Chemical Engineering Journal 144 (2008) 42–50

Adsorption of polluting substances on activated carbons

prepared from rice husk and sugarcane bagasse
Dimitrios Kalderis a,∗ , Dimitrios Koutoulakis a , Panagiota Paraskeva a , Evan Diamadopoulos a ,
Emilia Otal b , Joaquı́n Olivares del Valle b , Constantino Fernández-Pereira b
a Department of Environmental Engineering, Technical University of Crete, 73100 Chania, Greece
b Asociación de Investigación y Cooperación Industrial de Andalucı́a (AICIA), University of Seville, Seville, Spain
Received 2 August 2007; received in revised form 21 December 2007; accepted 4 January 2008

Abstract
Rice husk and sugarcane bagasse were chemically impregnated with ZnCl2 and carbonized at 700 ◦ C in a large-scale rotary furnace. The activated
carbons (ACs) obtained had BET surface area of 811 and 864 m2 /g, respectively, and were essentially microporous. The adsorption of arsenic,
humic acid, phenol and a municipal solid waste landfill leachate was examined. Both ACs showed the best adsorption behaviour towards phenol,
removing around 80% at the equilibrium time of 4 h. The adsorption isotherms for arsenic and humic acid were also favourable, although the
maximum loadings achieved were lower than that of phenol. Finally, the rice husk AC showed 60% and 70% removal efficiency for colour and
COD, respectively, when tested on a landfill leachate.
© 2008 Elsevier B.V. All rights reserved.

Keywords: Rice husk; Activated carbon; Sugarcane bagasse; Landfill leachate; Arsenic; Humic acid; Phenol; Adsorption

1. Introduction surface area can be obtained at lower temperatures, compared

to the temperatures required for physical activation. In a com-
Activated carbon (AC) products are produced from organic prehensive review regarding the use of agricultural residues as
materials that are rich in carbon. The choice of precursor is precursors for AC production, Ioannidou and Zabaniotou [1]
largely dependent on its availability, cost and purity. The pre- provide critical information with respect to the type of residue,
cursor and the method used for activation largely determine the process parameters, activation type and desirable use of the AC.
adsorption properties of the AC. Typical precursors for acti- Biowaste originated activated carbons with BET surface areas
vated carbon production are coal, lignite, wood, nut shells, peat, ranging from 250 to 2410 m2 /g and pore volumes ranging from
coke, etc. ACs can be prepared by either physical or chemi- 0.022 to 91.4 cm3 /g have been reviewed and compared.
cal activation. Physical activation is usually a two-stage process Many researchers have investigated the use of agricultural
(pyrolysis–activation) that employs gases such as nitrogen, by-products as precursors for the production of ACs with high
steam, carbon dioxide, air or mixtures of the above, without the adsorptive capacity. Date and olive stones, coconut and almond
presence of a catalyst. Activation occurs at temperatures above shells, rice husks (RH) and sugarcane bagasse (SB) are only few
700 ◦ C at residence times that vary between 30 min and sev- of the materials that have been successfully used for the prepa-
eral hours. Chemical activation involves a pre-treatment scheme ration of ACs [2–9]. Ng et al. [10], produced activated carbon
(impregnating step) where the material is mixed with a chemical from pecan shells by double-stage steam activation at 850 ◦ C for
before the activation process. The most widely used chemicals 2 h after carbonization at 700 ◦ C for 1 h. Surface areas resulted
include H3 PO4 , H2 SO4 , KOH, NaOH and ZnCl2 . Impregna- were in the range of 750–850 m2 /g. Physical activation of dry
tion can take up to 24 h depending on the chemical used, the bagasse in a two-stage carbonization/activation process within
precursor and the subsequent activation process. ACs with high a temperature range of 750–840 ◦ C produced activated carbon
with significant surface areas (404–607 m2 /g), the surface area
increasing with increasing temperature [11]. In a recent study,
∗ Corresponding author. Tel.: +30 28210 37820; fax: +30 28210 37847. Zabaniotou et al. [12] investigated the production of ACs from
E-mail address: [email protected] (D. Kalderis). olive kernels in two different modes: laboratory-scale pyrolysis

1385-8947/$ – see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.cej.2008.01.007
 D. Kalderis et al. / Chemical Engineering Journal 144 (2008) 42–50 43

