Arabian Journal of Chemistry (2016) 9, S1155–S1162

King Saud University

Arabian Journal of Chemistry

www.ksu.edu.sa
www.sciencedirect.com

ORIGINAL ARTICLE

Characterization of activated carbon prepared

by phosphoric acid activation of olive stones
S.M. Yakout *, G. Sharaf El-Deen

Hot Labs Center, Atomic Energy Authority, P.C. 13759 Cairo, Egypt

Received 26 June 2011; accepted 10 December 2011

Available online 22 December 2011

KEYWORDS Abstract The effects of activating agent concentration on the pore structure and surface chemistry
Activated carbon; of activated carbons derived from olive stone with chemical activation method using phosphoric acid
Yield; as the activating agent were studied. Mass changes associated with the impregnation, carbonization
Olive stones; and washing processes were measured. With H3PO4 dilute solutions (60, 70, and 80 wt% H3PO4), the
Phosphoric acid; loading of substance on CS increases with concentration. The concentration of the H3PO4 solution
Porous texture seems to control the processes of impregnation, carbonization and washing in the preparation of AC
from olive stones by H3PO4 chemical activation. ACs have been characterized from the results
obtained by N2 adsorption at 77 K. Moreover, the fractal dimension (D) has been calculated in order
to determine the AC surface roughness degree. Optimal textural properties of ACs have been
obtained by chemical activation with H3PO4 80 wt.%. The BET surface areas and total pore volumes
of the carbons produced at H3PO4 80 wt.% are 1218 m2/g and 0.6 cm3/g, respectively.
ª 2011 Production and hosting by Elsevier B.V. on behalf of King Saud University. This is an open access
article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).

1. Introduction tivated surface area, and 99% of the total olive fruit
production (Niaounakis and Halvadakis, 2006). The world
Olives are the most extensively cultivated fruit crop in the production of olives, for the year 2007, was 17.4 Mton, of
world. Olive cultivation is particularly widespread throughout which 12.6 Mton came from Europe (FAOSTAT, 2009).
the Mediterranean region and plays an important role in its Activated carbon is also among the potential conversions,
rural economy, local heritage, and environment protection. with applications in the removal of dyes, odors, tastes, and
The largest producing countries are located in the Mediterra- contaminants, in water purification and other decontamina-
nean and Middle East regions providing 98% of the total cul- tion processes (Matos et al., 2010). Olive stone is a lignocellu-
losic material, with hemicellulose, cellulose and lignin as the
* Corresponding author. Tel.: +20 0482807763.
main components. Olive stones could be a very adequate feed-
stock to obtain active carbons with good adsorptive properties
E-mail address: sobhy.yakout@gmail.com (S.M. Yakout).
and hardness, which could be of interest in future environmen-
Peer review under responsibility of King Saud University.
tal protection programs.
Activated carbon (AC) is widely used on an industrial scale
as an adsorbent mainly in the purification/separation of liquids
Production and hosting by Elsevier and gases and also as a catalyst and catalyst support

http://dx.doi.org/10.1016/j.arabjc.2011.12.002
1878-5352 ª 2011 Production and hosting by Elsevier B.V. on behalf of King Saud University.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).
S1156 S.M. Yakout, G. Sharaf El-Deen

