chemical engineering research and design 1 0 4 ( 2 0 1 5 ) 579–588

Contents lists available at ScienceDirect

Chemical Engineering Research and Design

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

Modeling adsorption rate of tetracyclines on

activated carbons from aqueous phase

R. Ocampo-Pérez a,∗ , R. Leyva-Ramos a , J. Rivera-Utrilla b , J.V. Flores-Cano b ,

M. Sánchez-Polo b
a Centro de Investigación y Estudios de Posgrado, Facultad de Ciencias Químicas, Universidad Autónoma de San
Luis Potosí, Av. Dr. M. Nava No. 6, San Luis Potosi 78210, SLP, Mexico
b Departamento de Química Inorgánica, Facultad de Ciencias, Universidad de Granada, 18071 Granada, Spain

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

Article history: The adsorption rate of three tetracyclines (TCs) (tetracycline, oxytetracycline, chlortetra-
Received 14 February 2015 cycline) on two activated carbons (ACs) were investigated in this work. The experimental
Received in revised form 23 June adsorption equilibrium data of the TCs on both carbons was obtained in a batch adsorber at
2015 T = 298 K and pH range of 4–5, and the Langmuir isotherm better interpreted the experimen-
Accepted 16 September 2015 tal data than the Freundlich isotherm. The adsorption of TCs on the ACs was mainly due to
Available online 26 September 2015 ␲–␲ interactions. The rate of adsorption of TCs was interpreted using kinetic models along
with diffusional models. The pseudo-first-order and pseudo-second-order were fitted to the
Keywords: experimental concentration decay curves of the TCs for the adsorption of TCs on ACs. The
Activated carbon first-order kinetic model matched reasonably well the experimental concentration decay
Adsorption rate data, but the rate constant, k1 , considerably decreased with time. Thus, the rate of adsorp-
Diffusional models tion of TCs on ACs cannot be interpreted by the first-order kinetic model. The pore volume
Surface diffusion diffusion model (PVDM) and the pore volume and surface diffusion model (PVSDM) were
Tetracyclines also applied to interpret the rate of adsorption of TCs. The PVDM overpredicted the experi-
mental concentration decay data indicating that intraparticle diffusion was due to both pore
volume and surface diffusion mechanisms. The PVSDM fitted quite well the experimental
concentration decay data of the TCs on both ACs, showing that the intraparticle diffusion
of TCs is due to the pore volume diffusion, as well as the surface diffusion. Furthermore,
the contribution of surface diffusion is directly dependent on the adsorption capacity of the
carbons because the concentration of TCs adsorbed on the surface is the driving force of
surface diffusion. Additionally, the contribution of surface diffusion is affected by the time
and radial position.
© 2015 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

1. Introduction manufacture and formulation of drugs, animal farms and dis-

Tetracyclines (TCs) are extensively used for human therapy
and the livestock industry for their antimicrobial activity
against a wide variety of bacteria (Mathers et al., 2011; Gao
et al., 2012a). Due to their low cost, TCs are also used as a
food additive to enhance the growth rate of animals. TCs are
released into the environment by water discharges from the
posal of unused or expired pharmaceutical products. Most
of the TCs (50–80%) administered to livestock are excreted
through the feces and urine as unmodified parent compounds.
As a result, TC residues are commonly found in wastewater
treatment plants (Lin et al., 2009).
TCs have been detected at concentrations ranging from
0.11 to 4.20 ␮g/L in surface waters, while concentrations in

∗
Corresponding author. Tel.: +52 4448132157.
E-mail address: raul.ocampo@uaslp.mx (R. Ocampo-Pérez).
http://dx.doi.org/10.1016/j.cherd.2015.09.011
0263-8762/© 2015 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
580 chemical engineering research and design 1 0 4 ( 2 0 1 5 ) 579–588

