Raman spectroscopy (LRS) methods. The carbon contents of the ash samples range from 0.4 to 4.1 mass%. The LRS analysis shows that the C consists of both amorphous and crystallized forms, and the proportion of the former is as large as 50−65 C%. Further, the TPD measurement exhibits that the C contains several types of surface oxygen species, such as carboxyl and lactone/ acid anhydride groups, which can readily be decomposed into CO2 up to 700 °C to provide active carbon sites. The results of the TPD also indicate that the ashes have surface CaCO3, and most of this species can be converted to CaO and CO2 around 600−700 °C. Interestingly, there is a significant correlation between organic fluorine concentrations and carboxyl/lactone/acid anhydride groups or surface CaCO3 contents in the ash samples. It might thus be possible that the formation of organic F forms proceeds through gas−solid−solid interactions among HF (and/or F2) in flue gas, active carbon sites and surface Ca species produced around 600−700 °C downstream of coal-fired boilers.





INTRODUCTION With regard to the fate of halogens in high-temperature combustion processes, the authors’ research group has recently shown that the fluorine in fly ashes formed during pulverized coal combustion is connected predominantly with unburned carbon through covalent CF bonds.1 In addition, we have found that the chlorine in dust particles discharged in iron ore sintering and electric arc furnace steelmaking processes is present mainly in chlorinated aromatic structures (organic C Cl forms).2,3 Depending on conditions, these CF and CCl species may be transformed into hazardous halogenated compounds. It may thus be important to understand halogen chemistry for developing superclean coal, iron- and steelmaking technologies that should be targeted on zero emissions. Since active carbon sites on unburned carbon surface play crucial roles in the formation of organohalogen structures,1,3,4 it is of interest to evaluate quantitatively the chemical forms of unburned carbon. However, most researchers have paid no attention to this topic. The main objective of the present work is, thus, to investigate in detail the functionalities of the unburned carbon present in coal fly ashes with temperatureprogrammed desorption (TPD)/temperature-programmed oxidation (TPO) and laser Raman spectroscopy (LRS) techniques. It has been accepted that TPD is one of the most powerful methods to determine surface oxygen-functional groups of carbonaceous materials,5−9 whereas TPO and LRS are both effective as the analytical methods to evaluate carbon structures quantitatively.10−12 © 2015 American Chemical Society

SAMPLES AND METHODS Fly Ash Samples. Six kinds of coal fly ashes, which were supplied from the Japan Coal Energy Center, were used in the present work. These samples were equally divided into approximately 500 mg at room temperature and stored in N2-purged plastic bags. The chemical composition of each ash sample is shown in Table 1. As expected, the main component was SiO2 in every case, followed by Al2O3, respectively within the ranges of 47−63 and 19−33 mass%. Amounts of carbon element ranged from 0.4 to 4.1 mass%. Of the halogens, fluorine element was present in the concentration range of 20− 130 μg/g, the vast majority in organic forms.1 Chlorine element was as low as less than 10 μg/g, regardless of the ash type. XPS Analysis. To examine surface carbon forms of the ash samples, the C 1s X-ray photoelectron spectroscopy (XPS) analyses were conducted with a Mg−Kα X-ray source. Each sample was kept in place on an In plate and analyzed under a high vacuum of 6 × 10−8 to 4 × 10−7 Pa, the number of scans being 30. The background of C 1s spectra was subtracted by the Shirley method, and the binding energy was referred to an In 3d5/2 peak of In2O3 at 444.9 eV.13 All of the C 1s spectra observed were also curve-fitted with the least-squares method using Gaussian peak shapes.1,2 The 1s peaks at 284.3, 285−286, Received: Revised: Accepted: Published: 5189

December March 18, March 20, March 20,

11, 2014 2015 2015 2015 DOI: 10.1021/es506023r Environ. Sci. Technol. 2015, 49, 5189−5194

