Chemical Engineering Journal 171 (2011) 883–896

Contents lists available at ScienceDirect

Chemical Engineering Journal

journal homepage: www.elsevier.com/locate/cej

Dye adsorption on autohydrolyzed pine sawdust in batch and fixed-bed systems

D. Sidiras a,∗ , F. Batzias a , E. Schroeder b , R. Ranjan c , M. Tsapatsis c
a
Laboratory of Simulation of Industrial Processes, Department of Industrial Management and Technology, University of Piraeus, 80 Karaoli & Dimitriou, GR 18534 Piraeus, Greece
b
Institut fuer Kern- und Energietechnik, Forschungszentrum Karlsruhe GmbH, Postfach 3640, 76021 Karlsruhe, Germany
c
Department of Chemical Engineering and Materials Science, University of Minnesota, 421 Washington Avenue SE, Minneapolis, MN 55455, USA

a r t i c l e i n f o a b s t r a c t

Article history: Batch and column kinetics of three basic dyes, methylene blue (MB), bismarck brown (BB) and acri-
Received 13 January 2011 dine orange (AO), adsorption on autohydrolyzed (160–240 ◦ C, 0–50 min) Scots pine (Pinus sylvestris L.)
Received in revised form 13 April 2011 sawdust were studied, using untreated pine sawdust as control, in order to explore the potential use
Accepted 14 April 2011
of these low-cost industrial byproducts (untreated autohydrolyzed sawdust) for wastewater cleaning.
The effect of the autohydrolysis time and temperature on the microstructure and the crystallinity of
Keywords:
pine sawdust was investigated be means of: SEM, FTIR, XRD and BET, in order to obtain information
Adsorption
at a higher granularity level, in comparison with previous studies. The BET surface area increased from
Autohydrolysis
Dye
0.89 to 19.3 m2 g−1 . In the case of MB, the Freundlich’s adsorption capacity KF increased from 5.60 to
Sawdust 15.7 (mg g−1 )(L mg−1 )1/n , the amount of dye adsorbed when saturation is attained (Langmuir constant qm )
Column increased from 38.7 to 88.0 mg g−1 , and the Bohart-Adams adsorption capacity coefficient N increased
from 8046 to 14157 mg L−1 , indicating the extent that the severity of the autohydrolysis treatment
enhanced the adsorption behavior of the material.
© 2011 Elsevier B.V. All rights reserved.

1. Introduction for the removal of dyes in aqueous solutions. The environmental

Autohydrolysis of lignocellulosic materials like sawdust is
already used in bioethanol industry for water soluble fermentable
sugars production [1,2]. The main idea in this manuscript is the
use of the autohydrolysis solid residue as an adsorbent. In fact, the
adsorbent is the solid by-product in the fermentable sugars produc-
tion by autohydrolysis. Moreover, sawdust is a by-product of the
wood industry and pure water is used as autohydrolysis reagent
can be recycled after the fermentation and the water/bioethanol
distillation processes. Consequently, the present work might also
be considered within an Industrial Ecology framework.
Autohydrolysis of lignocellulosic materials can be used as a
first step while enzymatic or acid hydrolysis is usually used as
second step in the bioethanol industry. Autohydrolysis might sig-
nificantly increase the enzymatic hydrolysis efficiency [1]. Many
industrial byproducts (e.g. sawdust, and wood-chips) and agricul-
tural residues, e.g. corncobs, almond shells, olive stones, rice husks,
wheat straw, and barley straw, can be used as feedstock for the
production of xylo-oligosaccharides by autohydrolysis [2].
According to many review papers [3–8] low-cost adsorbents
offer a lot of promising benefits for commercial purposes in the
future. They could be used in place of commercial activated carbon
∗ Corresponding author. Tel.: +30 2104142360; fax: +30 2104142392.
E-mail address: sidiras@unipi.gr (D. Sidiras).
impact of the dyes present in the wastewater streams of many
industrial sectors, such as textile, paper, dyeing, tannery and the
paint industry, has drawn a lot of attention, emphasizing the neces-
sity for their removal. In particular, wood sawdust [9], a relatively
abundant and inexpensive material, has been extensively inves-
tigated as an adsorbent for removing contaminants from water.
Other adsorbent materials that have been studied include indus-
trial byproducts and agricultural residues, untreated or pretreated
in various ways.
Among the untreated materials that have been investigated,
many agricultural residues, such as wheat straw, rice husk, corn-
cobs and wood chips, have been used successfully to adsorb
individual dyes and dye mixtures in textile effluent [10]. Removal of
MB and other basic dyes has been carried out using beech sawdust
[9,11], wheat straw [12], cedar sawdust [13], rubberwood sawdust
[14], kudzu [15], banana and orange peels [16] and palm kernel
fiber [17].
Acid and alkali pretreated lignocellulosic materials (wheat
straw, corncobs, barley husks, wood sawdust) were successfully
used as adsorbents for a variety of dyes [18–20]. Prehydrolysed
(with dilute sulphuric acid aquatic solution at 100 ◦ C) wheat straw
[12] and beech sawdust [21] and chloride salts treated (at 100 ◦ C)
beech sawdust [9,11] has been proven to be effective for basic dyes
adsorption in batch and fixed-bed systems.
The autohydrolysis was chosen as a pretreatment (a) because
compared with the other treatments with aquatic solutions uses

1385-8947/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.cej.2011.04.029
884 D. Sidiras et al. / Chemical Engineering Journal 171 (2011) 883–896