(Olivares-Marin et al., 2006). Furthermore, newer applications temperature) in inert atmosphere of nitrogen (flow
are ever emerging, particularly those concerning environmen- rate = 200 mL/min. The heating rate was 3 C/min and hold-
tal protection and technological development. ing time 2 h). After heating, the system was cooled down to
There are basically two methods for activated carbons’ pro- room temperature under the same flow of nitrogen and the
duction: physical and chemical activations. Physical activation product obtained was washed with distilled water until pH 6
includes the carbonization of the starting material and the acti- was reached in the residual liquid and finally oven-dried at
vation of the carbonizate using steam and carbon dioxide. In the 120 C. OS6, OS7, OS8 were used as a label for activated car-
chemical activation method, phosphoric acid and zinc chloride bon prepared from olive stones at H3PO4 60wt.%, 70wt.%,
are used for the activation of lignocellulosic materials, which 80wt.% concentrations, respectively.
have not been carbonized previously; whereas metal com- Low temperature-ramping rate of 3 C/min was used in this
pounds such as potassium hydroxide are used for the activation stage to minimize the temperature difference between the fur-
of coal precursors or chars. When compared to zinc chloride, nace chamber and activation mixture as well as maintain the
phosphoric acid is the most preferred because of the environ- mixture at a specific temperature range for enough time so that
mental disadvantages associated with zinc chloride. Problems perceptible activation extent could be obtained and an opti-
of corrosion and inefficient chemical recovery are also associ- mized activation temperature range could be determined.
ated with it. Moreover, the carbons obtained using zinc chloride
cannot be used in pharmaceutical and food industries as they 2.2. Active carbon characterization
may contaminate the product (Prahas et al., 2008). Although
potassium hydroxide develops large microporosity, the yield
Nitrogen adsorption/desorption isotherms were measured at
of activated carbon impregnated by potassium hydroxide is
77 K on an automatic adsorption instrument (Quantachrome
lower than those activated with zinc chloride or phosphoric
Instruments, Model Nova1000e series, USA) in relative pres-
acid, and at high temperature, i.e. > ±650 C, the carbon con-
sure ranging from 10 6 to 0.999. Prior to the measurement,
tent is less than fixed carbon in the initial precursor. The pres-
all the samples were crushed and powdered to shorten the time
ence of metallic potassium will intercalate to the carbon
required for reaching equilibrium in the isotherm study and de-
matrix (Prahas et al., 2008), yielding lower yield of activated car-
gassed at 250 C under nitrogen flow for 16 h.
bon, less than the carbon content of the raw material. H3PO4 is
The nitrogen adsorption–desorption isotherms were used to
the most widely used impregnation agent. Using this substance,
determine the following parameters: specific surface area
most of the numerous studies carried out so far on the prepara-
SBET (according to the BET equation), total pore volume Vtot
tion of AC have focused on the influence of concentration of the
(calculated from the nitrogen uptake at relative pressure of
impregnation solution and soaking temperature on the porous
0.95), total micropore volume Vmicro, according to simplified
structure. Because of the highly polar character of H3PO4 and
equations (Mehandjiev et al., 1994; Nickolov and Mehandjiev,
hence the control of the physical and chemical interactions
2000, 1995), total mesopore volume Vmes (determined by sub-
occurring in the bulk of the solution and with the substratum
tracting the micropore volume from the total pore volume),
during the impregnation treatment, the solution concentration
mean pore radius rp (calculated using the total pore volume
is likely to be the primary factor of the activation process. Here,
and the BET surface area, assuming a cylindrical pore model),
dilute and concentrated solutions of H3PO4 are used in the
the micropore size distribution parameter Xo (the half-width of
impregnation of an AC precursor, the influence of phosphoric
the distribution curve maximum) was calculated according to
acid concentration on the loading of H3PO4 is investigated. This
simplified equation of the Horvath–Kawazone (HK) method
work deals with the influence of the treatment methods of olive
(Ustinov and Do, 2002).
stones and SEOP on the structure and surface properties of the
The fractal dimension is often used as an index of rough-
produced activated carbons.
ness or irregularity of the surface of ACs (Dıaz-Dıez et al.,
2004) among other materials. In this work, the fractal dimen-
2. Materials and methods sion was determined by applying the FHH (Frenkel–Halsey–
Hill) equation (Halsey, 1948) to the adsorption isotherm.
2.1. Active carbon preparation Apparent (bulk) density of all samples was calculated as the
ratio between weight and volume of packed dry material.
As indicated above, olive stones (OS) were used as a precursor. The FTIR spectra of the samples were recorded between
OS as received was ground, dried and sieved, and only the 4000 and 450 cm 1 in a Perkin-Elmer 1720 spectrometer. Pel-
fraction of particle sizes comprised between 1 and 2 mm was lets were prepared by thoroughly mixing carbon and KBr at
selected for the preparation of activated carbons. Approxi- the 1:400 carbon/KBr weight ratio in a small size agate mortar.
mately, 50 g of conditioned OS was impregnated with 200 ml The resulting mixture was compacted in a Perkin-Elmer man-
of phosphoric acid solutions with concentrations of 60, 70 ual hydraulic press at 10 ton for 3 min.
and 80 wt.% mixture stirred at 85 C, maintaining for 4 h to The sample morphology was observed using a scanning
ensure the access of H3PO4 to the interior of the olive stones. electron microscope (SEM) model JEOL 5400. The specimens
After mixing, carbonization of acid-impregnated OS was car- for SEM observation were prepared by depositing the fibers
ried out in a stainless steel reactor provided with gas inlet onto specimen-stubs with conductive double sticky copper
and outlet, which was placed in a vertical cylindrical furnace. tapes, and then sputter-coating (model Polaron SC7640, Quo-
The temperature within the furnace was first calibrated and rum Technologies Ltd., UK) the sample surface with Au–Pd to
the length and position of the constant temperature hot zone prevent electrical charging during examination. Imaging was
determined. Heating from room temperature to the maximum done in the high vacuum mode under an accelerating voltage
heat treatment temperature 500 C (i.e., carbonization of 15 kV, using secondary electrons.
Characterization of activated carbon prepared by phosphoric acid activation of olive stones S1157

