Microporous and Mesoporous Materials 52 (2002) 105–117

www.elsevier.com/locate/micromeso

Porosity development in activated carbons obtained from

date pits under chemical activation with phosphoric acid
Badie S. Girgis *, Abdel-Nasser A. El-Hendawy
National Research Centre, Laboratory of Surface Chemistry, 12622 Dokki, Cairo, Egypt
Received 5 February 2001; received in revised form 15 November 2001; accepted 27 November 2001

Abstract

Date pits, a low-cost agricultural by-product, was tested as a precursor for the production of porous carbons in a
chemical scheme using phosphoric acid. The raw material was impregnated with increasing concentrations of H3 PO4
(30–70 vol.%) followed by pyrolysis at 300, 500 or 700 °C. Texture characteristics of the products were determined by
adsorption of N2 at 77 K, as well as iodine, phenol, and methylene blue values. Carbons obtained at 300 °C were very
poorly porous, although with anomalously high capacity for the uptake of probe molecules from solution. This was
attributed to the contribution of unsaturated chemical bonds initiated in the partially decomposed material. Carbons
obtained at 500 and 700 °C are good to excellent adsorbents and attain best developed porosity at 700 °C, contrary to
the earlier well-established temperature of 500 °C recommended for treatment of agricultural precursors. Thermo-
gravimetric/differential thermogravimetric tracings of the H3 PO4 -impregnated date pits indicated a delayed-decom-
position effect shifted to higher temperature as compared to the raw material. Phosphoric acid is suggested to inflict
physical and chemical modifications on the botanical structure by penetration, particle swelling, partial dissolution of
the biomass, bond cleavage and reformation of new polymeric structures resistant to thermal decomposition. In ad-
dition, raw date pits are proposed to be composed of a low-porosity and compact cellular structure that needs higher
acid concentrations and/or temperatures to attain the optimum effect normally reached at lower temperature in case of
other feedstocks of botanical origin. Ó 2002 Elsevier Science B.V. All rights reserved.

Keywords: Porosity; Activated carbon; Date pits

1. Introduction removing both organic and inorganic trace con-

taminants. Its use can be broken down into dis-
Activated carbon (AC) is perhaps one of the crete areas such as effluent treatment, potable
most widely used adsorbents. Over the last few water treatment, solvent recovery, air treatment,
decades, adsorption systems involving AC has decolorizing, metal ore processing, and many more
gained importance in purification and separation general domestic applications. This explains the
processes on an industrial scale. It is now consid- increasing market trend for the use of ACs [1].
ered as one of the best available technologies in They owe their distinguished properties to the ex-
tensive surface area and internal porosity, as well
as their developed surface structure. The inherent
*
Corresponding author. nature of the precursor, or starting material, as

1387-1811/02/$ - see front matter Ó 2002 Elsevier Science B.V. All rights reserved.
PII: S 1 3 8 7 - 1 8 1 1 ( 0 1 ) 0 0 4 8 1 - 4
106 B.S. Girgis, A.-N.A. El-Hendawy / Microporous and Mesoporous Materials 52 (2002) 105–117

well as the methods and conditions employed for heat treatment step and is achieved at lower
carbon synthesis, determine the final pore size temperatures (400–500 °C), higher yields are ob-
distribution and the adsorption properties of the tained and most of the phosphoric acid can be
AC [2,3]. recovered after the process is completed. Thus, it
ACs can be produced from virtually any car- is a material-, energy- and time-saving scheme for
bonaceous precursor, both naturally occurring the production of ACs. Its use as an activating
and synthetic. Process economics normally dictate agent is now more favored over ZnCl2 , not only
the selection of readily available, low-cost feed- because of the simplicity of the extraction process
stocks. Common examples of commercial carbon after carbonization and impregnant recovery (only
feedstocks are materials of botanical origin (e.g. water is required), but also on account of the de-
wood, coconut shells, and fruit stones) and de- cline of the ZnCl2 process due to problems of
graded or coalified plant matter (e.g. peat, lignite environmental contamination with zinc com-
and all ranks of coal). A list of the most common pounds [23].
raw materials, as given by Bansal et al. [4], reads as In earlier investigations, we dealt with activa-
follows: wood, 35%, coal, 28%, lignite, 14%, co- tion using H3 PO4 for the production of good ad-
conut shells, 10%, peat, 10% and others, 3%. Pre- sorbing carbons derived from various agricultural
cursors from fresh biomass origin offer the most by-products (sugar cane bagases [24], apricot stone
economical source because they are copious re- shells [25], rice hulls [26], corn cobs [27,28], cotton
newable supplies, with low mineral content and stalks [29], olive mill wastes [30], and peanut hulls
appreciable hardness, and of low cost. The lit- [31]). Date pits represent an equally abundant by-
erature pertaining to the use of agricultural product generated from food processing and jam
by-products (lignocellulosics) is very extensive, production industries. In Egypt, date palms exceed
however, coconut shells are still receiving renewed eight million trees which generate a considerable
attention [5,6], next we meet: almond shells, peach amount of low-cost, non-valuable waste material.
stones, apricot stones, plum stones, cherry stones Our previous experience with the above-mentioned
and nutshells (such as hazelnuts, walnuts, cashew plant wastes urged the present investigation, as
and macadamia nuts) [7–17]. In comparison, date date pits are hard, high-density low-ash fruit
pits received much less consideration as a source kernels. These qualities recommend them for the
material for the preparation of AC [18–21]. production of granular ACs suitable for fixed-bed
The methods of activation commonly employed purification systems, as well as in regeneration
can broadly be divided into two main types: schemes. In earlier studies date pits were essen-
thermal (or physical) activation and chemical tially subjected to thermal activation in two-step
activation. Thermal activation involves primary [18–21] or in one-step steam pyrolysis [32] proce-
carbonization (below 700 °C) followed by con- dures, which urged their treatment in the well-
trolled gasification under the action of oxidizing established chemical activation with H3 PO4 .
gases at high temperature (up to 1100 °C) [4]. In The present study describes the preparation
chemical activation the precursor is mixed with and porosity evaluation of ACs from date pits at
a chemical restricting the formation of tars (e.g. increasing concentrations of H3 PO4 and/or heat
ZnCl2 , H3 PO4 , etc.), after kneading carbonized treatment temperatures (HTTs). It aims at dem-
and washed to produce the final AC. The chemical onstrating the potential of utilizing this precursor,
incorporated into the interior of the precursor and the development of its porosity as a function
particles reacts with the thermal decomposition of varying conditions of treatment, by analysis
products reducing the evolution of volatiles and of the N2 adsorption isotherms at 77 K applying
inhibits the shrinkage of the particles. In this way, various procedures. Additional information is ob-
the conversion of the precursor to carbon is high, tained from adsorption experiments of various
and once the chemical is eliminated after the heat probe molecules from aqueous solution. Descrip-
treatment, a large internal porosity is formed [22]. tion of the impact of the activation process on
Phosphoric acid activation only involves a single porosity development is displayed, based on cur-
 B.S. Girgis, A.-N.A. El-Hendawy / Microporous and Mesoporous Materials 52 (2002) 105–117 107

