CARBON 94 (2015) 243–255

Contents lists available at ScienceDirect

CARBON
journal homepage: www.elsevier.com/locate/carbon

Spherical potassium intercalated activated carbon beads for pulverised

fuel CO2 post-combustion capture
Jingjing Liu a, Nannan Sun a,c, Chenggong Sun a,⇑, Hao Liu a,⇑, Colin Snape a, Kaixi Li b,⇑, Wei Wei c,
Yuhan Sun c
a
Faculty of Engineering, The Energy Technologies Building, Jubilee Campus, University of Nottingham, Nottingham NG7 2TU, UK
b
Institute of Coal Chemistry, Chinese Academy of Sciences, Taiyuan, Shanxi 030001, China
c
Shanghai Advanced Research Institute, Chinese Academy of Sciences, No. 99 Haike Road, Zhangjiang Hi-Tech Park, Pudong, Shanghai, China

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

Article history: Spherical carbon beads with a uniform diameter of ca. 0.6–0.8 mm and high mechanical strength can be
Received 17 March 2015 prepared by hydrothermal synthesis. To optimise the performance of these adsorbents for pulverised fuel
Received in revised form 15 June 2015 post-combustion capture, the efficacy of potassium intercalation via a KOH treatment has been investi-
Accepted 16 June 2015
gated, deliberately using nitrogen-free phenolic resin derived activated carbon (AC) beads so that the
Available online 23 June 2015
enhanced CO2 adsorption achieved by potassium intercalation could be delineated from any other effects.
At 25 °C and CO2 partial pressure of 0.15 bar, the adsorption capacity of K-intercalated ACs nearly dou-
bled from 0.79 mmol/g for the untreated carbons to 1.51 mmol/g whilst the effect on the morphology
and mechanical strength is relatively small. It was found that only slightly more than ca. 1 wt.% of K is
required to give the maximum benefit from intercalation that increases the surface polarity and the affin-
ity towards CO2. The notably increased CO2 uptake of the K-AC beads as a result of modest increase in
adsorption heat (32–40 kJ/mol compared to 27 kJ/mol for the original AC), coupled with the fast adsorp-
tion kinetics, suggest that the overall energy penalty is potentially superior to strongly basic polyethyle-
neimine and other amine-based solid adsorbent systems for carbon capture.
Ó 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://
creativecommons.org/licenses/by/4.0/).

1. Introduction (PCC), offering potentially improved process efficiency at signifi-

cantly reduced energy penalty, lower capital and operational costs
The International Energy Agency (IEA) and Intergovernmental and smaller plant footprints. Various solid adsorbents are under
Panel on Climate Change (IPCC) have identified carbon capture and investigation such as zeolites, supported/grafted amines, metallic
storage (CCS) as a critical greenhouse gas reduction solution [1–6]. organic frameworks (MOFs), functionalised carbon materials, and
However, the successful development and deployment of efficient calcium or alkali metal based sorbent materials [5,10–17], and some
and cost-effective carbon capture technologies plays a decisive role of these solid adsorbent-based capture systems have been demon-
in determining the viability of CCS as a whole because as it stands strated at varying scales, the largest being the pilot demonstration
now, the cost of carbon capture using the state-of-art of dry carbonate sorbent technology for CO2 capture at Hadong
energy-intensive CO2 amine scrubbing technology accounts for over power plant in South Korea [18].
70% of total CCS cost [7,8]. Consequently, alternative capture tech- Amongst the most studied materials as candidates for CO2
nologies have been under intensive development over recent years, adsorption [19,20], carbon-based adsorbents are characterised by
including advanced solvent scrubbing, oxyfuel combustion, chemi- the advantages of relatively low cost, ease of regeneration and gen-
cal looping combustion, membrane separation and solid adsorbent erally stable cyclic performance. Traditionally, activated carbons
looping technologies both at low and high temperatures [9]. Of these are considered as typical physical adsorbents because of their
capture technologies being developed, low temperature solid adsor- highly developed micro-porosity and large surface area, which
bents looping technology (SALT) has widely been recognised as hav- leads to the fact that activated carbons can normally only achieve
ing the potential of being viable for post-combustion CO2 capture better adsorption capacities at high CO2 partial pressures, limiting
their use for post-combustion capture where CO2 partial pressure
is usually low [21]. In order to enhance the surface affinity of acti-
⇑ Corresponding authors.
vated carbons towards the acidic CO2, many investigations have
E-mail addresses: [email protected] (C. Sun), liu.hao@
nottingham.ac.uk (H. Liu), [email protected] (K. Li). been carried out to modify the chemistry of carbon surface by