followed by chemical activation with KOH and (b) pilot/bench- al. [22] produced a series of activated carbons from almond
scale pyrolysis followed by physical activation with H2 O–CO2 . shells with surface areas from 200 to 600 m2 /g and compared
The maximum BET surfaces obtained were 1000–1200 m2 /g them to commercial activated carbons of 950 m2 /g surface area
and 3049 m2 /g at industrial scale with physical activation and at for the removal of polar and non-polar organic compounds.
laboratory scale with chemical activation, respectively. Results showed that nearly all carbons (except the one with BET
Apart from the two-stage physical activation process, a num- 200 m2 /g) performed, on average, equally well compared to the
ber of researchers have investigated a single-stage physical commercial ones. Recently, Skodras et al. [23] reviewed the
activation process, when carbonization/activation takes place adsorption of mercury in ACs prepared from biomass materi-
simultaneously. Savova et al. [13] applied a one-step pyroly- als and waste tires. They prepared ACs from pine/oak woods
sis process to almond shells, nut shells, peach stones, cherry through physical activation and from olive seeds through chem-
stones and grape seeds. The process involved direct heating of ical activation with KOH. One of the main conclusions was that
the precursors with steam at a rate of 15 ◦ C/min and 1 h retention an increase in the surface area of the ACs results in an almost
time at 800 ◦ C. The best results were obtained for apricot stones proportional increase in the Hg adsorption capacity. The olive
with a surface area of 1190 m2 /g, followed by almond shells at seed ACs showed better adsorption properties compared to the
998 m2 /g, cherry stones at 875 m2 /g, nut shells at 743 m2 /g and pine wood or oak wood ACs.
grape seeds at 497 m2 /g. Kula et al. [24] studied the adsorption of Cd(II) on ACs pre-
Girgis et al. [14], compared three activating agents, KOH, pared from olive stones by ZnCl2 activation. The adsorption
H3 PO4 and ZnCl2 for the activation of rice hulls. Of the experiments they performed indicated that the activated carbon
three, H3 PO4 produced carbons with the highest surface area from olive stone has a good adsorption capacity for Cd(II) ions
of 1177 m2 /g. KOH after pyrolysis and 3 h heating at 500 ◦ C from aqueous solutions. Specifically, at a Cd(II) concentration
resulted in activated carbon with 268 m2 /g surface area and of 45mg/L, the highest adsorption capacity was 1.654 mg/g.
ZnCl2 after 6 h pyrolysis at 300 ◦ C and 3 h heating at 750 ◦ C In addition, ACs were demonstrated to have excellent effec-
resulted in 420 m2 /g surface area. All chemicals performed bet- tiveness in the removal of organic matter and nitrogenous
ter than steam activation alone. Caturla et al. [15] impregnated substances in landfill leachates. Rivas et al. [25,26] studied
peach stones with concentrated ZnCl2 at ratios (ZnCl2 :material) some integrated processes treating landfill leachates. The sys-
0.4:1–2.5:1. The impregnated mixture was kept at 85 ◦ C from tems tested included adsorption onto activated carbon. COD
7 to 24 h under evaporation, partial evaporation and no evap- removals obtained ranged from 80% to 96% for initial COD val-
oration conditions. The material was then heated under N2 at ues close to 11000 mg/L, although none of the processes tested
500–800 ◦ C. The produced carbons had surface areas up to reduced the COD levels sufficiently to allow direct discharge.
3000 m2 /g. Increasing the ZnCl2 :material ratio from 0.2:1 to Recently, Kurniawan et al. [27] reviewed the main physico-
1:1 increased the surface area from 1000 to 2000 m2 /g. Further chemical treatments for landfill leachate. According to the
increase of the ratio resulted in a decrease of the surface area. authors, none of the individual physico-chemical techniques
An increase of temperature from 600 to over 1000 ◦ C leads to is universally applicable or highly effective for the removal of
a decrease of surface area. Through chemical activation with recalcitrant compounds from stabilized leachate. A combination
KOH, Stavropoulos and Zabaniotou [16] produced ACs with of physico-chemical and biological treatments has demonstrated
high surface area (>1300 m2 /g) from olive seed waste residues. its effectiveness for the treatment of stabilized leachate. Bio-
The highest surface area (3049 m2 /g) was achieved when acti- logical treatment alone usually results in low COD removals,
vation occurred at 900 ◦ C and 4 h residence time. Their carbons because of high COD, high ammonium-N content and presence
also showed high adsorptive capacity with respect to methylene of toxic compounds. A pre-treated leachate was subjected to
blue. adsorbent-supplemented biological treatment in an aeration tank
Due to their adsorption properties, ACs are used for the operated in fed-batch mode [28]. COD and NH4 + -N removal per-
removal of a wide range of pollutants from wastewaters, such formances of powdered activated carbon (PAC) and powdered
as heavy metals, phenols, dyes and other polar and non-polar zeolite were compared during biological treatment. Percent
organic substances [17–19]. Activated carbons produced from COD removals with PAC addition were significantly higher
almond shells, pecan shells and walnut shells either by chem- than those obtained with the zeolite. However, zeolite performed
ical or physical activation removed almost 100% of lead ion, better than the PAC in ammonium-N removal from the leachate.
90–95% of copper ion and 80–85% of zinc ions when used in Addition of landfill leachate to a domestic wastewater treat-
a “point of use” drinking water filter system [20]. Regarding ment system may have negative impact, since leachate is a highly
VOCs’ adsorption, activated carbons prepared from almond and concentrated wastewater both in terms of inorganic and organic
pecan shells by chemical, CO2 and steam activation were com- pollutants. A recent research revealed that the negative impact
pared to commercial products made from coal and coconut shells of leachate on a municipal wastewater treatment system could
[21]. Commercial products adsorbed 480–520 ppb of VOCs per be prevented by the addition of PAC [29].
gram of carbon, with the coconut shell product being superior. In all the above studies, the ACs employed for the adsorption
The adsorption capacity of the produced carbons ranged from tests had significant surface areas and adsorption capabilities, in
350 to 510 ppb VOCs/g of carbon. Although the chemically acti- many cases comparable to the commercially available carbons.
vated carbons had a larger surface area, their adsorption capacity However, their production methods involved several stages of
was inferior to that of the physically activated carbons. Toles et preparation, long retention times for pyrolysis and activation and
44 D. Kalderis et al. / Chemical Engineering Journal 144 (2008) 42–50

