Journal of Colloid and Interface Science 251, 242–247 (2002)doi:10.1006/jcis.2002.8412

Microporous Activated Carbons Prepared from Palm Shell by ThermalActivation and Their Application to Sulfur Dioxide Adsorption

Jia Guo∗,1 and Aik Chong Lua†∗Department of Chemical and Process Engineering, University of Strathclyde, James Weir Building, Glasgow G1 1XJ, Scotland, United Kingdom; and†School of Mechanical and Production Engineering, Nanyang Technological University, Nanyang Avenue, Singapore 639798, Republic of Singapore

E-mail: jia.guo@strath.ac.uk

Received January 23, 2002; accepted April 1, 2002; published online June 18, 2002

Textural characterization of activated carbons prepared frompalm shell by thermal activation with carbon dioxide (CO2) gasis reported in this paper. Palm shell (endocarp) is an abundant agri-cultural solid waste from palm-oil processing mills in many tropicalcountries such as Malaysia, Indonesia, and Thailand. The effectsof activation temperature on the textural properties of the palm-shell activated carbons, namely specific surface area (BET method),porosity, and microporosity, were investigated. The activated car-bons prepared from palm shell possessed well-developed porosity,predominantly microporosity, leading to potential applications ingas-phase adsorption for air pollution control. Static and dynamicadsorption tests for sulfur dioxide (SO2), a common gaseous pollu-tant, were carried out in a thermogravimetric analyzer and a packedcolumn configuration respectively. The effects of adsorption tem-perature, adsorbate inlet concentration, and adsorbate superficialvelocity on the adsorptive performance of the prepared activatedcarbons were studied. The palm-shell activated carbon was foundto have substantial capability for the adsorption of SO2, compara-ble to those of some commercial products and an adsorbent derivedfrom another biomass. C© 2002 Elsevier Science (USA)

Key Words: activated carbon; porosity; surface area; microporos-ity; SO2 adsorption; adsorption rate.

INTRODUCTION

The global production of sulfur dioxide (SO2) by humanactivities (e.g., combustion of high-sulfur-content fossil fuels,metal smelting, sulfuric acid production, and other industrialprocesses) is estimated to be 1.4 million tonnes per year. Theamount of emitted SO2 in the USA, Western Europe, and Japanhas decreased recently, while that from the developing countriesis increasing gradually (1). Sulfur dioxide can cause acid rain,which has acidified soils and streams, accelerated corrosion ofbuildings, and reduced visibility. Long-term exposure to SO2

also results in various respiratory diseases. Many efforts havebeen taken to eliminate SO2 emissions or to remove them from

1 To whom correspondence should be addressed. Fax: (44) 141 552 2302.

240021-9797/02 $35.00C© 2002 Elsevier Science (USA)All rights reserved.

flue gases before emitting into the atmosphere. Among thesetechniques, the dry removal of SO2 using activated carbon isa promising approach that has attracted much attention. Thisdry method offers distinct advantages of simplicity and econ-omy over wet scrubbing because the later requires high capitalinvestment and operating cost for wastewater treatment facili-ties (2).

Commercial activated carbons can be manufactured from avariety of carbonaceous precursors such as lignite and coal(∼42%), peat (∼10%), wood (∼33%), and coconut shell. Sincethe price of commercial activated carbon has dropped continu-ally over the past decade or so, interests are growing in the useof other low-cost and abundantly available lignocellulosic mate-rial as the precursor for the preparation of activated carbon (3).These renewable agricultural solid wastes are cost-effective al-ternatives to more expensive and polluting precursors like coalin the production of activated carbon. Some agricultural solidwastes, such as rockrose (4), kraft lignin (5), apricot stone (6, 7),walnut shell (8), and almond shell and hazelnut shell (9) havebeen successfully converted into activated carbons in the labo-ratories.

