Adsorption of sulfur dioxide on ammonia-treated activated carbon fibers

Adsorption of sulfur dioxide on ammonia-treated activated carbon fibers

Carbon 39 (2001) 1689–1696 Adsorption of sulfur dioxide on ammonia-treated activated carbon fibers C.L. Mangun a , *, J.A. DeBarr b , J. Economy a a ...

306KB Sizes 0 Downloads 33 Views

Carbon 39 (2001) 1689–1696

Adsorption of sulfur dioxide on ammonia-treated activated carbon fibers C.L. Mangun a , *, J.A. DeBarr b , J. Economy a a

Department of Materials Science and Engineering, University of Illinois, 1304 W. Green Street, Urbana, IL 61801, USA b Illinois State Geological Survey, 615 E. Peabody Drive, Champaign, IL 61820, USA Received 27 June 2000; accepted 31 October 2000

Abstract A series of activated carbon fibers (ACFs) and ammonia-treated ACFs prepared from phenolic fiber precursors have been studied to elucidate the role of pore size, pore volume, and pore surface chemistry on adsorption of sulfur dioxide and its catalytic conversion to sulfuric acid. As expected, the incorporation of basic functional groups into the ACFs was shown as an effective method for increasing adsorption of sulfur dioxide. The adsorption capacity for dry SO 2 did not follow specific trends; however the adsorption energies calculated from the DR equation were found to increase linearly with nitrogen content for each series of ACFs. Much higher adsorption capacities were achieved for SO 2 in the presence of oxygen and water due to its catalytic conversion to H 2 SO 4 . The dominant factor for increasing adsorption of SO 2 from simulated flue gas for each series of fibers studied was the weight percent of basic nitrogen groups present. In addition, the adsorption energies calculated for dry SO 2 were shown to be linearly related to the adsorption capacity of H 2 SO 4 from this flue gas for all fibers. It was shown that optimization of this parameter along with the pore volume results in higher adsorption capacities for removal of SO 2 from flue gases.  2001 Elsevier Science Ltd. All rights reserved. Keywords: A. Activated carbon; Carbon fibers; B. Activation; C. Adsorption; D. Surface properties

1. Introduction The introduction of sulfur dioxide into the atmosphere is of major concern because it is the principal cause of acid rain. Sulfur dioxide can react slowly with oxygen to form the trioxide that subsequently reacts with water to form sulfuric acid. Many methods have been devised to reduce emissions of SO 2 . Some utilities have switched to lowsulfur coal or they cleanse the coal to remove sulfur prior to burning. Another method is to remove the SO 2 from the flue gas using either wet or dry scrubbing. Wet methods, such as lime-limestone scrubbers, are highly efficient but they tend to require large capital investment and / or produce sludge, which must be disposed of. Recently, more attention has been invested in developing dry methods for removal of sulfur dioxide because of their potential

*Corresponding author. Tel.: 11-217-244-2523; fax: 11-217333-2736. E-mail address: [email protected] (C.L. Mangun).

simplicity and lower cost. One such method that shows great promise is using beds of adsorbent materials such as activated carbons [1]. Activated carbons have been shown to be effective adsorbents for SO 2 and in the presence of oxygen and moisture, they can catalyze the conversion of SO 2 to H 2 SO 4 thus increasing adsorption capacity [2]. Activated carbon fibers (ACFs) are especially interesting since they have a uniform micropore structure, faster adsorption kinetics, and a lower pressure drop as compared to granular activated carbons [3,4]. Several researchers have studied adsorption of SO 2 onto ACFs produced from pitch [5,6] and polyacrylonitrile (PAN) [7–9]. Additionally, it has been shown that specific chemical groups on the surface of carbon micropores can enhance adsorption capacity toward SO 2 . Mochida et al. found that using a fixed bed of PAN-based carbon fibers was very effective for removal of SO 2 from flue streams [7]. They suggested that surface nitrogen groups might be responsible for the high adsorption capacities by PAN fibers; however no further analysis was done [8]. Fei et al. [5] studied pitch fibers that had been heat treated to de-

