Journal of the Taiwan Institute of Chemical Engineers 52 (2015) 109–117

Contents lists available at ScienceDirect

Journal of the Taiwan Institute of Chemical Engineers

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

Adsorption of carbon dioxide using activated carbon impregnated with

Cu promoted by zinc
Soraya Hosseini a, Iman Bayesti a,b, Ehsan Marahel a,b, Farahnaz Eghbali Babadi a,
Luqman Chuah Abdullah a,c, Thomas S.Y. Choong a,c,∗
a
Department of Chemical and Environmental Engineering, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia
b
Pars Oil and Gas Company, Assaluyeh, Bushehr, Iran
c
INTROP, Universiti Putra Malaysia, 43400 UPM Serdang, Selangor, Malaysia

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

Article history: In the present study, modified activated carbon (AC) was used in a fixed-bed column for CO2 adsorption of
Received 24 September 2014 gaseous mixtures. The adsorbents were prepared by impregnation two metals Cu and Zn on the surface of
Revised 10 December 2014
the acid modified AC using a two-stage modification. At the first stage, the samples of AC were pretreated
Accepted 8 February 2015
by oxidizing agent (nitric acid) to increase the amount of oxygen surface groups and at the second stage,
Available online 4 March 2015
the acid modified AC were impregnated by two metal salts Cu and Zn on the surface to produce a superior
Keywords: CO2 adsorbent. Metal-loaded acid modified AC was prepared by using different ratios of Cu/Zn ranging from
CO2 adsorption 4 to 20%. The CO2 adsorbed have been measured over the temperature range of (30–50 °C), pressure (100–
Modified activated carbon 200 kPa) and CO2 concentrations from 5 to 50%. An increase of 49% CO2 adsorbed was resulted by using
Metal impregnation (Cu/Zn) modified activated carbon. The breakthrough curves indicated that the breakthrough time increased with
Deactivation model increasing the operating pressure, and decreased with increasing the temperature from 30 to 50 °C and CO2
concentration from 5 to 50%. The deactivation model was successfully applied to analyze the breakthrough
curves under various operating conditions.
© 2015 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

1. Introduction
The world is faced with the challenge of global warming and cli-
mate change issues due to increasing concentration of greenhouse
gases in the atmosphere [1,2]. Carbon dioxide (CO2 ) is one of the
main greenhouse gases responsible [3]. CO2 emission has rapidly in-
creased over the past few decades [4]. The increase of CO2 concentra-
tion has caused global warming and environmental impacts such as
floods, droughts, increased frequency of storms and a steady increase
of global sea levels [5]. Studies reported that the global CO2 emissions
have recently increased by about 30% and consequently, an increase
of temperature from 0.3 to 0.68 °C [6].
Many researches have been carried out in the field CO2 capture by
adsorption process, especially as a post-combustion treatment of flue
gas. Adsorption is a simple method with low cost operation for re-
moving process gases, and vapors from air [7]. A wide variety of solid
adsorbents is used on adsorption process. Activated carbons (ACs)
are versatile adsorbents which are mostly utilized as adsorbents of
liquid, gases and vapors. Their adsorptive properties are due to high
∗
Corresponding author. Tel.: +60389466293; fax: +60386567120.
E-mail address: csthomas@upm.edu.my (T.S.Y. Choong).
surface area, microporous structure and high surface reactivity [8]. Al-
though structural properties of activated carbons are effective on their
adsorptive capacity but specific interactions of adsorbate/adsorbent
play an important role in the adsorption process. Improvements of
adsorptive properties of activated carbons can be achieved by us-
ing chemical and physical methods. Metals loading such as salts of
chromium, copper, iron and zinc on activated carbon has been re-
cently reported [9–12].
Somy et al. [9] studied the effect of impregnation of activated
carbon with two metal oxides Cr2 O and Fe2 O3 on its adsorptive prop-
erties of carbon dioxide. The results indicated that impregnation of
Cr2 O increased the CO2 adsorbed by about 20% in compared to un-
modified activated carbon, while Fe2 O3 was not an effective impreg-
nating species. Co-impregnation of Cr2 O with zinc carbonate hydrox-
ide led to another 5% increase in CO2 capture capacity.
According to literatures, the presence of copper oxide on carbon
materials can increase the adsorption capacity of the acidic adsor-
bate [13,14], indicating that copper oxides serve as an electron donor.
Fenrong et al. [13] and Kim et al. [14] reported that activated carbon
modified by copper showed a higher adsorption capacity compared
to that of the activated carbon. Somy et al. indicated co-impregnation
Cr2 O/Zn on activated carbon enhanced CO2 adsorption. From tox-
icity viewpoint, Cr proved significantly more toxic than Cu. Thus,