Nomenclature
T␪a autohydrolysis temperature in ◦ C
AO acridine orange u linear velocity (cm min−1 ) in the case of column
BB bismarck brown adsorption process
BET Brunauer, Emmet and Teller x bed depth of the adsorption column (cm)
bKF empirical parameter of eq. (4) XRD X-ray diffraction
bqm empirical parameter of eq. (8)
C concentrations of MB or BB or AO in the bulk solution  XRD angle (◦ )
at time t in the case of batch adsorption process;  order of the adsorption kinetic model
also, the effluent concentration (mg L−1 ) in the case  spectrophotometer wavelength (nm)
of the column adsorption process
c intercept of the intra-particle diffusion equation
(mg g−1 )
C0 initial dye concentration in the case of batch adsorp- only pure water as a reagent, and (b) this treatment is already used
tion process (mg L−1 ) as first step in the fermentable sugars for bioethanol production
Ce equilibrium concentration of the adsorbate (mg L−1 ) industry [1,2].
for t→ ∞ The suitability of a range of materials as dye adsorbents is usu-
Ci influent concentration (mg L−1 ) in the case of col- ally established using the kinetics of MB (or other dye) adsorption
umn adsorption process during batch and continuous (column) processes [22–26].
E adsorption activation energy (kJ mol−1 ) In this work, the removal of MB, BB and AO by autohydrolyzed
FTIR Fourier transform infrared pine sawdust was studied using untreated pine sawdust as con-
K adsorption rate coefficient (L mg−1 min−1 ) in the trol. The effect of the autohydrolysis temperature and time on the
case of column adsorption process microstructure and the crystallinity of pine sawdust was deter-
k first order rate constant for the batch adsorption mined be means of: SEM, FTIR, XRD and BET, in order to obtain
process (min−1 ) information at a higher granularity level, in comparison with pre-
k2 second order rate constant for the batch adsorption vious studies. The batch and column adsorption kinetics of these
process (min−1 ) dyes were used to estimate and compare the adsorption capac-
KF Freundlich constant related to adsorption capacity ity of the untreated and treated pine sawdust. The contribution of
(mg g−1 )(L mg−1 )1/n this paper concerns the thorough investigation of the effect of pure
KF,0 empirical parameter of eq. (4) water treatment at 160–240 ◦ C on sawdust. The autohydrolysis
KL Langmuir constant related to the energy of adsorp- treated sawdust in this temperature range simulates the byproduct
tion (L mg−1 ) in the fermentable sugar production, part of the second generation
kp intra-particle diffusion rate constant bioethanol producing industry.
(mg g−1 min−0.5 )
k␬ -order rate constant for the batch adsorption pro- 2. Materials and methods
cess (min−1 ).
m weight of the adsorbent used (g) 2.1. Adsorbent processing
MB methylene blue
N adsorption capacity coefficient (mg L−1 ) in the case The Scots pine (Pinus sylvestris L.) sawdust used was obtained
of column adsorption process from a local furniture manufacturing company, as a suitable source
n inverse of the slope of the Freundlich isotherm, it is for full-scale/industrial applications. The moisture content of the
related to adsorption intensity material when received was 8.7% (w/w); after screening, the frac-
NLRA non-linear regression analysis tion with particle sizes between 0.2 and 1 mm was isolated. The
p frequency factor at the Arrhenius law (min−1 ) composition of the raw material was as follows (expressed in % w/w
q amount adsorbed per unit mass of the adsorbent for on a dry weight basis): 40.1% cellulose measured as glucan (with
t→ ∞ (mg g−1 ) 52.5% XRD degree of crystallinity); 28.5% hemicelluloses (16.0%
qm Langmuir constant related the amount of dye measured as manan, 8.9% measured as xylan and the rest 3.6% mea-
adsorbed (mg g−1 ) when the saturation is attained sured as arabinan); 27.7% Klason acid-insoluble lignin, 0.2% ash,
qm,0 empirical parameter of eq. (8) and 3.5% extractives and other acid soluble components (e.g. acid
qt amounts of dye adsorbed per unit mass of the adsor- soluble lignin).
bent (in mg g−1 ) at time t The autohydrolysis process was performed in a 3.75-L batch
R2 coefficient of determination reactor PARR 4843. The isothermal autohydrolysis time was tai = 0,
RL dimensionless constant called ‘equilibrium param- 10, 20, 30, 40 and 50 min (not including the non-isothermal pre-
eter’ or ‘separation factor’ expressing the essential heating and the cooling time-periods); the reaction was catalyzed
characteristics of the Langmuir isotherm by the organic acids produced by the pine sawdust itself during
SEE standard error of estimate autohydrolysis at a liquid-to-solid ratio of 10:1; the liquid phase
SEM scanning electron microscope volume (water) was 2000 mL and the solid material dose (pine
t adsorption time (in min) sawdust) was 200 g; stirring speed 150 rpm. The reaction ending
T adsorption temperature in K temperature values were T␪a = 160 ◦ C, 200 ◦ C and 240 ◦ C, reached
ta autohydrolysis time (in min), ta = tp + ta after tp = 42, 62 and 80 min preheating time values, respectively.
Ta autohydrolysis temperature in K. The autohydrolysis product was filtered using a Buchner filter
tai autohydrolysis isothermal time (in min) with Munktell paper sheet (grade 34/N) to separate the liquid
tp autohydrolysis preheating time (in min) phase and from the solid phase. The solid residue was washed
T␪ adsorption temperature in ◦ C with water until neutral pH (the initial filtrate pH was 2.90–4.76
depending on the autohydrolysis severity). The solid residue was
 D. Sidiras et al. / Chemical Engineering Journal 171 (2011) 883–896
Fig. 1. (a) Solid residue yield of the pretreated sawdust and (b) lignin percentage of
this residue vs. autohydrolysis time.
dried at 110 ◦ C for 10 days at room temperature to reach the
humidity of the untreated material. Then it was used as adsor-
bent.
2.2. Adsorption isotherm studies
Adsorption isotherms were derived from batch experiments.
Following the batch procedure, accurately weighed quantities of
adsorbent were transferred into 0.8-L bottles, where V = 0.5 L of
adsorbate solution were added. The sorbent weight varied from
m = 0.5 to 3 g, the temperature was T␪ = 23 ◦ C, the initial MB (MERCK,
C.I. 52015) concentration varied from C0 = 1.4 to 156 mg L−1 , the
initial BB (SIGMA-ALDRICH, C.I. 21000) concentration varied from
C0 = 2.2 to 11 mg L−1 , and the initial AO (SIGMA-ALDRICH, C.I.
46005) concentration varied from C0 = 1.8 to 9 mg L−1 . The bot-
tles were sealed and mechanically tumbled for a period of 7
days. This time period was chosen after experimental studies
(the time varied from 4 h to 14 days), to ensure that nearly
equilibrium conditions were achieved. The resulting solution con-
centrations were determined and the equilibrium data from each
bottle represented one point on the adsorption isotherm plots.
The initial pH of the solution was 7.8–8.0 and the final pH was
7.3–7.7.
2.3. Kinetic studies
Adsorption rate batch experiments were conducted in a 2-L
completely mixed glass reactor fitted with a twisted blade-type
stirrer, operating at 300 rpm for keeping the lignocellulosic mate-
rial in suspension. The reactor, containing V = 1 L aqueous dye
solution, was placed into a water bath to keep temperature constant
at the desired level. The sorbent weight varied from m = 1 to 6 g, the
temperature varied from T␪ = 10 to 55 ◦ C, the initial MB concentra-
tion varied from C0 = 1.4 to 156 mg L−1 , the initial BB concentration
varied from C0 = 2.2 to 11 mg L−1 , and the initial AO concentration
varied from C0 = 1.8 to 9 mg L−1 . The effect of stirring was studied
in the rage of 0–600 rpm. The pH effect was studied in the range of
885
1.5–13 (the initial pH of the dye solutions was adjusted using dilute
H2 SO4 or NaOH solutions, as appropriate).
2.4. Column studies
Fixed-bed up-flow adsorber studies were conducted in
15 cm × 2.5 cm and 25 cm × 2.5 cm stainless steel columns; the bed
height was x = 15 cm and 25 cm, respectively. The experimental
set-up consisted of one parallel column, fed by a multi-channel
peristaltic pump at constant flow rates Q = 10, 20, 30, 40 mL min−1 .
The initial constant MB concentration was at the ranges of
Ci = 14.0–14.3 mg L−1 and 156–192 mg L−1 . The initial constant BB
concentration was at the range of Ci = 11.0–11.1 mg L−1 . The initial
constant AO concentration was at the range of Ci = 8.9–9.0 mg L−1 .
Interconnective tubing and fittings were made of polytetrafluo-
roethylene (PTFE). Effluent samples were analyzed to yield output
concentration breakthrough curves.
2.5. Analytical techniques and instruments
The degree of crystallinity of pine sawdust cellulose was
measured by XRD [27] and by acid hydrolysis [21]. The resisting-
to-reaction hemicelluloses fraction (mainly xylan 1) was also
measured by acid hydrolysis [21]. The Saeman et al. [28] technique
was used for the quantitative saccharification of the untreated
lignocellulosic material and the autohydrolysis reaction solid
residues. The filtrates from the quantitative saccharification, were
analyzed for glucose, xylose and arabinose using high-performance
liquid chromatography (HPLC, Agilent 1200) with Aminex HPX-
87H Column, refractive index detector and 5 mM H2 SO4 in water
as the mobile phase. Cellulose was estimated as glucan and hemi-
celluloses were estimated as xylan and arabinan. Finally, the
acid-insoluble lignin (Klason lignin) was determined according to
the Tappi T222 om-88 method [29].
The BET (Brunauer, Emmet and Teller) surface area of the
untreated and the pretreated pine sawdust was measured from the
N2 adsorption isotherm with a Nova® Surface Area Analyzer (Quan-
tachrome Instruments) in accordance with DIN 66132 [30]. Prior to
this measurement, the samples were dried under vacuum at 150 ◦ C
overnight to clean the surface. The experimental data were ana-
lyzed using the “Quantachrome NovaWin2 - Data Acquisition and
Reduction for NOVA instruments” software.
The scanning electron microscope (SEM) used herein was a JEOL
JSM-6700F Field Emission Scanning Electron Microscopy. The SEM
tests were carried out on samples which were Pt coated (5 nm). The
magnifications were 750, 7500 and 50000, respectively.
Fourier transform infrared (FTIR) spectra were obtained
using a spectroscope (MAGNA-IR 750 Spectrometer, Serrie II,
Nicolet). The sampling technique used herein was diffuse