often elevated temperatures. The present study aimed at testing Table 1

the adsorption properties of ACs that were prepared in a sin- Properties of rice husk and sugarcane bagasse
gle activation process in relatively short retention times. Model Rice husk Sugarcane bagasse
pollutants such as phenol, humic acid and arsenic were used. Moisture (%) 4.2 6.1
Additionally, the adsorption of highly contaminated leachates Ash content (%) 16.1 3.3
generated in a municipal landfill onto three different adsorbents: Volatile matter (%) 62.0 65.9
two activated carbons (the rice husk AC produced in this work Fixed carbon (%) 17.7 24.7
and one commercial AC) were investigated, both in terms of
contaminant level reduction per unit mass of absorbent and in
terms of the process kinetics. Programme Project titled ‘Production of High Added Value
Products for Environmental Applications from Agricultural By-
products in India: Activated Carbon Production from Bagasse
2. Materials and methods and Rice Husks’ [30]. In these experiments, dried and ground
(≤0.5 mm) rice husk and bagasse were impregnated with ZnCl2
2.1. Precursor materials at ratios of 0.25 to 1. Distilled water was added at a quantity
amounting to 10 times the total weight of the solid mixture. The
Sugarcane bagasse was obtained from Ajnala Cooperative mixture was then stirred and heated to make homogeneous and
Sugar Mill on the district of Amritsar in India in the form of a impregnated at a temperature of 80–100 ◦ C. A thick uniform
sheaf of bagasse. Bagasse was dried in an oven at 110 ◦ C for 6 h. paste was obtained after 2 h. A sample of wet paste (around
It was then ground with a microhammer cutter mill and sieved 75% moisture) was weighed before being fed to the reactor for
to a 10 or 32 mesh (2.0 mm or 500 ␮m) particle size prior to a single step of carbonization/activation.
activation. Bagasse with a particle size of 500 ␮m was used for The optimum ZnCl2 -to-rice husk ratio was 1:1 and for
characterization and the production of activated carbons. bagasse this ratio was 0.75:1.The optimum experimental con-
Rice husk was obtained from Janta Rice Mill in Gurdaspur in ditions were found to be 700 ◦ C carbonization/activation
India. Rice husk was initially washed thoroughly with water to temperature and 30 min residence time. At these conditions, the
remove any impurities, dried at 110 ◦ C for 6 h and then ground AC surface areas obtained were 674 and 750 m2 /g for rice husk
with a microhammer cutter mill and sieved to a 10 or 32 mesh and bagasse, respectively. For this study, the optimum impreg-
(2.0 mm or 500 ␮m) particle size. The finer fraction was used for nating ratios were followed and the optimum temperature and
characterization and for the activation studies. The properties of residence time were applied to obtain ACs from a large-scale
the raw materials are shown in Table 1. reactor.
Fig. 1 shows the main components of the AC production
2.2. Activation method and apparatus unit and their dimensions (in cm). The experimental set up of
the reactor consisted of a mechanical screw feeder connected
The choice of the optimum experimental parameters for the to a cylindrical rotary furnace chamber, and an AC collection
impregnation and activation of rice husk and sugarcane bagasse box. A gas inlet line was fitted at the end of the screw feeder
were based on preliminary laboratory-scale experiments per- and two gas outlet lines were fitted on top of the collection box
formed under the 6th FP EU–India Economic Cross Cultural to avoid accumulation of gaseous by-products in the furnace.

Fig. 1. Diagram of the main components of the AC production unit.

 D. Kalderis et al. / Chemical Engineering Journal 144 (2008) 42–50 45

The rotation rate of the screw feeder (i.e. the speed at which 2.4. Adsorption of phenol
the paste entered the furnace) was controlled through a control
board. The required residence times were achieved by adjusting For the adsorption of phenol, a 10 ppm stock solution of phe-
the inclination of the furnace (through a mechanical lever shown nol was prepared. Seven Erlenmeyer flasks were used, each
in Fig. 1) in conjunction with the speed of furnace rotation. one of them containing 0.03, 0.05, 0.08, 0.1, 0.13, 0.15, and
Nitrogen gas was allowed to pass through the reactor 0.18 g of AC and 200 mL of the stock solution, respectively.
(4 L/min) in order to remove air from the system. The fur- The adsorption of phenol was initially tested on both rice husk
nace was heated to 700 ◦ C, and once the target temperature was AC and sugarcane bagasse AC. The flasks were mechanically
reached (approximately 30 min), the gas supply was changed shaken for 9 days and the solutions were then filtered through
from nitrogen to CO2 (5 L/min) and the feeding of the paste com- a 0.45 ␮m polypropylene membrane filter to separate the liquid
menced. The residence time required for activation was 30 min. from the AC. The percentage of phenol removal was determined
The ACs were obtained in the collection box in crude form by measuring the phenol remaining in solution using a UV spec-
(lumps of 0.2–1.5 cm) and were allowed to cool to room tem- trophotometer at a wavelength of 269 nm. The adsorption data
perature. Then they were mechanically crushed and thoroughly were analyzed in terms of the Langmuir and Freundlich model
washed with a solution of HCl 0.1 M to remove excess Zn and isotherms using the equations:
other impurities. Finally, the ACs were washed with distilled QbCe
water and dried at 105 ◦ C for 6 h. q= (2)
1 + bCe
and
2.3. Characterization of the activated carbons
q = KCe 1/n (3)
Each activated carbon’s surface area was determined
respectively, where q is the mass of adsorbate per mass of adsor-
from nitrogen adsorption data obtained with a gas sorption
bent, Ce is the equilibrium concentration of adsorbate in solution
analyzer (NOVA 2200, Quantachrome). The samples were
and Q, b, K, n are constants for a given adsorbate and adsorbent
degassed under vacuum at 300 ◦ C for 8 h prior to measure-
at a particular temperature.
ment. The adsorption equilibrium time was set at 90 s. The
To study the kinetics of phenol adsorption, a 100 ppm stock
Brunauer–Emmett–Teller (BET) method was used to calculate
solution of phenol was prepared. Six Erlenmeyer flasks were
the surface area, based on selected N2 adsorption data within the
used, each containing 100 mL of the stock solution and 0.2 g of
p/po range of 0.025–0.3 [31].
rice husk AC. The flasks were mechanically shaken and pro-
The total pore volume (Vtotal ) and the average pore diam-
gressively removed at appropriate time intervals.
eter (dav ) were derived from the Barrett, Joyner and Halenda
(BJH) method [32]. The micropore volume (Vo ) was calculated
2.5. Adsorption of arsenic
according to the Dubinin–Radushkevich (DR) method [33,34].
The particle size distributions were measured in a particle size
To test the adsorption capacities of the rice husk and sugar-
analyzer (Mastersizer S, Malvern) equipped with an optical mea-
cane bagasse ACs, a 10 ppm stock solution of As was prepared
surement unit and a sample suspension unit. The lens used
in the form of AsO4 3− . Nine separate Erlenmeyer flasks were
was type 300 RF, able to detect particle sizes in the range of
used, each one of them containing 0.2, 0.4, 0.6, 0.8, 1.1, 1.4, 1.8,
0.05–880 ␮m.
2, and 2.2 g of rice husk AC, respectively, and 200 mL of the
The ash content of the carbons was determined using the stan-
stock solution of As. The flasks were then mechanically shaken
dard ASTM procedure D2866-94. This method involves heating
for 96 h at 180 rpm to achieve equilibrium between As and the
the sample at 650 ◦ C for several hours until constant weight has
carbon surface [35]. At the end of the process, the solutions
been achieved. The ash is then weighed and the ash content of
were filtered through a 0.45 ␮m polypropylene membrane filter
the activated carbon is calculated as follows:
to remove the AC and the filtrates were prepared for analysis
 