http://dx.doi.org/10.1016/j.jtice.2015.02.015
1876-1070/© 2015 Taiwan Institute of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
110 S. Hosseini et al. / Journal of the Taiwan Institute of Chemical Engineers 52 (2015) 109–117
the present work was aimed at investigating the adsorption of CO2
on the activated carbon impregnated by Cu, Zn and binary mix-
tures of Cu/Zn with different ratios. However, no literature reported
the application of Zn and the mixture (Cu/Zn) for CO2 adsorption.
The cubic structure of copper is more isotropic and adding hexag-
onal structure of zinc forms dendrite morphology (small branches).
Therefore, more dendrite structure is expected by a mixture of Cu/Zn
[15]. The adsorption capacity of CO2 on unmodified AC was com-
pared with the modified AC that was prepared by binary mixtures
of metals (Cu/Zn). The deactivation model can be used to model the
breakthrough curves. The deactivation model (DM) is a simplified
model used to predict the breakthrough curve. In this model, it is as-
sumed that a dense product layer over the surface of the adsorbent
formed and caused a drop in the adsorption rate due to change the
number of active sites and the possible variations in the adsorption
of active sites. The deactivation model has been applied by many
researchers to model the breakthrough curves in CO2 adsorption
[16–19].
2. Experimental
2.1. Chemicals and materials
A commercial activated carbon (Norit@ SA2, decolorizing, Sigma–
Aldrich) was used as unmodified adsorbent. HNO3 (65%) was used
as oxidizing agent for pre-treatment of activated carbon and was
purchased from BIS chemicals company. Most of researchers be-
lieved that the presence of carbon–oxygen surface functional group
(due to oxidizing agent) on ACs surface is one of the most impor-
tant factors that affect deposition of metals. HCl (36.5%), NaHCO3 ,
Na2 CO3 and NaOH were used for Boehm’s titration in order to study
the functional groups of adsorbent surface that was prepared from
Sigma–Aldrich, Malaysia. High purity commercial powder metal ni-
trate, Cu(NO3 )2 3H2 O (99.5%, Friendemann-Schmidt) and sulfate,
ZnSO4 7H2 O (>99%, Sigma–Aldrich) were used as metal oxide pre-
cursor to modify AC. Carbon dioxide (99.99%) and purified helium
(99.99%) were purchased from Global Plus Solution Company. All
other reagents and chemicals used were of analytical grade or as
reported.
2.2. Modification of adsorbent
The initial stage of preparation method is pre-treatment of AC by
using HNO3 solution. 50 g of AC was mixed with a proper amount of
HNO3 solution (1 g carbon/3 mL HNO3 ). The mixture was continuously
stirred and heated for 2 h at temperature range 90–105 °C in a reflux
system to prevent loss of HNO3 . After oxidation, the AC was washed
several times with distilled water to remove the residual HNO3 and
the sample was then filtered. The sample was then dried in an oven
at 110 °C for 12 h and ready for loading of metals. The samples were
impregnated with Cu(NO3 )2 and ZnSO4 by an aqueous ion-exchange
process. The amounts of metal loading ranged from 2 to 10% of AC for
each pure metal. The AC was first impregnated by using an aqueous
solution of the precursor (Cu(NO3 )2 3H2 O). During the impregnation,
the solution was heated to 70 °C and stirred until the liquid was totally
evaporated. The samples were then dried at 110 °C for 24 h and
following calcination under N2 flow at 500 °C for 4 h. The procedure
was repeated for impregnation using ZnSO4 . The percentage of metal
loading was used to index samples. The prepared samples are (Cu/Zn
(4%), Cu/Zn (8%), Cu/Zn (12%), Cu/Zn (16%) and Cu/Zn (20%)) and the
total metal loading is reported as 4, 8, 12, 16 and 20%. The amount of
metals was equally considered for impregnation. The properties are
summarized in Table 1.
Table 1
Properties of original AC, HNO3 -treated and modified AC samples.
Micropore Mesopore
Sample SBET (m2 /g) volume (cm3 /g) volume (cm3 /g)
AC 1316.13 0.700 0.186
HNO3 -treated AC 1195.32 0.697 0.130
Cu/Zn–4% AC 925.07 0.597 0.122
Cu/Zn–8% AC 850.72 0.550 0.115
Cu/Zn–12% AC 790.05 0.502 0.107
Cu/Zn–16% AC 730.53 0.486 0.102
Cu/Zn–20% AC 599.41 0.399 0.089
Cu–20% AC 645.21 0.421 0.096
Zn–20% AC 621.90 0.407 0.094
2.3. Characterization methods
The prepared samples were characterized by different techniques.
The characterization of chemical structure of AC and modified AC was
performed using FT-IR spectroscopy method. Scanning electron mi-
croscopy (SEM) was used to determine surface morphology of sam-
ples. The element analysis and weight percentage of metals impreg-
nated on modified AC samples were characterized by EDX analysis
and ICP-AES. The impregnated metals amount of the prepared sam-
ple was measured by soaking 1 g of sample in 20 mL of 65% nitric
acid for 24 h. The metal or metal oxide was dissolved from the sup-
port (AC) and the filtrate was diluted in order to determine the metal
concentration using inductivity coupled plasma with atomic emis-
sion spectrophotometer (ICP-AES). The specific surface area and pore
volume of all samples were estimated by BET analysis. Prior to the
start of adsorption process, the sample was degassed at 250 °C for
6 h under vacuum so that its porous volume can be devoid of any
pre-adsorbed species. T-plot method was used to estimate the con-
tribution of surface area for micropores and micropore volume. The
micropore surface area can be subtracted from the overall surface
area measured by BET to evaluate the mesoporous or external surface
area.
Boehm’s titration method was used to determine the amount of
functional groups on the adsorbent surface. 0.5 g of each sample was
mixed with solutions NaOH (0.1 M), NaHCO3 (0.1 M), Na2 CO3 (0.1 M)
and HCl (0.1 M). Each solution contains 100 cm3 (0.1 M) of NaOH,
NaHCO3 , Na2 CO3 , HCl, 0.5 g adsorbent and then was shaken for 48 h.
The mixture was then separated from the aqueous solution by vacuum
filtration. 10 mL of each filtrate was titrated until the pH of solution
reach to the neutral point (pH 7). Samples contain NaOH (0.1 M),
NaHCO3 (0.1 M) and Na2 CO3 (0.1 M) were titrated by using HCl (0.1 M)
whereas NaOH (0.1 M) solution was used for titrating HCl (0.1 M).
2.4. CO2 adsorption procedure
The CO2 adsorption was carried out in a fixed-bed column. The
experimental apparatus used to measure CO2 adsorption is shown in
Fig. 1. The design of CO2 adsorption was carried out in our group
and fabricated by local company (Solution Engineering Sdn Bhd,
Malaysia). The stainless steel fixed-bed reactor was designed with di-
mension 33 cm height, 2.5 cm diameter and equipped with a porous
plate that is located 3 cm from the base of the column. The gas
manifold system consisted of two lines fitted with mass flow con-
trollers from Bronkhorst High-Tech with flows ranging between 1
and 1000 mL/min. The controllers demonstrated an accuracy of 1%
full scale and a repeatability of 0.1% full scale. Two feeds lines contain
CO2 and He was installed in the system. Different gas mixtures can be
prepared by adjusting the concentrations of both gases. Three tem-
perature sensors (type K thermocouple) are placed at the inlet, the
middle and the outlet of the adsorption column for temperature mea-
surement. The thermocouples provided continuously monitoring of
the column temperature with an accuracy of ±1.5 °C. The temperature
 S. Hosseini et al. / Journal of the Taiwan Institute of Chemical Engineers 52 (2015) 109–117 111