rent literature, and in light of the texture para- 2.2. Characterization of the porosity
meters obtained. This contribution is intended to
complement previous studies in a program dedi- This was carried out by determining the ad-
cated to the development of ACs from locally sorption isotherms of N2 at 77 K by the help of a
available low-cost agricultural by-products. Such vacuum apparatus of the conventional type. Some
a process would serve a double purpose. First, it adsorption data were obtained by an automatic
converts undesirable, surplus agricultural waste, of recording apparatus of the type ASAP 2010 (a
which millions of tons are produced annually, to product of Micromeritics, Inc.) which showed es-
useful value-added adsorbents. Second, the pro- sentially similar results. The tested carbons were
duced ACs find increasing applications in water out-gassed under vacuum at 200 °C for 2 h prior to
treatment, and other fields. This would contribute starting the measurements. Adsorption data were
to solving many current-day environmental pol- analyzed by applying the BET equation, the as -
lution problems. method [33] and the DR-plot [34]. Thus, the ad-
sorption data (Va , cm3 g1 ) were plotted against
the as -values reported by Selles-Perez and Martin-
2. Materials and methods Martinez [35] for the N2 adsorption on a
non-porous carbon material. The linear section
2.1. Activated carbons connecting to the origin corresponds to the specific
surface area (Sta ), and the linear relation at and
Washed-clean whole date pits were dried in an beyond as P 1:0 describes the non-microporous
air oven at 80 °C for 48 h, which proved effective surface area (Sna ) whereas its intercept with the Va -
to facilitate subsequent crushing and grinding. axis gives the micropore volume (V0a ). Plots for the
Sixty grams of the crushed pits (0.5–2.0 mm in DR equation are in the form ln na (na , mmol g1 )
size) were mixed with 100 cm3 of prediluted H3 PO4 vs. ½A=b2 , where A ¼ RT ln P 0 =P and b ¼ 0:35,
(85 wt.%) to varying concentrations in the range such that the intercept of the rectilinear relation-
30–70 vol.%. The slurry was left overnight at room ship measures the micropore volume (V0DR mmol
2
temperature, then transferred to a stainless steel g1 ) and the slope equals ð1=E0 Þ [36].
reactor (diameter of 40 mm and length of 60 cm) In addition, the mesopore volume (Vmeso ) was
with narrow ports of 10 mm diameter at both obtained from the volume of nitrogen, as liquid,
ends. Exhaust gases and liquid condensates are adsorbed for each sample between P =P 0 ¼ 0:1
vented through these openings. The reactor is ad- and 0.95, as suggested earlier [37].
mitted into an electric tube furnace and heated at
a slow rate (
5 K min1 ), to allow free evolution 2.3. Adsorption from solution
of volatiles, up to the hold temperatures 300, 500,
or 700 °C. Soaking at each final temperature was The iodine and phenol numbers were deter-
conducted for 2 h. The cooled mass was subjected mined by using the single bottle point uptake from
to thorough washing with hot water until the 0.02 M iodine solution, or 100 mg l1 solution
washings were neutral in effect, then dried at 110 of phenol, respectively. Adsorption isotherms for
°C for 24 h, and finally kept in tightly closed methylene blue (MB), from aqueous solution, were
bottles. Carbon yield ranged from 38% to 43%. determined by contacting fixed amounts of the
Fifteen samples were thus obtained and given predried powdered carbon sample (200 mg) with
notations including two numerals; the first denotes 100 cm3 of the dye solution of varying initial
the HTT (3 ¼ 300°, 5 ¼ 500° and 7 ¼ 700 °C) and concentrations. The stoppered bottles were agi-
the second number indicates the acid concentra- tated for 24 h, at room temperature, and the fil-
tion (3–7 denotes 30–70 vol.%). A carbon sample tered solution was measured for the residual color
55, thus, represents the product obtained by heat at k ¼ 664 nm by the help of a UV/VIS recording
treating the precursor at 500 °C and preimpreg- spectrophotometer produced by Shimadzu (2401
nated with 50 vol.% H3 PO4 . PC).
108 B.S. Girgis, A.-N.A. El-Hendawy / Microporous and Mesoporous Materials 52 (2002) 105–117

2.4. Thermogravimetric analysis of date pits

In order to detect the effect of phosphoric acid

on the raw material thermogravimetric (TG) and
differential thermogravimetric (DTG) recordings
were determined. The raw finely powdered date
pits and a sample impregnated with 50 vol.%
H3 PO4 (dried at 105 °C) were analyzed. An ap-
paratus of the type Perkin–Elmer 7 Series Thermal
Analysis system was used at a heating rate of 10
K min1 up to 900 °C.