0008-6223 / 01 / $ – see front matter  2001 Elsevier Science Ltd. All rights reserved. PII: S0008-6223( 00 )00300-6


C.L. Mangun et al. / Carbon 39 (2001) 1689 – 1696

compose surface oxygen functional groups resulting in increased activity for conversion of SO 2 to H 2 SO 4 . More recently, DeBarr and Lizzio have attempted to optimize activated carbons for SO 2 removal [10–12]. The fundamentals of the SO 2 –carbon reaction were studied in an attempt to learn the properties of activated carbons that influence their SO 2 removal. Carbons with varying pore structure and surface chemistry were prepared from an Illinois coal and tested for their ability to remove SO 2 from a simulated flue gas. A novel carbon preparation involving nitric acid treatment followed by thermal desorption of carbon–oxygen complexes was developed to increase SO 2 adsorption capacity to levels greater than that of the best available commercial carbon. A new mechanism was proposed that correlated SO 2 removal with the number of free sites available on the carbon surface. In a follow-up study, the SO 2 -removal properties of ACFs were determined [13–15] and compared to those of activated coal chars [14]. The uptake of SO 2 by both coal chars and ACFs was comparable. It was found that the number of free sites along with the pore size distribution and total pore volume were important factors in determining the carbon’s SO 2 removal behavior. Our research group has also been investigating the effect of pore size and surface chemistry on adsorption properties [16–19]. Understanding the effect that pore structure has on adsorption allows activated carbons to be tailored for selective and / or enhanced removal of specific contaminants. The present paper provides a systematic study of the role of basic nitrogen groups in ACFs on adsorption of SO 2 from both dry and flue environments. Basic groups are introduced either by reaction of ACFs with ammonia at high temperatures or by directly activating the phenolic fiber with ammonia.

2. Materials and methods

manner as above except that ammonia was used immediately upon heating.

2.2. Measurement of nitrogen isotherms The adsorption isotherms were measured with a Coulter Omnisorb 100 using nitrogen at 77 K. The activated fibers were outgassed for 24 h under vacuum at 1508C prior to testing. The adsorption experiments were performed in static mode, which allowed sufficient time to reach equilibrium. In all cases, Type I isotherms were observed which is indicative of microporous carbons. These data were then used to calculate the BET surface area using the standard method applied over a relative pressure range of 0.01 to 0.20. In addition, the average micropore size (x) and micropore volume (Wo ) were also calculated by applying the Dubinin–Radushkevich [20] equation to the experimental nitrogen isotherm: log (W ) 5 log (Wo ) 2 M 3 log 2 (Po /P)


where P/Po is the relative pressure of the adsorbate, and W is the amount of gas / vapor adsorbed in ml / g. The micropore volume Wo is calculated from the intercept of a 2 log (W ) vs. log (Po /P) plot, while the slope, M, of the best fit line is related to the interaction energy, E, by M 5 2 2.303 3 (RT /E)2


where R is the ideal gas constant and T is the adsorption temperature in Kelvin x 5 b k /E.


In Eq. (3), the micropore half width x is related to the adsorption energy, b is the similarity coefficient that is equal to 0.33 for nitrogen, and k is a structural constant. In previous ACF studies, this equation has shown excellent correlation when compared to direct observation of the porosity using STM [19].

2.1. Preparation of activated carbon fibers 2.3. Adsorption of sulfur dioxide The ACFs are commercially available from Nippon Kynol and are produced by steam / CO 2 activation of a woven phenolic fiber. In this study, three fiber types were used: ACF-10, ACF-15, and ACF-25 with N 2 BET surface areas of 730, 1585, and 1890 m 2 / g, respectively. The ammonia treatment of the ACFs consisted of placing a sample of fiber in a Lindberg tube furnace under flowing nitrogen. The furnace was heated to the desired temperature and then the gas was switched to ammonia. The samples were held for a predetermined time and then cooled to room temperature. The samples are denoted using the fiber type, the temperature, and time (e.g. A10500-60 is an ACF-10 treated at 5008C for 60 min). In addition, several samples were produced by direct ammonia activation of the phenolic precursor fabric (denoted as BCFs). These samples were produced in the same