Article

Environmental Science & Technology Table 1. Chemical Compositions of Fly Ash samples Used content (mass%-dry)

content (μg/g-dry)

ash sample

Ca

Na2Ob

MgOb

Al2O3b

SiO2b

K2Ob

CaOb

TiO2b

Fe2O3b

Fc

Clc

A B C D E F

0.41 0.81 1.2 1.7 1.7 4.1

2.6 1.7 1.0 0.34 0.18 0.14

1.3 1.4 1.4 2.3 1.9 1.4

28 19 20 29 31 33

55 63 63 51 49 47

0.66 1.0 1.8 0.78 0.68 0.51

4.2 6.0 5.8 7.4 8.9 8.2

0.94 0.76 0.94 1.5 1.8 1.6

4.9 4.8 4.0 5.1 4.3 3.5

20 (20)d 60 (60)d 130 (110)d 110 (100)d 60 (50)d 120 (110)d

nile nile nile nile nile nile

a

Determined with a conventional, combustion-type elemental analyzer. bDetermined by the XRF method. cDetermined according to the oxygen bomb method (ASTM D 2361). dOrganic CF forms estimated by the F 1s XPS method.1 eLess than 10 μg/g.

and 287.8 eV were assigned to graphitic carbon (CC),14,15 aliphatic carbon (CC and/or CH)16,17 and O-functional groups (CO, CO, and OCO),13,18 respectively. TPD and TPO Measurements. All runs were carried out with a flow-type fixed-bed quartz reactor. The details of the apparatus have been reported elsewhere.4 In a TPD run, approximately 100 mg of the ash charged into the reactor was heated at 5 °C/min up to 950 °C in a stream of high-purity He and quenched to room temperature. The ash in the reactor after the TPD was then subjected, without exposure to laboratory air, to the TPO experiment, in which the sample was reheated at 5 °C/min up to 950 °C under 10 vol % O2/He atmosphere. The amounts of CO2 and CO evolved in the TPD/TPO runs were analyzed online at 2.5 min intervals with a high-speed micro GC equipped with both PP-Q and MS-5A columns. Yields of these gas species were expressed in percent of total carbon in feed ash. The reproducibility was within ±1% in all cases. To quantify separately several sources of CO2 evolved in the TPD process, least-squares curve-fitting analyses of the rate profiles for the CO2 formation observed were performed with Gaussian peak shapes. 4 Upon deconvolution, the CO 2 formation was assumed to arise mainly from decomposition reactions of carboxyl groups, lactone/acid anhydride groups and surface CaCO3 around 300 °C,7 610 °C,8 and 650 °C,1,19 respectively, based on earlier work concerning surface functionalities of coal chars and carbons using the TPD technique. The details of this curve-fitting analysis have been described previously.4 LRS Analysis. The measurements were conducted to evaluate carbon structures in the ash samples after the TPD. In a typical analysis, the incident light beam of 488.0 nm from an Ar ion laser was focused into a spot with the size of 5 μm on each sample. The C LRS spectra observed were calibrated with respect to the signal position of graphite at 1575 cm−1,10 and the Raman intensities were normalized on the basis of carbon content in the sample. The background-removed C spectra were also deconvoluted into amorphous (A-) and crystallized (C-) carbon by the curve-fitting method using Gaussian peak shapes.20 In this method, the Raman peaks at 1355 and 1575 cm−1 were assigned to A-carbon20 and C-carbon,10 respectively.

in the bulk. All of the C 1s spectra observed were broad in the binding energy range of 281−291 eV, irrespective of the kind of the ash, and they had two distinct peaks around 285 and 288 eV. The latter peak was attributable to CO, CO, and O− CO bonds,13,18 and the C-signal deconvolution analysis1,2 indicated that the proportion of these O-functional groups was as high as 60−85 mol %. The O-functional groups might be formed through solid−gas interactions of the carbon on the ash surface with oxidizing gases (for examples, O2, NO, and SO2) in flue gas. Formation of CO2 and CO during TPD and TPO Runs. As well-known, surface O-functional groups on carbonaceous materials can release CO2 or CO upon gasification to provide active carbon sites. Since the carbon sites have been reported to exhibit approximately 100−1000 times higher reactivity than carbon atoms in condensed aromatic structures,21 they may work efficiently as the C-sources for the formation of organic CF forms (Table 1). It is thus of interest to investigate the behavior of the release of CO2 and CO in the TPD/TPO processes of the ash samples. Figure 1 shows the rate profiles of

Figure 1. Rate profiles for CO2 and CO evolved in the TPD runs of three fly ash samples with different carbon contents: (a) fly ash-B, (b) fly ash-E, and (c) fly ash-F.