reflectance. The powder samples were scanned for wavenumber
650–4000 cm−1 .
The powder XRD patterns of the origin and the pretreated sam-
ples were measured on by a SIEMENS D5005 X-ray Diffractometer
using Ni-filtered CuKa ( = 0.154 nm) radiation at 45 kV and 40 mA
and continuous scan mode. The XRD patterns were recorded in the
scan range 2 = 5–70, at scan rate step = 0.04◦ , dwell time = 3 s, i.e.
total scan time approximately 1 h and 30 min.
The concentration of MB, BB and AO in the solution was
obtained by measuring O.D. at  = 663, 460, and 490 nm, respec-
tively, using a HACH DR4000U UV–vis spectrophotometer. These
optical wavelengths, determined by performing the ‘scan ’ option
of the measuring instrument, were satisfactorily compared with
the reported ones in literature as well as with the corresponding
values quoted in the dye manufacturer’s manual.
886 D. Sidiras et al. / Chemical Engineering Journal 171 (2011) 883–896
Fig. 2. SEM micrographs for untreated (a–c) and autohydrolyzed (240 ◦ C, 40 min) sawdust (d–f).
3. Results and discussion
3.1. Autohydrolysis
3.1.1. Autohydrolysis solid residue
Cellulose crystallinity, accessible surface area, protection by
lignin, and sheathing by hemicelluloses, all contribute to lignocel-
lulosic biomass resistance to autohydrolysis [1,2]. Autohydrolysis
of cellulose and hemicelluloses produces glucose, manose, xylose
and degradation products as 5-HMF and furfural, respectively.
The autohydrolysis is affected by temperature and reaction
time. Cellulose fractions are hydrolyzed to water-soluble cellulo-
oligosaccharides and glucose. Hemicelluloses are hydrolyzed to
mano-oligosaccharides, xylo-oligosaccharides, manose and xylose.
The acid-insoluble lignin fraction is not significantly affected by
autohydrolysis.
The yield of the pretreated pine sawdust (dry weight of product
% w/w of the untreated dry material) is presented in Fig. 1, plotted
vs. autohydrolysis reaction time (ta in min, including the preheating
period).
The hemicelluloses content of the untreated pine sawdust was
significantly decreased by autohydrolysis. The cellulose fraction
decreased by its amorphous part (and partially by its crystalline
part) while the lignin quantity remained practically unchanged;
the lignin percentage on the solid residue increases as is shown in
Fig. 1(b).
3.1.2. Microstructure
The SEM micrographs for untreated pine sawdust and for auto-
hydrolyzed at 240 ◦ C for 40 min pine sawdust are given in Fig. 2
for three different magnifications. In Fig. 2(a) and (d) with magni-
fication 750 we observe that the pine sawdust particle has fibrous
structure. This structure is more open for the particle presented in
Fig. 2(d), i.e. the autohydrolyzed particle. Comparing Fig. 2(b) and
(c) with magnification 7500 we observe that the micro fibril struc-
ture of the autohydrolyzed particle is significantly different that
 D. Sidiras et al. / Chemical Engineering Journal 171 (2011) 883–896
Fig. 3. SEM micrographs for autohydrolyzed (200 ◦ C, 40 min) sawdust before (a–c) and after (d–f) MB adsorption (C0 = 156 mg L−1 , t = 190 min, m/V = 1 g L−1 , 300 rpm, 23 ◦ C).
the untreated one. In Fig. 2(b) we can see the micro fibrils, less than
1 ␮m, but in (d) we can see just a rough texture. For magnification
50,000 we clearly observe that the texture of the pretreated saw-
dust particle in Fig. 2(f) is rougher, like to be constructed by smaller
particles less than 1 nm, comparing to the still smooth surface of the
untreated one in (f). The picture of the texture in Fig. 2(c) is similar
with that of Klason lignin. This observation seems reasonable, tak-
ing into account that the pretreated material is 60.5% (w/w) lignin,
as is presented in Fig. 1(b). These changes can be explained by the
removal of the hemicelluloses that gave the look to the smooth
particle surface, and the appearance of the naked cellulose-lignin
complex which has a more similar to lignin look. The changed sur-
face of the pretreated sawdust explains the enhanced adsorption
properties and the higher BET surface area values. Moreover, these
particles seem to have cavities but they have not pores. Sawdust is
not a porous material.
In Fig. 3 we can see the SEM micrographs for autohydrolyzed at
200 ◦ C for 40 min pine sawdust before and after MB adsorption. The
887
sawdust particle in Fig. 3(a) has some pores. The surface texture of
this particle is less rough than that of the autohydrolyzed particle
in Fig. 2(f). That fact can be attributed to the less severe autohydrol-
ysis conditions. The adsorption conditions were C0 = 156 mg L−1 ,
t = 190 min, m/V = 1 g L−1 , agitation speed = 300 rpm. The texture of
the pretreated sawdust after MB adsorption in Fig. 3(f) is rougher
comparing to the same material before MB adsorption presented in
Fig. 3(c). This fact indicates the swelling effect on the lignocellulosic
material’s particles after MB adsorption.
3.1.3. Surface chemistry
The cell walls of pine sawdust mainly consist of cellulose, hemi-
celluloses, lignin and many hydroxyl groups, such as tannins or
other phenolic compounds. Lignin is a polymer material built up
from the phenyl propane nucleus, an aromatic ring with a three-
carbon side chain. Vanillin and syringaldehyde are the two other
basic structural units of lignin molecule. Tannins are complex poly-
hydric phenols, which are soluble in water and have the property
888 D. Sidiras et al. / Chemical Engineering Journal 171 (2011) 883–896
Fig. 4. FTIR spectra of (a) the untreated and the autohydrolyzed (240 ◦ C, 40 min)
sawdust, and (b) the autohydrolyzed (200 ◦ C, 40 min) sawdust before and after MB
adsorption (C0 = 156 mg L−1 , t = 190 min, m/V = 1 g L−1 , 300 rpm, 23 ◦ C).
of precipitating protein (gelatin). The FTIR spectra of the untreated
pine sawdust and of the autohydrolyzed (T␪a = 240 ◦ C, tai = 40 min)
pine sawdust are given in Fig. 4(a). The comparison of the FTIR
spectrums shows that some peaks were shifted. In the case of
the untreated pine sawdust, there is a strong peak at wavenum-
ber 3471 cm−1 representing the OH stretching of phenol group
of cellulose and lignin, and the peak at 2939 cm−1 indicates the
presence of CH2 stretching of aliphatic compound. The peak at
2146 cm−1 indicates the presence of NH stretching. The appear-
ance of peaks at 1741 cm−1 and 1612 cm−1 indicate the presence
of C O stretching of aldehyde group and C C stretching of phe-
nol group, respectively. Whereas the peak at 1516 cm−1 in the
spectrum of sawdust can be due to C C of aromatic ring. Peaks
at 1470 cm−1 and 1437 cm−1 can be due to CH2 bending and
OH bending, respectively. The peak at 1372 cm−1 shows C O H
bending. Peaks at 1286 cm−1 and 1140 cm−1 in the FTIR spectrum
of sawdust can be due to C O stretching of phenolic group and
six-member cyclic ether group of cellulose, respectively. These
wavenumber values are very close to those reported for pine saw-
dust by Wang et al. [31] and for meranti sawdust by Ahmad et al.
[32].
In the case of the pretreated with autohydrolysis (T␪a = 240 ◦ C,
tai = 40 min) pine sawdust, the shifted peaks are: 3359 cm−1 repre-
senting the OH stretching of phenol group, 2065 cm−1 indicating
the presence of NH stretching, 1713 cm−1 indicating the pres-
ence of C O stretching of aldehyde group, 1450 cm−1 due to OH
bending, 1065 cm−1 due to C O stretching of six member cyclic
ether group of cellulose, and a new one at 822 cm−1 indicating the
presence of aromatic C-H deformation (see Table 1). These shifts
observed in the FTIR spectrum indicated the changes of those func-
tional groups on the surface of autohydrolyzed pine sawdust.
In Fig. 4(b) the transmittance of the FTIR spectra of the
autohydrolyzed (T␪a = 200 ◦ C, tai = 40 min) pine sawdust before
and after MB adsorption are given. The adsorption conditions
were C0 = 156 mg L−1 , t = 190 min, adsorbent = 1 g L−1 , agitation
speed = 300 rpm. In the case of the autohydrolyzed pine sawdust
before MB adsorption, the shifts observed in the FTIR spectrum
Fig. 5. (a) XRD patterns of the untreated and the autohydrolyzed (240 ◦ C, 40 min)
sawdust and (b) BET surface area vs. autohydrolysis time.
indicated the changes of those functional groups on the surface of
autohydrolyzed pine sawdust comparing to the untreated one (see
Table 1). These shifts were less severe than those in the case of the
autohydrolyzed at T␪a = 240 ◦ C pine sawdust for the same time. In
the case of the autohydrolyzed (T␪a = 200 ◦ C, tai = 40 min) pine saw-
dust after MB adsorption, there was a shift at 1454 cm−1 due to OH
bending, at 1088 cm−1 due to C O stretching of six-member cyclic
ether group of cellulose, and at 889 cm−1 indicating the presence
of aromatic C H deformation (see Table 1). The shifts observed in
the FTIR spectrum of the sawdust before and after MB adsorption
indicated some changes of functional groups on the surface of the
sawdust due to bonding with the adsorbed MB.
3.1.4. Crystalline structure
XRD is applicable to analyze the crystallinity of cellulose in
wood. XRD patterns of (i) the untreated pine sawdust and (ii)
the autohydrolyzed (240 ◦ C, 40 min) pine sawdust are shown in
Fig. 5(a). Two broad peaks at the 2 values of around 16◦ and 22◦
for the untreated sawdust were due to the 1 0 1 and 0 0 2 lattice
spacing of cellulose in wood [31]. For the autohydrolyzed sawdust,
these two peaks became narrow, suggesting structural destruction
of crystalline cellulose, in accordance with a sharp weight loss at
these conditions (Fig. 1). The obtained rich-to-lignin material is
mostly amorphous.
3.1.5. BET surface area as affected by the autohydrolysis
conditions
The surface area (in m2 g−1 , measured by the BET method)
is given in Fig. 5(b) as a function of autohydrolysis time ta at
T␪a = 160–240 ◦ C. The effect of autohydrolysis is not significant at
160 ◦ C. The BET surface area is increasing significantly for autohy-
drolysis at 240 ◦ C for ta = 0–50 min.
The efficient removal of the hemicelluloses (in the form of water
soluble reducing sugars) and the amorphous part of cellulose (in the
form of water soluble glucose) results in ‘opening’ of the pores of
the structure of the lignocellulosic matrix [9,21]. This phenomenon
increases the BET surface area of the material, which accounts par-
tially for the advanced adsorption properties of the hydrolyzed
materials over the untreated one. Furthermore, the treatment of
 D. Sidiras et al. / Chemical Engineering Journal 171 (2011) 883–896 889