et al., 2007). Batt et al. (2007) and Ternes et al. (2002) found
Nomenclature that biodegradation and chlorination processes only slightly
degrade TCs, explaining their presence in the effluents from
B constant of the Langmuir isotherm related to treatment plants. It is important to point out that traces of
the heat of adsorption (L/mg) these antibiotics in the environment can make microorgan-
CA concentration of tetracycline in aqueous solu- isms more resistant to these antibiotics, and humans may be
tion (mg/L) exposed to antibiotics via drinking water.
CA0 initial concentration of tetracycline in solution The US Environmental Protection Agency recommended
(mg/L) the adsorption on activated carbon as the best available
CAe final concentration of tetracycline at equilib- technology for removing non-biodegradable toxic organic
rium in solution (mg/L) compounds from drinking water and industrial wastewa-
CA,r concentration of tetracycline in the pore vol- ter (USEPA, 1991). The TCs can be removed from aqueous
ume at a distance r (mg/L) solutions by adsorption on different adsorbents such as
CA |r=Rp concentration of tetracycline in the solution at MnFe2 O4 /activated carbon composites (Shao et al., 2012),
the external surface of GAC (r = Rp ) (mg/L) clays (Chang et al., 2012), graphene oxide (Gao et al., 2012b),
dp average particle diameter of GAC (cm) multiwalled carbon nanotubes (Zhang et al., 2011) and sludge-
DAB molecular diffusion coefficient at infinite dilu- derived adsorbents (Rivera-Utrilla et al., 2013).
tion (cm2 /s) The adsorption mechanism of TCs on carbonaceous mate-
Dep effective pore volume diffusion coefficient rials is mainly governed by dispersive interactions between
(cm2 /s) the ␲-electrons in the graphene layers of the carbon materials
Ds effective surface diffusion coefficient (cm2 /s) and the ␲-electrons in the aromatic rings of TCs (Rivera-Utrilla
k1 rate constant for the first-order reaction (h−1 ) et al., 2013). This type of interaction can be also affected by
k2 rate constant of the second-order kinetic model the electrostatic adsorbent–adsorbate interactions when the
(g/(mg h)) adsorption is occurring at pH values causing TC protona-
Kf parameter of Freundlich isotherm tion (Rivera-Utrilla et al., 2013). In a previous study, the high
(mg1−1/n L1/n /g) adsorption capacity of sludge-derived adsorbents toward TC
kL external mass transfer coefficient (cm/s) was attributed to an elevated content of metallic species and
m mass of GAC (g) the affinity of TCs to form complexes with the metals on the
N number of experimental data surface of these materials (Rivera-Utrilla et al., 2013). It was
n parameter of the Freundlich isotherm also found that the presence of bacteria reduced the adsorp-
qe mass of tetracycline adsorbed at equilibrium tion/bioadsorption of TCs on activated carbon by weakening
per mass of GAC (mg/g) the interactions between the adsorbate and the biofilm formed
qe,exp experimental mass of tetracycline adsorbed per on the carbon surface.
mass of GAC (mg/g) The overall adsorption rate of TCs on sludge-derived
qe,pred mass of tetracycline adsorbed per mass of GAC adsorbents was interpreted using kinetic models (first-order,
predicted with the isotherm model (mg/g) second-order and Langmuir) and diffusional models (pore vol-
q mass of tetracycline adsorbed per mass of GAC ume diffusion and surface diffusion) (Ocampo-Pérez et al.,
at time t (mg/g) 2012a). The first-order kinetic model provided the best inter-
r distance in radial direction of GAC (cm) pretation of TC adsorption rate on all adsorbents, and the rate
Rp radius of the particle (cm) constant was linearly dependent upon the macropores and
S external area of the solid particles per adsor- mesopores volume of the adsorbents. Additionally, the intra-
bent mass (cm2 /g) particle diffusion controlled the TC adsorption rate and the
t time (s or min) contribution of pore volume diffusion to the total intraparticle
T temperature (K) diffusion represented more than 80%. Hence, surface diffusion
V volume of the tetracycline solution (mL or L) did not play a significant role in the intraparticle diffusion of
VP pore volume per unit mass of GAC (cm3 /g) TC in the sludge-derived adsorbents.
Xm maximum mass of tetracycline adsorbed onto The design of an adsorption column requires informa-
GAC (mg/g) tion on both the adsorption rate and the adsorption capacity
of the adsorbent (adsorption isotherm). The adsorption rate
Greek letters
in porous solids can be predicted by kinetic and diffusional
εp void fraction of GAC
models. In the latter models, the overall adsorption rate is
A dimensionless concentration of tetracycline in
controlled by pore volume diffusion, surface diffusion or both
the solution
(Leyva-Ramos et al., 2009). The dependence of the porous
exp experimental dimensionless concentration of
structure on the adsorption rate has been also incorporated
tetracycline in the solution
in some diffusional models (Borrelli et al., 1996; Yang and Al-
pred dimensionless concentration of tetracycline in
Duri, 2001).
solution predicted with the model
The adsorption mechanism of TCs on different adsorbents
p particle density of GAC (g/cm3 )
has been extensively studied, but no work has been reported
on the diffusion mechanism controlling the intraparticle dif-
fusion of TCs in activated carbons.
effluents from wastewater treatment plants ranged from 46 The main objective of this work was to interpret the overall
to 1300 ng/L for tetracycline, 270 to 970 ng/L for chlortetracy- adsorption rate of three TCs (tetracycline, oxytetracycline and
cline, and 240 ng/L for oxytetracycline (Gao et al., 2012a; Lin chlortetracycline) on commercial activated carbons by using
et al., 2009; Yang et al., 2005; Ternes et al., 2002; Stackelberg diffusional and kinetic models. Besides, the overall rate of
 chemical engineering research and design 1 0 4 ( 2 0 1 5 ) 579–588 581

adsorption was also investigated to find out the intraparti- This equation can be also formulated in terms of CA and
cle diffusion mechanism controlling the overall rate, and the CA0 , yielding the following equation:
influence of the physicochemical properties of the TCs on the
 m  q2 k t
parameters for the kinetic and diffusional models was also e 2
CA = CA0 − (8)
studied in detail. V 1 + qe k2 t

3. Diffusional models
2. Kinetic models
The diffusional model used in this work is based on the follow-
In the kinetic models, the overall adsorption rate is con-
ing assumptions: (i) the intraparticle diffusion occurs by pore
trolled by the adsorption rate on the surface of the adsorbent,
volume diffusion (Fick diffusion) and surface diffusion, (ii)
and the intraparticle diffusion and external mass transfer are
the rate of adsorption on an active site is instantaneous, and
neglected. It is also assumed that the adsorption rate of a
(iii) the carbon particles are spherical. The model equations
solute on the adsorbent surface can be represented in the
and initial and boundary conditions are the following (Leyva-
same manner as the rate of a chemical reaction. Adsorp-
Ramos and Geankoplis, 1985; Ocampo-Pérez et al., 2010):
tion rate on several adsorbents is commonly interpreted using
first- and second-order kinetic models.
dCA  
V = −mSkL CA − CAr |r=Rp (9)
dt
2.1. First-order kinetic model
t = 0 CA = CA0 (10)
Lagergren advanced the first-order kinetic model to predict
∂CAr ∂q 1 ∂
 
∂CAr ∂q

the adsorption rate of oxalic and malonic acids onto char- εp + p = 2 r2 Dep + DS p (11)
∂t ∂t r ∂r ∂r ∂r
coal (Ho and Mckay, 1998). The first-order kinetic model has
been extensively applied to interpret the adsorption rate of CAr = 0 t = 0 0 ≤ r ≤ Rp (12)
solutes on different adsorbents (Srivastava et al., 2006). The

first-order kinetic model can be mathematically represented ∂CAr 
 =0
∂r r=0
(13)
by the following equation:
  