500

Volume Adsorbed cm3/g STP

400

300

200

Adsorption
100
Desorption

0
0 0.2 0.4 0.6 0.8 1
Relative Pressure(P/Po)

Fig. 2. N2 adsorption/desorption of HNO3 -treated activated carbon at 77 K.

can be caused the extra diffusion resistance that is expected to pro-

Fig. 1. A schematic of experimental set-up for CO2 adsorption.
was controlled by an electrical resistor coiled around the reactor in-
side the thermal insulation jacket. A back-pressure regulator was in-
stalled to control the column pressure with a repeatability of 0.5% full
scale. The system was also equipped with a CO2 analyzer (GFG, 50%)
at outlet of the column. The CO2 concentration of effluent gas was
continuously monitored as a function of time to obtain the break-
through curves. Before each run, the adsorbent is dried by flowing
He (50 mL/min) for 15 min at 100 °C at atmospheric pressure. Af-
ter the drying step, the bed temperature and pressure were adjusted
and gaseous mixture CO2 /He was fed through the column. The inlet
flow rate was kept constant (250 mL/min). Different operational con-
ditions such as pressures (100–200 kPa), temperatures (30, 40 and
50 °C) and CO2 concentrations (5–50%) were tested.
2.5. Regeneration experiments
After each adsorption run, the column temperature was increased
to the level of 80 °C. The regeneration process was carried out by
a compressor that installed at the end of the column. For all sam-
ples, desorption was completed at a higher temperature and pressure
(1 bar). For each experiment, the temperature was set up at 80 °C
for the desorption process. By purging inert gas (He) into the fixed
bed, the adsorbed CO2 was desorbed from the adsorbent (unmodified
AC and 20% Cu/Zn–AC) and the concentration of the effluent gas was
monitored continuously by online monitoring through CO2 analyzer.
The zero value was appeared on the CO2 analyzer indicator and the
process was then stopped.
2.6. Model description
Yasyerli et al. [20] suggested a deactivation model for the predic-
tion of breakthrough curves in packed columns. They concluded that
the model provided an excellent prediction for H2 S and CO2 break-
through curves. The deactivation model (DM) was subjected in the
sight of adsorption dynamics in order to fit the breakthrough curve
of adsorption process. A dense product layer around absorbent solid
vide a drop in the adsorption rate. The significant changes on textural
structure of adsorbent provide a decrease of vacant surface area of the
adsorbent with time. The effect of all parameters on the decreasing
rate CO2 capture can be considered in a deactivation rate term. The de-
activation model is based on assumptions such as the pseudo-steady
state, elimination of the axial dispersion in the fixed bed and ignor-
ing mass transfer resistances. The zeroth solution of the deactivation
models is given as:
C   
= exp −ks τ exp −kd t (1)
C0
where C0 and C are the inlet and outlet concentrations of CO2 , respec-
tively; parameters ks (m/s), kd (s−1 ) and t (s) are the surface reaction
rate constant, first-order deactivation rate constant and reaction time,
respectively. The surface time τ is defined as the ratio of adsorbent
surface area over volumetric flow rate of gas. Surface time with unit
(s/m) defined as:
WS0
τ= (2)
F
By plotting ln(ln(C/C0 )) versus t, values of ks τ and kd can be derived
from the intercept and slope of the straight line [21,22]. The value of
(C/C0 ) can be determined from the breakthrough curves by using
experimental data. For each experimental, the adsorbed amount of
CO2 on samples was obtained by using the breakthrough curve as
follows:
 tb  
Co F C
q= 1− dt (3)
W 0 C0
where F, W, S0 and tb are amount of the adsorbent (g), volumetric
flow rate of gaseous mixture (cm3 /min), specific initial surface area
of the adsorbent (m2 /g) and breakthrough time (min), respectively.
3. Results and discussion
3.1. Characterization of adsorbent
Nitrogen adsorption/desorption is a standard technique for the
characterization of porous materials. Figs. 2 and 3 showed N2 ad-
sorption/desorption for unmodified AC and modified samples (metal
loaded AC). The isotherms can be classified as type IV isotherm which
is typical of adsorption in micropores and mesopores [22]. The metal
loading was caused a decrease in the surface area of AC. This de-
crease may be due to some of pores blocked by the presence of oxide
functional groups introduced during the HNO3 treatment and load-
ing metals during the modification [23] so that N2 molecules cannot
112 S. Hosseini et al. / Journal of the Taiwan Institute of Chemical Engineers 52 (2015) 109–117

500 Table 2
Total acidic groups for original AC, HNO3 -treated and modified AC samples.

Sample Carboxylic Phenolic Lactonic Total acidity

400
AC 0.36 0.22 0.00 0.58
Volume Adsorbed cm3/g STP

HNO3 -treated AC 1.96 0.49 0.73 3.18

Cu/Zn–4% AC 1.20 0.80 0.64 2.64
300 Cu/Zn–8% AC 0.90 0.95 0.61 2.46
Cu/Zn–12% AC 0.79 1.05 0.60 2.44
Cu/Zn–16% AC 0.75 1.09 0.58 2.42
200 Cu/Zn–20% AC 0.59 1.15 0.55 2.29
Cu–20% AC 0.64 1.10 0.57 2.31
Zn–20% AC 0.71 1.02 0.61 2.34