Table 1
FTIR of (a) the untreated and the autohydrolyzed (240 ◦ C, 40 min) pine sawdust, and (b) the autohydrolyzed (200 ◦ C, 40 min) pine sawdust before and after MB adsorption.

Frequency (cm−1 ) Differences (cm−1 ) Assignment

Untreated Pretreated

(a) Effect of autohydrolysis at 240 ◦ C for 40 min on sawdust

3471 3359 112 OH stretching of phenol group
2939 2945 −6 CH2 stretching of aliphatic compound
2146 2065 81 NH stretching
1741 1713 28 C O stretching of aldehyde group
1612 1610 2 C C stretching of phenol group
1516 1518 −2 C C of aromatic ring
1470 1468 2 CH2 bending
1437 1450 −13 OH bending
1372 1375 −3 C O H bending
1286 1284 2 C O stretching of phenolic group
1140 1065 75 C O stretching of six-member cyclic ether group
– 822 – C H deformation

Frequency (cm−1 )

Before MB adsorption After MB adsorption

(b) Effect of methylene blue adsorption on autohydrolyzed (200 ◦ C, 40 min) sawdust

3392 3398 −6 OH stretching of phenol group
2916 2916 0 CH2 stretching of aliphatic compound
2156 2162 −6 NH stretching
– – – C O stretching of aldehyde group
1606 1605 1 C C stretching of phenol group
1516 1516 0 C C of aromatic ring
1466 1468 −2 CH2 bending
1433 1454 −21 OH bending
1338 1340 −2 C O H bending
1282 1286 −4 C O stretching of phenolic group
1101 1088 13 C O stretching of six-member cyclic ether group
901 889 12 C H deformation

the material leads to the activation of the internal surface of pine found to be 2.500 for untreated and 2.582 for pretreated sawdust
sawdust, thus increasing the number of active sites available for (see Table 2). The standard error of estimates (SEE)-values were
dye binding. calculated by the following equation

 n
3.2. Adsorption isotherms 
 (yi − yi,theor )2

The comparison of the adsorption capacity of the untreated and  i=1
SEE = (3)
pretreated pine sawdust samples was based on the Freundlich [33], n − p
Langmuir [34] and Sips [35] isotherm models. The first two models
are both widely used for investigating the adsorption of a plethora where yi is the experimental value of the depended variable, yi,theor
of dyes on various lignocellulosic materials and activated carbons. is the theoretical (estimated) value of the depended variable, n
The Freundlich [33] isotherm is given by the following equation: is the number of the experimental measurements and p is the
number of parameters, i.e., (n –p ) is the number of the degrees
q = KF · (Ce )1/n (1) of freedom.
where q is the amount adsorbed per unit mass of the adsorbent The KF for the removal of MB by adsorption on pine sawdust as
(mg g−1 ), Ce is the equilibrium concentration of the adsorbate affected by the autohydrolysis conditions was found in this work
(mg L−1 ) and KF [(mg g−1 )(L mg−1 )1/n ], n are the Freundlich con- to be as follows
stants related to adsorption capacity and intensity, respectively. KF = KF,0 ebKF ta (4)
Deriving the logarithmic form of eq. (1):
1 where ta is the autohydrolysis time measured in min, KF,0 and bKF
log q = log KF + log Ce (2) are empirical parameters; KF,0 = 5.602, coefficients of determina-
n
tion R2 = 0.8467–0.9836 for the corresponding linearized model,
The KF and n values were estimated by non-linear regression and
analysis (NLRA) from the experimental adsorption data obtained
at 23 ◦ C for MB, BB and AO. Fig. 6(a) presents, as an example, MB bKF = 4 × 10−7 Ta 2 − 0.0004Ta + 0.0854 (5)
adsorption isotherms by untreated (in its natural form) and auto-
hydrolyzed (240 ◦ C, 50 min) pine sawdust. The theoretical curves where Ta is the autohydrolysis temperature measured in Kelvin.
are estimated according to the Freundlich equation. The Freundlich The Langmuir isotherm equation [34] is based on the following
parameter values are shown in Table 2. The KF values estimated ‘pseudo-monolayer’ adsorption model.
for the autohydrolyzed samples were significantly higher com- KL qm Ce
paring to those of the untreated material, indicating an increased q= (6)
1 + KL Ce
adsorption capacity of the former; the KF values were exponen-
or
tially increased by increasing the treatment time for constant
1
 1   1
 1
autohydrolysis temperature (see Fig. 7(a)). The parameter n was = + · (7)
not significantly affected by the pretreatment conditions and was q qm KL · qm Ce
890 D. Sidiras et al. / Chemical Engineering Journal 171 (2011) 883–896

Table 2
Estimated values of SEE, when least squares are used, for the alternative isotherm models considered herein.

KF [(mg g−1 )(L mg−1 )1/n ] KL (L mg−1 ) qm (mg g−1 ) n SEE

Methylene blue
Untreated sawdust
Freundlich 5.602 2.500 2.900
Langmuir 0.06842 38.72 1.209
Sips 0.06055 40.42 1.084 1.198
Pretreated sawdust: autohydrolysis at 240 ◦ C for 40 min
Freundlich 15.68 2.582 3.334
Langmuir 0.1072 88.02 4.355
Sips 0.0245 138.1 1.691 1.525
Bismarck brown
Untreated sawdust
Freundlich 3.367 1.683 0.2365
Langmuir 0.3771 12.24 0.2971
Sips 0.02901 40.09 1.482 0.2494
Pretreated sawdust: autohydrolysis at 240 ◦ C for 40 min
Freundlich 15.20 1.797 0.8645
Langmuir 2.567 18.26 1.137
Sips 0.02619 138.2 1.721 0.9299
Acridine orange
Untreated sawdust
Freundlich 3.112 1.802 0.2590
Langmuir 0.5691 8.792 0.2628
Sips 0.02713 33.13 1.595 0.2653
Pretreated sawdust: autohydrolysis at 240 ◦ C for 40 min
Freundlich 15.52 1.435 0.4763
Langmuir 1.914 18.78 0.5210
Sips 0.04112 166.0 1.388 0.5030

where KL is the Langmuir constant related to the energy of adsorp-

tion (L mg−1 ) and qm the amount of dye adsorbed (mg g−1 ) when
saturation is attained. In cases where the isotherm experimental
data approximates the Langmuir equation, the parameters KL and
qm can be estimated either by plotting 1/q versus 1/Ce either by
NLRA. Table 2 presents the estimated (by NLRA) parameter values
for the data gathered in the present study. The qm -values, obtained
for untreated pine sawdust were significantly lower than the values
for the pretreated samples. The fitting of the Langmuir’s adsorption
model to the present data (see theoretical curves in Fig. 6(a)) was
also quite satisfactory but to a lesser degree comparing with the
Freundlich model in the case of the pretreated sawdust, as shown
by the corresponding SEE-values given in Table 2.

Fig. 6. Isotherms of (a) MB, (b) BB and (c) AO adsorption on untreated and autohy-
drolyzed (240 ◦ C, 40 min) sawdust.