CO2 and CO evolved during the TPD runs of fly ash-B, E, and F with different carbon contents. No appreciable amounts of CH4 and C2 hydrocarbons were detectable in all cases. With the ash-B (Figure 1a), CO2 formation started around 150 °C, exhibited a small, broad peak at about 325 °C and a large, sharp peak at 600 °C, and increased again after 850 °C. The ash-E (Figure 1b), F (Figure 1c), and the other three ashes gave almost the same rate profiles. However, CO started to evolve at a higher temperature than the case of CO2, regardless of the ash type, and the rate of CO formation tended to increase with increasing temperature in every case (Figure 1). Similar results



RESULTS Surface Carbon Forms of Fly Ash Samples. The XPS results for the ash samples used have been shown in detail elsewhere1 and are thus simply explained below. The XPS spectra of the six ashes revealed the surface enrichment of the carbon in all the samples by comparing atomic C/Si ratios obtained by the XPS and bulk analysis. The ratios were 80 × 10−2 to 120 × 10−2 at the surface and 3.6 × 10−2 to 40 × 10−2 5190

DOI: 10.1021/es506023r Environ. Sci. Technol. 2015, 49, 5189−5194

Article

Environmental Science & Technology

that the combination of the present TPD and TPO techniques can detect almost all of the carbon present in the ash. Structures of Unburned Carbon. In previous studies about the TPD analysis using coal chars and carbons,5−9 CO2 evolved came mainly from the decomposition of carboxyl groups, lactone/acid anhydride groups, and surface CaCO3. It is of interest to quantify separately different sources of CO2 evolved in the present TPD process. The rate profiles for the CO2 formation in Figure 1 were thus curve-fitted in a similar manner to that reported earlier.4 The typical deconvolution profiles for the ash-B, E, and F are given in Figure 3, where the

were also observed for the other three ashes. The evolution of CO2 and CO in the wide temperature range described above shows the presence of various O-functional groups on the unburned carbon in the ash samples, as indicated in the C 1s XPS spectra. As can also be expected, the behavior of the CO2 or CO evolution had no clear relationship with the bulk metal compositions shown in Table 1. The rates of CO2 and CO formed during the TPO runs after the experiments shown in Figure 1 are given in Figure 2, where

Figure 2. Rate profiles for CO2 and CO formed in the TPO runs after the experiments shown in Figure 1: (a) fly ash-B, (b) fly ash-E, and (c) fly ash-F.

the vertical scale of the rates is expressed with 1/24 times that in Figure 1. Compared with the TPD results (Figure 1), larger amounts of CO2 and CO were produced, irrespective of the kind of the ash. As seen in Figure 2, these gas species provided the rate profiles peaking at about 650 °C, and the rates of CO2 formation were always larger than those observed for CO. Such a trend was common with all of the ashes used. Cumulative yields of CO2 and CO evolved in the TPD/TPO experiments can be calculated by integrating the corresponding rate profiles shown in Figures 1 and 2. The results for all the ashes are compiled in Table 2. Yields of CO2 and CO evolved

Figure 3. Deconvolution of the rate profiles for CO2 formation shown in Figure 1: (a) fly ash-B, (b) fly ash-E, and (c) fly ash-F.

sources of the CO2 formation above 850 °C are denoted as other forms. The reproducibility of the present curve-fitting analysis fell within ±3% in every case. Table 3 summarizes the deconvolution results for all ash samples used. The extent of the contribution of carboxyl groups, lactone/acid anhydride groups, surface CaCO3, and other forms to the CO2 formation from the six samples depended strongly on the ash type and was estimated to be 3−8, 22−43, 26−43, and 17−48%, respectively. The sources of the CO2 observed at ≥850 °C remain still unknown, but the CO2 might be produced by the catalysis of some metal components present in the ashes. However, CO evolved during TPD has been reported to arise mostly from decomposition reactions of ketone, ether, and phenolic groups.5,9 As shown in Figure 2, the formation of CO2 and CO in the TPO took place predominantly between 550 and 750 °C, and the shape of each peak was symmetrical, such a trend being common for all the ashes. In other words, there was no significant difference in the TPO behavior among the ashes. Although the ashes were also subjected to the XRD measurements, no diffraction signals attributable to C species were detected in any cases. According to previous work,12 it has been reported that CO2 and CO formed during TPO originate mainly from carbon atoms in condensed heterocyclic ring structures remaining after the complete decomposition of Ofunctional groups on carbon surface. To investigate carbon structures in the ashes after the TPD, the samples were