∂CAr  ∂q
dqt
= k1 (qe − qt ) (1)
Dep  + DS p ∂r = kL CA − CAr |r=Rp
∂r r=R
(14)
dt
Eqs. (9)–(14) represent the pore volume and surface dif-
This equation can be integrated using the initial condition
fusional model (PVSDM). The parameters kL , Ds and Dep
qt = 0 when t = 0, and the resulting equation is
correspond to external mass transport, surface diffusion and
pore volume diffusion mechanisms, respectively. The PVSDM
qt = qe (1 − e−k1 t ) (2) model can be simplified by considering that the intraparticle
diffusion mechanism may be exclusively due to either pore
The above equation can also be expressed in terms of CA volume diffusion (PVDM) (Dep = / 0, Ds = 0) or surface diffusion
and CAe by using the mass balance equation at time t and (SDM) (Dep = 0, Ds =/ 0).
equilibrium as follows: If the adsorption rate on an active site is considered to be
instantaneous so that there is local equilibrium between the
V (CA0 − CA ) TC concentration in the solution within the pore of the carbon
qt = (3)
m and the mass of TC adsorbed on the surface of the pore. This
equilibrium relationship between CAr and q is represented by
V (CA0 − CAe ) the adsorption isotherm:
qe = (4)
m
q = f (CAr ) (15)
These mathematical relationships can be substituted into
Eq. (2), and the following equation can be obtained:
4. Experimental methods

CA = CAe + (CA0 − CAe ) e−k1 t (5)

4.1. Reagents

2.2. Second-order kinetic model Tetracycline (TCH2 ), oxytetracycline (OTC) and chlortetracy-
cline (CTC) were supplied by Sigma-Aldrich. The chemical
The second-order kinetic model can be represented by the structures of the TCs are depicted in Fig. 1(a), and the phys-
following differential equation (Blanchard et al., 1984): icochemical properties of TCs are summarized in Table 1.
Additionally, the speciation diagram of TCH2 as a function of
dqt solution pH is shown in Fig. 1(b).
= k2 (qe − qt )2 (6)
dt
4.2. Adsorbents
Integrating Eq. (6) and using the initial condition qt = 0 when
t = 0, the following equation can be obtained: The two commercial granular activated carbons (GACs) used
in this work were Sorbo Norit (S) and Merck (M). The particle
q2e k2 t sizes of the GACs ranged 0.60–1.00 mm. The GAC samples
qt = (7)
1 + qe k2 t were characterized by determining their surface area, pore
582 chemical engineering research and design 1 0 4 ( 2 0 1 5 ) 579–588

1.0
TC2-
0.9 TCH3+ TCH2
pKa,2/pKa,3 TCH -
0.8
H3C CH3 Tetracycline (TCH2)
C22H24N2O8 0.7
H CH3 H N
R1. H

Molar fraction
HO
H H 0.6
OH R2. H
Oxytetracycline (OTC) 0.5
C22H23N2O9
R1. H 0.4
NH2 R2. OH
Chlorotetracycline (CTC)
0.3
OH C22H24N2O8 0.2
OH O OH O O R1. Cl
R2. H 0.1
pKa,3/pKa,2 pKa,1
0.0
0 2 4 6 8 10 12 14
Solution pH

(a) (b)
Fig. 1 – Chemical structure of TCs (a) and speciation diagram of TCH2 as a function of solution pH (b).

volume accessible to water, pore size distribution, oxygen
surface groups and pH of the point of zero charge (PZC). The
techniques and methods used for the characterization were
previously described in detail (Bautista-Toledo et al., 2008).
4.3. Aqueous solutions of TCs and determination of
concentration of TCs
All TCs solutions were prepared by using ultrapure water
obtained from Milli-Q® equipment (Millipore). The concen-
trations of TCs in aqueous solutions were quantified using a
spectrophotometer, Thermo/Milton Roy Spectronic, Genesys
5. The absorbance was determined at a wavelength of
 = 350 nm and the concentration of a sample was estimated
with a calibration curves, concentration of TCs vs. absorbance.
4.5. Procedure for obtaining the adsorption kinetic
data
The adsorption kinetic data were obtained by adding 0.1 g of
GAC in Erlenmeyer flasks containing 100 mL of a TCs solu-
tion with an initial concentration of 700 mg/L. The flasks were
placed in a thermostatic bath under agitation at 298 K for eight
days (time to reach equilibrium), and samples were periodi-
cally extracted to determine the concentration of TC in the
solution and the amount of TC adsorbed as a function of time.
The volume of each sample was 0.5 mL, and the total variation
of the solution volume due to sampling was less than 7.5%.
5. Results and discussion
5.1. Characterization of activated carbons

The textural properties of the two commercial GACs used in

4.4. Procedure for obtaining the adsorption
equilibrium data
The adsorption equilibrium data were obtained by a batch
mode method. A portion of 0.1 g of GAC and 100 mL of an
aqueous solution of TCs with an initial concentration ran-
ging from 100 to 1000 mg/L were added into a flask, which
was placed in a thermostatic bath to maintain the temper-
ature constant at 25 ◦ C. Enough time (eight days) was allowed
to reach equilibrium and the concentration of TC at equilib-
rium was determined as described earlier. In the adsorption
experiments, the solution pH was not kept constant, but var-
ied slightly between 4 and 5. The mass of TCs adsorbed was
estimated by performing a mass balance of TCs represented
by Eq. (3).
this study are reported in Table 2. Both GACs had large surface
area (>1200 m2 /g) and highly-developed porosity. The exter-
nal surface area (corresponding to pore diameters > 6.6 nm) of
carbon S was slightly larger than that of carbon M.
The micropore volumes W0 (N2 ) for both GACs were con-
siderably higher than the micropore volumes W0 (CO2 ). This
result indicated that a very heterogeneous micropore distri-
bution because CO2 is only adsorbed in smaller micropores
(ultramicropores) while N2 is adsorbed on the surface of all
micropores (Lopez-Ramon et al., 1999). As expected, mean
micropore width L0 (N2 ) was higher than L0 (CO2 ) (Table 2).
The chemical properties of the GAC samples are also shown
in Table 2. Accordingly to the concentrations of basic and
acidic sites, the surface nature of carbons S and M is basic and
slightly basic, respectively. This result was also substantiated

Table 1 – Physicochemical properties of TCs.