http://dx.doi.org/10.1016/j.carbon.2015.06.036
0008-6223/Ó 2015 The Authors. Published by Elsevier Ltd.
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
244 J. Liu et al. / CARBON 94 (2015) 243–255

manipulating either the precursor materials and/or activation elsewhere [42]. Briefly, a solution of hexamethylenetetramine
methodologies [22–26]. Thermal or thermochemical and novolac-type phenolic resins were dissolved in methanol, fol-
post-preparation treatment has been investigated as the potential lowed by mixing in aqueous polyvinyl alcohol (PVA), and then the
means to introduce different surface functional groups such as mixture was heated to 130 °C under stirring (400 RPM) in an auto-
basic oxygen functionalities (ketone, pyrone, chromene, etc.), nitro- clave for 1 h. After washing with abundant deionised water, the
gen functional groups (–NH2–, –CN, pyridinic nitrogen, etc.), other resulted resin beads were dried at 110 °C for overnight before they
heteroatoms or even ionic liquids [27,28]. On the other hand, the were carbonised at 830 °C in N2 for 1 h followed by steam activa-
thermodynamic and kinetic properties of adsorption also play an tion for another 1 h at the same temperature to obtain the parent
important role in determining the ultimate overall performance raw AC beads (denoted as PR0).
of activated carbons materials for CO2 capture [29]. Liu and
Wilcox [30,31] evaluated how the realistic surface functional
2.2. KOH treatment
groups effected CO2 adsorption using plan-wave electronic struc-
ture calculations and found that the adsorption thermodynamics
5 g of the raw AC beads were impregnated with 50 ml aqueous
and kinetics can be effectively improved via re-addressing the sur-
solution of KOH for 24 h. After drying in a vacuum oven at 70 °C for
face chemistries of carbon materials. However, limited decisive
overnight, which ensures that all the samples were completely
progress has been achieved so far in this field.
evaporated, the samples were heated in a horizontal tube furnace
It is only recently that carbon-based materials with enhanced
from ambient to a pre-selected treatment temperature at
CO2 adsorption capacities at relatively low CO2 pressures (ca.
3 °C/min and maintained at the temperature for an hour. The sam-
1.0–1.8 mmol/g at 0.15 bar CO2 and 25 °C) have been reported
ples were then washed with deionized water until neutral filtrate
[32–37]. Most of the adsorbents in these investigations were pre-
was obtained. Different KOH/AC mass ratios for impregnation
pared by involving nitrogen-containing functionalities, and in most
and various chemical activation temperatures used in the prepara-
cases coupled with chemical activation by potassium hydroxide
tion are summarised in Table 1 along with sample designations.
(KOH) [19,38–41]. These protocols are effective in enhancing CO2
Some of the samples were further treated by exhaustive Soxhlet
adsorption capacities, but the samples obtained were typically por-
extraction using de-ionised water to obtain samples with variable
ous powders with very low bulk densities which will give low CO2
contents of intercalated potassium contents (PR3_700_xh, where x
uptakes on a volumetric basis. Further, for practical applications
stands for the extraction duration in hours). Detailed procedures of
where either moving or fluidized-bed adsorbers are used for CO2
Soxhlet extraction with deionised water include: (1) The extraction
capture, these fine powders need to be agglomerated to form pel-
thimble which contained the carbon bead sample was loaded into