(D − B) according to the molybdate blue method [36]. The percentage
Total ash (%) = × 100 (1) of As removal was determined by measuring the As remaining
(C − B)
in solution using a UV spectrophotometer at a wavelength of
where B is the weight of the crucible (g), C is the weight of 885 nm.
the crucible plus original sample (g) and D is the weight of the
crucible plus ash containing sample (g). 2.6. Adsorption of humic acid
The pH of the carbons was determined using the standard
test of ASTM D 3838-80. This procedure involves heating 10 g For the adsorption of humic acid, a 20 ppm stock solution of
of carbon in 100 mL of distilled water to its boiling point for humic acid was prepared. Eight Erlenmeyer flasks were used,
15 min. The slurry is then filtered and the pH of the filtrate is each one of them containing 0.1, 0.2, 0.3, 0.4, 0.7, 0.9, 1.2, and
measured at 50 ◦ C. 1.4 g of AC, respectively, and 200 mL of the stock solution. As
The adsorption capacities of the carbons were then tested in arsenic adsorption case, only the rice husk AC was used. The
using solutions of phenol, arsenic, and humic acid, as well as flasks were mechanically shaken for 7 days at 180 rpm and the
leachate from a municipal solid waste (MSW) landfill. solutions were then filtered through a 0.45 ␮m polypropylene
46 D. Kalderis et al. / Chemical Engineering Journal 144 (2008) 42–50

membrane filter to separate the liquid from the AC [35]. The per- Table 2
centage of humic acid removal was determined by measuring the Properties of the ACs used for the adsorption tests
humic acid remaining in solution using a UV spectrophotometer AC from rice husk AC from sugarcane
at a wavelength of 254 nm. bagasse

SBET (m2 /g) 811 864

Average pore diameter dav (nm) 2.02 2.05
2.7. Activated carbon treatment of a landfill leachate
Pore volume (cm3 /g) 0.41 0.43
Micropore volume (cm3 /g) 0.42 0.44
The landfill leachate used in this work was collected from the Average particle size (nm) 41.21 54.87
storing unit of the liquid waste generated by the municipal solid Ash content (%) 27.8 40.6
waste organic fraction (before composting), in an integrated pH 5.3 5.9
MSW plant located in Seville (Spain). Given the high content
of contaminants in the raw leachate sample, it was diluted up
to 10% using tap water for all the laboratory adsorption tests. 3. Results and discussion
After that, the diluted leachate sample was filtered through a
GF/C glass fibre filter. 3.1. Single-stage production of ACs from rice husk and
The diluted leachate sample (COD = 3167 mg/L and NH4 - sugarcane bagasse
N = 144 mg/L) was subjected to three alternate adsorption
treatments, using a stirred tank reactor of 1 L capacity and two The properties of the ACs that were prepared in single-stage
adsorbent concentrations (10 and 30 g/L). The three adsorbents chemical activation process are shown in Table 2. As indicated
used were: (a) the rice husk powdered AC carbon (RH AC), (b) by the average particle size, the ACs are in powdered form.
a commercial powdered AC from CPL Carbon Link (PAC) and They have comparable values for their properties, the excep-
(c) a zeolitised fly ash material (CV-Z), resulting from the alka- tion being the ash content, where the AC from rice husk and
line activation of fly ashes from the Narcea Power Plant (Spain). the AC from bagasse contain 27.8% and 40.6% ash, respec-
The CV-Z product is composed of different zeolites (mainly Na- tively. These ACs were used for the adsorption tests. As shown
P1, and lower levels of analcime and chabazite) and the fly ash earlier, ACs with comparable surface area values, microporos-
remaining after the alkaline activation [37]. The resulting sus- ity and ash content have yielded promising results with respect
pensions were then stirred for 30 min at 150 rpm, and finally to arsenic, phenol, humic acid and landfill leachate adsorption
were settled for 4 h. [17–22,25–29]. Table 3 provides a summary of the activating
Samples of influent and effluent water were taken for analy- conditions, surface area and total pore volume of the ACs pre-
ses. Chemical oxygen demand (COD) and NH4 -N analyses were pared for this study, and of carbons from agricultural by-products
based on standard methods [38]. COD was carried out by the reported in other papers.
dichromate reflux method, whereas NH4 -N concentration was
measured by using an ammonium ion selective electrode (Cri- 3.2. Adsorption of phenol
son Instruments). Colour was determined by spectrophotometric
method (measuring at 400 nm), which involves the prior filtra- As seen in Fig. 2, the adsorption isotherms of phenol on rice
tion of the samples through 0.45 ␮m cellulose acetate filter. All husk AC and bagasse AC, respectively, were favourable. Rice
the analyses were done in duplicate, taking the average of two husk AC exhibited a higher adsorption capacity compared to
values. bagasse AC, as indicated by the maximum loadings of 27.58

Table 3
Summary of optimal conditions for activation of various agricultural by-products
Material Impregnating chemical Activation temperature (◦ C) Activation time (h) BET (m2/g) Total pore volume (cm3/g) Reference

Rice husk ZnCl2 700 0.5 811 0.41 This work

Sugarcane bagasse ZnCl2 700 0.5 864 0.43 This work
Pecan shells H3PO4 450 1 1250 [63]
Olive stones KOH 900 4 3049 1.52 [16]
Olive kernels KOH 900 4 2159 – [12]
Walnut shells PO4 salts 750 – 850 – [64]
Hazel nuts ZnCl2 750 10 793 – [65]
Peanut hulls H3 PO4 500 3 1177 0.57 [14]
Peach stones ZnCl2 850 – 3000 – [15]
Corn cobs KOH 850 1 2700 – [66]
Olive seed waste KOH 800 3 1690 0.9 [23]
Oak wood – 900 – 684 0.45 [23]
Pine wood – 900 2.5 896 0.61 [23]
Almond shells H3 PO4 450 1 1340 – [21]
Olive stones ZnCl2 650 2 790.25 – [24]
Oil palm shell Steam 900 1 1183 0.69 [67]
 D. Kalderis et al. / Chemical Engineering Journal 144 (2008) 42–50 47

Fig. 3. Phenol kinetics for the rice husk and bagasse activated carbons.