Experiments were carried out using either dry SO 2 or a mixture of SO 2 / O 2 / H 2 O to mimic a typical flue gas. Two separate systems were used to measure the weight gain of the adsorbate with time. In both cases, 30–50 mg of the fiber were dried under nitrogen at 1508C to remove any adsorbed contaminants. For the experiments with dry SO 2 , a TA Instruments TGA-951 was used to determine the equilibrium weight gain at several concentrations. This was done by using mass flow controllers (Tylan General) to dilute a standard mixture (5000 ppmv SO 2 in nitrogen) to lower concentrations. For the experiments with simulated flue gas, a Cahn TG-131 TGA was used. The sample was held isothermally at 1208C and then the gas was switched to a mixture of 5% O 2 , 7% H 2 O, and the balance nitrogen. Once the weight of the sample stabilized, SO 2 at

C.L. Mangun et al. / Carbon 39 (2001) 1689 – 1696

a concentration of 2500 ppmv was introduced and the weight increase of the sample was recorded with time.

3. Results and discussion

3.1. Characterization of pore structure Previous researchers have studied ammonia treatment of carbon and shown that ammonia decomposes to free radicals such as NH 2 or NH which at higher temperatures attack the carbon leading to pore formation and creation of nitrogen-containing functional groups [21,22]. Table 1 lists the surface area, micropore volume, and average pore width of all the samples used in SO 2 adsorption experiments. Generally, for each series of activated carbon fibers, there is an increase in all three parameters indicating that ammonia is acting as an activating agent. Longer times and / or increased temperatures result in more activation thus increasing the pore size. For the ammonia-activated phenolics, however, the pore width decreases until a temperature of 8008C is reached. Presumably, these samples are not fully carbonized at the lower temperatures resulting in a partially charred carbon structure. As the treatment temperature is raised, there is moderate shrinkage of the carbon matrix to compensate for the weight loss during charring resulting in a narrowing of the pores. By

Table 1 Physical characteristics of activated carbon fibers Sample

BET surface 2 area (m / g)

Wo (ml / g)

Average pore ˚ width (A)

ACF-10 ACF-15 ACF-25

730 1585 1890

0.319 0.670 0.763

A10-500-10 A10-500-60 A10-800-10 A10-800-60

766 734 886 1356

A15-500-10 A15-500-60 A15-600-60 A15-700-60 A15-800-10 A15-800-60


8008C, significant activation has taken place; thus the pores become much wider. The results of elemental analysis on these samples are shown in Table 2. Since there is little nitrogen present in the precursor material, it can be assumed that the increase in nitrogen is due to addition of groups at the carbon edge sites where they can directly affect adsorption properties. The results indicate that in each series of fibers the amount of nitrogen introduced increases with increased time and / or temperature of treatment. The type of nitrogen present is dependent on the temperature of reaction. The 5008C treatment yields amides, aromatic amines, and nitrile groups while higher temperatures result in formation of aromatic amines and pyridines (a more thorough analysis of the functional groups can be found in Ref. [23]). Comparing the ACFs one notes that for similar treatments, more nitrogen is incorporated into the ACF-10 structure followed by the ACF-15 and then the ACF-25. This trend can be explained by examining the amount of pore widening that occurs for each set of samples. For ACF-10, the pore size increases considerably more for a given treatment as compared to the other fibers. The activation process allows nitrogen to become incorporated into the carbonaceous structure. By the same argument, one notes that the ammonia activated phenolics have even larger amounts of nitrogen present (up to 9.27 wt.%) since all of the pore formation had to occur through ammonia treatment. The elemental analysis measures only carbon, hydrogen,