lets or beads with the aid of binders or other additives, which can
the main chamber of Soxhlet extractor; (2) Connect the Soxhlet
dramatically reduce the adsorption capacities and kinetic perfor-
exactor with distillation flask and reflux condenser; (3) Heat the
mance. The use of spherical carbon beads with high physical
distillation flask to 100 °C and maintain at this temperature for dif-
strength avoids this problem where we have reported on their
ferent periods of extraction. By controlling the extraction time, car-
potential for both pre- and post-combustion capture [42] with
bons containing different levels of intercalated potassium can be
CO2 uptakes at 1 bar being similar to those of many other carbons
obtained in order to evaluate the importance of intercalated potas-
[43–45]. However, attempts to significantly increase CO2 uptakes,
sium at different levels.
both by using nitrogen precursors and post-treatment with ammo-
nia, have met with limited success [46–48].
Although the basis to enhance CO2 uptakes at relatively low 2.3. Characterization of the samples
partial pressures on carbon-based adsorbents has been established
using KOH activation [49,50], only powdered samples have been Physical adsorption of N2 at 196 °C was carried out on a
reported. In this study, the use of potassium (K) intercalation via Micromeritics ASAP 2420 analyser. Prior to any measurements,
mild KOH activation as an effective means to boost the adsorption all samples were degased at 120 °C for overnight. The apparent
performance of activated carbons for post-combustion CO2 capture surface area (SBET) was calculated according to the method sug-
has been investigated, using activated-carbon beads with desirable gested by Parra et al. [52]. The cumulative pore volumes (Vtotal)
spherical diameters (ca. 0.6–0.8 mm) suitable for direct practical were calculated from the amount of nitrogen adsorbed at P/P0 of
applications. It is noteworthy that the KOH/AC mass ratios we used ca. 0.99, and the average pore volume was calculated by
are much lower (from 0.1:1 to 1:1) than those used in typical stan- 4Vtotal/SBET. The micropore volume (Vmicro) and surface area
dard KOH activations (often 2:1 or even higher [41,51]). The nov- (Smicro) were determined by the t-plot method.
elty of this study lies in that the potentially vital role of alkaline
metal intercalation as a means to enhance the CO2 adsorption
capacity and strength of carbons have been investigated and suffi- Table 1
ciently strong K-AC beads suitable for use in practical applications Preparation conditions and designation of the AC bead samples.
are obtained for the first time, with the beneficial effects of
Number Sample Initial KOH/AC mass Chemical activation
K-intercalation being quantified free from potential interference ratio for impregnation temperature (°C)
from nitrogen functionalities by using a phenolic resin as the pre-
1 PR0 0 None
cursor and, finally, the minimum amount of K required to enhance 2 PR1_600 0.1:1 600
CO2 adsorption at a partial pressure of 0.15 bar CO2 is identified to 3 PR1_700 0.1:1 700
simulate PCC in pulverised fuel (PF) power plant. 4 PR1_800 0.1:1 800
5 PR2_600 0.3:1 600
6 PR2_700 0.3:1 700
7 PR2_800 0.3:1 800
2. Experimental 8 PR3_600 0.5:1 600
9 PR3_700 0.5:1 700
2.1. Preparation of the raw activated carbon (AC) beads 10 PR3_800 0.5:1 800
11 PR4_600 1:1 600
12 PR4_700 1:1 700
The AC beads using phenolic resins as precursor were synthe-
13 PR4_800 1:1 800
sised with a hydrothermal method which has been described
 J. Liu et al. / CARBON 94 (2015) 243–255 245