100 Cu/Zn-20%

0 were appeared at lower frequencies around 750 cm−1 .
0 0.2 0.4 0.6
Relative Pressure(P/Po)
Fig. 3. N2 adsorption/desorption of unmodified (AC) and modified activated carbon
(Cu/Zn-20%) at 77 K.
100
90
80
70
60
50
40
30
20
10
0 from carboxylic acid on the activated carbon surface [11].
4000 3665 3330 2995 2660 2325 1990
wavenumber (cm-1)
Fig. 4. FTIR pattern for unmodified AC (solid line) and acid modified AC (dashed line).
transport into the adsorptive sites. Also commercial ACs may contain
inorganic species which can be removed by HNO3 oxidation. Fig. 3
demonstrates that the volume of N2 adsorbed in sample Cu/Zn–
20% is much less unmodified AC. As seen in Table 1, after treat-
ment, the surface area decreased slightly from 1316 to 1195 m2 /s
and more diminish from 925 to 599 m2 /s was appeared when the
metals were loaded from 4 to 20%. However, no change was oc-
curred in the micropore volume of the activated carbon, suggesting
that the HNO3 treatment does not affect significantly in the microp-
ores [24,25]. The reduction in surface area, total pore volume, micro
and mesopore volumes were dependent on the amount of loaded
metals.
Fig. 4 shows FTIR spectra of both unmodified and modified AC
in the range 600–4000 cm−1 . The wide peaks at 3457 cm−1 and
3451 cm−1 illustrated the expanding vibration of O–H bond involving
hydrogen bonds for unmodified and modified AC, respectively. The
band appearing around 3000 cm−1 is assigned to alkenyl C–H stretch.
The band at 2850 cm−1 is revealed as alkyl C–H stretch. The most sig-
nificant differences in the surface functional groups were appeared
during the acidic treatment and it can be easily seen in the range
of 1600–1800 cm−1 . The appeared peak at 1750 cm−1 in modified
AC spectrum is assigned to stretching vibrations C═O bond in car-
boxylic acid functional group. The bands at 1589 cm−1 , 1374 cm−1 and
1232 cm−1 are appeared due to C═C aromatic, C−OH and the C−O−C
stretch, respectively. Moreover, the intense peak at 1056 cm−1 is as-
cribed to C−O stretching vibrations which may be caused by ether,
AC
phenol, carboxylic acid, and carboxylic anhydride. The metal bands
0.8 1
The Boehm titration is used to describe the principle of oxygen
groups on carbon surfaces which have different degrees of acidity
and can be neutralized by basic of different strengths. Sodium hy-
droxide (NaOH) is generally utilized in order to neutralize all Brøn-
sted acids, while sodium carbonate (Na2 CO3 ) neutralizes carboxylic
and lactonic groups, and sodium bicarbonate (NaHCO3 ) neutralizes
carboxylic acids. Acidify and basicity of all samples was summarized
in Table 2. The unmodified sample demonstrated a small amount of
acidity groups while the acidity groups increased via HNO3 treatment.
Based on Boehm’s titration, the total acidity of the HNO3 treated sam-
ple is 3.18 mmol/g and it is slightly decreased when metals loaded
on the surface AC. Furthermore, from FT-IR spectra of the sample, the
most abundant oxygen groups in the sample are ketone, ester and
ether-type structures. On the other hand, the HNO3 oxidization treat-
ment and then Zn/Cu impregnation of the activated carbons showed
a decrease in carboxylic surface groups. As it is shown in Table 2, this
decrease may be arise due to the ability of Zn and Cu ions in the zinc
sulfate and copper nitrate solutions that is exchanged with H ions
1655 1320 985 650
The fine and aggregated metallic Cu/Zn particles were observed
on the surface of carbon (Fig. 5). As shown in Fig. 5, the aggregated
metallic Cu/Zn particles are clearly seen and loaded on activated car-
bon. Comparing unmodified samples, the modified AC revealed the
distribution of fine Cu/Zn particles. The EDX and ICP results proved
that the content of Cu/Zn on activated carbon increased with increas-
ing the percentage of Cu/Zn in the impregnating solution (ions), as
seen in Table 3. ICP analysis of the samples revealed that the metal
ratios obtained are very close to those predicted from the prepa-
ration (Table 3). The quantitative measurements (ICP-AES) are not
in accordance with those found by EDX. It may be to select fewer
points to find the metals loading on modified activated carbon via EDX
method.
Acid treatment of carbon was employed to oxidize the porous
carbon surface as it increases the acidic property and improves the
hydrophilic nature of surface. Oxygen functional groups containing
proton donors (acidic functional groups) on carbon surface was found
to be highly favorable to link with metal ions to form metal com-
plexes with the negatively charged acid groups. In other word, after
HNO3 treatment (Table 3), more oxygen atoms were formed and the
oxygen functional groups are more polar than the original carbon
surface; which caused to increase the wettability of activated carbon
samples.
A variety of interactions between cationic species (Cu2+ ) and the
activated carbon surface might occur, such as formation of surface
complexes, redox reactions and ion-exchange processes. This process
is considered to be an ion exchange reaction with the participation of
strong acidic surface groups which could be described by the follow-
ing mechanism [26].
AC(nH+ ) + Mn+ (aq) = AC(Mn+ ) + nH+ (aq) (4)
 S. Hosseini et al. / Journal of the Taiwan Institute of Chemical Engineers 52 (2015) 109–117 113

Fig. 5. SEM images of (a) activated carbon (b) modified Cu/Zn-20%.

Table 3
EDX and ICP analysis of original AC, HNO3 -treated and modified AC samples.