Table 2 Elemental analysis of activated carbon fibers Sample

Wt.% carbon

Wt.% nitrogen

Wt.% hydrogen

8.1 12.2 13.4

ACF-10 ACF-15 ACF-25

91.4 93.2 95.2

0.16 0.29 0.24

0.33 0.17 0.06

0.327 0.313 0.382 0.572

8.3 8.1 9.0 12.3

A10-500-10 A10-500-60 A10-800-10 A10-800-60

92.4 92.8 87.3 84.0

0.92 1.44 4.06 6.06

0.43 0.45 0.38 0.37

1619 1594 1602 1598 1711 2044

0.684 0.674 0.677 0.676 0.706 0.792

12.2 12.2 12.3 12.3 12.8 14.2

A15-500-10 A15-500-60 A15-600-60 A15-700-60 A15-800-10 A15-800-60

95.6 96.2 96.8 94.0 92.7 91.0

0.93 0.94 1.14 2.42 3.46 4.25

0.20 0.14 0.16 0.10 0.05 0.13

A25-500-10 A25-500-60 A25-800-10 A25-800-60

1893 1965 1999 2195

0.768 0.791 0.796 0.831

13.4 13.6 13.8 14.4

A25-500-10 A25-500-60 A25-800-10 A25-800-60

97.2 96.4 94.4 90.5

0.55 0.61 2.78 3.99

0.17 0.40 0.06 0.07

BCF-500 BCF-600 BCF-700 BCF-800

582 579 718 1846

0.247 0.246 0.308 0.657

9.6 9.0 8.3 15.0

BCF-500 BCF-600 BCF-700 BCF-800

88.7 89.3 84.0 80.4

1.98 2.76 6.71 9.27

2.71 1.89 0.83 0.94


C.L. Mangun et al. / Carbon 39 (2001) 1689 – 1696

Fig. 1. Adsorption isotherms of dry SO 2 at 258C on as-received ACFs.

Fig. 3. Adsorption isotherms of dry SO 2 at 258C on ammoniatreated ACF-15.

and nitrogen thus leaving an additional 5–10 wt.% unaccounted. Presumably, this difference is due to experimental error from adsorbed water in combination with the presence of surface oxygen groups such as phenolic hydroxyls and quinones (note that lower temperature treatment with ammonia produces amides which also contribute). These types of groups are neutral to slightly acidic in nature and thus should have little effect on the adsorption capacity of a strong acid such as sulfur dioxide.

3.2. Adsorption of dry SO2 Figs. 1–5 show the adsorption capacity in mg of SO 2 per gram of carbon over a concentration range of 500– 5000 ppmv for each set of ACFs. Fig. 1 compares the as-received ACFs where ACF-10 shows the largest adsorption capacity. The smaller pore size of the ACF-10 results in a higher overlap in potential between the pore

Fig. 2. Adsorption isotherms of dry SO 2 at 258C on ammoniatreated ACF-10.

Fig. 4. Adsorption isotherms of dry SO 2 at 258C on ammoniatreated ACF-25.

Fig. 5. Adsorption isotherms of dry SO 2 at 258C on ammoniaactivated phenolics (BCFs).

C.L. Mangun et al. / Carbon 39 (2001) 1689 – 1696

walls so the SO 2 molecule is more tightly bound at these lower concentrations. At much higher concentrations, the adsorption capacity for the ACF-10 would level out due to limited pore volume thus allowing the higher surface area fibers to display a larger capacity. Indeed Kaneko et al. [6] reported that higher surface area pitch fibers adsorbed more SO 2 because their study was done in this high concentration regime (.10 5 ppm). In our work, the SO 2 concentration range is more typical of actual concentrations from coal-fired utilities. Fig. 2 shows that as basic nitrogen groups are introduced, there is an enhancement in the adsorption of SO 2 . However, it does not scale with nitrogen content since the best fiber is A10-800-10. Although A10-800-60 has more basic groups, the pore has been widened significantly thus decreasing this fiber’s effectiveness for adsorbing SO 2 . The same argument holds for ACF-15 (Fig. 3); in this case the 5008C treatment has a deleterious effect on adsorption. For ACF-25 (Fig. 4) the slight widening of the pore for A25-500-60 is enough to reduce adsorption but here the increase in basic groups is more important than pore widening for the 8008C treatments. Finally, for the ammonia-activated phenolics (Fig. 5), BCF-700 has the optimal combination of nitrogen groups and pore size thus giving it the advantage over the other fibers. To better compare these results, Eqs. (1) and (2) were used to calculate an overall adsorption energy for dry SO 2 . Table 3 shows that the adsorption energy (ESO 2 ) for each