Palm shell (also known as endocarp), an abundant agriculturalby-product from palm-oil processing mills in tropical countriessuch as Malaysia, Indonesia, and Thailand, is a prospective pre-cursor for the preparation of high-quality activated carbon due toits high density, relatively high carbon content, and low ash con-tent (10). However, there are few publications in the scientificliterature relating to the preparation and characterization of ac-tivated carbons from palm shell, particularly those of micropo-rous adsorbents used for the adsorption of gaseous pollutants.Preparation of activated carbons from this cheap and abundantbiomass will eliminate the costly problem of solid waste dis-posal while at the same time derive economic benefits fromsuch value-added products.

In this paper, the feasibility of developing activated carbonswith appropriate porosity from palm shell by thermal activationwith carbon dioxide (CO2) was studied. The effects of activationtemperature on the textural properties of the palm-shell activatedcarbons were investigated. Static and dynamic adsorption tests

2

R

PALM-SHELL ACTIVATED CA

of SO2 onto the activated carbon were carried out. The effects ofadsorption temperature, inlet adsorbate concentration, and ad-sorbate superficial velocity on the adsorptive performance wereinvestigated. The amount of SO2 adsorbed onto the palm-shellactivated carbon was compared to those of some commercialproducts and an adsorbent derived from another biomass.

EXPERIMENTAL

Materials

As-received palm shells from Selangor palm-oil processingmill in Malaysia were dried, crushed, and sieved to a particlesize fraction of 2.0–2.8 mm. Both carbonization and activationprocesses were carried out in a stainless-steel reactor (550 mmlength and 38 mm i.d.), which was placed in a vertical tube fur-nace (818P, Lenton). During carbonization, the starting materialwas placed on a metal mesh in the reactor under a nitrogen (N2)flow (150 cm3 min−1). The furnace temperature was raised fromroom temperature (298 K) to 873 K, held at this temperature for3 h, and then cooled down to room temperature. The resultingchars were activated at 773 to 1173 K for 0.5 h under a CO2

flush of 100 cm3 min−1 to produce the final product.

Methods

The solid density (ρs) and apparent density (ρa) of the sam-ple were measured by an ultra-pycnometer (UPY-1000, Quan-tachrome) and a mercury intrusion porosimeter (PoreSizer-9320, Micromeritics), respectively. The total porosity of thesample (ε) was calculated as

ε = [(ρs − ρa)/ρs] × 100%. [1]

Textural characteristics of the samples were determined byN2 adsorption at 77 K with an accelerated surface area andporosimeter (ASAP-2000, Micromeritics). The BET surfacearea (SBET) was estimated from the adsorption isotherms usingthe BET equation (11). The Dubinin–Radushkevich equationwas used to calculate the micropore volume (Vm), from whichthe micropore surface area (Sm) was determined (12). The textu-ral characteristics of raw palm shell, char, and activated carbonare given in Table 1.

Static adsorption of SO2 onto the activated carbon sample

was carried out using a thermogravimetric analyzer (TA-50,Shimadzu). SO gas of various concentrations (balanced by N )

activation temperature. This was attributed to the strong depen-dency of the carbon–CO reaction on the reaction temperature.

2 2

TABLE 1Textural Characteristics of Raw Palm Shell, the Resulting Char, and the Activated Carbon Prepared

Solid density, ρs Apparent density, ρa Total porosity, ε BET surface area, SBET Micropore surface area, Sm

Sample (g cm−3) (g cm−3) (%) (m2 g−1) (m2 g−1)

Raw palm shell 1.53 1.47 3.9 1.6 0.2Char 1.63 1.35 17.2 176 108

2

Activated carbon 2.03 0.69

BON FOR SO2 ADSORPTION 243

was introduced into the analyzer chamber where a platinumsample holder containing about 20 mg sample was suspended.The subsequent sample weight gain due to adsorbed SO2 wasrecorded. Tests were conducted at different temperatures in orderto understand the temperature effect on adsorption.