Element AC HNO3 -treated Cu/Zn–4% Cu/Zn–8% Cu/Zn–12% Cu/Zn–16% Cu/Zn–20% Cu 20% Zn 20%

Weight % (DEX analysis)

C 79.9 76.01 85.62 82 79.84 77.19 75.36 74.3 78.32
O 16.1 22.92 11.23 11.67 12.16 13.24 10.98 12.29 14.67
Cu 0 0 1.76 2.10 3.97 4.46 6.70 12.43 0.0
Zn 0 0 0.76 3.38 2.83 4.40 6.18 0.0 6.16
Others 3.85 1.07 0.63 0.85 1.20 0.71 0.78 0.98 0.85

Weight % (ICP analysis)

C _ _ _ _ _ _ _ _ _
O _ _ _ _ _ _ _ _ _
Cu 0.00 0.00 1.79 3.49 5.45 7.34 8.93 18.01 0.0
Zn 0.00 0.00 1.68 3.28 5.1 6.68 7.35 0.0 16.38
Others 3.67 2.15 0.94 1.01 1.25 1.78 2.21 2.14 1.96

Here, two hydrogen (H+ ) ions from the surface of the activated car-
bon can be replaced by a copper ion (Cu2+ ), which leads to a decrease
in the acidity of the AC as shown in Table 2. The negatively charged
acid groups on AC and Cu-loaded AC are surrounded with Zn2+ ions.
These ions are adsorbed on to the substrate surface, which allows
surface diffusion and consequently, the reduction of ions at more fa-
vorable sites. The reduction of Zn2+ involves gaining two electrons to
form zero-valent Zn metal deposition the Cu particles.
Zn2+ + 2e− = Zn
The porous structure and surface chemistry govern the adsorption
properties of carbonic adsorbents. CO2 can be physically adsorbed on
the surface sites and pores of porous carbons and the appropriate sur-
face functional groups is usually used to adsorb CO2 via chemisorption
process [27]. Generally, a catalytic surface has surface sites, on each
of which a single gaseous molecule can adsorb. Two metals Cu and
Zn indicated catalytic activity and the surface coverage of CO2 chem-
ically adsorbed on the surface of Cu and Zn of AC. Smykowski et al.
reported analysis of the adsorption process of the CO2 molecule on
the cationic sites Cu, Zn, Ni, Pd over zeolite [28]. They concluded a
chemical structure in which each of the two oxygen atoms of the ad-
sorbing CO2 molecule is assumed to contact Cu or Zn atoms of the
support. The shorter the distance between CO2 and cations, the lower
adsorption energy would be favored in adsorption process. Metals
over acid modified AC were caused to decrease surface area. How-
ever, the porous structure indicated physically to contribute to CO2
capture by physical adsorption. Catalytic activity of two cations can
also considered as a promising parameters of CO2 adsorption over
modified AC.
3.2. CO2 adsorption breakthrough curve
The CO2 adsorption capacity was studied in a fixed-bed setup.
The breakthrough curves were obtained by passing the standard gas
mixture (CO2 /He) at different temperatures 30, 40, and 50 °C, dif-
ferent pressures (100, 150 and 200 kPa), CO2 concentration from 5
to 50% and the gas flow rate was 250 mL/min. The results showed
(5)
that activated carbons would be practical for use in CO2 adsorption.
However, CO2 could be considerably adsorbed by modified activated
carbon due to the heterogeneity of the adsorbent surface. A variety of
surface complexes appeared via combining carbon atoms with other
elements such as oxygen, hydrogen, nitrogen and metals. After modi-
fication, the metals loaded are linked to functional groups in order to
provide active sites for adsorbing CO2 .
Two parameters were calculated to estimate the CO2 column dy-
namics: the experimental CO2 adsorption capacity and the break-
through time tb , which is the time it takes for adsorbent material that
is not fully saturated with CO2 . In a typical operation, the flow would
be stopped when the CO2 concentration reached to saturation state.
During the desorption step, the adsorbent was fully regenerated so
that the repeatability of the breakthrough curves could be assessed. In
the present study, breakthrough curves of CO2 for unmodified ACs at
different conditions are given in Figs. 6–8. The breakthrough times and
adsorption capacities of various operating conditions are also sum-
marized in Table 4. The breakthrough curves were analyzed using the
Deactivation model. The parameters of ks τ and kd were obtained by
114 S. Hosseini et al. / Journal of the Taiwan Institute of Chemical Engineers 52 (2015) 109–117

1 1

0.8 0.8

0.6
0.6

C/C0
200 Kpa
30C
C/C0

0.4
150 Kpa 0.4 40C

0.2 100 Kpa 50C

0.2

0
0 20 40 60 80 100 120
0
t(min) 0 20 40 60 80
t(min)
Fig. 6. Breakthrough curves (15% CO2 concentration, temperature 30 °C with varying
pressure from 100 to 200 kPa, unmodified AC). Fig. 7. Breakthrough curves (15% CO2 concentration, temperature 30–50 °C, pressure
100 kPa, unmodified AC).