3. Results

3.1. Modifications in the course of pyrolysis under

the action of phosphoric acid

During the pyrolysis of lignocellulosic materi-

als, three stages can be distinguished [22]. These
are: (a) loss of water in the 300–573 K range, (b)
primary pyrolysis in the 573–773 K range with
Fig. 1. TG and DTG for date pits (DP) and H3 PO4 -impreg-
evolution of most gases and tars with formation of nated date pits (DPP).
the basic structure of the char, and (c) consolida-
tion of the char structure at 773–1123 K with a
small weight loss. Despite the influence of the in- date pits (see Fig. 1 and Table 1): (i) it shifts de-
trinsic chemical composition (relative content of gradation to considerably higher temperatures, (ii)
the constituents: cellulose, hemicellulose and lig- it enhances the early dehydration stage combined
nin), the extrinsic factor of the chemical impreg- with an evolution of low-boiling volatiles (cf. 12%
nant would be equally effective in directing the and 5%, respectively), (iii) it strongly retards the
course of pyrolysis. evolution of volatiles, where up to 600 °C only
According to Fig. 1, raw date pits (DP) exhibit around 50% weight is lost, whereas in the absence
only one prominent wave of weight loss (between of H3 PO4 complete degradation is finished before
200 and 400 °C) with a maximum centered at 270
°C, followed by a slow weight loss with decreas-
ing rate. A considerably different pattern is exhib- Table 1
ited by the H3 PO4 -treated lignocellulosic material Weight loss of date pits (DP) and H3 PO4 -impregnated date pits
(DPP) within different ranges of temperatures
(DPP). A regular small weight loss covers the early
stage between ambient and 400 °C, with a broad Temperature range D wt.%
weak peak around 250 °C, followed by another DP DPP
slow higher weight loss with increasing tempera- r.t.–200 4.9 11.9
ture from 400 to 700 °C. Beyond this temperature 200–300 41.5 10.9
the decomposing mass is still showing loss in 300–400 17.0 5.6
400–500 7.1 6.0
weight at a slower rate up to 900 °C. Through this 500–600 4.5 14.2
temperature range of 400–900 °C two DTG peaks 600–700 2.7 29.1
appear with maxima at 525 and 700 °C. 700–800 1.8 14.8
Thus, impregnation with H3 PO4 brings about r.t.–800 79.5 83.5
considerable effects on the course of pyrolysis of r:t: ¼ Room temperature.
 B.S. Girgis, A.-N.A. El-Hendawy / Microporous and Mesoporous Materials 52 (2002) 105–117 109

450 °C, (iv) it effectively suppresses the principal

degradation stage (200–400 °C) as it is accompa-
nied by a weight loss of only 17% in presence and
58% in the absence of H3 PO4 , and (v) it promotes
the carbon yield within the range of 400–700 °C.
These observations could be reasonably associated
with the suggested mechanism, involving dissolu-
tion of some chemical components with bond
cleavage, followed by recombination at other sites
forming new polymeric structures that are ther-
mally more resistant [38].

3.2. Variations in porosity characteristics as a

function of the conditions of preparation

3.2.1. General features of the N2 adsorption iso-

therms
Figs. 2–4 give illustrative examples for the
shape and behavior of the N2 adsorption iso-
therms, and of the as - and DR-plots, for the series
of carbons obtained at 300–700 °C at varying
H3 PO4 concentrations. The isotherms (Fig. 2) ap-

Fig. 3. as -Plots of N2 /77 K isotherms for ACs obtained at

300–700 °C.

proach the Type I (Langmuir) with small upward

bending at higher pressure, indicating an essen-
tially microporous character with some contribu-
tion of wider pores (meso- and macropores), as
will be demonstrated by the evaluated parameters.
The as -plots (Fig. 3) show two linear sections: one
connecting the early points up to as  0:8 and
another one connecting the high-pressure data
with a considerable slope corresponding to the
Fig. 2. Representative adsorption isotherms of N2 at 77 K on non-microporous surface area (type a-1b, [39]). In
ACs obtained at 300 ° C. many cases it was not possible to join the origin
110 B.S. Girgis, A.-N.A. El-Hendawy / Microporous and Mesoporous Materials 52 (2002) 105–117

On the other hand, the DR-plots (Fig. 4) indi-

cate a type C behavior, with a small upward
2
bending at high pressures (low ð A=bÞ values)
which results from adsorption in pores wider than
those associated with micropore filling and are
therefore a consequence of a bimodal distribution
of A with V [36]. However, the estimated micro-
pore volumes (V0DR ) are mostly higher than the V0a
values (Tables 2–4), which is generally attributed
to an overlap between adsorption in narrow and
wide micropores [37]. This also leads to higher es-
timates of S KDR as compared to Sta values, whereas
the latter values are mostly concordant with the
SBET surface areas (Tables 2–4).
It is of interest to compare the P =P 0 values
that realize the V0a and V0DR estimates which are