Fig. 7. (a) KF and (b) qm for MB adsorption on sawdust vs. autohydrolysis time.
 D. Sidiras et al. / Chemical Engineering Journal 171 (2011) 883–896
Fig. 8. (a) MB, (b) BB and (c) AO adsorption kinetics for untreated and autohy-
drolyzed (240 ◦ C, 40 min) sawdust (C0 = 12.5 mg L−1 for MB, 3.7 mg L−1 for BB and
6.7 mg L−1 for AO; m/V = 1 g L−1 ; 300 rpm, 23 ◦ C).
The qm -values, for the removal of MB by adsorption on pine
sawdust as affected by the autohydrolysis conditions (see Fig. 7(b)),
were found in this work to be
qm = qm,0 ebqm ta
where ta is the autohydrolysis time measured in min, qm,0 and
bqm are empirical parameters; qm,0 = 29.62 mg L−1 , coefficients of
determination R2 = 0.6592–0.8742, and
bqm = 8 × 10−5 Ta − 0.0314
Fig. 9. k values for (a) MB, (b) BB and (c) AO adsorption on sawdust vs. agitation
speed (C0 = 12.5 mg L−1 for MB, 3.7 mg L−1 for BB and 6.7 mg L−1 for AO; m/V = 1 g L−1 ;
300 rpm, 23 ◦ C).
891
Fig. 10. (a) MB, (b) BB and (c) AO concentration for adsorption on sawdust vs. initial
pH of the dye solution (t = 190 min; C0 = 12.5 mg L−1 for MB, 3.7 mg L−1 for BB and
6.7 mg L−1 for AO; m/V = 1 g L−1 ; 300 rpm; 23 ◦ C).
where Ta is the autohydrolysis temperature measured in Kelvin.
The essential characteristics of the Langmuir isotherm can be
described by a dimensionless constant called ‘equilibrium param-
eter’ or ‘separation factor’ RL , defined by the following equation:
1
RL = (10)
1 + KL · C0
where C0 is the initial dye concentration (mg L−1 ) and KL is the
Langmuir constant (L mg−1 ). The value of RL indicates the type of the
isotherm to be either unfavorable (RL > 1), linear (RL = 1), favorable
(0 < RL < 1) or irreversible (RL = 0). At the present work, the RL values
were found 0.0564–0.9126, i.e. 0 < RL < 1 for all dye concentrations
C0 in the range of 1.4–156 mg L−1 for MB, for all adsorbents (pre-
treated and untreated sawdust) studied. They confirmed that the
pine sawdust is favorable for adsorption of dye under conditions
used in this study.
The Sips (Langmuir–Freundlich) [35] isotherm equation, also
examined in the present work, is based on the following adsorption
model.
qm · (KL · Ce )1/n
q= (11)
1 + (KL · Ce )1/n
where KL , qm is the Langmuir constants and n the Freundlich con-
stant.
The parameters of the three isotherm models can be obtained
by NLRA. Table 2 presents the estimated parameter values for the
experimental data obtained in the present study. The fitting of the
Sips’ adsorption model to the present data (see theoretical curves
(8)
in Fig. 6(a)) was the most satisfactory for MB adsorption, better
than the other two isotherm models, in both cases of the untreated
and pretreated sawdust, as shown by the corresponding SEE-values
given in Table 2. For BB and AO adsorption the Freundlich model
was the best in all cases (see Fig. 6(b) and (c), and Table 2).
(9)
3.3. Kinetics of adsorption
The kinetics of adsorption of MB on several lignocellulosic mate-
rials has been extensively studied using various kinetic equations.
However, information on the kinetics of BB [14] and AO on such
materials is scarce. At first place, we used Lagergren equation [36]:
q − qt = q · e−k·t (12)
where q and qt are the amounts of dye adsorbed per unit mass of
the adsorbent (mg g−1 ) at equilibrium time (t→ ∞) and adsorption
time t, respectively, while k is the pseudo-first order rate constant
for the adsorption process (min−1 ). Moreover, q = (C0 − Ce )V/m and
qt = (C0 − C)V/m, where C, C0 , Ce are the concentrations of methylene
blue in the bulk solution at time t, 0, and ∞, respectively, while m is
892 D. Sidiras et al. / Chemical Engineering Journal 171 (2011) 883–896
Fig. 11. (a) MB, (b) BB and (c) AO amount adsorbed on sawdust vs.
(C0 = 12.5 mg L−1 for MB, 3.7 mg L−1 for BB and 6.7 mg L−1 for AO; m/V = 1 g L−1 ;
300 rpm; 23 ◦ C).
the weight of the adsorbent used (g), and V is the solution volume
(mL). Further modification of eq. (16) in logarithmic form gives:
ln(q − qt ) = ln q − k · t
The plots of ln(q − qt ) vs. t for all dyes adsorbent systems were
found to be linear, indicating the possibility of first order nature
of the adsorption process. MB, BB and AO adsorption kinetics
by untreated and pretreated (autohydrolysis, 240 ◦ C, 40 min) pine
sawdust are presented in Fig. 8(a–c) respectively. The theoret-
ical curves are estimated according to the Lagergren equation.
In the case of MB, the estimated by NLRA values of k were
0.0040–0.0184 min−1 and the SEE-values were 0.100–0.301. All
SEE-values were found low, indicating the applicability of this
kinetic equation to the adsorption of MB on pine sawdust. In the
case of BB, the k values were 0.0317–0.1396 min−1 and the SEE-
values were 0.067–0.140, and in the case of AO, the k values were
0.0199–0.1594 min−1 and the SEE-values were 0.126–0.233, indi-
cating the applicability of this kinetic equation to the adsorption of
BB and AO on pine sawdust.
According to the Lagergren model and the Arrhenius law
k = p · exp(− E/RT), the activation energy for the adsorption of MB
on untreated and pretreated pine sawdust was estimated by lin-
ear regression of lnk on 1/T. The MB adsorption activation energy E
was found to be 14–15 kJ mol−1 for the treated materials, approx-
imately equal to the activation energy of the untreated sawdust
(15 kJ mol−1 ). This indicates that a physical process, i.e. the intra-
particle diffusion, is the controlling step of the adsorption process.
Assuming a -order kinetic model:
dq
= k (q − qt ) (14)
dt
From the differential eq. (14) for  =
/ 1 we obtain:
 1/(1−)
qt = q − q1− + ( − 1) k t (15)
The order of the kinetic model was found to be  = 2.87 ± 1.00
for untreated sawdust and  = 1.87 ± 0.22 for pretreated (autohy-
drolysis, 240 ◦ C, 40 min) sawdust.
For  = 2 we have the pseudo-second order kinetic model intro-
duced [37,38] as follows
 −1 1
qt = q − q−1 + k2 t or qt = q − (16)
(1/q) + k2 t
The diagrams for second order kinetics are presented in Fig. 8(a)
for MB, in Fig. 8(b) for BB and in Fig. 8(c) for AO, for untreated and
pretreated (autohydrolysis, 240 ◦ C, 40 min) sawdust. In the case of
MB, the estimated by NLRA values of the second order rate con-
stant k2 were 0.00091–0.00515 min−1 and the SEE-values were
0.157–0.322. All SEE-values were found low, but higher than the
SEE-values of the first-order kinetic model, indicating the higher
applicability of the first-order kinetic equation to the adsorption
of MB on pine sawdust. In the case of BB, the estimated k2 values
were 0.0055–0.4733 min−1 and the SEE-values were 0.013–0.039.
All SEE-values were found lower than the SEE-values of the first-
order kinetic model, indicating the higher applicability of the
second-order kinetic equation to the adsorption of BB on pine
sawdust. The same was found in the case of AO, where k was
0.0030–0.1594 min−1 and SEE was 0.031–0.111.
In all cases, the k-values of the first-order and second-order
kinetic model were not significantly enhanced by the autohydrol-
√
t ysis pretreatment of pine sawdust. Despite this, the adsorption
rate dqt /dt is significantly enhanced due to the (q − qt ) difference
increase (see Fig. 8).
Agitation is an important parameter in sorption phenomena,
influencing the distribution of the solute in the bulk solution and
the formation of the external boundary film. The effect of stir-
ring speed (rpm) on the adsorption rate constant k (min−1 ) of
the untreated material was investigated (see Fig. 9). The kinetics
(13) seemed to be affected by the agitation speed for values between
0 and 200 rpm, thus confirming that the influence of external dif-
fusion on the sorption kinetic control plays a significant role. In
contrast, the small effect of agitation in the range of 200–600 rpm
showed that external mass transfer was not the rate limiting step,
and implied that intraparticle diffusion resistance needed to be
included in the analysis of overall sorption [21].
The effect of the pH of the dye solution on the amount of dye
adsorbed was studied for MB, BB and AO by varying the initial pH
under constant process parameters. The final concentration C of
dyes solutions after an adsorption period of 190 min was signifi-
cantly higher for pH values between 1.5 and 4 for both untreated
and pretreated materials than the relevant concentration (C for
t = 190 min) for pH 8 (see Fig. 10). The lower adsorption of dye
at acidic pH was due to the presence of excess H+ ions that com-
peted with the dye cation for adsorption sites. As the pH of the
system increased (pH > 8), the number of positively charged avail-
able sites decreased while the number of the negatively charged
sites increased. The negatively charged sites favored the adsorption
of dye cation due to electrostatic attraction. The increase in initial
pH from 8 to 13, slightly increased the amount of dye adsorbed.
The final pH of the solution was found to decrease only slightly (by
0.3–0.5 pH units) after adsorption of dye (in cationic form) with the
release of H+ ions from the active site of the adsorbent surface. The
results were in agreement with other literature reports [24–26].
 D. Sidiras et al. / Chemical Engineering Journal 171 (2011) 883–896 893