series of samples increases with increasing nitrogen content. These adsorption energies are affected by both physical and chemical parameters. The physical contribution increases as the pore size is decreased while the chemical contribution is expected to increase as more basic nitrogen groups are introduced into the micropores. Therefore, the ammonia treatment results in a competition between increasing the chemical interaction from increased nitrogen versus decreasing the physical interaction because the pore is being widened. Since the adsorption energies always increase, the chemical contribution must be increasing faster than the physical portion is decreasing. In Fig. 6 the adsorption energy (kJ / mol) for SO 2 is plotted versus the wt.% nitrogen for each set of fibers. For the smaller pore fibers, the increased nitrogen content does not increase the adsorption energy as quickly as compared to the larger pore materials. This is because the smaller pores already have a high affinity for adsorption of SO 2 so pore widening with ammonia treatment is much more detrimental to the physical energy contribution thus making the overall energy increase more slowly. The higher surface area fibers show that a moderate increase in nitrogen content greatly increases the adsorption energy for SO 2 . This is because the pore size does not increase significantly with treatment so the physical energy contribution is not reduced substantially. Therefore the introduction of basic groups is thought to dominate by increasing the chemical energy contribution.

3.3. Adsorption of flue SO2 Table 3 Correlation of adsorption properties Sample

ESO 2 (kJ / mol)

H 2 SO 4 adsorbed (ml / g)

% of pore volume utilized

ACF-10 ACF-15 ACF-25

11.42 11.01 10.81

0.170 0.080 0.056

53.2 11.9 7.4

A10-500-10 A10-500-60 A10-800-10 A10-800-60

11.59 12.00 13.08 14.03

0.216 0.225 0.267 0.388

66.1 71.9 69.8 67.8

A15-500-10 A15-500-60 A15-600-60 A15-700-60 A15-800-10 A15-800-60

11.58 11.59 11.74 13.20 13.65 15.46

0.091 0.099 0.150 0.253 0.337 0.460

13.3 14.6 22.2 37.4 47.7 58.1

A25-500-10 A25-500-60 A25-800-10 A25-800-60

10.90 10.89 12.18 14.66

0.066 0.089 0.163 0.336

8.6 11.3 20.5 40.5

BCF-500 BCF-600 BCF-700 BCF-800

11.51 12.03 13.15 15.37

0.159 0.179 0.241 0.381

64.5 72.6 78.1 58.0

Figs. 7–10 show the adsorption of SO 2 from a typical flue stream over a period of |15 h. For each series of ACF, the adsorption capacity of SO 2 scales with increasing nitrogen content. The sample with the highest capacity turns out to be the ACF-15 sample treated for 1 h at 8008C. This is presumably due to having the right combination of

Fig. 6. Effect of nitrogen content on adsorption energies of dry SO 2 .


C.L. Mangun et al. / Carbon 39 (2001) 1689 – 1696

Fig. 7. Adsorption uptake curve for flue SO 2 on ammonia-treated ACF-10.

Fig. 8. Adsorption uptake curve for flue SO 2 on ammonia-treated ACF-15.

Fig. 9. Adsorption uptake curve for flue SO 2 on ammonia-treated ACF-25.