Table 2
Texture characteristics of ACs obtained at 300 °C in the pres-
ence of varying H3 PO4 concentrations
Parameter 30% 40% 50% 60% 70%
C33 C34 C35 C36 C37
SBET (m2 g1 ) 28 33 68 71 86
Vp (cm3 g1 ) 0.034 0.017 0.056 0.082 0.040
Vmeso (cm3 g1 ) 0.024 0.005 0.013 0.057 0.009
Sta (m2 g1 ) 28 28 72 68 79
Sna (m2 g1 ) 12 5 6 – 5
Smic 16 23 66 68 –
V0a (cm3 g1 ) 0.005 0.011 0.028 – 0.033
V0DR (cm3 g1 ) 0.017 0.015 0.031 0.030 0.038
E0 (kJ mol1 ) 7.0 8.5 11.2 8.5 9.5
S KDR (m2 g1 ) 49 42 86 83 106
Xp 0.72 0.96 1.20 1.44 1.68

Table 3
Texture characteristics of ACs obtained at 500 °C in the pres-
ence of varying H3 PO4 concentrations
Parameter 30% 40% 50% 60% 70%
C53 C54 C55 C56 C57
2 1
SBET (m g ) 546 498 556 647 635
Fig. 4. DR-linear plots for the N2 /77 K isotherms for ACs. Vp (cm3 g1 ) 0.304 0.321 0.296 0.437 0.459
Vmeso (cm3 g1 ) 0.030 0.188 0.086 0.222 0.221
Sta (m2 g1 ) 549 484 547 629 627
Sna (m2 g1 ) 80 174 83 238 197
with the points located at the lower as -values, and Smic 469 370 464 391 430
a negative deviation appears. This kind of devia- V0a (cm3 g1 ) 0.199 0.131 0.227 0.187 0.212
tion was described as due to the presence of a V0DR (cm3 g1 ) 0.243 0.229 0.248 0.288 0.266
primary and secondary micropore filling mecha- E0 (kJ mol1 ) 11.4 9.1 10.8 7.6 13.1
S KDR (m2 g1 ) 679 642 695 808 747
nism [39].
 B.S. Girgis, A.-N.A. El-Hendawy / Microporous and Mesoporous Materials 52 (2002) 105–117 111

Table 4 chars and ACs has been observed in many cases as

Texture characteristics of ACs obtained at 700 °C in the pres- mentioned by Carrasco-Marin et al. [42]. In such
ence of varying H3 PO4 concentrations
cases adsorption of CO2 at temperatures around
Parameter 30% 40% 50% 60% 70%
ambient showed different estimates for the micro-
C73 C74 C75 C76 C77 porosity than the N2 /77 K results. It should be
SBET (m2 g1 ) 740 733 945 520 522 noted that this trend applied to the data from the
Vp (cm3 g1 ) 0.372 0.482 0.545 0.415 0.329 conventional vacuum system as well as the auto-
Vmeso (cm3 g1 ) 0.075 0.129 0.187 0.181 0.125
mated ASAP unit. An equilibrium duration of
Sta (m2 g1 ) 810 695 934 560 554
Sna (m2 g1 ) 14 118 155 242 112 10–15 min in the early isotherm region seems to
Smic 796 477 779 318 442 be insufficient to satisfy the activated diffusion of
V0a (cm3 g1 ) 0.317 0.258 0.346 0.218 0.191 nitrogen into the micropores.
V0DR (cm3 g1 ) 0.331 0.336 0.424 0.239 0.229 Mixed micro-mesopore adsorbents are thus
E0 (kJ mol1 ) 13.8 8.9 10.6 11.5 13.1
apparent, although the contribution of mesopo-
S KDR (m2 g1 ) 928 942 1188 671 643
rosity is lower in comparison to the microporosity
(cf. Sna and Sta , Tables 2–4). This appears more
evident by comparing V0 , Vmeso and Vp , V0DR , S KDR ,
observed as 0.03–0.14 and 0.18–0.80, respectively, as well as the low E0 values (denoting wide
where the latter values are definitely too high and micropores) as well as the average pore widths
unreasonable. Filling of micropores, as postulated (13–25 A ).
by the theory of volume filling of micropores
cannot be believed to continue up to such very 3.2.2. Influence of heat treatment temperature on
high-pressure ranges. This phenomenon (V0DR > the development of porosity
V0a ) has been observed earlier with several ACs It has been established by several investigators
obtained by H3 PO4 activation from various pre- that, in case of H3 PO4 treatment of agricultural
cursors [28,29,40,41]. Such a behavior seems materials, temperatures around 450 °C are most
thus particular to the H3 PO4 -activated carbons, suitable to obtain optimum properties of the ACs
whereas in case of steam-activated carbons small [7,24–30,43–47]. Even a temperature of 300 °C was
differences appear in both surface area and pore found sufficient, in some cases, to produce active
volume estimations [32]. This may be associated carbons with developed porosity (e.g. apricot
with the highly developed porosity with a wide stones with SBET of
1200 m2 g1 ). For the present
range of pore sizes in which pore filling overlaps to precursor, date pits, a temperature of 300 °C has
result in such a behavior. Another possible view no influence on the poorly adsorbing char at this
may be proposed which explains the effect and the temperature (Table 1). Increased acid concentra-
negative intersection in the early region of the as - tion, or the acid: precursor ratio (Xp ), slightly
plots (see above). It seems that, at very low pres- raises the porosity from 38 to 86 m2 g1 , which is
sures, diffusion of the adsorbed N2 molecules at 77 due to the aggressive action on the cellular struc-
K is very slow, due to the presence of constrictions ture.
at the entry of the micropores and/or to deposi- Raising the HTT to 500 °C (the previously es-
tions of the unleached residual phosphates. The tablished optimum range) has a good influence on
low volume of adsorbed nitrogen will deform the the raw material with a generation of moderately
initial linear section of the DR relationship by developed porosity (SBET ¼ 500–630 m2 g1 , Vp ¼
shifting it downwards and consequently leads to 0:30–0.46 ml g1 ). This effect is enhanced with the
high slopes measuring low E0 values (cf. 7.0–13.8 concentration of H3 PO4 which indicates the suit-
kJ mol2 ) as well as to the observed high intercepts ability of this temperature for the evolution of
corresponding to the ln V0DR values (Tables 2–4). porosity. It is astonishing that, at a still higher
Probably, much longer times are needed to attain temperature of 700 °C, development of better ad-
equilibrium conditions at the early adsorption sorbing carbons is attained with a SBET reaching
pressures. Restricted diffusion of N2 at 77 K in 950 m2 g1 with a corresponding total pore volume
112 B.S. Girgis, A.-N.A. El-Hendawy / Microporous and Mesoporous Materials 52 (2002) 105–117