Fig. 2. Adsorption isotherms of phenol on rice husk and sugarcane bagasse

activated carbons. and chlorophenols on several commercial ACs. They concluded
that their adsorption was not affected by the physical properties
and 24.68 mg phenol/g AC, respectively. The data were analyzed of activated carbons but mainly by chemical interaction.
using the Langmuir and Freundlich model isotherms and the In a recent study, Mukherjee et al. [47] examined the adsorp-
coefficients are shown in Table 4. tion of phenol on commercial AC, wood charcoal and bagasse
These results agree with those obtained by several researchers ash. They used a phenol solution of 30–50 mg/L and adsorbent
that observed favourable adsorption of phenol on ACs compared at 50 g/L. They concluded that wood charcoal and bagasse ash
to other adsorbents such as peat, fly ash and bentonite [39–41]. were comparable to commercial grade AC in the removal of phe-
Due to the microporous nature of our carbons, phenol, being a nol from water. Their experimental results showed that ca. 98%,
smaller size molecule, adsorbed better than HA. It is known that 90% and 90% removal was achieved in 1 h by the commercial
the rate of adsorption often depends on the diffusion coefficient, AC, the wood charcoal and the bagasse ash, respectively.
which is strictly associated with the molecular size of the adsor- Similar results were obtained by Özkaya [48], who studied
bate. Diffusion coefficients of HA and phenol are sufficiently the adsorption and desorption of phenol on commercial activated
different to ensure superior kinetics of the smaller compound carbon using 100 mg/L solutions. For the adsorption experi-
[42,43]. ments, the AC dosage ranged from 0.05 to 0.9 g/100 mL. His
In kinetic terms, it was found that the phenol–rice husk AC results showed that after the equilibrium time of 2 h, 96% of phe-
and phenol–bagasse AC adsorption system reached equilibrium nol was removed from solution. Kennedy et al. [49] tested the
in ∼4 h, as shown in Fig. 3. At this time, ∼80% of the initial phenol adsorption on AC prepared from rice husk in a two-stage
phenol concentration was removed. This is comparable to the process using H3 PO4 as the impregnating agent. The optimum
equilibrium times achieved by other researchers that used phenol AC was obtained at 900 ◦ C and had a surface area of 438.9 m2 /g.
and ACs in similar conditions. Zogorski et al. [44] observed that This AC showed the maximum phenol uptake at pH of 2.7. In our
the adsorption of phenolic compounds by granular carbon was studies, the pH remained unaffected throughout the adsorption
extremely rapid and within the first hour of contact, 60–80% process.
of the ultimate adsorption could occur. Sachan et al. [45] also Several researchers have investigated the effect of particle
reported that within 2 h of adsorption time, about 80% phenol size on the equilibrium kinetics of phenol and agreed that the
was adsorbed by AC. Our observations also agree with those of smaller the particle size the faster the diffusion of the phenol
Jung et al. [46] that studied the adsorption behaviour of phenol molecules [40,50,51,53]. Roostaei and Tezel [40] studied the

Table 4
Langmuir and Freundlich adsorption model coefficients
Freundlich isotherm

K 1/n

Phenol HA Arsenic Phenol HA Arsenic

Rice husk AC 0.4080 4.9496 0.8768 0.4303 0.1619 0.1433

Bagasse AC 0.3414 0.5435

Langmuir isotherm

a b
Phenol HA Arsenic Phenol HA Arsenic

Rice husk AC 19.455 333.33 3.795 1.8423 0.0005 0.228

Bagasse AC 12.33 2.4727
48 D. Kalderis et al. / Chemical Engineering Journal 144 (2008) 42–50