X-ray Fluorescence (XRF) on selected samples was carried out on compared with coal and biomass-derived ACs [42]. Information
a Bruker S8 Tiger Spectrometer. Their morphologies were observed from XPS, SEM-EDX and XRF on the distribution of the K is pre-
on a FEI Quanta 600 Scanning Electron Microscope (SEM) and JEOL sented in SI1 and SI2. XRF indicates that the K-AC samples acti-
2100F Transmission Electron Microscope (TEM), respectively. The vated at 700 °C have concentrations of K in the range of 14–
Energy Dispersive X-ray analysis (EDX) software is Esprit 1.9 by 22 wt.%. However, XPS data (SI1) suggests that the surface K con-
Bruker. X-ray Photoelectron Spectroscopy was measured on a centrations are considerably lower, being slightly less than
Kratos Analytical Ultra-2008 spectrometer, and X-ray Diffraction 10 wt.%. This reveals that most of the K has been effectively inter-
(XRD) was carried out on a Bruker D8 instrument. To determine calated within the carbon beads. It was difficult to precisely link
the quantity of remaining potassium for K-ACs after washing, the the amount of intercalated K with the KOH/carbon mass ratios
samples were ashed using a thermogravimetric analyser (TGA, used in the preparation method as the formation of some crys-
Q600, TA instruments; samples were first dehydrated at 120 °C for talline potassium compound clusters on the outer surface of the
20 min, and then heated to 600 °C with a ramping rate of carbon beads is evident (see Fig. 1). However, it must be stressed
20 °C/min, followed by an isothermal period for 40 min). The resul- that the materials prepared using the same procedure and condi-
tant ashes were analysed by an Inductive Coupled Plasma Optical tions are fairly reproducible (SI3).
Emission Spectrometer (Perkin-Elmer Optima 33-DV ICP-OES, USA).

3.1.2. Textural properties

2.4. CO2 adsorption measurements The N2 adsorption isotherms are illustrated in Fig. 2 and SI4.
Despite the small hysteresis loops observed for some of the sam-
The CO2 adsorption isotherms were measured by the same ples due to the relatively minor development of mesoporous struc-
instrument used for physical adsorption of N2 (Micromeritics tures, all isotherms can be generally classified as type I according
ASAP 2420). Similar to the nitrogen adsorption isotherm tests, to IUPAC classification [53], suggesting that the spherical carbons
samples were degased at 120 °C overnight. are mainly microporous.
CO2 adsorption of the K-AC beads was also investigated by Table 2 summarises the specific BET surface areas, pore vol-
using a thermogravimetric analyser (TGA, Q500, TA instruments). umes, microporosities for all the samples derived from N2 (77 K)
The sample was first dried at 150 °C in pure N2 for 45 min to adsorption data. The initial AC bead sample (PR0) exhibits a BET
remove any physisorbed moisture and/or CO2, the temperature surface area of 1128 m2/g, which is typical for phenolic resin based
was then cooled to the adsorption temperature (25 °C). After the activated carbons [54,55]. The micropore surface area (Smicro) cal-
temperature stabilized, a flow of 100 mL min1 of a simulated flue culated by the t-plot method is fairly close to the BET surface area,
gas containing 15% CO2 in N2 was introduced into the sample consistent with the dominant microporosity. However, it is inter-
chamber at the adsorption temperature for 60 min and the sample esting to note that despite prolonged washing with water, consid-
weight was recorded in order to calculate the CO2 uptake. After erably lower BET surface areas and total pore volumes were
adsorption, the gas atmosphere was switched to N2 and the tem- observed for most of the K-ACs, particularly the PR3 series. These
perature was increased to 150 °C at a rate of 30 °C/min to desorb results are in sharp contrast to those reported in the literature
the CO2. Up to 50 adsorption–desorption cycles were performed where enhanced porosity of carbons was obtained via KOH activa-
to evaluate the stability of the adsorbent samples. tion [19,41,56]. Therefore, it appears that the carbon matrix struc-
tures inherited from the earlier steam activation has been further
2.5. Heat of adsorption measurements modified leading to intercalation of K species but with the loss of
some microporosity. Previous investigations [19,57–61] have
The heat of adsorption and the specific heat capacity of the K-AC revealed that KOH activation involves a variety of chemical reac-
carbon materials were determined using a SENSYS evo TG-DSC tions as detailed below, with the significance of the individual
instrument (Setaram) under the conditions similar to those used reactions varying with temperatures and KOH/carbon mass ratios.
in adsorption tests using a mixture of CO2 and nitrogen. Since this During the activation, KOH acts in three simultaneous/consecutive
instrument simultaneously provides mass changes and heat flows, ways: (1) the catalyst to accelerate the gasification reactions which
the heat of adsorption can then be deduced in terms of the heat
released per mole of CO2 adsorbed.