Fig. 10. Adsorption uptake curve for flue SO 2 on ammoniaactivated phenolics (BCFs).

pore volume and nitrogen content. The initial uptake of SO 2 is dependent on both the pore size and the amount of nitrogen present. For the most part, the pore size is not the dominant feature except for a few cases where one notes higher initial uptake for the smaller pore fiber followed by a crossover with the fiber containing more nitrogen groups. It is known that the SO 2 is converted to H 2 SO 4 upon ¨ ¨ [2] suggested adsorption into the ACFs. Juntgen and Kuhl that this type of catalytic activity is due to the presence of delocalized electrons at the edge sites of the carbons where oxygen can chemisorb and then react with SO 2 to form SO 3 which then reacts with water to form H 2 SO 4 . They state that the presence of nitrogen-containing groups should suppress this reaction based on an earlier reported mechanism. However, our data indicate that just the opposite is occurring. One can interpret these results on the basis that SO 2 is preferentially adsorbed on the basic pore surface thus allowing for eventual conversion to H 2 SO 4 . These results support the scheme put forth by Kisamori et al. [9] where SO 2 is adsorbing at the basic sites, followed by reaction with O 2 to form the trioxide, and then with water to form sulfuric acid. This type of scheme is also supported by the fact that adsorption capacity scales with the amount of basic nitrogen groups present. The SO 2 (H 2 SO 4 condensed) capacity of the fibers at 15 h is shown in Table 3. The percent of pore volume utilized was then calculated by dividing the amount of H 2 SO 4 adsorbed by the total micropore volume. For ACF-10 one notes that the percentage of pore filling changes very little with ammonia treatment. The small pore size of ACF-10 provides enough driving force to fill most of its pore volume even with no treatment. The introduction of basic groups allows ACF-10 to maintain its high percentage of pore filling as the pore size and volume increase, thus increasing the adsorption capacity. In contrast, the larger pore ACFs which do not provide as much overlap in adsorption potentials have a significant increase in the

C.L. Mangun et al. / Carbon 39 (2001) 1689 – 1696


different temperatures. This will give additional insight into using ammonia as an activating agent and be correlated to determine if specific groups are more advantageous than others in adsorption of acidic gases such as SO 2 .

Acknowledgements This research work was supported by the National Science Foundation (Grant[ DMR-9208545). We would like to thank Joe Hayes of American Kynol Inc. for his continuing support of this program and for providing the activated carbon fibers.

Fig. 11. Increased adsorption energies for dry SO 2 result in greater capacity for flue SO 2 .

amount of SO 2 that is adsorbed as the nitrogen content is increased. Thus, a higher portion of their pore volume is utilized.

3.4. Comparison of adsorption properties In Fig. 11 the adsorption energy of dry SO 2 is plotted against the adsorption capacity of flue SO 2 . This gives a good linear relationship indicating that, by optimizing the adsorption energy, one can increase the adsorption capacity for SO 2 from flue gases. Optimization of ESO 2 is a function of pore size and pore surface chemistry. However, the other concern is the amount of pore volume that is available. Although ACF-10 and BCF have high nitrogen contents and thus a high contribution from the chemical energies, they do not possess enough pore volume for high capacity. The result is that A15-800-60, with the highest adsorption energy and a significant pore volume, is the most advantageous fiber for adsorption of SO 2 from flue gases.

4. Conclusions The adsorption capacities for dry SO 2 are dependent on both the pore size and the amount of basic groups incorporated into the carbonaceous structure. Calculation of the adsorption energies from the DR equation is a good indicator of the effectiveness of the ACFs to adsorb SO 2 from flue gases. Therefore, by optimizing the adsorption energy and the pore volume simultaneously, one can achieve a high removal of SO 2 from flue gases using ammonia-treated ACFs. Recently, we have utilized various techniques (XPS, FTIR, and temperature programmed desorption) to analyze the types of nitrogen functional groups which are formed with ammonia treatment at