Table 2. Cumulative Yields of CO2 and CO Evolved during TPD and TPO Runs TPD in high-purity He

TPO in 10% O2/ He

CO2

CO

CO2

CO

carbon mass balance

ash sample

(C%)a

(C%)a

(C%)a

(C%)a

(%)

A B C D E F

9.5 9.6 11 6.5 3.7 3.8

9.6 4.6 3.7 2.9 2.1 1.6

62 61 63 61 63 63

23 22 26 27 29 32

104 97 104 97 98 100

a

Based on carbon content.

during the TPD up to 950 °C ranged from 3.7 to 11 C% and from 1.6 to 9.6 C%, respectively. These observations point out the difference in surface O-functional groups on the unburned carbon present in the ash samples. In the TPO runs, however, yields of CO2 and CO were 61−63 and 22−32 C%, respectively, and the former yield was thus higher than the latter one in all cases. As given in Table 2, carbon mass balances fell within the reasonable range of 97−104%, which indicates 5191

DOI: 10.1021/es506023r Environ. Sci. Technol. 2015, 49, 5189−5194

Article

Environmental Science & Technology Table 3. Deconvolution Results of the Rate Profiles for CO2 Formation Observed in the TPD Process contribution to CO2 formation (%) ash sample

carboxyl groups

lactone/acid anhydride groups

surface CaCO3

other forms

A B C D E F

8 4 4 3 5 3

23 41 22 33 43 39

43 37 26 35 35 37

26 18 48 29 17 21

On the basis of the above-mentioned results, the functional forms of the unburned carbon present in the ashes used are estimated and summarized in Figure 5, where the proportion of

supplied to the LRS measurements. Figure 4 presents typical LRS spectra for C element in the ash-B, E and F samples. The

Figure 4. Typical LRS spectra for carbon element in fly ash-B (a), E (b) and F (c) samples. Figure 5. Functional forms of unburned carbon estimated by the TPD/TPO and LRS methods.

original Raman intensity tended to be larger for the ash with a higher carbon content. As shown by a solid line in Figure 4, each spectrum observed had two distinct peaks around 1355 and 1575 cm−1. This tendency was common for all the samples. The peaks at 1355 and 1575 cm−1 can be assigned to A-carbon and C-carbon, respectively.10,20 To evaluate quantitatively these forms, all of the C LRS spectra observed were deconvoluted into contributions due to A-carbon and T-carbon in the same manner as that reported previously.20 The deconvolution results are shown as the dashed lines in Figure 4 and also summarized in Table 4. The reproducibility of this analytical method was within ±1% in every case, and the proportion of Acarbon or C-carbon was calculated to be in the range of 58−70 or 30−42%, respectively.

each component is presented as a percentage of total carbon content (Table 1) in each sample. As can be expected from the results of the TPD/TPO and LRS measurements, A-carbon was the main form of the unburned carbon, and the proportion was in the range of 51−66 C%, whereas C-carbon ranged from 25 to 37 C%. The proportion of carboxyl groups and ketone/ ether/phenolic groups were the highest with the ash-A. However, lactone/acid anhydride groups were the highest with the ash-B. Surface CaCO 3 was present in the concentration range of 1.3−4.1 C%. These results point out that the chemical forms of the unburned carbon depend strongly on the kind of the ash. When the proportion of each C-form was plotted as a function of the content of Na, Mg, Al, Si, K, Ca, Ti, or Fe element in fly ash given in Table 1, no distinct relationship could be observed in any cases. It may thus be reasonable to conclude that these metal contents are not important as the factors for determining the functionalities of the unburned carbon. Although some issues to be solved in future work may remain, the present TPD/TPO and LRS methods may be effective as the techniques to quantitatively evaluate the functional forms of unburned carbon in coal fly ashes.