3. Result and discussion A slightly lower bulk density of activated carbon product
was also observed at high phosphoric acid content due to the
3.1. Yield of activated carbon difference in the pore structure of the carbon product as dis-
cussed later on.
The nomenclature, operation conditions, density and yield of
3.2. Characterizing pore structure of the carbons
the different samples prepared by the method of chemical acti-
vation with phosphoric acid are summarized in Table 1. The
results obtained clearly show that, as expected, as H3PO4 con- The shape of the adsorption isotherm can provide qualitative
centration increases the total yield of the process decreases. information on the adsorption process and the extent of the
Essentially, Olive stone is a lignocellulosic material, with hemi- surface area available to the adsorbate. Fig. 1 shows the N2
cellulose, cellulose and lignin as the main components. In acti- adsorption–desorption isotherms at 196 C of OS-based car-
vation or carbonization, these polymeric structures decompose bon prepared by phosphoric acid activation at different con-
and liberate most of the non-carbon elements, mainly hydro- centration. It is apparent that all isotherms of all carbon
gen, oxygen and nitrogen in the form of liquid (called as tars) series belonging to type I according to the IUPAC classifica-
and gases, leaving behind a rigid carbon skeleton in the form tion (Sing et al., 1985), showing a significant increase in the
of aromatic sheets and strips. In other words, the reaction of adsorption at low P/P0 values (<0.1), with barely defined knee
lignocellulose with phosphoric acid begins as soon as the com- and long plateau which extends to P/P0  1.0, indicate that the
ponents are mixed, the acid first attacks hemicellulose and lig- ACFs are essentially microporous solids. An absence of hyster-
nin because cellulose is known to be more resistant to acid esis indicated the absence of mesoporosity, and suggested that
hydrolysis (Jagtoyen and Derbyshire, 1998). Here the acid will the carbon products contained mostly micropores with only a
hydrolyze glycosidic linkages in lignocellulosic and cleave aryl small contribution of mesopore.
ether bond in lignin. These reactions are accompanied by fur- In the present investigation, although the adsorption iso-
ther chemical transformations that include dehydration, degra- therms for all samples are similar, the adsorption capacities
dation, and condensation. As the acid concentration increases, are significantly different according to the H3PO4 concentra-
the aromatic condensation reactions also take place among the tion. It was observed that increasing H3PO4 concentration
adjacent molecules, which result in the evolution of gaseous shifts the isotherm upward compared to that of other carbons.
products from the hydroaromatic structure of carbonized char To take full advantages of the information provided by
leading to decreased yield of carbon (Timur et al., 2006). Also, high-resolution nitrogen isotherm, these isotherms are repre-
here the excess phosphoric acid will promote gasification of sented on a semilogarithmic scale (Li et al., 1998). The adsorp-
char and increased the total weight loss of carbon. The same tion isotherms expressed in a logarithmic pressure scale are
result was also observed by other researchers (Jagtoyen and often useful to investigate the differences in the structural
Derbyshire, 1998; Timur et al., 2006). properties especially at lower P/P0 region (Park and Jung,
As seen from table, the bulk density decreases with increas- 2003). Adsorption processes can be divided into several dis-
ing acid concentration. This means that increasing acid con- tinct stages at relative pressures of 10 6–10 4, 10 4–10 2,
centration results in carbon richer in carbon and is more 10 2–10 1, 0.1–0.9, and 0.9–1.0 which can be attributed to
porous. Also, as seen the weight after impregnation increases multiple stage pore filling processes (Ryu et al., 2000), such
as the acid concentration increases. This means that the weight as: (i) filling of narrow micropores; (ii) monolayer formation
of residue after phosphoric acid impregnation shows a great on the surfaces of wider micropores such as supermicropores;
influence of the H3PO4 concentration on the amount of (iii) filling of these wider micropores; (iv) monolayer formation
H3PO4 loaded on OS. and filling of mesopores by capillary condensation, and (v) fill-
From these results it is clear that with the concentrated ing of macropores by capillary condensation, which takes
H3PO4 solution a synergic effect occurred in the impregnation place at relative pressures close to unity.
of OS which was likely connected with the strong dependence Fig. 2 shows the adsorbed volume of N2 on the carbons vs.
of the molecular association of H3PO4 on the concentration of the logarithmic pressure. As seen from these semilogarithmic
its aqueous solutions. Perhaps as a result of the loading of plots all samples show no uptake of nitrogen under a relative
H3PO4 on OS, the supernatant diluted and this facilitated the pressure of 10 3, indicating the presence of a few ultra microp-
vaporization of water to dryness. For the dilute solutions, the ores. There is significant uptake of nitrogen in the relative pres-
solute–solvent interactions occurring in the bulk of the solution sure range of 10 3–10 1 indicating that all prepared samples
should be stronger than for the concentrated solution because of have great part of its pores as super micropores and wider
the smaller degree of H3PO4 association, and this should have micropores.
an unfavorable effect on the loading of H3PO4 on OS. The effect of acid concentration on the BET surface area,
Close inspection of the weight of carbon after washing com- micropore, mesopore volume, and total pore volume are given
pared after carbonization indicates that loaded phosphoric in Table 2. The total pore volume, Vtotal, was calculated from
acid species were very sensitive to the subsequent washing. nitrogen adsorption data as volume of liquid nitrogen at a rel-
This was expected as the mass loss produced by washing of ative pressure of 0.95. The micropore volume, Vmicro, was
the OS-derived carbon with distilled water is attributable to determined by DR method, and the mesopore volume, Vmeso,
lixiviation of phosphorus species, which are very polar ions/ was obtained by the subtraction of micropore volume and
molecules and therefore highly soluble in water. Perhaps, the total pore volume. As can be seen, the BET surface area and
solution process was favored as a result of a decrease in the size total pore volume increase with the increasing acid concentra-
of phosphorus species present in the OS carbon because of the tion. The increase in porosity with temperature can be attrib-
carbonization process. uted to the release of tars from cross-linked framework
S1158 S.M. Yakout, G. Sharaf El-Deen