TCs Molecular volumea (Å3 ) Cross-sectional areaa (Å2 ) Water solubilityb (mg/L) Log Kow b pKab

TCH2 403.72 396.90 22 −1.30 3.32-7.78-9.58

OTC 413.13 407.49 17 −0.90 3.22-7.46-8.94
CTC 416.90 410.50 4.2 −0.62 3.33-7.55-9.33

a
Determined by applying a Monte-Carlo algorism with the Spartan08 program.
b
Determined by means of the Advanced Chemistry Development (ACD/Labs) Software v8.14 program. pKa corresponding to the successive
deprotonation reactions. Kow = Octanol–water partition coefficient.
 chemical engineering research and design 1 0 4 ( 2 0 1 5 ) 579–588 583

by that fact that the pH of the point of zero charge was

pHPZC
pHPZC = 9.0 for carbon S and pHPZC = 7.7 for carbon M.

9.0
7.7
5.2. Adsorption equilibrium isotherm
Basic groups (meq/g)
The adsorption equilibrium data of the three TCs on the two
1.08 GACs were obtained to assess the capacity of the ACs for
0.44
adsorbing the TCs. The Langmuir and Freundlich isotherms
were fitted to the adsorption equilibrium data, and the equa-
tions representing these isotherm models are the following:

qm kCAe
qe = (16)
Acidic groups (meq/g)

1 + kCAe

qe = kCAe
1/n (17)
0.45
0.40

The average percentage deviation was chosen as the good-

ness of fit criterion and was appraised from the subsequent
equation:

 
V3 h (cm3 /g)

1  qe, exp − qe, pred 

N

%D =   × 100% (18)
0.48
0.28

N qe, exp
i=1

The parameters of the adsorption isotherm were evalu-

V2 g (cm3 /g)

ated by a least-squares method, and their values are listed

in Table 3. Even though, both models interpreted the experi-
0.04
0.10

mental data quite well; the Langmuir isotherm better fitted the
experimental data because the values of %D for the Langmuir
isotherm were smaller than those for the Freundlich isotherm.
Sext f (m2 /g)

The Langmuir isotherms and the experimental adsorption

46.90
41.90

equilibrium data of TCs on carbons M and S are plotted in

Fig. 2(a and b), respectively.
The values of the maximum adsorption capacity of the GAC
toward TCs (qm ) indicate that the adsorption capacity of the
Lo (CO2 )e (nm)

GACs was very high. The values of qm varied from 471.1 mg/g
0.70
0.70

for the TCH2 -carbon M system to 65.1 mg/g for the CTC–carbon
S system. Furthermore, the values of Kqm ranged from 4.20 to
Volume of pores with diameters ranging 6.6–50 nm determined by mercury porosimetry.
Table 2 – Textural and physicochemical properties of the activated carbons.

142.0 revealing that the TCs presented high chemical affinity

toward the carbons M and S. Hence, the TCs can be effec-
Lo (N2 )d (nm)

tively removed from water solutions by adsorption on these

Volume of pores with diameters >50 nm determined by mercury porosimetry.

commercial activated carbons.

1.02
1.69

Volume of micropores determined by N2 and CO2 adsorption, respectively.

Volume of micropores determined by N2 and CO2 adsorption, respectively.

In Table 3 can be noted that the adsorption capacity of car-

bon M toward the three TCs was greater than that of carbon
Mean widths of micropores determined with the Dubinin equation.
Mean widths of micropores determined with the Dubinin equation.

S, and the relative affinity of carbon M toward these TCs was

Surface area determined from N2 adsorption isotherm at 77 K.

also considerably superior compared to those of carbon S. For

Wo (CO2 )c (cm3 /g)

both carbons, the capacity for adsorbing these TCs increased

External surface area determined by mercury porosimetry.

in the following order: CTC < OTC < TCH2 .

0.29
0.29

The adsorption equilibrium experiments were carried out

at the pH range of 4–5 and at this pH range, the surface of
both carbons is positively charged and the molecules of TCs
are either protonated species (cationic molecules) or neutral
Wo (N2 )b (cm3 /g)

species because the pKa values for the protonation reactions

ranged from 3.22 to 3.33, and the pKa values for the first
0.39
0.42

deprotonation reactions varied from 7.46 to 7.78. To illus-

trate the species of TCs present in aqueous solutions, the
speciation diagram of the TCH2 is shown in Fig. 1(b). Then,
there is repulsion between the positively charged GAC sur-
SBET a (m2 /g)

face and the protonated TCs, and there is not electrostatic

attraction between the neutral TCs molecules and the GAC
1225
1301

surface. Hence, the adsorption mechanism of TCs on GAC is

mainly controlled by the dispersive interactions between the
␲-electrons of the aromatic rings of TCs and the aromatic
ACs

rings of the graphene layers of the GACs (Rivera-Utrilla and

M
S

c
d

h
g
f

Sánchez-Polo, 2002).
584 chemical engineering research and design 1 0 4 ( 2 0 1 5 ) 579–588