2.6. Mechanical strength testing

The mechanical strengths of the samples were measured using

a DMAQ800 dynamic mechanical analyser (TA instruments, USA).
The dynamic force exerted on the sample allows the change in
deformation with temperature to be monitored. Each sample was
tested ten times during which one carbon bead was randomly
selected every time to ensure the accuracy of the results. The aver-
age values for the 10 repeat tests for each sample are reported.

3. Results and discussion

3.1. Characterization of the K-containing activated carbon beads

3.1.1. Chemical composition

Our previous work has already showed that the parent carbon
bead sample (PR0) was free from nitrogen (less than 0.01 wt.%) Fig. 1. SEM image of the PR1_700 sample showing the formation of clustered
and inorganics, which was expected as the sample was derived potassium compounds on the outer surface of the treated carbons during the drying
from pyrolysis of a phenolic resin containing negligible impurities process.
246 J. Liu et al. / CARBON 94 (2015) 243–255

Fig. 2. N2 isotherms of the raw parent sample PR0 and KOH-intercalated samples PR1_600, PR2_600, PR3_600 and PR4_600. (A color version of this figure can be viewed
online.)

to the carbon matrix during the porosity development, largely

Table 2
due to the high volatility of metallic potassium (the melting and
Texture properties of the carbon beads.
boiling point for metallic K are 63 and 759 °C, respectively), see
Sample SBET Vtotal Average pore Smicro Vmicro in Reactions (5) and (7)(9); and (3) the formation of potassium
(m2/g) (cm3/g) diameter (nm) (m2/g) (cm3/g)
carbonate layer as shown in Reactions (4) and (5), which can effec-
PR0 1128 0.45 1.59 1115 0.43 tively prevent the carbon from over consumption, although the
PR1_600 972 0.39 1.60 958 0.37
carbonate can also be partly or wholly consumed in Reactions (6)
PR1_700 943 0.38 1.61 930 0.36
PR1_800 1047 0.42 1.61 1029 0.40 and (9), depending on the KOH/carbon ratios used. It is believed
PR2_600 758 0.31 1.63 748 0.29 that the intercalated K is most likely associated with the formed
PR2_700 902 0.36 1.61 892 0.35 oxygen functionalities (as shown in Reaction (10)) within the car-
PR2_800 963 0.41 1.70 939 0.38 bon matrix structures via the formation of quasi-chemical bonds,
PR3_600 857 0.35 1.64 843 0.34
PR3_700 826 0.34 1.63 815 0.32
giving rise to the residual potassium species that cannot be readily
PR3_800 820 0.35 1.69 804 0.32 removed by the routine washing with excessive water.
PR4_600 983 0.40 1.64 963 0.38
PR4_700 1071 0.44 1.65 1050 0.42 2KOH ! K2 O þ H2 O ð1Þ
PR4_800 1171 0.50 1.76 1150 0.47 C þ H2 OðsteamÞ ! CO þ H2 ð2Þ
CO þ H2 O ! CO2 þ H2 ð3Þ
are shown in Reactions (1)–(3) and (5). Otowa et al. [61] proposed K2 O þ CO2 ! K2 CO3 ð4Þ
the reactions of carbon with CO2 and H2O (steam) produced from
6KOH þ 2C ! 2K þ 3H2 þ 2K2 CO3 ð5Þ
KOH decomposition at temperatures from ca. 400 °C, as shown in
Reactions (1)–(3). However, recent work by Linares-Solano and K2 CO3 ! K2 O þ CO2 ð6Þ
his colleagues [56,57,60] has observed that carbon can react K2 O þ H2 ! 2K þ H2 O ð7Þ
directly with KOH as shown in Reaction (5) at temperatures from K2 O þ C ! 2K þ CO ð8Þ
ca. 400 °C as the standard Gibbs free energy of this reaction turns K2 CO3 þ 2C ! 2K þ 3CO ð9Þ
negative from this temperature, and this has been vindicated by
K2 O þ C ! K þ CAOAK ð10Þ
the formation of H2 and K2CO3 at temperatures below 700 °C with
no appreciable amount of CO and CO2 being produced [56]; (2) the The surface area or microporosity of KOH-activated carbons is
formation of metallic potassium particularly at temperatures determined by a combination of activation temperature and the
higher than 700 °C that can readily be mobilised and intercalated amount of KOH used. As shown in Table 2, however, the
 J. Liu et al. / CARBON 94 (2015) 243–255 247