Fig. 12. Breakthrough curves for (a) MB with Ci = 190 mg L−1 , (b) MB with Ci = 14 mg L−1 , (c) BB and (d) AO adsorption on fixed beds (x = 15 cm) of untreated and autohydrolyzed
(240 ◦ C, 40 min) sawdust.

Adsorbate species are probably transported from the bulk of
the solution into the solid phase through an intra-particle diffu-
sion/transport process, which is frequently the rate-limiting step
in many adsorption processes, especially in a rapidly stirred batch
reactor. The possibility of intra-particle diffusion was explored by
using the intra-particle diffusion model [39]:
√ R2 = 0.9071 for pretreated (autohydrolysis, 240 ◦ C, 40 min) sawdust
qt = c + kp · t (17)
where qt is the amount (mg g−1 ) of dye adsorbed at time
t, c is a constant (mg g−1 ) and kp is the intra-particle diffusion
rate constant in mg g−1 min−0.5 . In the case of MB, the esti-
mated by NLRA values of the kp were 0.315–0.805 mg g−1 min−0.5
and the SEE-values were 0.308–0.596, assuming c = 0. All SEE-
values were found higher than the SEE-values of the first-order
and second-order kinetic models, indicating the lower applicabil-
ity of the intra-particle diffusion equation to the adsorption of
MB on pine sawdust. On the other hand, using linear regression
analysis, qt = 1.2921 + 0.6241t0.5 with coefficient of determination
R2 = 0.9264 for untreated sawdust, and qt = 2.3447 + 0.946t0.5 with
(see Fig. 11(a)). Analogous findings were in the case of BB, where
kp was 0.057–0.160 mg g−1 min−0.5 and SEE was 0.062–0.179, and
in the case of AO, where kp was 0.342–0.534 mg g−1 min−0.5 and
SEE was 0.293–1.317. Moreover, using linear regression analy-
sis, in the case of BB, qt = −0.0004 + 0.0758t0.5 with R2 = 0.9739 for
untreated sawdust, and qt = 0.1475 + 0.1571t0.5 with R2 = 0.9464 for

Table 3
Estimated values of SEE, when least squares are used, for the alternative column models considered herein.

Ci (mg L−1 ) Q (mL min−1 ) n N (mg L−1 ) K (L mg−1 min−1 ) SEE

Methylene blue
Untreated sawdust
Bohart-Adams 192 46 8046 0.00050 7.58
Bohart-Adams 14.0 10 8722 0.00134 0.194
Clark 192 46 2.500 8723 0.00058 8.75
Clark 14.0 10 2.500 8781 0.00159 0.201
Pretreated sawdust: autohydrolysis at 240 ◦ C for 40 min
Bohart-Adams 187 41 14157 0.00034 7.13
Bohart-Adams 14.3 10 14313 0.00072 0.179
Clark 187 41 2.582 15132 0.00042 8.47
Clark 14.3 10 2.582 14429 0.00087 0.217
Bismarck brown
Untreated sawdust
Bohart-Adams 11.0 10 6576 0.00045 0.300
Clark 11.0 10 1.683 6394 0.00040 0.340
Pretreated sawdust: autohydrolysis at 240 ◦ C for 40 min
Bohart-Adams 11.1 10 8918 0.00093 0.150
Clark 11.1 10 1.797 8866 0.00086 0.156
Acridine orange
Untreated sawdust
Bohart-Adams 8.9 10 4452 0.00168 0.242
Clark 8.9 10 1.802 4425 0.00156 0.262
Pretreated sawdust: autohydrolysis at 240 ◦ C for 40 min
Bohart-Adams 9.0 10 7161 0.00114 0.124
Clark 9.0 10 1.435 6998 0.00093 0.149
894 D. Sidiras et al. / Chemical Engineering Journal 171 (2011) 883–896

Table 4
The Freundlich and Langmuir parameters of adsorption isotherms for various adsorbents according to the literature.

Freundlich Langmuir Reference

KF [(mg g−1 )(L mg−1 )1/n ] n qm (mg g−1 ) KL (L mg−1 )

Methylene blue
Bamboo dust activated carbon 0.68 1.96 143 0.12 [23]
Banana peels 1.34 3.0 20.8 16.5 [16]
Beech sawdust 6.05 1.59 9.78 1.60 [21]
Beech sawdust CaCl2 -treated 11.99 1.61 16.05 2.13 [9]
Beech sawdust MgCl2 -treated 2.91 1.41 14.2 0.288 [12]
Beech sawdust prehydrolyzed 23.42 1.60 30.50 2.09 [21]
Beech sawdust NaCl-treated 3.18 1.61 9.2 0.613 [12]
Beech sawdust ZnCl2 -treated 2.85 1.35 11.7 0.292 [12]
Brazil nut shells 0.216 0.876 7.81 0.15 [42]
Cedar sawdust 92.78 3.94 111.97 20.89 [13]
Cherry 39.84 1.08 [24]
Coconut bunch waste 13.45 3.66 70.92 0.029 [25]
Commercial activated carbon 6.7 3.4 980 0.48 [23]
Crushed brick 44.93 2.83 80.60 1.73 [13]
Hazelnut 76.9 0.36 [24]
Hazelnut shell 21.44 0.65 [26]
Meranti sawdust 12.14 1.441 120.48 0.042 [32]
Oak 29.94 0.30 [24]
Orange peels 1.75 3.8 18.6 19.9 [16]
Palm kernel fiber 8.67 1.77 95.4 0.0317 [17]
Peanut hull dehydrated 81.6 22.3 108.6 0.121 [47]
Pine sawdust 5.602 2.500 38.72 0.06842 In this work
Pine sawdust autohydrolyzed 15.68 2.582 88.02 0.1072 In this work
Pitch-pine 27.78 0.43 [24]
Tea (rejected) 3.15 2.19 147 0.047 [22]
Tea (rejected) NaOH-modified 40.87 2.11 242.11 0.13 [48]
Walnut 59.17 0.15 [24]
Wheat straw 1.59 0.97 2.23 0.758 [11]
Wheat straw prehydrolyzed 11.95 1.33 20.41 0.663 [11]
Wheat straw citric acid-modified 69.14 2.76 396.9 0.0685 [46]
Wood apple shell 20.4 3.23 95.20 0.059 [43]

Bismarck brown
Chemical activated carbon 54 3.07 164 0.22 [14]
Commercial activated carbon 103 4.02 303 0.19 [14]
Fodder yeast cells magnetically modified 8.10 0.394 75.71 0.025 [45]
Grain magnetically modified 7.36 0.418 72.4 0.029 [44]
Pine sawdust 3.367 1.683 12.24 0.3771 In this work
Pine sawdust autohydrolyzed 15.20 1.797 18.26 2.567 In this work
Rubberwood sawdust chemical/steam-activated carbon 598 9.68 1111 0.11 [14]
Rubberwood sawdust steam-activated carbon 1003 8.58 2000 0.07 [14]