Table 3 – Parameters of the Langmuir and Freundlich isotherms for the adsorption of TCs (Gomez-Pacheco, 2011).
TCs ACs Langmuir Freundlich

qm (mg/g) Kqm (L/g) %D 1/n k (mg1−1/n L1/n /g) %D

TCH2 S 375.4 52.0 2.8 0.17 178 12.3

M 471.1 142.0 2.6 0.10 263 7.1

OTC S 252.6 6.2 3.1 0.45 19 11.7

M 413.2 15.0 5.2 0.23 100 10.2

CTC S 65.1 4.2 3.6 0.52 3 10.6

M 309.9 10.0 3.2 0.21 86 5.8

The accessibility of the TCs to the GACs surface can affect
the adsorption capacity and the accessibility is dependent
upon the TCs molecular size and pore diameters of the GACs.
Carbon M has larger mean width of micropores and mesopore
volume (Table 2) resulting in greater accessibility to the surface
area and consequently, the adsorption capacity of carbon M for
TCs was higher than that of carbon S. Besides, the adsorption
capacity of both carbons toward these TCs decreased in the
following order TCH2 > OTC > CTC, whereas the molecular size
(see molecular volume and cross-sectional area in Table 1) of
the TCs increased very slightly in the same order. It is impor-
tant to point out that the variation of the molecular size among
the TCs is not significant so that the variation of the adsorption
capacity of GACs due to the nature of TCs cannot be ascribed
to differences in the molecular size of TCs.
5.3. Kinetic models for adsorption of TCs on ACs
The adsorption rate of the TCs on the GACs was studied to
analyze the influence of the chemical characteristics of the
adsorbate on the TC adsorption rate. Fig. 3(a and b) depicts
the dimensionless concentration decay curves, (CA /CA0 ) vs.
time, for the adsorption of three TCs on the carbons M and
S, respectively.
The kinetic models were fitted to the experimental con-
centration decay curves, and the best values of the kinetic
parameters were estimated by a least-squares method. The
adsorption rate constants (k1 , k2 ) are reported in Table 4. The
1.0

0.8

500
Mass of TCs adsorbed, mg/g

0.6
CA/CA0

400

0.4
300

200 0.2 TCH2

OTC
CTC
100 TCH2
OTC
0.0
0 20 40 60 80 100 120 140 160 180 200
CTC

0 Time, h
0 100 200 300 400 500 (a)
Concentration of TCs at equilibrium, mg/L
1.0
(a)

450
TCH2 0.8
400 OTC
CTC
Mass of TCs adsorbed, mg/g

350
0.6
CA/CA0

300
250
0.4
200
150
0.2 TCH2
100 OTC
CTC
50
0.0
0 20 40 60 80 100 120 140 160 180 200
0
0 100 200 300 400 500 Time (h)
Concentration of TCs at equilibrium, mg/L (b)
(b)
Fig. 3 – Concentration decay curves for TCs adsorption
Fig. 2 – Adsorption isotherms of TCs on carbons M (a) and S kinetics on activated carbons M (a) and S (b) at pH = 4, initial
(b) at T = 25 ◦ C and pH range of 4–5 (Gomez-Pacheco, 2011). concentration of TCs = 700 mg/L and T = 298 K. The lines
The lines represent the prediction of the Langmuir represent the prediction of the pseudo-first-order kinetic
isotherm. model.
 chemical engineering research and design 1 0 4 ( 2 0 1 5 ) 579–588 585

Table 4 – Adsorption rate parameters for the pseudo-first and pseudo-second-order models.
TCs ACs First-order Second-order
−1
k1 × 10 (h ) %D k2 × 10 (L mg/h)
3
%D

TCH2 S 0.180 14.2 0.104 22.6

M 0.194 8.2 0.079 35.6

OTC S 0.182 6.4 0.139 33.8

M 0.185 4.7 0.097 23.5

CTC S 0.195 2.6 0.190 17.3

M 0.199 9.6 0.113 42.3

%D was calculated using Eq. (18), and its values are also given indicates that the values of k1 are strongly time-dependent.
in Table 4. The fitting results revealed that the pseudo-first- Similar results were obtained for the other TCs-GAC systems.
order model better fitted the experimental adsorption kinetic Therefore, the overall TC adsorption rate cannot be interpreted
data because the %D for the first-order model were always less using the first-order kinetic model.
than those for the second-order model. In a previous work,
Ocampo-Pérez et al. (2012a) reported that the pseudo-first- 5.4. Diffusional models
order model reasonably well interpreted the rate of adsorption
of TCH2 on various sludge-derived adsorbent materials. The diffusional models have been successfully applied for
Analyzing the values of k1 for any TC, it can be observed interpreting the rate of adsorption of organic compounds from
that the k1 values for carbon M are greater than those for car- aqueous solutions on activated carbon. The PVSDM, PVDM
bon S. Thus, the adsorption rate of TCs on carbon M was faster and SDM models can explain quite well the rate of adsorp-
than that for carbon S (Table 4). This trend can be explained tion of phenolic compounds on GAC. In a previous work, the
recalling that carbon M had larger mean width of micropores overall adsorption rates of various phenolic compounds on
and volume of mesopores (Table 2). two commercial activated carbons were analyzed using the
Nieszporek (2013) proposed a method to find out whether PVDM (Ocampo-Pérez et al., 2012b). The results revealed that
or not the first-order kinetic model can interpret the rate of the contribution of pore volume diffusion represented more
adsorption data. This method is based on dividing the whole than 92% of overall intraparticle diffusion, confirming that
time range into various consecutive time intervals and fitting pore volume diffusion was controlling the overall adsorption
the first-order kinetic model to the experimental data in each rate, and that surface diffusion can be neglected. Similar find-
time interval. The constant rate k1 can be estimated for each ings were reported for the adsorption rate of TCH2 on various
time interval, and if k1 is independent of the time interval sludge-derived adsorbents (Ocampo-Pérez et al., 2012b).
then, the adsorption rate can be represented by the first-order Based on the above results, the experimental concentration
kinetic model. Otherwise, if the values of k1 vary with time, decay curves of TCs were first analyzed using the PVDM model.
the overall rate of adsorption is controlled by intraparticle dif- The TCs concentration decay curves for adsorption on S and
fusion. M samples were predicted with the numerical solution of the
The adsorption rate data were analyzed using the above PVDM. The external mass transfer coefficient (kL ) and effective
procedure, and the mass uptake as function of time, qt vs. time, diffusion coefficients (Dep ) were required to solve the PVDM.
is shown in Fig. 4 for the adsorption kinetic data of TCH2 on Furusawa and Smith (1973) suggested the following equation
carbon M. The values of k1 decreased considerably from 0.0333 to predict the experimental values of kL :
to 0.0199 h−1 while the time was raised. This trend clearly
d 
(CA /CA0 ) −mSkL
= (19)
dt t=0 V
500
k1 = 0.0194 h-1 The experimental values of kL are shown in Table 5.
Mass uptake of TCH2, mg/g