Table 1 Nomenclature, preparation conditions and yields of the activation procedure.

Carbon H3PO4 Carbonization Weight after Weight after Weight after Bulk Yield
name (wt.%) temperature (C) impregnation (g) carbonization washing/drying density (g/ (%)
(g) (g) cm3)
OS6 60 500 50.5 23.2 18.4 1.0 36.8
OS7 70 500 54 28.5 16.9 0.84 33.8
OS8 80 500 57 32.7 15.6 0.83 31

Similar effects of the amount of phosphoric acid on poros-

ity development have been reported for other lignocellulosic
precursors (Baquero et al., 2003; Molina-Sabio et al., 1995).
It is generally accepted that the porosity is generated with
phosphoric acid remaining intercalated in the internal struc-
ture of lignocellulosic materials. As the amount of H3PO4 used
increases, the volume filled by it and various polyphosphates
will increase, resulting in larger pore volume and pore size.
Fig. 3 shows as-plots drawn from the isotherms in Fig. 1. It
has been shown that the as-plots of all activated carbons show
two straight points (one for as = 1.0 and other for as > s1.0).
The second linear section is almost parallel to the x-axis. This
indicates the presence or absence of mesoporosity. The exten-
sion of the points at lower P/P0 values does not pass through
the origin but cuts the y-axis for a positive value. This kind of
deviation has been ascribed to the presence of narrow micro-
porosity in these activated carbons (Seller-Perez and Martin-
Figure 1 Nitrogen adsorption isotherms for olive stone carbons. Martinez, 1991). Regarding the as-plot analysis, Kaneko
et al. (1992) found that the microporous carbon could show
two upward swings from linearity below the downward bend-
ing due to saturated filling at higher as, which they designated
as filling swing FS and cooperative swings CS, corresponding
to as smaller and greater than 0.5, respectively. Both swings
(FS and CS) come from enhanced adsorbent–adsorbate inter-
action as a function of the pore size and can be ascribed to dif-
ferent types of filling: FS results from primary micropore
filling, which takes place in the ultramicropore (<0.8 nm)
and CS results from cooperative adsorption in larger micro-
pore (supermicropores of diamater > 1 nm). The as-plot given
in Fig. 3 shows that almost all samples show cooperative
swings (0.5 < as < 1) with little evidence for filling swing
(0 < as < 0.5) further confirming that wider micropores are
the dominant pores in these samples.
Our results confirm the data obtained by (Terzyk et al. 2002)
who studied the relation between DA-method parameters (E
and n) and concluded that carbon of the same n values show
as-plot with swings (FS or CS) depend on E values. In this con-
cern, for high E values (>30 kJ/mol) the obtained as-plot be-
Figure 2 Semi-logarithmic scale of nitrogen adsorption iso- longs to the FS group. For E in the range 20–25 kJ/mol, the
therms for olive stone carbons. as-plot belongs to the FS/CS-type. For the smallest energies,
the as-plot becomes the CS type. For our carbons n value of
DA-method is 1.8 and the value of E is small (5.0–5.5 kJ/mol).
generated by the treatment of phosphoric acid (Hsu and Teng, Then the resultant as-plot is CS type and nitrogen adsorption
2000). The results are consistent with the findings of other mechanism is related to the cooperative swing of secondary
workers (Molina-Sabio et al., 1995) who suggested that micropore filling (in supermicropores).
increasing the amount of phosphorus leads to an increase in In our opinion, the picture of narrow microporosity from
the volumes of micro- and mesopores (table). The experimen- nitrogen adsorption is not obvious. There is lower cutoff of
tal data in Table 2 also show an increase in the average pore pore width less than two nitrogen molecule diameters (around
diameter with increasing concentration, showing that the 0.4 nm) is generally assumed for nitrogen adsorption at 77 K
development of porosity is also accompanied by a widening owing to restricted diffusion of nitrogen in narrow micropo-
of the porosity as the amount of H3PO4 is increased. rosity. Thus, it has been suggested that the adsorption of these
Characterization of activated carbon prepared by phosphoric acid activation of olive stones S1159