Table 4. Carbon Structures in Fly Ash Samples after TPD Runs type of carbon (%) ash sample

amorphous

crystallized

A B C D E F

70 66 58 58 67 70

30 34 42 42 33 30 5192

DOI: 10.1021/es506023r Environ. Sci. Technol. 2015, 49, 5189−5194

Article

Environmental Science & Technology



DISCUSSION It is well-known that surface O-functional groups on coal chars and carbons are decomposed into CO2 and CO during gasification to give active carbon sites, which can work as reactive sites with gasifying agents, such as O2, H2O, and CO2.22,23 It may thus be possible that active sites from the Ofunctional groups given in Figure 5 play important roles in the formation of organic halogen compounds. To discuss this point, the concentration of organic CF forms shown in Table 1 was plotted against the amount of O-functional groups determined by the TPD method. Figure 6 presents the

plotted against the content of Na, Mg, Al, Si, K, Ca, Ti, or Fe element in fly ash given in Table 1, and the results obtained showed that organic fluorine forms tended to be richer in the ash with a higher content of Ca element. This observation may suggest that part or all of the Ca in the samples controls the formation of covalent CF bonds. The concentration was then plotted as a function of the amount of surface CaCO3 shown by TPD. The results are illustrated in Figure 7, where

Figure 7. Relationship between organic CF forms and surface CaCO3 in fly ash samples.

the amount is calculated from the CO2 peak observed at 650 °C in Figure 3. Although the number of data points is insufficient, it seemed that the ash with a larger amount of the CaCO3 had a higher concentration of organic F functionalities. The present authors’ research group has recently found that HCl can react readily with active carbon sites in activated carbon at 500 °C to yield organic CCl forms,27,28 and the extent of the formation increases by addition of Ca2+ cations to the carbon.29 It may thus be possible that the CaCO3 shown by TPD exists as CaO around 650 °C downstream after combustion and promotes the formation of covalent CF bonds at the surface. As shown in Figures 6 and 7, the concentration of organic CF forms tended to level off when carboxyl/lactone/acid anhydride groups and surface CaCO3 exceeded about 400 μg-C/g and 400 μg-Ca/g, respectively. These observations may indicate the importance of HF concentration in flue gas in the organic fluorine formation. On the basis of the above-described results and discussion, active carbon sites derived from carboxyl/lactone/acid anhydride groups may react with HF in flue gas to give Fcontaining intermediates (eqs 1 and 2), and the transformation of C(HF) and C(F) into organic CF forms might occur around 600−700 °C where the decomposition of surface CaCO3 in the ashes into CaO proceeds (Figure 3). The role of surface Ca species in the formation of organic fluorine forms is not clear at present and should be clarified in future work. In this work, the functionalities of the unburned carbon present in six kinds of coal fly ashes have been studied mainly by means of TPD/TPO and LRS methods. The conclusions are summarized as follows: (1) Amorphous carbon is the main Cform in the ash samples, and crystallized carbon is present in the range of approximately 25−40% on a carbon basis. (2) The results of the TPD show the formation of carboxyl/lactone/ acid anhydride/ketone/ether/phenolic groups on the surface of the unburned carbon. (3) It is also indicated that the ashes contain small amounts of surface CaCO3. (4) Interestingly, there is a significant correlation between organic fluorine concentrations and carboxyl/lactone/acid anhydride groups or surface CaCO3 contents in the ashes.

Figure 6. Relationship between organic CF forms and carboxyl/ lactone/acid anhydride groups in fly ash samples.

relationship between organic F forms and carboxyl/lactone/ acid anhydride groups in the ash samples. Interestingly, there was a significant correlation between the two. When the F concentration was plotted against the amount of ketone/ether/ phenolic groups, on the other hand, no distinct relationship could be deduced. In addition, no occurrence of organic fluorine in coal has been reported.24 Thus, these observations strongly suggest that carboxyl/lactone/acid anhydride groups can affect the formation of organic F functionalities. Part or all of active sites derived from the groups may react secondarily with HF (and/or F2) evolved during coal combustion to provide covalent CF bonds. The groups are likely to be present as active sites around 700 °C downstream of pulverized coal-fired boilers. When a free active site is denoted as C( ), the reaction with HF might be expressed as follows: C( ) + HF → C(HF)

(1)

where C(HF) may mean HF-containing intermediates, such as chemisorbed HF and surface HF species. In this process, HF might undergo dissociation reactions to provide two surface species, C(F) and C(H), through the following equation: 2C( ) + HF → C(F) + C(H)