of 0.55 ml g1 . With many earlier studies this efficient at 500 °C. However, the intensive action in
temperature was mostly associated with an ap- generating porosity at 500 °C is not correspond-
parent reduction in porosity which was ascribed to ingly related with the increase of the amount of
the breakdown of the microporous structure re- impregnant. An optimum value of 1 Xp ¼ 1:44
sulting in an evolution of wider porosity charac- seems to be sufficient for a maximum effect, which
terized by lower internal surface area. The present is somewhat higher than that observed before with
precursor, however, exhibits a generation of po- other precursors [24,25,27,30]. Additional acid is
rosity essentially in the micropore range at 700 °C. not benificial, as it does not lead to further action
It may be suggested that the abnormal response and probably forms an insulating layer (or skin),
of date pits towards activation with H3 PO4 can be covering the particles, thus reducing the activation
attributed to its special compact botanical struc- process and the contact with the surrounding at-
ture. The relatively loose cellular structure of mosphere [45].
several lignocellulosic materials, exhibiting low For carbons obtained at 700 °C, an increased
particle density and high inherent porosity, impregnation ratio seems to generate more poro-
facilitates better contact and penetration of the sity. An optimum effect of acid occurs at Xp ¼
impregnant. Such action leads to considerable 1:20 which yields the best adsorbing carbon with
physical and chemical structural modifications SBET ¼ 945 m2 g1 and Vp ¼ 0:545 cm3 g1 . Intro-
which is accordingly reflected by the properties of duction of more acid into the particles degrades
the carbons developed at relatively lower temper- the porosity, which might be explained as above:
atures. This specific nature of date pits would the surplus dehydrated acid forms an insulating
explain their inertness at 300 °C, as well as the layer that limits the action of the activant
extended activating action of H3 PO4 up to 700 °C, considerably and reduces the formation of area-
rather than up to 500 °C . It is worth mentioning responsible micropores as observed in the con-
that simple carbonization of date pits at 500 and siderable decay in Smic as well as V0a values
700 °C produced chars of negligible porosity (Table 4).
(SBET < 20 m2 g1 , Vp < 0:01 cm3 g1 ) [32] which
demonstrates the important effect of H3 PO4 in 3.3. Adsorption capacity from aqueous solutions
porosity generation.
3.3.1. Iodine and phenol numbers
3.2.3. Influence of the H3 PO4 concentration on It has been established that the iodine number
porosity development (in mg g1 ) gives an estimate of the surface area (in
To demonstrate the effect of the amount of acid m2 g1 ) [8], and it measures the porosity for pores
introduced into the particles after impregnant de- with dimensions P 1.0 nm. Here, a difference be-
hydration, the so-called ‘‘degree of impregnation’’ tween the iodine number and SBET is observed for
is calculated. It expresses the weight of acid re- ACs obtained from date pits at and beyond 500 °C
tained per gram of raw precursor and is denoted (Table 5). A lower adsorption of iodine can be
by Xp [23]. As this ratio is increased it will intensify ascribed to the presence of pores narrower than 1.0
the action of the acid in two ways: (a) by in- nm, which largely makes up most of the structure
creasing the aggressive physico-chemical effect on of these carbons. For carbons pyrolyzed at 500 °C
the raw lignocellulosic material and (b) upon the accessible ratio of surface is generally high
dehydration, inside and between the original bo- (0.85 up to 1.00), whereas carbons produced at
tanical cellular structure, the acid creates much 700 °C exhibit a much lower capacity indicating
porosity that is prevented from contraction or considerable ultra-microporosity (ratio of SI =
collapse. Subsequent leaching by exhaustive SNa ¼ 0:47–0.67). These carbons contain mostly
washing will evacuate and free this internal po- micropores with a small contribution of mesopo-
rosity and leave behind an extensive porosity. rosity. This is confirmed by the various texture
The impregnation ratios applied (Xp ¼ 0:72– parameters (Table 4). In particular, sample 76
1.68) are hardly effective at 300 °C, but more clearly shows a high SI =SNa ratio of 0.85 and contains
 B.S. Girgis, A.-N.A. El-Hendawy / Microporous and Mesoporous Materials 52 (2002) 105–117 113

Table 5
Iodine and phenol numbers for the tested ACs
Sample notation pH Iodine no. SI =SNa Phenol no. SPh SPh =SNa
(mg/g) (mg g1 ) (m2 g1 )
33 6.23 183 (5.9) 59 165 (5.32)
34 6.1 193 (8.0) 75 210 (8.75)
35 5.9 203 (2.9) 64 179 (2.56)
36 5.63 213 (3.4) 44 123 (1.98)
37 5.23 228 (2.7) 64 179 (2.1)
53 5.48 508 0.925 132 370 0.673
54 5.48 518 1.00 146 409 0.807
55 5.23 518 0.921 132 370 0.657
56 5.01 523 0.977 127 355 0.663
57 4.02 528 0.853 118 330 0.553
73 4.71 366 0.472 120 336 0.434
74 4.68 477 0.675 110 308 0.435
75 4.4 508 0.537 166 465 0.492
76 4.18 457 0.856 151 423 0.792
77 4.01 345 0.616 162 453 0.809