Table 2 Pore structure parameters of olive stones carbons.

Carbon SBET (m2/g) r (nm) Vtotal (cc/g) Vmicro (cc/g) Vmeso (cc/g)
OS6 257 0.954 0.123 0.11 0.012
OS7 779 1.0 0.35 0.32 0.03
OS8 1218 1.1 0.6 0.5 0.1

Figure 3 a-plot of various olive stones’ carbon samples.

Figure 4 H–K pore size distribution of olive stones activated
carbon samples.
two molecules should be used in a complementary way in order
to characterize microporosity more fully such as carbon diox- the micropore volumes, since the area under the HK curve is
ide adsorption. In this way, data obtained from nitrogen larger in this latter case. The fact that the pore size distribution
adsorption isotherms should only be taken as a qualitative esti- plots show Downward end in the zone of the wider micropores
mation of the presence, to a lesser or larger extent, of narrow close to the mesopores (i.e., around 2 nm) suggests no develop-
microporosity (Bello et al., 2002). ment of the mesoporosity in our samples.
Bansal et al., (1988) have fit adsorption data for strongly
activated and heterogeneous carbons with n values (of DA 3.3. External surfaces of resulting carbon using SEM
equation) between 1.5 and 2. For the activated carbons studied
herein, it has been found that n is around 1.8. This indicates
that carbons under study are strongly activated with great het- Scanning electron micrographs of the surface morphology of
erogenous micropore structure. different samples of the activated carbons are given in Fig. 5.
The structural heterogeneity of porous material is generally This figure shows the differences of the surface and porosity
characterized in terms of the pore size distribution. This pore of the activated carbons prepared at 60wt.% and 70wt.%
size distribution represents a model of solid internal structure, H3PO4. From these figures, it is obvious that the carbons pro-
which assumes that an equivalent set of non-interacting and duced at 60wt.% are non-porous carbons, while the activated
regularly shaped model pores can represent the complex void carbons produced at 70wt.% have cavities and cracks on their
spaces within the real solid (Ismadji and Bhatia, 2001). The external surfaces. It seems that the cavities on the surfaces of
pore size distribution is closely related to both kinetic and carbons resulted from the evaporation of the activating agent
equilibrium properties of porous material and perhaps is the which in this case is phosphoric acid during carbonization,
most important aspect for characterizing the structural hetero- leaving the space previously occupied by the activating agent.
geneity of porous materials used in industrial application. The The impregnation with H3PO4 followed by a thermal treat-
pore size distribution of the samples was evaluated by Horv- ment under inert atmosphere involves a remarkable degrada-
arth–Kawazoe (HK) method (Ustinov and Do, 2002). The tion of the microstructure. Such degradation is coupled to
HK pore size distribution of activated carbons studied is relatively important mass losses. This mass effect becomes
shown in Fig. 4. more and more noticeable as the acid increases, which results
The HK plots are in all cases bimodal, with two maxima in a progressive decrease of the global yield of the process.
centered around 0.8 and 1.0 nm for OS6 and shifted to new
two peak maxima 1.2 and 1.6 for OS7 and OS8 with the 3.4. FTIR analysis
increase of acid concentration. The predominant pore width
indicates here that the presence of semicropore (diame- The carbon matrix does not consist of carbon atoms alone, but
ter<0.7 nm) is not observed. OS8 exhibits larger values of is also formed by other heteroatoms like hydrogen, oxygen,
S1160 S.M. Yakout, G. Sharaf El-Deen