References [1] Lu GQ, Do DD. Preparation of economical sorbents for SO 2 and NO x removal using coal washery reject. Carbon 1991;29(2):207–13. ¨ ¨ H. Mechanisms and physical properties of [2] Juntgen J, Kuhl carbon catalysts for flue gas cleaning. In: Thrower PA, editor, Chemistry and physics of carbon, Vol. 22, New York: Marcel Dekker, 1989, pp. 145–95. [3] Lin RY, Economy J. Preparation and properties of activated carbon fibers derived from phenolic precursor. Appl Polym Symp 1973;21:143–52. [4] Economy J, Lin RY. Adsorption characteristics of activated carbon fibers. Appl Polym Symp 1976;29:199–211. [5] Fei Y, Sun Y, Givens E, Derbyshire F. Continuous removal of sulfur oxides at ambient temperature using activated carbon fibers and particulates. ACS Preprints 1995;40(4):1051–5. [6] Kaneko K, Nakahigashi Y, Nagata K. Microporosity and adsorption characteristics against NO, SO 2 , and NH 3 of pitch-based activated carbon fibers. Carbon 1988;26(3):327– 32. [7] Kisamori S, Kuroda K, Kawano S, Mochida I, Matsumura Y, Yoshikawa M. Oxidative removal of SO 2 and recovery of H 2 SO 4 over poly(acrylonitrile)-based active carbon fiber. Energy Fuels 1994;8(6):1337–40. [8] Mochida I, Hirayama T, Kisamori S, Kawano S, Fujitsu H. Marked increase of SO 2 removal ability of poly(acrylonitrile)-based active carbon fiber by heat treatment at elevated temperatures. Langmuir 1992;8(9):2290–4. [9] Kisamori S, Mochida I, Fujitsu J. Roles of surface oxygen groups on poly(acrylonitrile)-based active carbon fibers in SO 2 adsorption. Langmuir 1994;10(4):1241–5. [10] Lizzio AA, DeBarr JA. Effect of surface area and chemisorbed oxygen on the SO 2 adsorption capacity of activated char. Fuel 1996;75(13):1515–22. [11] Lizzio AA, DeBarr JA. The mechanism of SO 2 removal by carbon. ACS Preprints: Fuel Chem Div 1996;41(1):201–5. [12] Lizzio AA, DeBarr JA. Mechanism of SO 2 removal by carbon. Energy Fuels 1997;11(2):284–91. [13] DeBarr JA, Lizzio AA, Daley MA. Adsorption of SO 2 on bituminous coal char and activated carbon fiber. Energy Fuels 1997;11(2):267–71. [14] DeBarr JA, Lizzio AA, Daley MA. Adsorption of SO 2 on







C.L. Mangun et al. / Carbon 39 (2001) 1689 – 1696 bituminous coal char and activated carbon fiber prepared from phenol formaldehyde. ACS Preprints: Fuel Chem Div 1996;41(1):339–43. Daley MA, Mangun CL, DeBarr JA, Riha JA, Lizzio AA, Donnals GL, Economy J. Adsorption of SO 2 onto oxidized and heat-treated activated carbon fibers. Carbon 1997;35(3):411–7. Andreopoulos AG, Economy J. Design of high-capacity, reversible polymeric adsorbents for acidic gases. Polym Adv Technol 1991;2:87–91. Economy J, Foster KL, Andreopoulos AG, Jung J. Tailoring carbon fibers for adsorbing volatiles. CHEMTECH 1992;22(10):597–603. Foster KL, Fuerman RG, Economy J, Larson SM, Rood MJ. Adsorption characteristics of trace volatile organic compounds in gas streams onto activated carbon fibers. Chem Mater 1992;4(5):1068–73. Daley M, Tandon D, Economy J, Hippo EJ. Elucidating the





porous structure of activated carbon fibers using direct and indirect methods. Carbon 1996;34(10):1191–200. Dubinin MM. Fundamentals of the theory of adsorption in micropores of carbon adsorbents: characteristics of their adsorption properties and microporous structures. Carbon 1989;27(3):457–67. Tomlinson JB, Freeman JJ, Theocharis CR. Preparation and adsorptive properties of ammonia-activated viscose rayon chars. Carbon 1993;31(1):13–20. Stoehr B, Boehm HP, Schloegl R. Enhancement of the catalytic activity of activated carbons in oxidation reactions by thermal treatment with ammonia or hydrogen cyanide and observation of a superoxide species as a possible intermediate. Carbon 1991;29(6):707–20. Mangun CL, Benak KR, Economy J, Foster KL. Surface chemistry, pore sizes and adsorption properties of activated carbon fibers and precursors treated with ammonia. Carbon, submitted.