Acridine orange
Fodder yeast cells magnetically modified 21.38 0.252 62.22 0.443 [45]
Pine sawdust 3.112 1.802 8.792 0.5691 In this work
Pine sawdust autohydrolyzed 15.52 1.435 18.78 1.914 In this work

pretreated sawdust (see Fig. 11(b)). Similarly, in the case of AO, expressed as non-linearized equation has the form of the ‘simple
qt = −0.1134 + 0.3955t0.5 with R2 = 0.9522 for untreated sawdust, logistic function’:
and qt = 2.0742 + 0.4211t0.5 with R2 = 0.7366 for pretreated sawdust
Ci
(see Fig. 11(c)). In all cases, i.e. MB, BB and AO adsorption on treated C= (19)
1 + Ae−rt
and untreated sawdust, the high R2 values indicate intra-particle
diffusion as the controlling step. where A = eKNx/u ; r = K · Ci .
Clark [41] has developed an alternative to the ‘simple logistic
function’, called the ‘generalized logistic function’, which incorpo-
rates the parameter n of the Freundlich adsorption isotherm:
3.4. Column studies

1/(n−1)
Ci n−1
The ‘bed depth service model’ developed by Bohart and Adams C= (20)
1 + Ae−rt
[40] is as follows
where n = inverse of the slope of the Freundlich isotherm. This
C  K ·N·x expression coincides with Eq. (19) for n = 2. Evidently, we can esti-
i
ln −1 = − K · Ci · t (18) mate the parameters K and N by applying linear and non-linear
C u
regression through expressions (18) and (19), respectively. Since
the SEE is given by expression (3), we may conclude that this
in which C = effluent concentration (mg L−1 ); Ci = influent con- statistic measure of goodness of fitting has a lower value when per-
centration (mg L−1 ); K = adsorption rate coefficient (L mg−1 min−1 ); forming non-linear regression. The equations used to calculate the
N = adsorption capacity coefficient (mg L−1 L); x = bed depth (cm); parameter values after having performed non-linear regression are
u = linear velocity (cm min−1 ); and t = time (min). This model, K = r/Ci and N = u lnA/(xK)= Ci u lnA/(xr).
 D. Sidiras et al. / Chemical Engineering Journal 171 (2011) 883–896
The values of A and r according to the Bohart–Adams model can
thus be estimated by NLRA from the column effluent data for all the
cases presented in Fig. 12(a) and (b) for MB, (c) for BB and (d) for
AO. The effluent dye solution volume V (in L) is V = Q. t, where Q is
the dye solution flow rate. The theoretical estimations, according
to the Bohart–Adams model, sufficiently simulate the experimen-
tal data in all the examples given in Fig. 12. The N and K values are
calculated from A and r values and presented in Table 3. It can be
mentioned that N = 14157 mg L−1 for the MB adsorption on auto-
hydrolyzed (240 ◦ C, 40 min) sawdust higher than N = 8046 mg L−1
for the untreated one. In addition, K = 0.00034 L mg−1 min−1 for
the treated material close to the K = 0.00050 L mg−1 min−1 for the
untreated one. The SEE-values are shown in Table 3. According to
them, the fitting of the Bohart–Adams model to experimental data
was found to be better than the fitting of the Clark model for the
three dyes used herein. According to both models and referring to
the three dyes used, N for the treated pine sawdust was higher than
that for the untreated one. The incorporated in the Clark model
parameter n of the Freundlich adsorption isotherm, presented in
Table 3, was taken from Table 2 results.
3.5. Discussion
The results presented herein showed to which extent the batch
and column kinetics of MB, BB and AO adsorption on autohy-
drolyzed (160–240 ◦ C, 0–50 min) pine sawdust was improved,
comparing to those of the untreated material as control. The
effect of the autohydrolysis on the microstructure and the crys-
tallinity of pine sawdust was significant, as determined by using
SEM, FTIR, XRD and BET surface area techniques; more specifi-
cally, the increased and rough surface of the pretreated sawdust
(shown in SEM micrographs with magnification 50,000) may be
a main contributor to the enhanced adsorption properties. The
estimated values of BET surface area, the KF , the qm and the Bohart-
Adams adsorption capacity coefficient N indicate quantitatively
how the specific pretreatment process presented herein enhanced
the adsorptivity of the corresponding untreated material.
In Table 4 we compare the MB, BB and AO adsorption capacity
KF and qm of pretreated and untreated pine sawdust with other
adsorbents derived from agricultural or waste materials, according
to the available in the literature data [42–48].
The sawdust particles mainly consist of cellulose and lignin, and
many hydroxyl-rich compounds, such as tannins or other phenolic
compounds. All these components exhibit properties contribut-
ing to ion exchange, which is a significant stage in several kinds
of adsorption, favoring also some penetration of the adsorbate
(or its product) beyond the surface of the adsorbent. Since ion
exchange is combined to adsorption, studies in particular kinetics
and isotherms, provide information on the mechanism of sorp-
tion. These mechanisms are, in general, complicated because they
implicate the presence of different interactions. In addition, a
wide range of chemical structures, pH, salt concentrations and the
presence of ligands often add to the complication. Some of the
reported interactions include: ion-exchange, complexation, coordi-
nation/chelation, electrostatic interactions, acid–base interactions,
hydrogen bonding, hydrophobic interactions, physical adsorption
and precipitation [3,4].
4. Conclusions
Autohydrolysis comparing other pretreatment methods has the
advantage that is using pure and easily recyclable water; no chem-
icals (acids, salts, based organic solvents) are needed. The new
results provide experimental evidence for enhanced adsorption
properties of the pretreated material. The gain is the understanding
895
of how and to what extent autohydrolysis affects the solid residue
properties. It is a cost-effective method, because the pretreat-
ment expenses are covered by the produced fermentable sugars for
the bioethanol production industry, which is subsidised in the EU
Countries. Moreover, the exclusive use of water is cost- and quality-
effective. The environmental implications might be the relatively
high temperatures used (e.g. compared with acid hydrolysis) mak-
ing autohydrolysis an energy intensive process. However thermal
energy demand can be covered by using cylindrical parabolic con-
centrators of solar radiation which is adequate for this purpose,
even in the Central-Northern European Countries. Thus, autohy-
drolyzed pine sawdust could be made widely available for use as an
alternative to commercial activated carbons for the removal of basic
dyes from water/wastewater effluents. Furthermore, as sawdust is
an industrial waste and no addition of chemicals is required, we
argue that this process of adsorbent modification may be consid-
ered to take place within an ‘Industrial Ecology’ framework, since a
(solid) waste is used to treat another (aquatic) waste, contributing
to pollution abatement without entailing excessive cost. This is an
argument a fortiori, when additives (e.g., chloride salts) are used to
enhance adsorptivity on condition they are available in the vicinity
as waste or low value by-products of proper quality.
Acknowledgments
The authors acknowledge the financial support from (a) the
Research Center of the University of Piraeus, (b) the Institut fuer
Kern- und Energietechnik and (c) the University of Minnesota.
Technical and economic support has been also provided within the
framework of the independent project “Rational Design of Inno-
vative Catalytic Technologies for Biomass Derivative Utilization”
running within the University of Minnesota. Moreover, the authors
would like to thank the anonymous reviewers for their insightful
suggestions in improving the initial manuscript.
References