Figure 5 Scanning electron microscope of OS6 (a) and OS7 (b).

nitrogen, halogen, sulfur, phosphorus, etc. These heteroatoms ally assigned to C‚O stretching vibrations of ketones,
are bonded to the edges of the carbon layers, which govern the aldehydes, lactones or carboxyl groups.
surface chemistry of the activated carbon. The spectra of the prepared activated carbons also show a
Infrared spectroscopy provides information on the chemi- strong band at 1600–1580 cm 1 due to C–C vibrations in
cal structure of the carbon material. Fig. 6 shows FTIR spectra aromatic rings. Broad band at 1000–1300 cm 1 (maxima at
of the synthetic carbons obtained by phosphoric acid activa- 1190–1200 cm 1) is usually found with oxidized carbons and
tion at different concentrations. All spectra show a wide trans- has been assigned to C–O stretching in acids, alcohols, phe-
mittance band at 3200–3600 cm 1 with a maximum at about nols, ethers and/or esters groups (Zawadzki, 1989). Neverthe-
3420–3440 cm 1. This band can be assigned to the O–H less, it is also a characteristic of phosphorous and phosphor
stretching mode of hydroxyl groups and adsorbed water. The carbonaceous compounds present in the phosphoric acid acti-
position and asymmetry of this band at lower wave numbers vated carbons (Puziy et al., 2002). Assignment in this region is
indicate the presence of strong hydrogen bonds (from carbox- difficult because absorption bands are overlapped. The peak at
yls, phenols or alcohols) (Solum et al., 1995). A weak sharp 1190–1200 cm 1 may be also assigned to the stretching mode
transmittance band at 3733 cm 1 present in spectra of all car- of hydrogen-bonded P‚O, to O–C stretching vibrations in
bons and decreasing in carbons treated with high H3PO4 con- P–O–C (aromatic) linkage and to P‚OOH (Puziy et al.,
centration (OS7 and OS8). This peak may be ascribed to 2002). The shoulder at 1100 cm 1 was ascribed to ionized link-
isolated O–H groups. age P–O– in acid phosphate esters, and to symmetrical vibra-
The FTIR spectrum of synthetic carbons OS6 and OS8 tion in a P–O–P chain (Bourbigot et al., 1995). To conclude
shows absorption bands due to aliphatic (2921 and on IR characterization, the most important changes
2855 cm 1: C–H stretching in –CH–; 1450 cm 1: –CH– defor- introduced by the increase of the acid concentration are the
mation). In Fig. 6, the bands at 885, 840, and 775 cm 1 are due development of C–H vibrations (possibly because of the loss
to out-of-plane deformation mode of C–H for different substi- of oxygen at the surface of the carbon material) as well as
tuted benzene rings. The small band at about 1700 cm 1 is usu- the increase of phosphorous group content (1100 cm 1).

Figure 6 FTIR spectra of activated olive stones at different H3PO4 concentrations.

Characterization of activated carbon prepared by phosphoric acid activation of olive stones S1161