plotting Eq. (1). The deactivation model parameters were evaluated 1

by analysis of the experimental breakthrough data and were sum-
marized in Table 4. The standard deviations of parameters ks τ and 0.8
kd were listed in Table 4. The regression analysis of the experimental
breakthrough data indicated good agreement with Eq. (1) with re-
0.6
gression coefficient more than 0.996. The results of the present work 50%
C/C0

showed that the deactivation model was a good representation of the 35%
CO2 breakthrough curves obtained in a fixed bed column. 0.4 25%
To express the effect of pressure, the breakthrough curves were 15%
obtained in the range of pressure from 100 to 200 kPa (Fig. 6). As 10%
0.2
seen, the breakthrough curve was shifted toward the left when the 5%
pressure of gaseous mixture is at lower value 100 kPa and a shift of the
breakthrough curves was observed with an increase in pressure. This 0
0 20 40 60 80
shift corresponds to an increase in the CO2 adsorption capacity from
t(min)
1.52 to 3.61 mol/kg. The gaseous molecules are packed more tightly
together with increasing pressure. Moreover, more molecules have a Fig. 8. Breakthrough curves (15–50% CO2 concentrations, temperature 30 °C, pressure
better chance to adsorb the available adsorption sites, and increasing 100 kPa, unmodified AC).
the number of molecules adsorbed [29].With increase in pressure in
the bulk phase, the kd values are observed to decrease and the ks
values increased. As shown in Fig. 6, the CO2 adsorption reached to shorter breakthrough times (Fig. 7) and the adsorbed amounts on
the bed outlet after 30.2 min and this time increased to 40.4 min when the AC decreased significantly since CO2 adsorption is an exothermal
the pressure increased from 100 to 200 kPa at temperature 30 °C, 15% process and the molecule’s speed also enhance due to increase the
CO2 concentration and flow rate 250 mL/min. temperature. In fact, the breakthrough capacities obtained at higher
To study the effect of temperature, breakthrough curves of CO2 temperature were also very low (Table 4). The deactivation constant
were measured for three different temperatures, i.e. 30, 40 and 50 °C. kd is a function of temperature as given by Arrhenius equation [30].
The effect of temperature on CO2 adsorption was studied at pressure With increasing temperature from 30 to 50 °C, deactivation constant
100 kPa, flow rate 250 mL/min and 15% CO2 feed concentration. Fig. 7 started to rise from 0.1826 to 0.446 s−1 . The activity of the adsorbent
shows that the breakthrough curves shifted toward the left with an for CO2 removal was significantly decreased and a steady decrease
increase in temperature. A higher adsorption temperature caused a was observed in the deactivation rate constants with an increase in

Table 4
Deactivation model parameters for original AC.

Temperature (°C) Pressure (kPa) CO2 concentration (%) Deactivation model

kd (s−1 ) ks τ (m/s)(s/m) Breakthrough time (s) Total capacity (mmol/g)

30 100 15 0.1668 5.2282 30.2 1.52

30 150 15 0.1587 5.8759 34.6 2.33
30 200 15 0.1451 6.2131 40.4 3.61
30 100 5 0.1587 5.8564 35.1 0.58
30 100 10 0.162 5.4781 32.4 1.08
30 100 25 0.1723 5.022 28.3 2.37
30 100 35 0.1779 4.7867 26.4 3.10
30 100 50 0.1826 4.4667 24.4 4.09
40 100 15 0.304 5.5861 18.3 0.88
50 100 15 0.446 6.105 12.9 0.60
 S. Hosseini et al. / Journal of the Taiwan Institute of Chemical Engineers 52 (2015) 109–117 115

1 1

0.8
0.8

0.6
20%
C/C0

16% 0.6
0.4 12% 20%Zn

C/C0
8%
4% 20%Cu/Zn
0.2 0.4

20%Cu
0
0 40 80 120 160 0.2
t(min)

Fig. 9. Breakthrough curves (15% CO2 concentration, temperature 30 °C, pressure from
100 kPa and different amount of metal loading on modified AC). 0
0 50 100 150
t(min)
temperature due to the less favorable thermodynamics of adsorption
process at higher temperatures. The breakthrough times are 24.4, 18.3 Fig. 10. Breakthrough curves (15% CO2 concentration, temperature 30 °C, pressure
and 12.9 min for temperatures 30, 40 and 50 °C, respectively. from 100 kPa and modified AC with single and bi-metals).

The effect of different feed concentrations was studied by mixing

the CO2 with helium in different proportions. The testes were carried
out in the range of CO2 concentration from 5 to 50%, temperature
30 °C, flow rate 250 mL/min and pressure of 100 kPa. It is observed
that as the inlet adsorbate concentration increases from 5 to 50%,
the breakthrough time decreases from 35.1 to 24.4 min. As shown
in Fig. 8, for higher feed concentration, steeper breakthrough curves
are resulted because of the higher mass-transfer flux from the bulk
solution to AC surface due to a higher driving force along the pores
[19,31]. Thus, the equilibrium is attained faster for values of higher
adsorbate concentration. The adsorption capacity increased from 0.58
to 4.09 mol/kg for CO2 concentration from 5 to 50%, respectively.

3.3. Effect of modification

Breakthrough curves of CO2 adsorption for modified ACs with dif-

ferent ratios Cu/Zn are shown in (Fig. 9). Fig. 9 compared the break- Fig. 11. CO2 adsorption capacity on unmodified and modified AC.