removal of phenol from aqueous solution by adsorption on to

various commercial adsorbents. They particularly examined the
influence of the adsorbent particle size on the adsorption process
and concluded that the particle size did not affect the capacity
of adsorption, however, the adsorption kinetics got faster with
decreasing particle size. Therefore, the smaller the particle size
the faster is the diffusion of phenol into adsorption sites within
the adsorbent. The fastest adsorption kinetics were exhibited
by HiSiv 1000 (zeolite Y-structure), which had a particle size
of 50 mesh × 70 mesh (212–297 ␮m). Using a 100 mg/L phenol
solution (as in our case), equilibrium was achieved in 3 h, which
is comparable to the 4 h achieved in this study. Fig. 4. Adsorption isotherms of arsenic on rice husk activated carbon.
Juang et al. [52] studied the liquid phase adsorption
and desorption of phenol onto granular activated carbons Huang and coworkers [54,55] also studied the adsorption of
(GACs) prepared from coconut shell. Their GACs had sur- As by commercial ACs and concluded that lignite-based ACs
face areas between 732 and 802 m2 /g, total pore volumes of (which are characterized by high ash and CaO/Fe2 O3 content)
0.49–0.64 cm3 /g and average pore size ∼3.90 nm. Adsorption were more effective in removing As(V), from water as opposed
was tested on different AC particle sizes, ranging from 1.15 to low-ash ACs. The predominant role of ash in arsenic adsorp-
to 4 mm. The results showed that as the carbon particle size tion by ACs was also confirmed by Lorenzen et al. [56]. They
decreases, so does the diffusion path. Therefore, quicker equilib- used a coconut-shell AC, a peat-based AC and a coal-based AC
rium times were achieved, although the loading values remained to adsorb arsenic and showed that the AC with the higher ash
very similar between the AC with different particle sizes. The content (but not necessarily the largest surface area) exhibited
equilibrium times achieved by Juang et al. ranged between 8 and the highest arsenic loading of 4.09 mg/g carbon.
10 h whereas in this study equilibrium times were considerably These results indicate that there is a strong interaction
lower (4 h). Since the GACs characteristics mentioned earlier between arsenic and mineral matter. Therefore, the ash con-
are comparable to the values of our ACs, the longer equilibrium tent and the CaO and Fe2 O3 percentages therein play a more
times can be attributed to the larger particle size of the GACs. important role in arsenic adsorption compared to the surface area
[35,41]. Both the rice husk and sugarcane bagasse ACs have
3.3. Adsorption of arsenic relatively high ash contents (27.8% and 40.6%, respectively);
however, the CaO and Fe2 O3 contents are very low (0.73% and
The result of the arsenic adsorption on rice husk AC is shown 0.28% for rice husk and 2.73% and 1.19% for bagasse, respec-
in Fig. 4. The adsorption isotherm is favourable, showing a max- tively). This possibly accounts for the low As loading achieved
imum loading of 1.22 mg As/g AC. This result essentially agrees in the rice husk AC. Since bagasse has comparable CaO and
with the findings of Diamadopoulos et al. [35], who studied the Fe2 O3 contents to rice husk, and the AC produced from bagasse
effect of AC properties on the adsorption of phenol, fulvic acid has very similar characteristics to the rice husk AC, we thought
and arsenic. In the case of As, they showed that ACs with high it would be redundant to perform arsenic adsorption tests on
ash content were able to remove up to 5 times as much As from the bagasse AC. Table 5 shows a comparison of the RH AC
water, compared to ACs without any ash. This was due to the with other carbons of similar surface area, with respect to As
high CaO and Fe2 O3 content in the ash. Mohan and Pittman [53] adsorption capacity.
reviewed the removal of arsenic from a wide range of adsorbents,
such as commercial ACs, agricultural by-products, clay minerals 3.4. Adsorption of humic acid
and zeolites. They showed that ACs produced from agricultural
wastes offered a low-cost, efficient alternative to commonly used The adsorption isotherm shown in Fig. 5 is favourable exhibit-
arsenic adsorbents. ing a maximum loading of 11.59 mg HA/g AC. The rice husk

Table 5
Comparison of the rice husk AC adsorption characteristics with other adsorbents, with respect to arsenica
Adsorbent Surface area (m2 /g) Arsenic concentration (mg/L) Capacity As(III) (mg/g)

RH AC used in this study 811 10 1.22

Activated alumina 370 1 0.18
Tea fungal biomass – 1.3 1.11
FeCl3 treated tea-fungal biomass – 1.3 5.4
FeO3 impregnated GAC 840 1 4.5
Olive-stone and olive pulp ACs 732–1850 5–20 0.21–1.39
Modified calcined bauxite – 0.5–0.8 1.37
Goethite 12.7 10–1000 7.5
a Data obtained from Ref. [53].
 D. Kalderis et al. / Chemical Engineering Journal 144 (2008) 42–50 49

Fig. 6. Landfill leachate treatment using a commercial activated carbon (PAC),

a rice husk activated carbon (RH AC) and a zeolitised fly ash material (CV-Z).
Fig. 5. Adsorption isotherm of humic acid on rice husk activated carbon.

the COD level from 1300 to 200 mg O2 /L after 2 h of contact,

AC behaved in a similar manner as in the case of As adsorption. which corresponds to a reduction of 85%.
The adsorption of humic acid is mainly determined by CaO and Regarding NH4 -N, CV-Z removed just 10% at 10 g/L, but at
Fe2 O3 content of the adsorbent [35] and the molecular size of 30 g/L the removal of ammonium-N significantly increased to
the HA [57]. The low CaO and Fe2 O3 of the precursor materials 74%. CV-Z did not remove any colour. The colour of the leachate
and the microporous nature of our carbons do not promote high was efficiently removed using PAC at 30 g/L under the same
HA adsorption. Han et al. [57] studied the adsorption of humic conditions. Removal efficiencies obtained were 79% and 98%,
acids on two commercial microporous ACs and one mesoporous for 10 and 30 g/L of PAC, respectively. The RH AC removed
AC. They clearly showed that regardless of the surface area of 31% and 60% of colour when 10 and 30 g/L were used, respec-
the carbons and the pH conditions, the adsorption capacity of tively. Therefore, RH AC showed at 30 g/L a similar adsorption
the mesoporous carbon was much higher than that of micro- capacity to the commercial AC in the case of COD.
porous carbons. At all the pH values they employed, the HA
loadings for the microporous carbons did not exceed 20 mg/g, 4. Conclusions
whereas the mesoporous carbon achieved a maximum loading of
100 mg/g. The production of added value products such as ACs from
Similar results were obtained by Lorenc-Grabowska and Gry- biowaste is very important from both environmental and eco-
glewicz [58], who studied the adsorption of HAs on coal-based nomical scope. In developing countries, the majority of rice
mesoporous ACs. They found that as mesoporosity increased, husk and bagasse are used as an alternative energy source to
so did the percentage removal of HAs. Since the ACs used in fossil fuels. Despite the wide consumption of bagasse and rice
this study were basically microporous, the low removal percent- husks as a fuel for the mill boilers, for electricity and steam
ages of HA from solution were rather expected. This observation generation, animal feed or as a raw material for the manufac-
was also confirmed by Tomaszewska et al. [59] who studied the ture of paper and board, the residues still remain as a surplus
removal of organic matter by coagulation enhanced with adsorp- which poses a disposal problem for mill owners. The processing
tion on powdered activated carbons. Daifullah et al. [60] studied and transformation of agricultural residues into activated car-
the factors affecting the removal of HA by AC prepared from rice bon with good adsorption properties would alleviate problems
husk impregnated with H3 PO4 . They found that acidic oxides of disposal and management of these waste by-products, while
such as SiO2 and P2 O5 (commonly present in ash) enhance the providing a high-quality end product for water and wastewater
adsorption of HA by lowering pH. For the same reasons as in treatment that could potentially expand the carbon market.
the case of arsenic, humic acid adsorption on bagasse AC was Our results showed that the ACs prepared from rice husk and
not performed. sugarcane bagasse had good adsorption capacities with respect
to arsenic, humic acid and phenol. They removed ∼80% of phe-
3.5. Activated carbon treatment of a landfill leachate nol at the equilibrium time of 4 h. Regarding the study on the
treatment of landfill leachate using the powdered activated car-
Fig. 6 shows the results obtained when the landfill leachate bon, the results showed that, with 30 g/L of AC it was possible to
was treated by the three different adsorbents. The percentages of remove 70% and 60% of COD and colour, respectively. There-
COD removal obtained with CV-Z were very low, both for 10 and fore, this material may be suitable in pre- or post-treatment steps
30 g/L of CV-Z used, reaching values of 5% and 14%, respec- of landfill leachate, in combination with other processes in order
tively. These results are in agreement with those obtained in to meet regulatory limits.
a previous work [61] using physico-chemical-adsorption treat-
ment. However, the RH AC treatment achieved a COD removal Acknowledgements
of 34% and 70% for the same concentrations. Rodriguez et al.
[62] used a granulated AC with a surface area of 1000 m2 /g and This work was sponsored by the European Union under
three resins to remove COD from landfill leachate. Of the four the European Union’s EU–India Cross Cultural Programme
adsorbents, the AC showed the highest COD removal, reducing (Contract: ALA/95/23/2003/077-124). Ms Olga Pantelaki,
50 D. Kalderis et al. / Chemical Engineering Journal 144 (2008) 42–50