Dynamic adsorption of SO2 was conducted in a copper col-umn (10 mm i.d. and 5–30 cm long) filled with activated carbonparticles. Columns were operated in the up-flow mode. SO2 gasfrom a cylinder was introduced to the bottom of the adsorptioncolumn (inlet) at different volumetric flow rates ranging from30 to 90 cm3 min−1. The respective superficial velocities were38.2 to 114.6 cm min−1. All the runs were carried out at roomtemperature of 298 K. The concentration time history at theexit of the column (outlet) was continually monitored (samplingtime <2 s) by a SO2 gas analyzer (MLT1, Fisher-Rosemount)equipped with a nondispersive infrared photometer, followedby a data recording system (LTVTE008, Trendview). The ad-sorption was performed until saturation, i.e., when the outletconcentration equaled the inlet one. A breakthrough curve wasthen obtained, and some characteristic parameters were derivedfrom the curve.

RESULTS AND DISCUSSION

Textural Characterization

Figure 1 shows the effects of activation temperature on thetotal porosity and final yield of the activated carbon preparedfrom palm shell by thermal activation. As the activation tem-perature increased from 773 to 1173 K, the sample weight (orthe final yield) reduced significantly due to a combination of therelease of volatile matters in a continual carbonization processand carbon burn-off through carbon–CO2 weak oxidation. Thishad resulted in the development of porosity from 23.6 to 66.0%.For the activated carbon prepared from the extracted rockroseby CO2 activation, Pastor-Villegas et al. (4) reported that thetotal porosity of the sample was 62.9%, which was close to thatof the palm-shell activated carbon prepared in this study.

Generally, the higher the pore surface area of the activatedcarbon, the larger its adsorptive capacity. The effects of acti-vation temperature on BET and micropore surface areas of thepalm-shell activated carbons are shown in Fig. 2. The BET andmicropore surface areas increased progressively with increasing

66.0 1366 985

244 GUO AN

773 873 973 1073 1173

15

25

35

45

55

65

75

Tot

al P

oros

ity, %

10

15

20

25

Fin

al Y

ield

, %

Total PorosityFinal Yield

Activation Temperature, K

FIG. 1. Effects of activation temperature on the total porosity and final yieldof the palm-shell activated carbon.

The fairly high BET and micropore surface areas of the activatedcarbons prepared from palm shells render them to be suitableas effective adsorbents. The maximum BET surface area of thepalm-shell activated carbon was 1366 m2 g−1 as compared tothose prepared from rockrose (4), kraft lignin (5), and apricotstone (6), which were 722, 1343, and 1175 m2 g−1, respectively.

Pores within porous materials are normally classified as mi-cropores (<2 nm diameter), mesopores (2–50 nm), or macro-pores (>50 nm) (13). This classification is important becausemost molecules of gaseous pollutants vary from about 0.4 to

0.9 nm in diameter. Therefore, gas-phase activated carbons usu- soil (20), separation and purification of hydrogen and light hy- ally consist predominantly of micropores while liquid-phase

773 873 973 1073 1173Activation Temperature, K

0

300

600

900

1200

1500

Sur

face

Are

a, m

2 g-1

0

25

50

75

100

Mic

ropo

re F

ract

ion,

%

BET Surface AreaMicropore Surface AreaMicropore Fraction

drocarbons (21), and storage of fuel gases (22). However, so

0 30 60 90 120

Time, min

0

20

40

60

80

Am

ount

of S

O2

Ads

orbe

d, m

g

353K

333K

313K

298K

FIG. 2. Effects of activation temperature on the BET surface area, microporesurface area, and micropore fraction of the palm-shell activated carbon.

D LUA

activated carbons have significantly more mesopores becauseof the larger sizes of liquid molecules. The micropore fractionsof the palm-shell activated carbons versus activation tempera-ture are also shown in Fig. 2. With increasing activation temper-ature, the micropore fraction increased continually, indicatingcontinuous formation of micropore structures. The predominantmicroporosity in palm-shell activated carbons will lead to ap-plications in gas-phase adsorption for the removal of gaseouspollutants.