through curves of CO2 adsorption for different modified adsorbent

(with metal loading 4–20%) at constant temperature 30 °C, pressure
100 kPa and 15% of CO2 concentration in the feed stream. The figure
showed that a higher mixture of Cu/Zn was caused more delay to
achieve the breakthrough time (tb ). For instance, the breakthrough
time of sample Cu/Zn–4% is 32 min and increased to 34 min when the
adsorption was carried out with metal loading Cu/Zn–8%. The break-
through time of samples Cu/Zn–12%, Cu/Zn–16%, and Cu/Zn–20% are
37, 39 and 45 min, respectively. Fig. 10 compared the breakthrough
curves of CO2 adsorption for modified ACs with single metals Cu and
Zn. It can be seen that a longer breakthrough time with higher ad-
sorption was resulted when the mixture of both metals was used for
modification of ACs compared to modified ACs with single metal. The
similar results were also reported by Somy et al. [9,32].
Fig. 11 shows the adsorption capacity for various AC samples. The
sample Cu/Zn–20% increased the CO2 adsorption capacity by approx-
imately 49% compared to unmodified AC. The modified samples with
single metal (Cu or Zn) indicated higher adsorption capacity than
unmodified AC. The Cu loaded AC revealed higher capacity than Zn
loaded AC. The comparison indicated the modification process is ef-
fective for increasing adsorption capacity. All the experiments were
carried out in the same operating condition (100 kPa, 30 °C and 15%
CO2 concentration). The CO2 adsorption capacity is greatly enhanced
because of the reaction of CO2 with the supported metals occurred
when they were physically adsorbed onto the micropores of the car-
bon [33]. Many researchers reported similar conclusion for activated
carbon modified with different metals [10–12,34].
The values of the rate constants for different modified AC samples
are summarized in Table 5. As given in Table 5, the CO2 adsorption
capacity of co-impregnation Cu/Zn–20% was around 11% higher com-
pared to single metal impregnation Cu-20%. This clearly indicates that
surface modifications by metal impregnation increase the affinity of
carbon toward CO2 .
3.4. Desorption process
In order to study the regeneration process, the adsorption pro-
cess was carried out by the column packed with the adsorbent and
saturated with a stream of the CO2 /He mixture at atmospheric pres-
sure and the flow rate of 250 mL/min. The desorption process of
two samples unmodified AC and Cu/Zn–20% modified AC was shown
(Fig. 12). Fig. 12 demonstrated variation of the CO2 concentration with
time during desorption process. The CO2 concentration was sharply
decreased with time and the initial value of C/C0 is started from the
value of 1. As seen that unmodified AC indicated a regeneration pro-
cess time less than Cu/Zn–20% modified AC.
116 S. Hosseini et al. / Journal of the Taiwan Institute of Chemical Engineers 52 (2015) 109–117

Table 5
Deactivation model parameters for modified AC.

Adsorbent Temperature (°C) Pressure (kPa) CO2 concentration (%) Deactivation model

kd (s−1 ) ks τ (m/s)(s/m) Breakthrough time (s) Total capacity (mmol/g)

Cu/Zn–4% AC 30 100 15 0.1658 3.7376 32.9 1.64

Cu/Zn–8% AC 30 100 15 0.1723 4.9591 34.8 1.74
Cu/Zn–12% AC 30 100 15 0.1734 5.5641 37.7 1.88
Cu/Zn–16% AC 30 100 15 0.1758 6.0238 39.6 1.98
Cu/Zn–20% AC 30 100 15 0.1799 7.5779 45.1 2.26
Cu–20% AC 30 100 15 0.1723 5.5641 39.9 1.99
Zn–20% AC 30 100 15 0.1799 6.3428 35.4 1.79