Department of Minerals and Resources Engineering, Techni- from Agricultural By-products in India: Activated Carbon Production from
cal University of Crete, Greece performed the BET surface area Bagasse and Rice Husks’, Final Technical Report, 2007.
determination. [31] S. Brunauer, P.H. Emmett, E. Teller, J. Am. Chem. Soc. 60 (1938)
309.
We also acknowledge CPL Carbon Link and Xavier Querol [32] E.P. Barrett, L.G. Joyner, P.H. Halenda, J. Am. Chem. Soc. 73 (1951)
from the Institute of Earth Sciences Jaume Almera of the Spanish 373.
CSIC for supplying the commercial AC and the synthetic zeolite, [33] M.M. Dubinin, H.F. Stoeckli, J. Colloid Interface Sci. 75 (1) (1980) 34.
respectively. [34] M.M. Dubinin, J. Colloid Interface Sci. 23 (1967) 487.
[35] E. Diamadopoulos, P. Samaras, G.P. Sakellaropoulos, Water Sci. Tech. 25
(1) (1992) 153.
References [36] V. Lenoble, V. Deluchat, B. Serpaud, J.-C. Bollinger, Talanta 61 (3) (2003)
267.
[1] O. Ioannidou, A. Zabaniotou, Renew. Sustain. Energy Rev. 11 (9) (2007) [37] N. Moreno, X. Querol, C. Ayora, C. Fernández Pereira, M. Janssen-
1966. Jurkovicova, Environ. Sci. Technol. 35 (2001) 3526.
[2] M.J. Mitchell, E.M.l Wayne, A.T Christopher, J. Chem. Technol. Biotech- [38] Standard Methods for the Analysis of Water and Wastewater, APHA,
nol. 71 (1998) 131. AWWA, WEF, 1989.
[3] A.A.M. Daifullah, B.S. Girgis, Water Res. 32 (4) (1998) 1169. [39] T. Viraraghavan, J. Alfaro, J. Environ. Eng. 5 (1992) 311.
[4] W.T. Tsai, Y.H. Chou, Y.M. Chang, Renew. Sustain. Energy Rev. 8 (2004) [40] N. Roostaei, F.H. Tezel, J. Environ. Manage. 70 (2004) 157.
461. [41] M. Ahmaruzzaman, D.K. Sharma, J Colloid Interface Sci. 287 (2005) 14.
[5] P.A. Quinlivan, D. Lei Li, R.U. Knappe, Water Res. 39 (2005) 1663. [42] G. Newcombe, J. Morrison, C. Hepplewhite, D.R.V. Knappe, Carbon 40
[6] F. Li, A. Yuasa, K. Ebie, Y. Azuma, T. Hagishita, Y. Matsui, Water Res. 36 (2002) 2147.
(2002) 4592. [43] B. Ericsson, G. Trägärdh, Desalination 108 (1996) 117.
[7] C.L. Aik, Y. Ting, J. Colloid Interface Sci. 290 (2005) 505. [44] J.S. Zogorski, S.D. Faust, J.H. Hass Jr., J. Colloid Interface Sci. 55 (2)
[8] M. Valix, W.H. Cheung, G. McKay, Chemosphere 56 (2004) 493. (1976) 329.
[9] E.F. Jaguaribe, L.L. Medeiros, M.C.S. Barreto, L.P. Araujo, Braz. J. Chem. [45] V.K. Sachan, K.M. Sujatha, S. Kumar, Indian J. Environ. Prot. 16 (4) (1996)
Eng. 22 (1) (2005) 41. 284.
[10] C. Ng, W.E. Marshall, R.M. Rao, R.R. Bansode, J.N. Losso, Ind. Crops [46] M.-W. Jung, K.-H. Ahn, Y. Lee, K.-P. Kim, J.-S. Rhee, J.T. Park, K.-J.
Prod. 17 (2003) 209. Paeng, Microchem. J. 70 (2001) 123.
[11] R.S. Juang, F.C. Wu, R.L. Tseng, Colloids Surfaces A: Physicochem. Eng. [47] S. Mukherjee, S. Kumar, A.K. Misra, M. Fan, Chem. Eng. J. 129 (2007)
Aspects 201 (2002) 191. 133.
[12] A.A. Zabaniotou, G.G. Stavropoulos, V. Skoulou, Bioresour. Technol. 99 [48] B. Özkaya, J. Hazard. Mater. B 129 (2006) 158.
(2008) 320. [49] L.J. Kennedy, J.J. Vijaya, K. Kayalvizhi, G. Sekaran, Chem. Eng. J. 132
[13] D. Savova, E. Apak, E. Ekinci, F. Yardim, N. Petrov, D. Budinova, M. (2007) 279.