Static Adsorption Tests

The amounts of SO2 adsorbed from 2000 ppm SO2 flow atvarious temperatures onto the palm-shell activated carbon inthe static tests are shown in Fig. 3. With increasing adsorptiontemperature from 298 to 353 K, the amounts of SO2 adsorbeddecreased significantly. The reason is that during the adsorptionprocess, the SO2 molecules loose their kinetic energies, makingadsorption an exothermic process (14). Therefore, the higherthe adsorption temperature, the lesser would be the amount ad-sorbed. If the temperature increases to the gas critical temper-ature or above, supercritical adsorption occurs. The results ofexperiments performed at above critical temperatures showedthat the behavior of supercritical fluids was fundamentally dif-ferent from the standard (IUPAC) classification schemes (15).Typical experimental isotherms for the Gibbs adsorption of su-percritical fluids have a maximum in the adsorption with in-creasing pressure. At room temperature this maximum is in therange of 100 to 200 bar for different systems: methane on carbonadsorbents (16), nitrogen on alumina (17), and carbon monoxideon zeolites (18). Supercritical adsorption has wide applicationsin the regeneration of adsorbents (19), remedy of contaminated

FIG. 3. Effects of adsorption temperature on the amount of SO2 adsorbedfor static adsorption test.

R

PALM-SHELL ACTIVATED CA

0.0 0.2 0.4 0.6 0.8 1.0

Micropore Volume, cm3 g-1

0

30

60

90

120A

mou

nt o

f SO

2 A

dsor

bed,

mg

Palm ShellApricot StoneCarbochem Filtracarb

FIG. 4. Amount of SO2 adsorbed versus the micropore volumes of variousactivated carbons.

far there are very few papers that present either data or discussthe mechanisms of physisorption of fluids above their criticaltemperatures.

The amounts of SO2 adsorbed onto the activated carbons pre-pared from palm shell and apricot stone as well as commercialactivated carbons versus the micropore volume are shown inFig. 4. For both the palm-shell and apricot-stone activated car-bons, with increasing micropore volume, the amount of SO2

gas adsorbed increased progressively. A linearly proportionalrelationship between the adsorptive capacity and the microporevolume was observed. For the apricot-stone activated carbon (7),a relatively larger adsorptive capacity was seen possibly due toa wider range of micropores within the adsorbents. Therefore,some submicropores (less than 0.5 nm diameter) were acces-sible to SO2 at normal room temperatures, but not detectableduring micropore measurements by N2 adsorption at a low tem-perature of 77 K due to “activated diffusion” problems since theadsorbates have insufficient kinetic energy to penetrate into thetortuous internal pores (23). In addition, it could be seen fromFig. 4 that the amount of SO2 adsorbed onto the palm-shellactivated carbon was comparable to those of two commercial

activated carbons, Filtracarb and Carbochem, which might have adsorbed at breakthrough. The adsorptive capacity for the dy- slightly different pore structures.

TABLE 2Effects of Inlet Adsorbate Concentration on the Characteristic Parameters for Both Static and Dynamic Adsorption Tests

Amount adsorbed for Equilibrium time forInlet SO2 concentration, C0 static adsorption static adsorption Breakthrough time, τ0.05 Exhaustion time, τ0.95

(ppm) (mg g−1) (min) (min) (min)