considerably lower surface areas obtained for most of the samples functionalities can be formed at over 700 °C and efficiently interca-
from the secondary KOH activation of the original steam-activated lated into the carbon framework structures. When washing with
carbon bead sample (PR0), which has a considerably higher surface excessive water, most of free surface potassium species, such as
area of 1128 m2 g1, tend to suggest that the potential extra micro- K2CO3 and K2O formed during the activation at high temperatures,
porosity developed from the secondary KOH activation was offset can be readily removed while intercalated potassium species or
by the even larger porosity loss as a consequence of the simultane- related functionalities (e.g. the –O–K chemical or quasi chemical
ous potassium intercalation that can potentially lead to complete bonds) remain in the carbon matrix structures. The element map-
or partial occlusion of the micropores. It was found that at the ping shown in Fig. 6 illustrates a high consistency between the sur-
same KOH/AC mass ratios, the surface areas of all the K-AC bead face distributions of K and O, further suggesting that the
samples increased significantly with increasing activation temper- intercalated K is most likely associated with the formed oxygen
atures, highlighting the temperature-dependent reactivity of car- functionalities via the formation of quasi-chemical bonds, giving
bon with KOH. At the same activation temperature, however, the rise to highly polarized carbon surfaces.
surface areas of the K-ACs were found to decline first with elevat-
ing KOH/AC mass ratios from 0.1:1 to 0.5:1 and then increase with 3.2. CO2 adsorption on the carbon beads
further rise in KOH/AC ratios from 0.5:1 to up to 1:1, indicating the
relative significance of porosity development and simultaneous
potassium intercalation during the chemical activation process. 3.2.1. Static adsorption measurement
As the most used method for adsorbent evaluation, the CO2 iso-
3.1.3. Morphology therms of the carbon beads were measured both at 0 and 25 °C by a
The morphology of the carbon beads was investigated by both Sieverts apparatus (volumetric method) to obtain the CO2 adsorp-
SEM and TEM. Fig. 3 shows some of the images from SEM and more tion capacities. According to Myers and others [66,67], absolute
can be found in SI5. Fig. 3a–c confirms that the AC samples have loadings are needed for thermodynamic processing such as iso-
well-developed spherical forms with uniform diameters of 0.6– therm fitting and adsorption enthalpy calculation, therefore the
0.8 mm. More importantly, the macroscopic morphology of the obtained excessive CO2 uptakes were converted to absolute
samples shows little change after the incorporation of K; the uptakes by using Peng–Robinson equation of state (SI7) [68,69],
desired spherical form has been preserved. Specifically, cracks and some of the results are presented in Fig. 7 while the others
and randomly distributed large holes are present on the outer sur- are provided in the Supporting information (SI8).
face. Fig. 3d–f shows the cross-sectional images for the carbon All samples display type I isotherms for CO2 adsorption with a
beads where the presence of large number of lm-scale pores and sharp uptake at the early adsorption stage in the low relative pres-
some interior hairline channels are evident. TEM images in Fig. 4 sure region, suggesting that these carbon beads are mainly micro-
and SI5 reveal the amorphous nano-structures with graphite layers porous materials. At atmospheric pressure, sample PR0 adsorbs
being observed. The fact that no crystallized K species were found 4.45 and 2.80 mmol CO2/g at 0 and 25 °C, respectively, being con-