References

[1] Lund H. The Kyoto mechanisms and technological innovation. Energy

Fig. 12. Regeneration curve of the saturated unmodified AC and modified 20% Cu/Zn
(T = 80 °C, He gas 250 cm3 /min).
4. Conclusions
The aim of this work is the effect of binary mixture of metals on ac-
tivated carbon and investigation CO2 adsorption capacity of modified
AC in different range of loading which can be used for flue gas treat-
ment. Cu/Zn-modified samples adsorbed more than the unmodified
sample and the adsorption was increased when the higher percent-
age of metal was loaded. The blend of metals improved the adsorption
capacity around 49% when 20% Cu/Zn was loaded on AC. The adsorp-
tion capacity for unmodified AC in the temperature of 30 °C, pressure
100 kPa and 15% CO2 concentration is 1.52 mol/kg and after modifica-
tion (20% of metal loading) the capacity increased to 2.25 mol/kg. One
the most important factor affecting the adsorption process is temper-
ature. In the case of physisorption, the adsorption time increases with
decreasing temperature. In the higher concentration of CO2 take less
time to reach the adsorption time and increasing pressure caused a
longer breakthrough time. The amount of adsorbed CO2 in the pre-
treated sample is less than unmodified sample. The acid pretreatment
indicated that the acidic groups increased from 0.58 to 3.18 mmol/g
for unmodified AC and pretreated sample, respectively. Since CO2 is
an acidic gas and the treatment by using HNO3 provided more acidic
functional groups and consequent a reduction on CO2 adsorption due
to increase acidic groups. The mechanism of CO2 removal is princi-
pally due to physical adsorption and chemisorption process. The small
pore sizes and large surface area of activated carbon play the major
role in gas adsorption.
Acknowledgments
This work was financially supported by the Ministry of Higher
Education, Government of Malaysia (via vot. 9416900 and 5523843)
and Universiti Putra Malaysia.
2006;31:2325–32.
[2] Socolow R, Roberta H, Greenblatt JB, Pacala S. Solving the climate problem. Envi-
ronment 2004;46:8–19.
[3] Sayari A, Belmabkhout Y, Serna-Guerrero R. Flue gas treatment via CO2 adsorption.
Chem Eng J 2011;171:760–74.
[4] Goodall C. How to live a low carbon life: the individuals guide to stopping climate
change. London: Academic Press; 2009.
[5] Mulgundmath VP, Jones RA, Tezel FH, Thibault J. Fixed bed adsorption for the
removal of carbon dioxide from nitrogen: breakthrough behaviour and modelling
for heat and mass transfer. Sep Purif Technol 2012;85:17–27.
[6] Nugent P, Belmabkhout Y, Burd SD, Cairns AJ, Luebke R, Forrest K, et al. Porous ma-
terials with optimal adsorption thermodynamics and kinetics for CO2 separation.
Nature 2013;495:80–4.
[7] Cheah W, Hosseini S, Khan MA, Chuah TG, Choong TSY. Acid modified
carbon coated monolith for methyl orange adsorption. Chem Eng J 2013;215:747–
54.
[8] Allwar M, Noor AB, Nawi MA. Textural characteristics of activated carbons pre-
pared from oil palm shells activated with ZnCl2 and pyrolysis under nitrogen and
carbon dioxide. Phys Sci 2008;19:93–104.
[9] Somy A, Mehrnia MR, Amrei HD, Ghanizadeh A, Safari M. Adsorption of carbon
dioxide using impregnated activated carbon promoted by zinc. Int J Greenhouse
Gas Control 2009;33:249–54.
[10] Jianying W, Fengyun Z, Yongqi H, Ruihong Z, Runjing L. Modification of activated
carbon fiber by loading metals and their performance on SO2 removal. Chin J
Chem Eng 2006;14:478–85.
[11] Chaedi M, Ansari A, Sahraei R. ZnS:Cu nanoparticles loaded on activated carbon
as novel adsorbent for kinetic, thermodynamic and isotherm studies of Reactive
Orange 12 and Direct yellow 12 adsorption. Spectrochim Acta A: Mol Biomol
Spectrosc 2013;114:687–94.
[12] Al Bahri M, Calvo L, Gilarranz MA, Rodriguez JJ, Epron F. Activated carbon sup-
ported metal catalysts for reduction of nitrate in water with high selectivity to-
wards N2 . Appl Catal B: Environ 2013;138–139:141–8.
[13] Fenrong L, Honghong Y, Xiaolong T, Ping N, Qiongfen Y, Dongjuan K. Adsorption
of carbon dioxide by coconut activated carbon modified with Cu/Ce. J Rare Earths
2010;28:334–7.
[14] Kim BJ, Cho KS, Park SJ. Copper oxide-decorated porous carbons for carbon dioxide
adsorption behaviors. J Colloid Interface Sci 2010;342:575–8.
[15] Eba H, Sakurai K. Pattern transition in Cu–Zn binary electrochemical deposition.
J Electroanal Chem 2004;571:149–58.
[16] Tseng H, Wey MY, Kiang YS, Chen KH. Catalytic removal of SO2 , NO and HCl
from incineration flue gas over activated carbon-supported metal oxides. Carbon
2003;41:1079–85.
[17] Ficicilar B, Dogu T. Breakthrough analysis for CO2 removal by activated hydrotal-
cite and soda ash. Catal Today 2006;115:274–8.
[18] Dahlan I, Mohamed AR, Kamaruddin AH, Lee KT. Dry SO2 removal process using
calcium/siliceous-based sorbents: deactivation kinetics based on breakthrough
curves. Chem Eng Technol 2007;30:663–6.
[19] Oh KJ, Park DW, Kim SS, Park SW. Breakthrough data analysis of adsorption of
volatile organic compounds on granular activated carbon. Korean J Chem Eng
2010;27:632–8.
[20] Yasyerli S, Ar I, Dogu G, Dogu T. Removal of hydrogen sulfide by clinoptilolite in a
fixed bed adsorber. Chem Eng Process 2002;41:785–92.
[21] Do DD. Adsorption analysis: equilibria and kinetics. London: Imperial College
Press; 1998.
[22] Hosseini S, Marahel E, Bayesti I, Abbasi A, Chuah Abdullah L, Choong TSY. CO2
adsorption on modified carbon coated monolith: effect of surface modification by
using alkaline solutions. Appl Surf Sci 2015;324:569–75.
[23] Oh WC, Lim CS. Comparison of metal loading effect for the activated carbons
pre-treated with various acids. Ceram Process Res 2005;6:73–80.
[24] Liu L, Deng QF, Liu YP, Ren TZ, Yuan ZY. HNO3 -activated mesoporous carbon
catalyst for direct dehydrogenation of propane to propylene. Catal Commun
2011;16:81–5.
[25] Muniz G, Fierro V, Celzard A, Furdin G, Gonzalez G, Ballinas ML. Synthesis, char-
acterization and performance in arsenic removal of iron-doped activated carbons
prepared by impregnation with Fe(III) and Fe(II). J Hazard Mater 2009;165:893–
902.
 S. Hosseini et al. / Journal of the Taiwan Institute of Chemical Engineers 52 (2015) 109–117
[26] Abdedayem A, Guiza M, Ouederni A. Copper supported on porous acti-
vated carbon obtained by wetness impregnation: effect of preparation condi-
tions on the ozonation catalyst’s characteristics. Comptes Rendus Chim 2014
http://dx.doi.org/10.1016/j.crci.2014.07.011.
[27] Houshmand AH, Shafeeyan MS, Arami-Niya A, Wan Daud WMA. Anchoring a
halogenated amine on the surface of a microporous activated carbon for carbon
dioxide capture. J Taiwan Inst Chem Eng 2013;44:774–9.
[28] Smykowski D, Szyja B, Szczygieł J. DFT modeling of CO2 adsorption on Cu, Zn, Ni,
Pd/DOH zeolite. J Mol Graph Model 2013;41:89–96.
[29] Plaza MG, Rubiera F, Pis JJ, Pevida C. Ammoxidation of carbon materials for CO2
capture. Appl Surf Sci 2010;256:6843–9.
117
[30] Shafeeyan MS, Daud WM, Houshmand A. Ammonia modification of activated
carbon to enhance carbon dioxide adsorption: effect of pre-oxidation. Appl Surf
Sci 2011;257:3936–42.
[31] Molina M, Perez V, Rodriguez F. Impregnation of activated carbon with chromium
and copper salts: effect of porosity and metal content. Carbon 1994;32:1259–65.
[32] Zhang Z, Xu M, Wang H, Li Z. Enhancement of CO2 adsorption on high surface area
activated carbon modified by N2 , H2 and ammonia. Chem Eng J 2010;160:571–7.
[33] Plaza MG, Pevida C, Arenillas A, Rubiera F, Pis JJ. CO2 capture by adsorption with
nitrogen enriched carbons. Fuel 2007;86:2204–12.
[34] Cavenati S, Grande CA, Rodrigues AE. Adsorption equilibrium of methane, car-
bon dioxide, and nitrogen on zeolite 13X at high pressures. J Chem Eng Data
2004;49:1095–101.