considerable mesoporosity (Sn and Vmeso ) with low porosity characteristics which determine the de-
V0a . gree of accessibility of these molecules. The degree
An anomalous behavior is observed for the 300 of coverage for these carbons ranges from 43% to
°C products with very poor porosity which show a 81%. The determined values are comparable to
high capacity for iodine. Thus, they take up 183– those for many commercial carbons and appear
228 mg g1 which correspond to 2.7 up to 8.0 times even higher even than those of some products
the N2 -surface areas. These values might suggest [49–51].
a non-adsorptive process, and are probably as-
sociated with an addition reaction of iodine to 3.3.2. Adsorption of methylene blue
available double bonds in the partially pyro- MB is one of the most widely recognized probe
lyzed lignocellulosic material. This process would molecules for assessing the removal capacity of a
contribute largely to the consumption of high specific carbon for moderate-size pollutant mole-
amounts of iodine, and such a phenomenon was cules ( P 1.5 nm). Fig. 5 illustrates the adsorption
currently observed with other plant precursors isotherms of MB on the carbons under consider-
(e.g. peanut hulls [31]). ation, which are of the type L (Langmuir) tending
It is worthwhile mentioning that these specific to a plateau parallel to the Ce -axis. Both the
carbon products also exhibit a high capacity for Langmuir and Freundlich equations were appli-
the removal of phenol from aqueous solutions (44– cable and showed satisfactory straight lines
75 mg g1 ) (Table 5). Considering that the area through the whole concentration range. Table 6
covered by an adsorbed phenol molecule in a flat summarizes the adsorption parameters derived
orientation equals 0.52 nm2 [48], anomalously high from both adsorption models. Adopting the well-
surface areas are evaluated which correspond to established value of 1.20 nm2 covered by a flat
degrees of coverage similar to the iodine values adsorbed MB molecule [52], the estimated mono-
(2.0 up to 8.8). Such values cannot be accounted layer capacities (qm ) were converted into surface
for by simple physical adsorption by the poorly areas accessible to MB (SMB ).
porous adsorbents. It confirms, furthermore, the Again, the 300 °C carbons seem to adsorb such
above observation with regard to the chemical amounts of MB that appear to measure much
nature of the surface of these carbon adsorbents. higher surface areas. However, this trend may be
On the other hand, removal of phenol by the explained if MB is imagined to be adsorbed in
ACs obtained at 500 and 700 °C is related to their a compact state in micelles with an aggregate
114 B.S. Girgis, A.-N.A. El-Hendawy / Microporous and Mesoporous Materials 52 (2002) 105–117

For the ACs obtained at 500 or 700 °C, MB is

adsorbed in a manner determined by the texture
characteristics of each adsorbent. The prevailing
microporosity, thus, limits considerably the uptake
of MB by most carbons so as to attain values
lower than those recommended for adsorbing car-
bons of 200 mg g1 (cf. most of the present values
range between 84 and 160 mg g1 ). Only three
carbons conform to the required limit. These are
samples nos. 74, 75 and 76 obtained at 700 °C.
This confirms the above-described observations
concerning the high surface area and developed
porosity particularly within the wide micro- and
mesopore range. The effective factor is the im-
pregnant concentration which lies within 50%
and 60% H3 PO4 (corresponding to a degree of
impregnation equivalent to 1.20–1.44).

4. Discussion

The action of phosphoric acid in the activation

of the lignocellulosic material may be visualized
as to take place in three stages in the process of
preparation: impregnation (or soaking), pyrolysis
and final leaching of the impregnant. During the
course of soaking, the acid introduced into the
lignocellulose produces chemical changes and
structural alterations, involving dehydration and
redistribution of biopolymers possibly by partial
dissolution in the acid solution, together with the
cleavage of ether linkages between the lignin and
cellulose, followed by recombination reactions in
which larger structural units are formed, with the
net result of a rigid crosslinked solid. Since H3 PO4
intensely modifies the structure of the precursor,
both the concentration of the impregnant and the
conditions under which impregnation is carried
Fig. 5. Adsorption isotherms of MB for the investigated out, must play important roles in the process
carbons.
[22,23]. This proposed effect is confirmed by the
thermal modifications inflicted on the course of
pyrolysis under the action of the acid (Fig. 1 and
number n ¼ 2 as suggested earlier by Giles et al. Table 1). Thermal decomposition is considerably
[52]. This brings the area values to within the shifted to higher temperatures with slow weight
estimated N2 -surface areas, i.e. 36, 41, 47, 49 and loss up to higher temperatures, due to the more
82 m2 g1 (cf. Tables 2 and 4), which means a more stable macromolecular structures, indicating thus
or less complete coverage of the charred surface a delayed decomposition of the impregnated ma-
with a monolayer of aggregated dye molecules. terial in comparison to the raw precursor.
 B.S. Girgis, A.-N.A. El-Hendawy / Microporous and Mesoporous Materials 52 (2002) 105–117 115

Table 6
Adsorption characteristics of MB by the various ACs
Sample Langmuir equation Freundlich equation SMB (m2 g1 ) SMB =SNa
Xm (mg g1 ) KL KF n
33 30.3 0.033 2.5 2.3 (73) 2.61
34 34.6 0.028 4.0 2.8 (83) 2.96
35 39.3 0.025 10.0 4.3 (94) 1.30
36 41.4 0.016 12.6 5.4 (99) 1.57
37 57.0 0.025 12.5 4.0 (168) 2.23
53 83.8 0.024 19.9 3.8 200 0.36
54 125.0 0.026 22.4 3.4 299 0.62
55 160.4 0.018 39.4 5.0 383 0.70
56 142.0 0.047 44.7 4.6 340 0.54
57 133.0 0.025 39.8 5.4 317 0.50