The values of Dep were estimated by the tortuosity factor

400
equation as follows:

300 DAB εp
Dep = (20)

200
As suggested by Leyva-Ramos and Geankoplis (1994), the
k1=0.0333 h-1 tortuosity factors for carbons M and S were assumed to be
100 k1=0.0279 h-1
3.5. The molecular diffusivities of the TCs were calculated
k1=0.0199 h-1
using the correlation proposed by Wilke and Chang (Poling
0
0 20 40 60 80 100 120 140 160 180 200 et al., 2006), and the values of Dep estimated for TCH2 , OTC,
Time, h and CTC were 6.92 × 10−7 , 6.85 × 10−7 , and 6.72 × 10−7 cm2 /s,
respectively.
Fig. 4 – The experimental mass uptake curve of TCH2 The rate of adsorption of TCH2 on carbons M and S were
during adsorption on carbon M. The solid lines represent predicted with the PVDM and using the Dep value estimated
the predictions of the first-order kinetic model for a specific by Eq. (20). The experimental and the predicted concentration
time interval and dashed line represents the prediction of decay curves for the adsorption of TCH2 on carbon S and M are
the first-order kinetic model for the whole time range. depicted in Fig. 5(a and b), respectively. In both carbons, the
586 chemical engineering research and design 1 0 4 ( 2 0 1 5 ) 579–588

Table 5 – Mass transfer parameters for the adsorption of TCs on ACs at T = 25 ◦ C, V = 0.1 L and m = 0.1 g.
TCs DAB × 106 (cm2 /s) ACs kL × 107 (cm/s) Ds × 109 (cm2 /s) %Da

TCH2 4.87 S 3.0 1.16 8.23

M 2.7 26.0 6.35

OTC 4.85 S 3.2 2.60 9.69

M 2.0 2.16 7.32

CTC 4.83 S 1.9 0.13 4.89

M 2.4 1.82 3.24

a
Average percentage deviation obtained by using Eq. (18) with CA,exp and CA,pred instead qe,exp and qe,pred , respectively.

PVDM considerably overpredicted the concentration decay of optimal value of Ds was obtained by minimizing the following
TCH2 . This result implies that the rate of adsorption of TCH2 objective function:
is much faster than that predicted with the PVDM. Similar
1 
N
results were observed for the adsorption rates of OTC and 2
CTC on both ACs. These findings suggest that pore volume Minimum = CA, exp − CA, pred (21)
N
diffusion is not the only intraparticle diffusion mechanism i=1

occurring in the adsorption rate of TCs.

In the PVSDM model, the intraparticle diffusion is assumed
to be due to both the pore volume diffusion and surface dif-
fusion. The values of Dep and kL were estimated as argued
earlier. Hence, the surface diffusion coefficient, Ds , was the
sole unknown mass transfer parameter and can be calculated
by fitting the numerical solution of the PVSDM model to the
experimental concentration decay curve data. In this case, the
1.0
0.8
0.6
CA/CA0
The concentration decay curves predicted with the PVSDM
are plotted in Fig. 5 (a and b) for the adsorption of TCH2 on
carbons M and S, respectively. As it can be noted in these fig-
ures, the PVSDM model satisfactorily fitted the experimental
concentration decay data. Similar results were observed for
the adsorption rate of the OTC and CTC on both carbons. The
optimal values of Ds are given in Table 5 and increased in
the following order: Ds,OTC > Ds,TCH2 > Ds,CTC for carbon S and
Ds,TCH2 > Ds,OTC > Ds,CTC for carbon M. The Ds values for carbon
M were almost one order of magnitude higher than those for
carbon S, except in the case of OTC.
The contribution of each diffusion mechanism to the intra-
particle mass transfer of the TCs was estimated by computing
the mass flux due to the pore volume diffusion, NAP , and sur-

face diffusion, NAS , using the following equations:

0.4 ∂CAr
NAP = −Dep (22)
∂r
0.2 Experimental data
PVDM, Dep = 6.92×10-7 cm2/s, τ = 3.5 ∂q
PVSDM, Dep = 6.92×10-7 cm2/s, Ds = 2.6×10-8 cm2/s
NAS = −Ds p (23)
∂r
0.0
0 30 60 90 120 150 180 Besides, the percentage contribution of surface diffusion
Time, h to the intraparticle mass transfer can be calculated with the
(a) subsequent equation:
1.0
NAS Ds p (∂q/∂r)
× 100% = × 100%
(NAS + NAP ) Ds p (∂q/∂r) + Dep (∂CAr /∂r)
0.9
(24)
0.8
CA/CA0