(2)

Although one may also think that part of the HF reacts with O2 in the processes of pulverized coal combustion and/or subsequent flue gas treatment to form H2O and F2 with much stronger fluorination ability, this reaction is unfavorable thermodynamically, because the standard Gibbs free energy changes at 300−1500 °C are as large as +81 to +99 kcal/mol. It is well-known that alkali, alkaline earth, and transition metals in unburned carbon work as the catalysts for de novo synthesis of chlorinated aromatic structures including dioxins.25,26 It is possible that some minerals in the ash samples may affect the transformation of C(HF) and C(F) into organic C F forms. To examine this point, the concentration was first 5193

DOI: 10.1021/es506023r Environ. Sci. Technol. 2015, 49, 5189−5194

Article

Environmental Science & Technology

(13) Naumkin, A. V.; Kraut-Vass, A.; Gaarenstroom, S. W.; Powell, C. J. NIST X-ray photoelectron spectroscopy database; NIST standard reference database 20, version 4.1 (web version); NIST: Gaithersburg, MD, 2012. (14) Johansson, G.; Hedman, J.; Berndtsson, A.; Klasson, M.; Nilsson, R. Calibration of electron spectra. J. Electron Spectrosc. Relat. Phenom. 1973, 2, 295−317. (15) Kieser, J.; Kleber, R. A new approach for the determination of the C 1s binding energy. Appl. Phys. 1976, 9, 315−319. (16) Puziy, A. M.; Poddubnaya, O. L. The properties of synthetic carbon derived from nitrogen- and phosphorus-containing polymer. Carbon 1998, 36, 45−50. (17) Bébin, P.; Prud′homme, R. E. Comparative XPS study of copper, nickel, and aluminum coatings on polymer surfaces. Chem. Mater. 2003, 15, 965−973. (18) Moulder, J. F.; Stichle, W. F.; Sobol, P. E.; Bomben, K. D. Handbook of X-ray Photoelectron Spectroscopy; Chastain, J., Ed.; PerkinElmer: Eden Praine, MN, 1992. (19) Ohtsuka, Y.; Tomita, A. Calcium catalyzed steam gasification of Yallourn brown coal. Fuel 1986, 65, 1653−1657. (20) Filik, J.; May, P. W.; Pearce, S. R. J.; Wild, R. K.; Hallam, K. R. XPS and laser Raman analysis of hydrogenated amorphous carbon films. Diam. Relat. Mater. 2003, 12, 974−978. (21) Radovic, L. R.; Lizzio, A. A.; Jiang, H. Reactive surface area: An old but new concept in carbon gasification. In Fundamental Issues in Control of Carbon Gasification Reactivity; Lahaye, J., Ehrburger, P., Eds.; Kluwer Academic Publishers: Dordrecht, The Netherlands, 1991; pp 235−255. (22) Laine, N. R.; Vastola, F. J.; Walker, P. L. The importance of active surface area in the carbon-oxygen reaction. J. Phys. Chem. 1963, 67, 2030−2034. (23) Radovic, L. R.; Walker, P. L.; Jenkins, R. G. Importance of carbon active sites in the gasification of coal chars. Fuel 1983, 62, 849− 856. (24) Davidson, R. M. Chlorine and Other Halogens in Coal; IEA Coal Research: London, UK, 1996 (IEAPER/28). (25) Stieglitz, L.; Vogg, H.; Zwick, G.; Beck, J.; Bautz, H. On formation conditions of organohalogen compounds from particulate carbon of fly ash. Chemosphere 1991, 23, 1255−1264. (26) Kuzuhara, S.; Kasai, E. Formation of PCDD/Fs during oxidation of carbonaceous materials at low temperatures. Tetsu To HaganeJ. Iron Steel Inst. Jpn. 2003, 89, 811−818. (27) Tsubouchi, N.; Saito, T.; Ohtaka, N.; Ohtsuka, Y. Evolution of hydrogen chloride and change in the chlorine functionality during pyrolysis of Argonne premium coal samples. Energy Fuels 2013, 27, 87−96. (28) Tsubouchi, N.; Saito, T.; Ohtaka, N.; Nakazato, Y.; Ohtsuka, Y. Chlorine release during fixed-bed gasification of coal chars with carbon dioxide. Energy Fuels 2013, 27, 5076−5082. (29) Ohtaka, N.; Tsubouchi, N.; Sato, M.; Suzuki, N.; Ohtsuka, Y. Secondary reactions of HCl during coal pyrolysis: Studies on reactions of HCl with model carbons prepared from phenol resin. In Proceedings of the 43rd Conference of the Japan Institute of Energy for Coal Science; The Japan Institute of Energy: Tokyo, Japan, 2006; pp 101−102.