500 14.8 115.0 — —1000 35.2 87.4 203.8 245.6

namic adsorption was much lower than that in a static system,

2000 76.3 5

BON FOR SO2 ADSORPTION 245

Dynamic Adsorption Tests

Table 2 shows the effects of inlet SO2 concentration (C0) onthe characteristic parameters for static and dynamic adsorptiontests. For a low SO2 concentration of 500 ppm, the amount ofSO2 adsorbed at equilibrium was small (14.8 mg g−1) and a longtime (115.0 min) was required to reach equilibrium. For the SO2

concentration of 2000 ppm, the amount adsorbed was as highas 76.3 mg g−1, and the equilibrium time was only 54.1 min.Such a short equilibrium time suggests a feasible application ofthese activated carbons for dynamic adsorption in a packed col-umn configuration. The characteristic parameters derived frombreakthrough curves obtained from dynamic adsorption of 1000and 2000 ppm of SO2 are also listed in Table 2. A shorter break-through time (τ0.05) and exhaustion time (τ0.95) were observedas the inlet SO2 concentration increased. This means that by in-creasing the inlet concentration, the service time of the packedcolumn will be reduced because of a higher loading factor.

For gas-phase adsorption in the packed column, the break-through time can be expressed using the semi-empirical equation(24)

τ0.05 = ρbW L

C0u− ρbW

C0 Kln

(C0 − C0.05

C0.05

). [2]

Equation [2] may be simplified to

τ0.05 = AL + B, [3]

where

A = ρbW/C0u and B = −ρbW

C0 Kln

(C0 − C0.05

C0.05

).

Accordingly, τ0.05 plots versus column length (L) should give astraight line (Fig. 5). The parameters W and K can be obtainedfrom the values of slope and intercept, respectively (Table 3).In Fig. 5, the slope of the fitted lines decreased with increas-ing SO2 superficial velocity. A smaller gradient implies a lesseramount of SO2 adsorbed at breakthrough, which was confirmedby the values of W as shown in Table 3. This was as expectedbecause the contact time between SO2 gas and the activatedcarbons in the packed column decreased with increasing super-ficial velocity, thereby reducing the amount of gas treated and

4.1 181.4 224.7

246 GUO AN

5 10 15 20 25 30

Column Length, cm

0

300

600

900

1200

1500B

reak

thro

ugh

Tim

e, m

in

114.6 cm min-1

89.1 cm min-1

63.7 cm min-1

38.2 cm min-1

FIG. 5. A plot of breakthrough time versus column length for the parkedcolumn operating at various adsorbate superficial velocities.

where equilibrium was attainted between the adsorbate and theactivated carbon. A similar observation was reported by Youssefet al. (25) when they studied the adsorption of SO2 onto coal-based activated carbons. It also could be seen from Table 3 thatincreasing the SO2 superficial velocity from 38.2 to 114.6 cmmin−1 increased the rate constant from 62.4 to 134.0 min−1. Thissuggested that the rate-limiting step was the external mass trans-fer of adsorbate molecules to the surface of the activated carbonparticle. In fact, the relationship between the adsorbate superfi-cial velocity (u) and the adsorption rate constant (K ) could begiven as

K = 5.6u0.68. [4]

The same dependence of the rate constant on the adsorbate super-ficial velocity was also found in other packed-column adsorptiontests (26, 27). However, theoretically, the rate constant shouldbe proportional to the square root of the adsorbate gas velocity(28).

The overall mass transfer resistance can be approximatelyexpressed by the combination of the bulk fluid resistance (1/βu),

TABLE 3Effects of Adsorbate Superficial Velocity on the Characteristic

Parameters Derived from Breakthrough Curves for DynamicAdsorption Tests

SO2 superficial velocity, u Adsorption capacity, W Adsorption rate(cm min−1) (mg g−1) constant, K (min−1)

38.2 19.3 62.463.7 16.0 110.289.1 12.3 118.9

114.6 6.8 134.0

D LUA

5 10 15 20 25 30

Reciprocal of Gas Velocity, X10-3 min cm-1

0

10

20

30

40

Rec

ipro

cal o

f Rat

e C

onst

ant,

X10

-3 m

in

2000 ppm1000 ppm

FIG. 6. A plot of the reciprocal of rate constant versus the reciprocal of gasvelocity for different inlet adsorbate concentrations.