0.7 At various times, the percentage contribution of surface

diffusion during adsorption of TCH2 on carbon S as a func-
0.6 tion of the particle radius is depicted in Fig. 6. As shown in
this figure, the contribution of surface diffusion was markedly
Experimental data
0.5
PVDM, Dep = 6.92×10-7 cm2/s, τ = 3.5
dependent on the time and the radial position in the parti-
PVSDM, Dep = 6.92×10-7 cm2/s, Ds = 1.16×10-9 cm2/s cle. For example, at a time of 33 h, the pore volume diffusion
0.4 of TCH2 is the main intraparticle diffusion mechanism at the
0 30 60 90 120 150 180
entrance of the pore (750 < particle radius < 800 ␮m); however,
Time, h
the surface diffusion of TCH2 was almost the only mechanism
(b)
for particle radius less than 600 ␮m.
Fig. 5 – Concentration decay curves of TCH2 for adsorption The contribution of surface diffusion was further analyzed
on carbon M (a) and carbon S (b). The lines represent the by calculating the radial average of the percentage contri-
predictions of PVDM and PVSDM models. Initial bution of surface diffusion (%RACSurfDiff) at a certain time,
concentration of TCH2 = 700 mg/L, pH = 4, and T = 298 K. which was estimated by the following equation:
 chemical engineering research and design 1 0 4 ( 2 0 1 5 ) 579–588 587

are important since %RACSurfDiff ranged from 4 to 100%

Fig. 6 – Percentage contribution of surface diffusion to the
total intraparticle diffusion at various times and radius
during TCH2 adsorption on carbon S.
for carbon S and from 15 to 100% for carbon M. It can also
be noticed that the %RACSurfDiff decreased in the follow-
ing order: TCH2 > OTC > CTC, which is directly related to the
adsorption capacity of the carbons toward these TCs. In other
words, the higher the adsorption capacity, the greater is the
contribution of surface diffusion, %RACSurfDiff. The surface
diffusion contribution for carbon M was higher than that for
carbon S, and this trend is largely attributed to the higher
adsorption capacity of the former, producing higher concen-
tration gradients at the carbon surface and increasing the
driving force for surface diffusion. The contribution of pore
volume diffusion for carbon S was greater than that for carbon
M. This result can be explained considering that the carbon M
has larger volume of pores with the pore diameters greater
than 50 nm, avoiding the influence of restrictive effects at the
entrance of pores on the diffusion of the CTs.

6. Conclusions

0
4 ((Ds p (∂q/∂r)) / (Ds p (∂q/∂r) + Dep (∂CAr /∂r))) r2 dr The first-order kinetic model better fitted the experimen-
r
%RACSurfDiff =
4/3RP3 tal data for the adsorption rates of TCs on ACs than the
second-order kinetic model; however, the rate constant, k1 ,
× 100% (25)
was dependent on time. The values of k1 for carbon M were
greater than those for carbon S because the adsorption rates
Fig. 7(a and b) depicts the %RACSurfDiff as a function of of the TCs were faster in carbon M than that in carbon S.
time for the three TCs on carbons S and M, respectively. As The overall adsorption rate of TCs on the activated carbons
seen in these figures, both intraparticle diffusion mechanisms is controlled by the intraparticle diffusion. The concentration
decay data of the TCs during the adsorption on both ACs
100 was interpreted quite well using the PVSDM model, and it
was demonstrated that the pore volume diffusion and sur-
face diffusion are important in the adsorption rate of the TCS .
80
The contribution of surface diffusion is directly related to the
adsorption capacity of the ACs. In general, the higher the
%RACSurfDiff

60 adsorption capacity, the greater is the contribution of surface

diffusion.
40
Acknowledgements
20 TCH2
OTC Dr. Raul Ocampo Pérez acknowledges the support provided
CTC
by PROMEP. Dr. Jose Valente Flores Cano acknowledges the
0
0 20 40 60 80 100 120 140 160 180 financial support provided by Consejo Nacional de Ciencia y
Time, h Tecnologia, CONACyT, Mexico, through a postdoctoral fellow-
(a) ship No. MOD-ORD-12-2013PCI-0721113.

100
References
80
Batt, A.L., Kim, S., Aga, D.S., 2007. Comparison of the occurrence
of antibiotics in four full-scale wastewater treatment plants
%RACSurfDiff