It may thus be reasonable to suppose that the amounts of the O-functional groups and CaCO3 are key factors in controlling the formation of organic CF forms. This information may help us to develop an efficient method for inhibiting organic fluorine compounds formed in pulverized coal-fired plants. To investigate the behavior of the formation and decomposition of the O-containing species in the presence of flue gas components should be the subject of future work from a practical point of view.



AUTHOR INFORMATION

Corresponding Author

*Phone: +81-11-706-6850; fax: +81-11-726-0731; e-mail: [email protected] (N.T.). Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This study was supported in part by a Grant-in-Aid for Scientific Research (B) from the Ministry of Education, Culture, Sports, Science, and Technology, Japan. The authors acknowledge the supply of fly ashes from the Japan Coal Energy Center.



REFERENCES

(1) Tsubouchi, N.; Hayashi, H.; Kawashima, A.; Sato, M.; Suzuki, N.; Ohtsuka, Y. Chemical forms of the fluorine and carbon in fly ashes recovered from electrostatic precipitators of pulverized coal-fired plants. Fuel 2011, 90, 376−383. (2) Tsubouchi, N.; Kasai, E.; Kawamoto, K.; Noda, H.; Nakazato, Y.; Ohtsuka, Y. Functional forms of carbon and chlorine in dust samples formed in the sintering process of iron ores. Tetsu To HaganeJ. Iron Steel Inst. Jpn. 2005, 91, 751−756. (3) Tsubouchi, N.; Hashimoto, H.; Ohtaka, N.; Ohtsuka, Y. Chemical characterization of dust particles recovered from bag filters of electric arc furnaces for steelmaking: Some factors influencing the formation of hexachlorobenzene. J. Hazard. Mater. 2010, 183, 116− 124. (4) Tsubouchi, N.; Kuzuhara, S.; Kasai, E.; Hashimoto, H.; Ohtsuka, Y. Properties of dust particles sampled from windboxes of an iron ore sintering plant: Surface structures of unburned carbon. ISIJ. Int. 2006, 46, 1020−1026. (5) Marchon, B.; Carrazza, J.; Heinemann, H.; Somorjai, G. A. TPD and XPS studies of O2, CO2, and H2O adsorption on clean polycrystalline graphite. Carbon 1988, 26, 507−514. (6) Otake, Y.; Jenkins, R. G. Characterization of oxygen-containing surface complexes created on a microporous carbon by air and nitric acid treatment. Carbon 1993, 31, 109−121. (7) Zhuang, Q.-L.; Kyotani, T.; Tomita, A. The change of TPD pattern of O2-gasified carbon upon air exposure. Carbon 1994, 32, 539−540. (8) Zhuang, Q.-L.; Kyotani, T.; Tomita, A. DRIFT and TK/TPD analyses of surface oxygen complexes formed during carbon gasification. Energy Fuels 1994, 8, 714−718. (9) Zielke, U.; Huttinger, K. J.; Hoffman, W. P. Surface-oxidized carbon fibers: I. Surface structure and chemistry. Carbon 1996, 34, 983−998. (10) Tuinstra, F.; Koenig, J. L. Raman spectrum of graphite. J. Chem. Phys. 1970, 53, 1126−1130. (11) Tuinstra, F.; Koenig, J. L. Characterization of graphite fiber surfaces with Raman spectroscopy. J. Compos. Mater. 1970, 4, 492− 499. (12) Aso, H.; Matsuoka, K.; Sharma, A.; Tomita, A. Evaluation of size of graphene sheet in anthracite by a temperature-programmed oxidation method. Energy Fuels 2004, 18, 1309−1314. 5194

DOI: 10.1021/es506023r Environ. Sci. Technol. 2015, 49, 5189−5194