Razvigorova, V. Minkova, Biomass Bioenergy 21 (2001) 133. [50] J. Rivera-Utrilla, E. Utrera-Hidalgo, M.A. Ferro-Garcia, C. Moreno-
[14] B.S. Girgis, S.S. Yunis, A.M. Soliman, Mater. Lett. 57 (2002) 164. Castilla, Carbon 29 (4–5) (1991) 613.
[15] F. Caturla, M. Molina-Sabio, F. Rodriguez-Reinoso, Carbon 29 (7) (1991) [51] V.K. Gupta, D. Mohan, Suhas, K.P. Singh, Ind. Eng. Chem. Res. 45 (2006)
999. 1113.
[16] G.G. Stavropoulos, A.A. Zabaniotou, Microporous Mesoporous Mater. 82 [52] R.-S. Juang, S.-H. Lin, C.-H. Cheng, Ultrason. Sonochem. 13 (2006)
(2005) 79. 251.
[17] F.-C. Wu, R.-L. Tseng, J. Colloid Interface Sci. 294 (2006) 21. [53] D. Mohan, C.U. Pittman, J. Hazard. Mater. 142 (1–2) (2007) 1.
[18] F. Villacañas, M.F.R. Pereira, J.M. Órfão José, L. José Figueiredo, J. Colloid [54] C.P. Huang, P.L. Fu, J. Wat. Poll. Contr. Feder. 56 (1984) 233.
Interface Sci. 293 (2006) 128. [55] C.P. Huang, L.M. Vane, J. Wat. Poll. Contr. Feder. 61 (1989) 1596.
[19] F.Sh. Mohamed, W.A. Khater, M.R. Mostafa, Chem. Eng. J. 116 (2006) [56] L. Lorenzen, J.S.J. van Deventer, W.M Landi, Miner. Eng. 8 (4–5) (1995)
47. 557.
[20] M. Ahmedna, W.E. Marshall, A.A. Husseiny, R.M. Rao, I. Goktepe, Water [57] S. Han, S. Kim, H. Lim, W. Choi, H. Park, J. Yoon, T. Hyeon, Microporous
Res. 38 (2004) 1062. Mesoporous Mater. 58 (2003) 131.
[21] R.R. Bansode, J.N. Losso, W.E. Marshall, R.M. Rao, R. Portier, J. Biore- [58] E. Lorenc-Grabowska, G. Gryglewicz, J. Colloid Interface Sci. 284 (2005)
sour. Technol. 90 (2003) 175. 416.
[22] C.A. Toles, W.E. Marshall, L.H. Wartelle, A. Mc Aloon, Bioresour. Tech- [59] M. Tomaszewska, S. Mozia, A.W. Morawski, Desalination 16 (1) (2004)
nol. 75 (2000) 197. 79.
[23] G. Skodras, Ir. Diamantopoulou, A. Zabaniotou, G. Stavropoulos, G.P. [60] A.A.M. Daifullah, B.S. Girgis, H.M.H. Gad, Colloids Surfaces A: Physic-
Sakellaropoulos, Fuel Process. Technol. 88 (8) (2007) 749. ochem. Eng. Aspects 235 (2004) 1.
[24] I. Kula, M. Ugurlu, H. Karaoglu, A. Celik, Bioresour. Technol. 99 (3) [61] E. Otal, L.F. Vilches, N. Moreno, X. Querol, J. Vale, C. Fernández Pereira,
(2008) 492. Fuel 84 (2005) 1440.
[25] F.J. Rivas, F.J. Beltran, O. Gimeno, J. Frades, F. Carvalho, J. Hazard. Mater. [62] J. Rodriguez, L. Castrillon, Marañon, H. Sastre, E. Fernandez, Water Res.
131 (1–3) (2006) 170. 38 (2004) 3297.
[26] F.J Rivas, F. Beltran, F. Carvalho, O. Gimeno, J. Frades, Ind. Eng. Chem. [63] M. Ahmedna, W.E. Marshall, R.M. Rao, Bioresour. Technol. 71 (2000) 3.
Res. 44 (8) (2005) 2871. [64] A.A. Bagreev, A.P. Broshnik, V.V. Strelko, A. Tarasenko Yu, Sorption Ion
[27] T.A. Kurniawan, Wai-hung Lo, G. YS Chan, J. Hazard. Mater. 129 (1–3) Exch. Process. 74 (2001) 205.
(2006) 80. [65] A. Aygun, S. Yenisoy-Karakas, I. Duman, Microporous Mesoporous Mater.
[28] F. Kargı, M.Y. Pamukoglu, Bioresour. Technol. 94 (3) (2004) 285. 74 (2003) 189.
[29] Ö. Aktaş, F. Çeçen, Environ. Eng. Sci. 21 (3) (2004) 303. [66] Q. Cao, K.-C. Xie, Y.-K. Lv, W.-R. Bao, Bioresour. Technol. 97 (2006)
[30] 6th FP EU–India Economic Cross Cultural Programme Project titled ‘Pro- 110.
duction of High Added Value Products for Environmental Applications [67] Q. Jia, A.C. Lua, Adsorption 13 (2007) 129.