60 with varying designs and operations. Chemosphere 68,

428–435.
Bautista-Toledo, M.I., Méndez-Díaz, J.D., Sánchez-Polo, M.,
40
Rivera-Utrilla, J., Ferro-García, M.A., 2008. Adsorption of
sodium dodecylbenzenesulfonate on activated carbons:
20 TCH2
effects of solution chemistry and presence of bacteria. J.
OTC Colloid Interface Sci. 317, 11–17.
CTC Blanchard, G., Maunaye, M., Martin, G., 1984. Removal of heavy
0 metals from waters by means of natural zeolites. Water Res.
0 20 40 60 80 100 120 140 160 180
18, 1501–1507.
Time, h Borrelli, S., Giordano, M., Salatino, P., 1996. Modelling
(b) diffusion-limited gasification of carbons by branching pore
models. Chem. Eng. J. Biochem. Eng. J. 64, 77–84.
Fig. 7 – The radial average of the percentage contribution of Chang, P., Li, Z., Jean, J., Jiang, W., Wang, C., Lin, K., 2012.
surface diffusion %RACSurfDiff as function of time for Adsorption of tetracycline on 2:1 layered non-swelling clay
carbons M (a) and S (b). mineral illite. Appl. Clay Sci. 67–68, 158–163.
588 chemical engineering research and design 1 0 4 ( 2 0 1 5 ) 579–588
Furusawa, T., Smith, J.M., 1973. Fluid-particle and intraparticle
mass transport rates in slurries. Ind. Eng. Chem. Fundam. 12,
197–203.
Gao, P., Mao, D., Luo, Y., Wang, L., Xu, B., Xu, L., 2012a. Occurrence
of sulfonamide and tetracycline-resistant bacteria and
resistance genes in aquaculture environment. Water Res. 46,
2355–2364.
Gao, Y., Li, Y., Zhang, L., Huang, H., Hu, J., Shah, S.M., Su, X.,
2012b. Adsorption and removal of tetracycline antibiotics
from aqueous solution by graphene oxide. J. Colloid Interface
Sci. 368, 540–546.
Gomez-Pacheco, V.C., 2011. Eliminación de Tetraciclinas de las
Aguas Mediante Procesos Avanzados de Oxidación, Carbones
Activados y Adsorbentes Obtenidos a Partir de lodos de
Depuradora. Universidad de Granada (Tesis Doctoral).
Ho, Y.S., Mckay, G., 1998. Kinetic models for the sorption of dye
from aqueous solution by wood. Process Saf. Environ. Prot. 76,
183–191.
Leyva-Ramos, R., Geankoplis, C.J., 1985. Model simulation and
analysis of surface diffusion in liquids in porous solids. Chem.
Eng. Sci. 40, 799–807.
Leyva-Ramos, R., Geankoplis, C.J., 1994. Diffusion in liquid-filled
pores of activated carbon. I. Pore volume diffusion. Can. J.
Chem. Eng. 72, 262–271.
Leyva-Ramos, R., Bernal-Jacome, L.A., Mendoza-Barron, J.,
Hernandez-Orta, M.M.G., 2009. Kinetic modeling of
pentachlorophenol adsorption onto granular activated
carbon. J. Taiwan Inst. Chem. Eng. 40, 622–629.
Lin, A.Y., Yu, T., Lateef, S.K., 2009. Removal of pharmaceuticals in
secondary wastewater treatment processes in Taiwan. J.
Hazard. Mater. 167, 1163–1169.
Lopez-Ramon, M.V., Stoeckli, F., Moreno-Castilla, C.,
Carrasco-Marin, F., 1999. On the characterization of acidic and
basic surface sites on carbons by various techniques. Carbon
37, 1215–1221.
Mathers, J.J., Flick, S.C., Cox Jr., L.A., 2011. Longer-duration uses of
tetracyclines and penicillins in U.S. food-producing animals,
indications and microbiologic effects. Environ. Int. 37,
991–994.
Nieszporek, K., 2013. The balance between diffusional and
surface reaction kinetic models: a theoretical study. Sep. Sci.
Technol. 48, 2081–2089.
Ocampo-Pérez, R., Leyva-Ramos, R., Alonso-Davila, P.,
Rivera-Utrilla, J., Sanchez-Polo, M., 2010. Modeling adsorption
rate of pyridine onto granular activated carbon. Chem. Eng. J.
165, 133–141.
Ocampo-Pérez, R., Rivera-Utrilla, J., Gómez-Pacheco, C.,
Sánchez-Polo, M., López-Peñalver, J.J., 2012a. Kinetic study of
tetracycline adsorption on sludge-derived adsorbents in
aqueous phase. Chem. Eng. J. 213, 88–96.
Ocampo-Pérez, R., Abdel daiem, M.M., Rivera-Utrilla, J.,
Méndez-Díaz, J.D., Sánchez-Polo, M., 2012b. Modeling
adsorption rate of organic micropollutants present in landfill
leachates onto granular activated carbon. J. Colloid Interface
Sci. 385, 174–182.
Poling, B.E., Prausnitz, J.M., O’Connell, J.P., 2006. The properties of
gases and liquids, 5a. Edición. McGraw-Hill, New York, NY.
Rivera-Utrilla, J., Sánchez-Polo, M., 2002. The role of dispersive
and electrostatic interactions in the aqueous phase
adsorption of naphthalenesulphonic acids on ozone-treated
activated carbons. Carbon 40, 2685–2691.
Rivera-Utrilla, J., Gómez-Pacheco, C.V., Sánchez-Polo, M.,
López-Peñalver, J.J., Ocampo-Pérez, R., 2013. Tetracycline
removal from water by adsorption/bioadsorption on activated
carbons and sludge-derived adsorbents. J. Environ. Manage.
131, 16–24.
Shao, L., Ren, Z., Zhang, G., Chen, L., 2012. Facile synthesis,
characterization of a MnFe2 O4 /activated carbon magnetic
composite and its effectiveness in tetracycline removal.
Mater. Chem. Phys. 135, 16–24.
Srivastava, V.C., Mall, I.D., Mishra, I.M., 2006. Equilibrium
modelling of single and binary adsorption of cadmium and
nickel onto bagasse fly ash. Chem. Eng. J. 117, 79–91.
Stackelberg, P.E., Gibs, J., Furlong, E.T., Meyer, M.T., Zaugg, S.D.,
Lippincott, R.L., 2007. Efficiency of conventional
drinking-water-treatment processes in removal of
pharmaceuticals and other organic compounds. Sci. Total
Environ. 377, 255–272.
Ternes, T.A., Meisenheimer, M., McDowell, D., Sacher, F., Brauch,
H., Haist-Gulde, B., Preuss, G., Wilme, U., Zulei-Seibert, N.,
2002. Removal of pharmaceuticals during drinking water
treatment. Environ. Sci. Technol. 36, 3855–3863.
USEPA, 1991. Granular activated carbon treatment. In: Report
EPA-540/2-91/024 US. Environmental Protection Agency U.S.
Government Printing Office, Washington, DC.
Yang, X., Al-Duri, B., 2001. Application of branched pore diffusion
model in the adsorption of reactive dyes on activated carbon.
Chem. Eng. J. 83, 15–23.
Yang, S., Cha, J., Carlson, K., 2005. Simultaneous extraction and
analysis of 11 tetracycline and sulfonamide antibiotics in
influent and effluent domestic wastewater by solid-phase
extraction and liquid chromatography–electrospray ionization
tandem mass spectrometry. J. Chromatogr. A 1097, 40–53.
Zhang, L., Song, X., Liu, X., Yang, L., Pan, F., Lv, J., 2011. Studies on
the removal of tetracycline by multi-walled carbon
nanotubes. Chem. Eng. J. 178, 26–33.