the fluid film resistance (1/Kf), and the intraparticle diffusionresistance (1/Ki), as

1

K= 1

βu+ 1

Kf+ 1

Ki[5a]

or

1

K= 1

βu+ 1

Kn[5b]

if the fluid film resistance and the intraparticle diffusion resis-tance were considered as a nonexternal resistance given by 1/Kn.Accordingly, 1/K (the overall mass transfer resistance) plot ver-sus the reciprocal of the velocity (1/u) should give a straight line.The value of 1/Kn can be obtained from the intercept. Figure 6shows the relationship between the overall mass transfer resis-tance and the reciprocal of the adsorbate superficial velocity forthe dynamic adsorption system. It was clear that the overall masstransfer essentially depended on the gas velocity for the two gasconcentrations studied here. The results of the best curve-fittingfor the above data are given as

1/K = 1.10 1/u + 4.08 × 10−3 for the case of 1000 ppm

[6a]

1/K = 0.49 1/u + 2.58 × 10−3 for the case of 2000 ppm.

[6b]

For the SO2 superficial velocity of 63.7 cm min−1, the ratios of1/Kn to 1/K were 0.28 for 2000 ppm and 0.20 for 1000 ppm.This means that external mass transfer resistance is predomi-nating, contributing mainly to the overall mass transfer resis-

tance (1/K ). Even when the superficial velocity increased upto 114.2 cm min−1, the above ratios were 0.35 for 2000 ppm

R

PALM-SHELL ACTIVATED CA

and 0.29 for 1000 ppm, showing that the external mass transferresistance remains predominate. However, if the superficial ve-locity increased large enough (i.e., the reciprocal of gas velocitywas very small), then there are no barriers for external masstransfer. Then the nonexternal (either fluid film or intraparticle)resistance will become the main factor to the overall mass trans-fer resistance. If this is the case, the dashed lines were supposedto be parallel to the x axis. Alternatively, if the adsorbate con-centration was high enough (2000 ppm), the drive force forexternal mass transfer will also be large. Then the external masstransfer was no longer a rate-limiting process. The dashed linewill be parallel to the x axis. Actually, as noticed in Fig. 6,the gradient for the case of 2000 ppm was much smaller thanthat of 1000 ppm. In brief, the external mass transfer resistancewas found to be predominant factor in the overall mass transferresistance under the operating conditions used in this study.

SUMMARY

Based on the experimental and theoretical studies on SO2

adsorption onto the palm-shell activated carbons, the followingconclusions can be drawn:

(1) During the CO2 activation process, as the activation tem-perature increased, the sample weight loss was due to the releaseof volatile matters in a continual carbonization process and car-bon burn-off through carbon–CO2 weak oxidation, resulting inthe development of porosity. The maximum BET surface areaof the palm-shell activated carbon was 1366 m2 g−1, which washigher than that of rockrose, kraft lignin, and apricot stone, whichhad 722, 1343, and 1175 m2 g−1, respectively. The predomi-nant micropore development in palm-shell activated carbons willlead to applications in gas-phase adsorption for the removal ofgaseous pollutants. The adsorptive capacity of SO2 for the palm-shell activated carbon was comparable to that of the apricot-stoneactivated carbon and commercial activated carbons.

(2) With increasing adsorption temperature in the range of298 to 353 K, the amounts of SO2 adsorbed onto the palm-shellactivated carbon in the static test decreased significantly due tothe exothermic nature of the adsorption process. Temperatureshigher than the critical temperature of the adsorbate resultedin no adsorption due to Brownian movement of the adsorbatemolecules.

(3) For dynamic adsorption in a packed column, a shorterbreakthrough time and exhaustion time were observed as theinlet SO concentration increased. In addition, the relationship

2

between the adsorption rate constant and the adsorbate super-

BON FOR SO2 ADSORPTION 247

ficial velocity was obtained. This confirmed that the externalmass transfer was the rate limiting step of the whole adsorptionprocess.