[1] J.M. Lee, J. Shi, R.A. Venditti, H. Jameel, Autohydrolysis pretreatment of Coastal
Bermuda grass for increased enzyme hydrolysis, Biores. Techn. 100 (24) (2009)
6434–6441.
[2] D. Nabarlatz, A. Ebringerová, D. Montané, Autohydrolysis of agricultural by-
products for the production of xylo-oligosaccharides, Carbohyd. Polym. 69 (1)
(2007) 20–28.
[3] A. Shukla, Y.-H. Zhang, P. Dubey, J.L. Margrave, The role of straw in the removal
of unwanted materials from water, J. Hazard. Mater. B95 (2002) 137–152.
[4] G. Crini, Non-conventional low-cost adsorbents for dye removal: A review,
Biores. Technol. 97 (9) (2006) 1061–1085.
[5] V.K. Gupta, Suhas. Application of low-cost adsorbents for dye removal–A
review, J. Env. Manag. 90 (2009) 2313–2342.
[6] V.K. Gupta, P.J.M. Carrott, M.M.L. Ribeiro Carrott, Suhas. Low-Cost Adsorbents:
Growing Approach to Wastewater Treatment—a Review, Cr. Rev. Env. Sc. Techn.
39 (2009) 783–842.
[7] A. Demirbas, Agricultural based activated carbons for the removal of dyes from
aqueous solutions: A review, J. Hazard. Mater. 167 (2009) 1–9.
[8] M. Rafatullaha, O. Sulaiman, R. Hashim, A. Ahmad, Adsorption of methylene
blue on low-cost adsorbents: A review, J. Hazard. Mater. 177 (1–3) (2010)
70–80.
[9] F.A. Batzias, D.K. Sidiras, Dye adsorption by calcium chloride treated beech
sawdust in batch and fixed-bed systems, J. Hazard. Mater. B114(1–3) (2004)
167–174.
[10] P. Nigam, G. Armour, I.M. Banat, D. Singh, R. Marchant, Physical removal of
textile dyes from effluents and solid-state fermentation of dye-adsorbed agri-
cultural residues, Biores. Technol. 72 (2000) 219–226.
[11] F. Batzias, D. Sidiras, E. Schroeder, C. Weber, Simulation of dye adsorption on
hydrolyzed wheat straw in batch and fixed-bed systems, Chem. Eng. J. 148 (2–3)
(2009) 459–472.
[12] F.A. Batzias, D.K. Sidiras, Simulation of methylene blue adsorption by salts-
treated beech sawdust in batch and fixed-bed systems, J. Hazard. Mater. 149
(1) (2007) 8–17.
[13] O. Hamdaoui, Batch study of liquid-phase adsorption of methylene blue using
cedar sawdust and crushed brick, J. Hazard. Mater. 135 (1–3) (2006) 264–273.
[14] B.G.P. Kumar, L.R. Miranda, M. Velan, Adsorption of Bismark Brown dye on acti-
vated carbons prepared from rubberwood sawdust (Hevea brasiliensis) using
different activation methods, J. Hazard. Mater. 126 (1–3) (2005) 63–70.
896 D. Sidiras et al. / Chemical Engineering Journal 171 (2011) 883–896
[15] S.J. Allen, Q. Gan, R. Matthews, P.A. Johnson, Comparison of optimised isotherm
models for basic dye adsorption by kudzu, Biores. Technol. 88 (2) (2003)
143–152.
[16] G. Annadurai, R.-S. Juang, D.-J. Lee, Use of cellulose based wastes for adsorption
of dyes from aqueous solutions, J. Hazard. Mater. B92 (2002) 263–274.
[17] G.O. El-Sayed, Removal of methylene blue and crystal violet from aqueous
solutions by palm kernel fiber, Desalination 272 (2011) 225–232.
[18] T. Robinson, B. Chandran, P. Nigam, The effect of pretreatments of three waste
residues, wheat straw, corncobs and barley husks on dye adsorption, Biores.
Technol. 85 (2002) 119–124.
[19] V.K. Garg, R. Gupta, A.-B. Yadav, R. Kumar, Dye removal from aqueous solutions
by adsorption on treated sawdust, Biores. Technol. 89 (2003) 121–124.
[20] X.Y. Shi, B. Xiao, X.Y. Yang, X.P. Zhou, J.F. Li, Batch study of dye removal from
aqueous solutions by adsorption on NaOH-treated firry sawdust, Fres. Env. Bull.
16 (12A) (2007) 1583–1587.
[21] F.A. Batzias, D.K. Sidiras, Dye adsorption by prehydrolysed beech sawdust in
batch and fixed-bed systems, Biores. Technol. 98 (6) (2007) 1208–1217.
[22] N. Nasuha, B.H. Hameed, T. Azam, Din. Mohd, Rejected tea as a potential low-
cost adsorbent for the removal of methylene blue, J. Hazard. Mater. 175 (2010)
126–132.
[23] N. Kannan, M.M. Sundaram, Kinetics and mechanism of removal of methylene
blue by adsorption on various carbons–a comparative study, Dyes Pigments 51
(2001) 25–40.
[24] F. Ferrero, Dye removal by low cost adsorbents: Hazelnut shells in comparison
with wood sawdust, J. Hazard. Mater. 142 (1–2) (2007) 144–152.
[25] B.H. Hameed, D.K. Mahmoud, A.L. Ahmad, Equilibrium modeling and kinetic
studies on the adsorption of basic dye by a low-cost adsorbent: Coconut (Cocos
nucifera) bunch waste, J. Hazard. Mater. 158 (1) (2008) 65–72.
[26] M. Dogan, H. Abak, M. Alkan, Biosorption of methylene blue from aqueous solu-
tions by hazelnut shells: Equilibrium, parameters and isotherms, Water Air Soil
Pollut. 192 (1–4) (2008) 141–153.
[27] L. Segal, J.J. Greely, A.E. Martin, C.M. Conrad, An empirical method for estimating
the degree of crystallinity of native cellulose using the x-ray diffractometer,
Textile Res. J. 29 (1959) 786–795.
[28] J.F. Saeman, J.F. Bubl, E.E. Harris, Quantitative saccharification of wood and
cellulose, Ind. Eng. Chem. Anal. Ed. 17 (1945) 35–37.
[29] Tappi Standards, Tappi Tests Methods, T222 om-88, Atlanta (1997).
[30] DIN 66132, Determination of specific surface area of solids by adsorption of
nitrogen; single-point differential method according to Haul and Dümbgen
(1975).
[31] Z. Wang, J. Cao, J. Wang, Pyrolytic characteristics of pine wood in a slowly
heating and gas sweeping fixed-bed reactor, J. Anal. Appl. Pyrol. 84 (2) (2009)
179–184.
[32] A. Ahmad, M. Rafatullah, O. Sulaiman, M.H. Ibrahim, R. Hashim, Scavenging
behaviour of meranti sawdust in the removal of methylene blue from aqueous
solution, J. Hazard. Mater. 170 (1) (2009) 357–365.
[33] H.M.F. Freundlich, Über die adsorption in lösungen, Zeitschrift für Physikalische
Chemie 57 (1906) 385–471.
[34] I. Langmuir, The constitution and fundamental properties of solids and liquids,
J. Am. Chem. Soc. 38 (1916) 2221–2295.
[35] R. Sips, Structure of a catalyst surface, J. Chem. Phys. 16 (1948) 490–495.
[36] S. Lagergren, Zur theorie der sogenannten adsorption gelöster stoffe, Kungliga
Svenska Vetenskapsakademiens, Handlingar 24 (1898) 1–39.
[37] Y.S. Ho, J.C.Y. Ng, G. McKay, Kinetics of pollutants sorption by biosorbents:
review, Sep. Purif. Methods 29 (2000) 189–232.
[38] F.-C. Wu, R.-L. Tseng, S.-C. Huang, R.-S. Juang, Characteristics of pseudo-second-
order kinetic model for liquid-phase adsorption: A mini-review, Chem. Eng. J.
151 (1–3) (2009) 1–9.
[39] W.J. Weber, J.C. Morris, Kinetics of adsorption on carbon from solution, J. Sanit.
Eng. Div. Am. Soc. Civ. Eng. 89 (1963) 31–60.
[40] G. Bohart, E.N. Adams, Some aspects of the behavior of charcoal with respect
to chlorine, J. Am. Chem. Soc. 42 (1920) 523–544.
[41] R.M. Clark, Modeling TOC removal by GAC: The general logistic function, J. Am.
Wat. Works Assoc. 79 (1) (1987) 33–37.
[42] S.M.O. Brito, H.M.C. Andrade, L.F. Soares, R.P. Azevedo, Brazil nut shells as a
new biosorbent to remove methylene blue and indigo carmine from aqueous
solutions, J. Hazard. Mater. 174 (2010) 84–92.
[43] S. Jain, R.V. Jayaram, Removal of basic dyes from aqueous solution by low-
cost adsorbent: Wood apple shell (Feronia acidissima), Desalination 250 (2010)
921–927.
[44] I. Safarik, K. Horska, M. Safarikova, Magnetically modified spent grain for dye
removal, J. Cereal Sci. 53 (2011) 78–80.
[45] I. Safarik, L.F.T. Rego, M. Borovska, E. Mosiniewicz-Szablewska, F. Weyda,
M. Safarikova, New magnetically responsive yeast-based biosorbent for the
efficient removal of water-soluble dyes, Enzyme Microb. Tech. 40 (2007)
1551–1556.
[46] R. Han, L. Zhang, C. Song, M. Zhang, H. Zhu, L. Zhang, Characterization
of modified wheat straw, kinetic and equilibrium study about copper ion
and methylene blue adsorption in batch mode, Carbohyd. Polym. 79 (2010)
1140–1149.
[47] D. Ozer, G. Dursun, A. Ozer, Methylene blue adsorption from aqueous
solution by dehydrated peanut hull, J. Hazard. Mater. 144 (2007) 171–
179.
[48] N. Nasuha, B.H. Hameed, Adsorption of methylene blue from aqueous solu-
tion onto NaOH-modified rejected tea, Chem. Eng. J. 166 (2011) 783–
786.