When concentrated H3PO4 is mixed with date stones at high pyrolysis of coffee bean husks in presence of phosphoric acid. J.
temperature, it appears to function both as an acid catalyst to Anal. Appl. Pyrolysis 70, 779–784.
promote bond cleavage reactions and formation of cross-links Bello, G., Garcia, R., Arriagada, R., Sepulveda-Escribano, A.,
via processes such as cyclization and condensation and to Rodriguez-Reinoso, F., 2002. Carbon molecular sieves from
eucalyptus globulus charcoal. Micropor. Mesopor. Mat. 56, 139–
combine with organic species to form phosphate and polyphos-
145.
phate bridges that connect and crosslink biopolymer fragments Benaddi, H., Legras, D., Rouzaud, J.N., Beguin, F., 1998. Influence of
(Jagtoyen and Derbyshire, 1998). Dastgheib and Rockstraw the atmosphere in the chemical activation of wood by phosphoric
(2001) proposed that various surface acidic functional groups acid. Carbon 36, 306–309.
(oxygen- and/or phosphorus-containing groups) are developed Bourbigot, S., Bras, M.L., Delobel, R., 1995. Carbonization mecha-
through the surface oxidation as well as attachment of different nisms resulting from intumescence. II. Association with an ethylene
oxygen/phosphorous groups to the surface, while developing re- terpolymer and the ammonium polyphosphate-pentaerythritol fire
quired porosity. Benaddi et al. (1998) suggested that dehydra- retardant system. Carbon 33 (3), 283–294.
tion of cellulose by phosphoric acid is similar to dehydration Dastgheib, S.A., Rockstraw, D.A., 2001. Pecan shell activated carbon:
of alcohols and at higher temperatures the phosphorous oxides synthesis, characterization, and application for the removal of
copper from aqueous solution. Carbon 39, 1849–1855.
act as Lewis acids and can form C–O–P bonds.
Dıaz-Dıez, M.A., Gomez-Serrano, V., Fernandez Gonzalez, C.,
Cuerda-Correa, E.M., Macıas-Garcıa, A., 2004. Porous texture of
3.5. Fractal dimension of samples activated carbons prepared by phosphoric acid activation of
woods. Appl. Sur. Sci. 238, 309–313.
Fractal analysis has become a new and powerful tool to de- FAOSTAT, 2009. Website of Food and Agriculture organization of
scribe the surface heterogeneity, geometric and structural the United Nations. Available from: <http://faostat.fao.org/>.
Halsey, G.D., 1948. J. Chem. Phys. 16, 93.
properties of fractal surfaces and pore structures. The fractal
Hayashi, J., Horikawa, T., Muroyama, K., Gomes, V.G., 2002.
dimension is the number used to quantify these properties:
Micropor. Mesopor. Mat. 55, 63–68.
the larger the value of the surface fractal dimension, the more Hsu, L.Y., Teng, H., 2000. Influence of different chemical reagents on
irregular and rougher the pore surface (Pyun and Rhee, 2004). the preparation of activated carbons from bituminous coal. Fuel
According to Frenkel, Halsey and Hill, a perfectly smooth Process Technol. 64, 155–166.
surface has the fractal dimension of 2, whereas a very rough or Ismadji, S., Bhatia, S.K., 2001. Characterization of activated carbons
irregular surface has the fractal dimension of 3. An activated using liquid phase adsorption. Carbon 39, 1237–1250.
carbon sample consists of very small crystallites such as graph- Jagtoyen, M., Derbyshire, F., 1998. Activated carbons from yellow
ite and disorganized carbon, because of the cross-linked struc- poplar and white oak by H3PO4 activation. Carbon 36, 1085–1097.
ture between the crystallites. If activated carbon has a Kaneko, K., Ishii, C., Ruike, M., Kuwabara, H., 1992. Origin of super
high surface area and microcrystalline graphitic structures of
structure in which the crystallites are located parallel with re-
activated carbons. Carbon 30, 1075.
spect to each other and has no cross-linked structure between
Khalili, N.R., Pan, M., Sandi, G., 2000. Determination of fractal
the crystallites, then the fractal dimension is close to 2. On the dimensions of solid carbons from gas and liquid phase adsorption
contrary, if the crystallites are located in a disorderly manner isothermes. Carbon 38, 573–588.
and cross-linked structures develop, then the fractal dimension Li, Z., Kruk, M., Jaroniec, M., Ryu, S.K., 1998. Characterization of
is close to 3. Thus, there is a close relationship between changes structure and surface properties of activated carbon fiber. J.
in the fractal dimension with that of the structure of activated Colloid. Interface Sci. 204, 151–156.
carbon (Hayashi et al., 2002). Matos, M., Barreiro, M.F., Gandini, A., 2010. Olive stone as a
The D values obtained for all products prepared in this renewable source of biopolyols. Ind. Crop. Prod. 32, 7–12.
study are 2.95, 2.97 and 2.99 for OS6, OS7 and OS8, respec- Mehandjiev, D., Bekyarova, E., Nickolov, R., 1994. Micropore size
distribution by a simplified equation. Carbon 32, 372–374.
tively. Increase in D values is observed as H3PO4 concentration
Molina-Sabio, M., Rodrı́guez-Reinoso, F., Caturla, F., Sellés, M.J.,
grows which is indicative of a development of the porous tex-
1995. Porosity in granular carbons activated with phosphoric acid.
ture, with the increasing concentration of acid. Generally all Carbon 33, 1105.
samples have high values of fraction dimension near 3, which Niaounakis, M., Halvadakis, C.P., 2006. Olive Processing Waste
suggests that the surfaces are irregular and have strong micro- Management: Literature Review and Patent Survey. Elsevier,
pore structures (Khalili et al., 2000). Amsterdam.
Nickolov, R., Mehandjiev, D., 2000. The simplified equation for
4. Conclusions micropore size distribution in adsorbents with different texture and
chemical nature. In: Proceeding of 9th International Symposium of
Catalysis, Varna, pp. 193–198.
Activated carbons with a well-developed pore structure are Nickolov, R.N., Mehandjiev, D.R., 1995. Application of the simplified
prepared from Olive stone by chemical activation with phos- equation for micropore size distribution to the study of water
phoric acid. In general, as the phosphoric acid concentration vapour adsorption on activated carbon. Adsorpt. Sci. Technol. 12,
increases the porous structure developed gives high surface 203–209.
and high total pore volume carbons. Olivares-Marin, M., Fernandez-Gonzalez, C., Macias-Garcia, A.,
Gomez-Serrano, V., 2006. Preparation of activated carbon from
cherry stones by chemical activation with ZnCl2. Sur. Sci. 252,
References
5967–5971.
Park, S.J., Jung, W.Y., 2003. KOH activation and characterization of
Bansal, R.C., Donnet, J.B., Stoeckli, F., 1998. Active Carbon. Marcel glass fibers-supported phenolic resin. J. Colloid. Interface Sci. 265,
Deckker, New York. 245–250.
Baquero, M.C., Giraldo, L., Moreno, J.C., Suárez-Garcı́a, F., Martı́- Prahas, D., Kartika, Y., Indraswati, N., Ismadji, S., 2008. Activated
nez-Alonso, A., Tascón, J.M.D., 2003. Activated carbons by carbon from jackfruit peel waste by H3PO4 chemical activation:
S1162 S.M. Yakout, G. Sharaf El-Deen