73 145.8 0.068 63.1 7.2 348 0.43

74 197.0 0.050 56.0 5.0 471 0.68
75 240.0 0.177 100 4.0 574 0.61
76 211.0 0.190 100 4.5 505 0.90
77 141.6 0.028 59.8 4.6 339 0.61

In the pyrolysis stage, the introduced acid development. An acid concentration of 50–60
dehydrates gradually with continuation of the vol.% represents the critical value (corresponding
chemical changes involving crosslinking, water to Xp of 1.2–1.44) in case of date pits, which is not
elimination and polymerization. The dehydrated different from previous lignocellulosic materials.
impregnant here reduces tar and volatiles forma- Only a higher thermal treatment temperature of
tion, inhibits shrinkage and collapse of the parti- 700 °C, instead of 500 °C, is required to attain
cles, and develops an extensive pore structure, with high and comparable surface areas (around 1000
a consequential increase in the carbon yield. Po- m2 g1 ). Still a low total pore volume is generated
rosity development in this stage is ascribed to the at this temperature (viz. 0.545 cm3 g1 , as com-
phosphoric oxides formed at the carbonization pared to 0.60–1.10 cm3 g1 for other precursors).
temperatures, which act as local oxidizing agents This indicates the role of the precursor in de-
performing controlled gasification [45]. This pro- termining the porosity characteristics developed
cess actually requires the presence of very limited under action of the H3 PO4 activation scheme. AC
supply of air. It may be speculated that higher acid obtained from date pits at 700 °C is essentially
concentrations, or higher impregnation ratios (Xp ), microporous which results in a high surface area
would enhance porosity development. However, product but with a low contribution of micropores
in previous reports [24,25,30,31,53], it was dem- to the pore volume (cf. Tables 2–4).
onstrated that higher acid concentrations (or Xp Finally, the third step in the preparation of AC,
values) were not accompanied by a respective im- that is the washing and acid recovery, plays a
provement in porosity. Rather, a loss in surface very important role of porosity evolution. After
area and pore volume was noticed, which may carbonization, most of the activant is still in the
be attributed to the formation of a layer of particle and intense washing to eliminate it pro-
polyphosphate (a skin) over the developing pore duces the porosity. It was found by Molina-Sabio
structure protecting it from an excessive gasifica- et al. [7] that there is a good agreement between the
tion [46]. High amounts of dehydrated, precipi- volume of micropores and the volume occupied by
tated phosphorus oxides will consequently inhibit the acid phase existing at the carbonization tem-
further burn-off of carbon and prevent activant perature. The entrapped polyphosphates in the
attack and volatiles release, thus reducing porosity final product will thus depend on its availability to
116 B.S. Girgis, A.-N.A. El-Hendawy / Microporous and Mesoporous Materials 52 (2002) 105–117