REFERENCES

1. Mochida, I., Korai, Y., Shirahama, M., Kawano, S., Hada, T., Seo, Y.,Yoshhikawa, M., and Yasutake, A., Carbon 38, 227 (2000).

2. Rubio, B., and Izquierdo, M. T., Carbon 35, 1005 (1997).3. Derbyshire, F., Jagtoyen, M., and Andrews, R., in “Proceedings of the

2nd Asia Pacific Conference on Sustainable Energy and EnvironmentalTechnologies” (G. Q. Lu, V. Rudolph, and P. F. Greenfield, Eds.), p. 457.University of Queensland, Brisbane, 1998.

4. Pastor-Villegas, J., Valenzuela-Calahorro, C., Bernalte-Garcia, A., andGomez-Serrano, V., Carbon 31, 1061 (1993).

5. Rodriguez-Mirasol, J., Cordero, T., and Rodriguez, J. J., Carbon 31, 87(1993).

6. Gergova, K., and Eser, S., Carbon 34, 879 (1996).7. Molina-Sabio, M., Munecas, M. A., Rodriguez-Reinoso, F., and McEnaney,

B., Carbon 33, 1777 (1995).8. Hu, Z. H., and Vansant, E. F., Carbon 33, 561 (1995).9. Balci, S., Dogu, T., and Yucel, H., J. Chem. Tech. Biotechnol. 60, 419 (1994).

10. Lua, A. C., and Guo, J., Carbon 36, 1663 (1998).11. Brunauer, S., Emmet, P. H., and Teller, F., J. Am. Chem. Soc. 60, 309 (1938).12. Dubinin, M. M., in “Progress in Surface and Membrane.” p. 25, Academic

Press, New York, 1975.13. IUPAC, “Manual of Symbols and Terminology of Colloid Surface.” p. 2,

Butterworths, London, 1982.14. Yang, R. T., “Gas Separation by Adsorption Processes.” p. 10, Butterworths,

Boston, 1987.15. IUPAC Commission on Colloid and Surface Chemistry Including Catalysis,

Pure Appl. Chem. 57, 603 (1985).16. Bose, T. K., Chahine, R., Marchildon, L., and St-Arnaud, J. M., Rev. Sci.

Instrum. 58, 2279 (1987).17. Michels, A. M., Menon, P. G., and Seldam, C. A. T., Rev. Trav. Chim.

Pays-Bas. 80, 483 (1961).18. Wakasugi, Y., Ozawa, S., and Ogino, Y., J. Colloid Interface Sci. 79, 399

(1981).19. Recasens, F., McCoy, B. J., and Smith, J. M., AIChE J. 35, 951 (1989).20. Erkey, C., Madras, G., Orejuela, M., and Akgerman, A., Environ. Sci. Tech-

nol. 27, 1225 (1993).21. Juntgen, H., Carbon, 15, 273 (1977).22. Barton, S. S., Dacey, J. R., and Quinn, D. F., in “Fundamentals of Adsorp-

tion” (A. L. Myers and G. Belfort, Eds.), p. 65. Engineering Foundation,New York, 1984.

23. Pandolfo, A. G., Aminz-Amoli, M., and Killingley, J. S., Carbon 32, 1015(1994).

24. Wood, G. O., and Stampfer, J. F., Carbon 31, 195 (1993).25. Youssef, A. M., Mostafa, M. R., and Dorgham, E. M., Adsorption Sci. Tech.

15, 803 (1997).26. Pota, A. A., and Mathews, A. P., J. Environ. Eng. 125, 705 (1999).27. Takeuchi, Y., Hino, M., Yoshimura, Y., Otowa, T., Izuhara, H., and

Nojima, T., Sep. Purifi. Technol. 15, 79 (1999).28. Busmundrud, O., Carbon 31, 279 (1993).