Pore structure and surface chemistry characterization. Chem. Eng. Solum, M.S., Pugmire, R.J., Jagtoyen, M., Derbyshire, F., 1995.
J. 140, 32–42. Evolution of carbon structure in chemically activated wood.
Puziy, A.M., Poddubnaya, O.I., Martınez-Alonso, A., Suarez-Garcıa, Carbon 33, 1247–1254.
F., Tascon, J.M.D., 2002. Synthetic carbons activated with Terzyk, A.P., Gauden, P.A., Kowalczyk, P., 2002. What kind of pore
phosphoric acid I. Surface chemistry and ion binding properties. size distribution is assumed in the Dubinin–Astakhov adsorption
Carbon 40, 1493–1505. isotherm equation? Carbon 40, 2879–2886.
Pyun, S.-I., Rhee, C.-K., 2004. An investigation of fractal character- Timur, S., Kantarli, I.C., Ikizoglu, E., Yanik, J., 2006. Preparation of
istics of mesoporous carbon electrodes with various pore structures. activated carbons from oreganum stalks by chemical activation.
Electrochim. Acta 49, 4171–4180. Energ. Fuel. 20, 2636–2641.
Ryu, Z., Zheng, J., Whang, M., Zhang, B., 2000. Nitrogen adsorption Ustinov, E.A., Do, D.D., 2002. Adsorption in slit-like pores of
studies of PAN-based activated carbon fibers prepared by different activated carbons: improvement of the Horvath and Kawazoe
activation methods. J. Colloid. Interface Sci. 230 (2), 312–319. method. Langmuir 18, 4637–4647.
Seller-Perez, M.J., Martin-Martinez, J.M., 1991. J. Chem. Soc. Farady Zawadzki, J., 1989. Infrared spectroscopy in surface chemistry of
Trans. 87, 1237–1243. carbons. In: Thrower, P.A. (Ed.), . In: Chemistry and Physics of
Sing, K.S.W., Everett, D.H., Haul, R.A.W., Moscow, L., Pinerotti, Carbon, vol. 21. Marcel Dekker, New York, pp. 147–386.
R.A., Rouquerol, J., Siemieniewska, T., 1985. Pure Appl. Chem.
57, 603.