leaching, and is a function of both impregnation References

ratio and HTT. Both Xp and HTT are the variables
of direct incidence since all carbons are post-trea- [1] B. El Duri, in: G. McKay (Ed.), Use of Adsorbents for the
Removal of Pollutants from Wastewaters, Chapter 1, p. 2
ted under the same ‘‘standard’’ washing procedure.
and S.J. Allen, Chapter 5, p. 61, CRC Press Inc., Boca
The residual entrapped dehydrated acid products Raton, 1996.
will appear in the form of a high ash content (5– [2] J.F. Byrne, H. Marsh, Introductory overview, in: J.W.
9%) as well as impart an acidic character on the Patrick (Ed.), Porosity in Carbons, John Wiley and Sons
carbon products (pH ¼ 6:2–4.0). Inc., London, 1995.
[3] A. Bota, K. Laszlo, L.G. Nagy, T. Capitzky, Langmiur 13
(1997) 6502.
[4] R.C. Bansal, J.B. Donnet, F. Stoeckli, Active Carbon,
5. Conclusions Marcel Dekker, New York and Basel, 1988, p. 3.
[5] A.E. Pandalfo, M. Amini-Amali, J.S. Killingly, Carbon 30
Date pits are potential raw materials for pro- (1994) 1015.
ducing ACs via chemical activation with H3 PO4 . [6] J.C. Gonzalez, M.T. Gonzalez, M. Molina-Sabio, F.
Rodriguez-Reinoso, A. Sepulveda-Escribano, Carbon 33
These exhibit, however, two abnormal qualities:
(1995) 1175.
first, no detectable effect of H3 PO4 at 300 °C and [7] M. Molina-Sabio, F. Rodriguez-Reinoso, F. Caturla, M.J.
second, an extended (delayed) effect of acid at 700 Selles, Carbon 34 (1996) 457.
°C. Low porosity carbons obtained at 300 °C show [8] K. Gergova, N. Petrov, V. Minkova, J. Chem. Tech.
good capacity to remove substrates from solution Biotechnol. 56 (1993) 77.
[9] F. Guzel, Sep. Sci. Technol. 10 (1993) 221.
(iodine, phenol, MB). Here, the removal seems to
[10] A. Bota, K. Laszlo, Per. Polytech. Ser. Chem. Eng. 41
be associated with an absorption process by the (1997) 19.
unsaturated free bonds and radicals of the par- [11] M.G. Lussier, J.C. Shull, D.J. Miller, Carbon 32 (1994)
tially decomposed components irrespective of the 149.
available porosity, and the amounts taken up are [12] K. Raveendran, A. Ganesh, K.C. Kilar, Fuel 74 (1995)
1812.
many-fold higher than the N2 -surface area. ACs
[13] R. Leboda, W. Grezerczyk, A. Lodyga, A. Dabrowski,
obtained at 500 and 700 °C are essentially micro- Adsorpt. Sci. Technol. 10 (1993) 221.
porous with some contribution of mesoporosity. [14] X. Dai, M.J. Antal Jr., Ind. Eng. Chem. Res. 38 (1999)
The optimum activant concentration, to get the 3386.
best adsorbing carbons, lies within 50 and 60 wt.% [15] M.S. Tam, M.J. Antal Jr., Ind. Eng. Chem. Res. 38 (1999)
268.
which corresponds to an acid to precursor weight
[16] C. Nguyen, D.D. Do, Carbon 33 (1995) 1717.
ratio of 1.20–1.44. Excess acid seems to form a [17] A. Ahmadpour, D.D. Do, Carbon 35 (1997) 1723.
protective solid layer that reduces the activation of [18] R.I. Razouk, Sh. Nashed, J. Chem. UAR 8 (1965) 181.
the raw material. Due to the mixed micro-/meso- [19] T.M. El-Akkad, Egypt. J. Chem. 21 (1978) 152,157.
porosity, the active carbons are not freely acces- [20] T.W. Ahmad, H.U. Tanzil, M. Mumtaz, Pak. J. Sci. Ind.
Res. 34 (1991) 121.
sible to the used probe molecules from aqueous
[21] A.A. Attia, Adsorpt. Sci. Technol. 15 (1997) 707.
solution. The textural characteristics are the gov- [22] F. Rodriguez-Reinoso, M. Molina-Sabio, Carbon 30
erning factor in determining the degree of uptake (1992) 1111.
of these molecules. ACs obtained from date pits [23] M. Molina-Sabio, F. Rodriguez-Reinoso, F. Caturla, M.
seem to be fairly suitable to take up contaminants Selles, Carbon 33 (1995) 1105.
[24] B.S. Girgis, L.B. Khalil, T.A.M. Tawfik, J. Chem. Tech.
from liquid effluents, and they are probably much
Biotechnol. 61 (1994) 87.
more efficient for smaller molecules from gas [25] C.A. Philip, B.S. Girgis, J. Chem. Tech. Biotechnol. 67
streams. Activation with H3 PO4 presents several (1996) 248.
advantages: it occurs in a single thermal treatment [26] L.B. Khalil, Adsorpt. Sci. Technol. 13 (1996) 317.
process, leads to a high carbon yield, the acid can [27] B.S. Girgis, Bull. NRC Egypt 22 (1997) 89.
[28] A.A. El-Hendawy, S.E. Samra, B.S. Girgis, Colloids Surf.
be economically recovered, it is effective at rela-
A 180 (2001) 209–221.
tively moderate temperatures, and it results in [29] B.S. Girgis, M.F. Ishak, Mater. Lett. 39 (1999) 107.
good adsorbents with mixed porosity that widen [30] L.B. Khalil, B.S. Girgis, T.A.M. Tawfik, Adsorpt. Sci.
the scope of its application. Technol. 18 (2000) 373.
 B.S. Girgis, A.-N.A. El-Hendawy / Microporous and Mesoporous Materials 52 (2002) 105–117 117

[31] B.S. Girgis, S.S. Yunis, A.M. Soliman, J. Mater. Lett., [43] M. Jagtoyen, F. Derbyshire, Carbon 36 (1998) 1085.
in press. [44] J. Laine, S. Yunes, Carbon 30 (1992) 601.
[32] M.A. Alaya, B.S. Girgis, W.E. Mourad, J. Porous Mater. [45] J. Laine, A. Calafat, M. Labady, Carbon 27 (1989)
7 (2000) 509. 191.
[33] K.S.W. Sing, Chem. Ind. (1968) 1520. [46] J. Laine, A. Calafat, Carbon 29 (1991) 949.
[34] M.M. Dubinin, J. Colloid Interf. Sci. 75 (1980) 39. [47] C.A. Toles, W.E. Marshall, M.M. Johns, Carbon 35 (1997)
[35] M.J. Selles-Perez, J.M. Martin-Martinez, J. Chem. Soc. 1407.
Faraday Trans. 87 (1991) 1237. [48] D.M. Nevskaia, A. Santianes, V. Munoz, A. Guerrero-
[36] R.H. Bradley, B. Rand, J. Colloid Interf. Sci. 169 (1995) Ruiz, Carbon 37 (1999) 1065.
168. [49] P.W. Collings, M. Streat, The 1996 ICHEME Event, 1996,
[37] V. Gomez-Serrano, C. Valenzuela-Calahorro, J. Pastor p. 696.
Pillegas, Biomass Bioenergy 4 (1993) 355. [50] A.A.M. Daifullah, B.S. Girgis, Wat. Res. 32 (1998)
[38] M. Jagtoyen, M. Thwaites, J. Stencel, B. McEnaney, F. 169.
Derbyshire, Carbon 30 (1992) 1089. [51] A.R. Khan, R. Atullah, A. Al-Haddad, J. Colloid Interf.
[39] M.J. Selles-Perez, J.M. Martin-Martinez, Fuel 70 (1991) Sci. 194 (1997) 154.
877. [52] C.H. Giles, A.P.D. D’Silva, A.S. Trivedi, Proceedings of
[40] Tarek A.M. Tawfik, M.Sc. Thesis, Cairo University, 1994. the International Symposium on Surface Area Determina-
[41] Tarek A.M. Tawfik, Ph.D. Thesis, Cairo University, 1999. tion, Bristol 1969, 1970, p. 317.
[42] F. Carrasco-Marin, M.A. Alvarez-Merino, C. Mareno- [53] J.B. Castro, P.R. Bonelli, E.G. Carella, A.L. Cukierman,
Castilla, Fuel 75 (8) (1996) 966. Ind. Eng. Chem. Res. 38 (2000) 4166.