reflects their superior uniform distribution, which is of great sistent with those reported elsewhere for phenolic resin derived
importance to form highly polarized surfaces to enhance CO2 cap- carbons not modified by nitrogen treatments [42]. For better com-
ture. This will be further discussed in the following sections. parison, the CO2 adsorption capacities measured at different tem-
peratures and partial pressures are summarised in Table 3 while
3.1.4. Element mapping a comparison with those reported in the literatures for KOH acti-
The distribution of the K species was obtained from SEM ele- vated carbons is presented in Table 4. As expected, all of the
ment mapping which is depicted in Fig. 5, where red colour repre- KOH-activated carbon bead samples showed substantially higher
sents the existence of K with the brightness indicating adsorption capacities at both temperatures, especially at relative
concentration. As can be seen from Fig. 5a, no K was detected for low pressures. It was found that at the same activation tempera-
the initial AC, PR0, as expected. As shown in Fig. 5b and c, the tures, the adsorption capacity of K-ACs ascended with increasing
K-intercalation treatment led to relatively even distributions of K KOH/AC mass ratios used in the chemical activation. However, no
with no obvious segregation, implying that the incorporated K is obvious correlation was found between the adsorption capacity
highly dispersed throughout the carbon beads. and the surface area or porosity of the carbon samples derived
For KOH activated carbons, it is generally recognised that all the from different conditions (Table 2), especially at lower CO2 partial
residual K species can be removed by subsequent washing with pressures. At 25 °C and a CO2 partial pressure of 0.15 bar, the orig-
acid (normally HCl solution), by which the porosity in the carbon inal steam-activated sample (PR0), which has the highest surface
samples can be recovered [56,62–65]. In the present case, however, area of 1128 m2/g and micropore volume of 0.43 cm3/g, adsorbed
deionised water was used as a ‘‘soft’’ washing procedure to succes- approximately only half of the amount of CO2 adsorbed by the
sively remove first the ‘‘free’’ and then the intercalated K. In other K-intercalated PR3_700 sample (1.51 mmol/g), which has a consid-
words, we are attempting to enhance the surface polarity of the erably lower surface area of 826 m2/g and micropore volume of
carbon beads by potassium intercalation into the carbon frame- 0.32 cm3/g, highlighting the importance of potassium intercalation
works as will be discussed later. Therefore, the distribution of for enhanced CO2 adsorption performance of the carbons. Similar
potassium and oxygen on the surface in microscopic scale plays phenomenon was also observed for the K-AC samples. For each ser-
an important role in promoting CO2 adsorption capacities (Fig. 5). ies of samples obtained from using the same KOH/AC ratios in acti-
To further investigate the distributions of K and O in relation to vation, the adsorption capacity of the activated samples at 800 °C
C, TEM mapping was carried out, and the results (Fig. 6) suggest was found to be considerably lower than that of the samples acti-
that the C, K and O elements are evenly distributed and appear vated at 700 °C, despite the higher surface areas of the activated
to be interlinked to each other, which confirms that the surface samples at the higher activation temperatures. For all the
modification is achieved successfully. The high distribution of KOH/AC mass ratios examined, the K-AC samples activated at
potassium species can also be evidenced by the weak diffraction 700 °C exhibit the best performance for CO2 uptake.
peak in the XRD pattern (SI6). Although carbon beads derived from a nitrogen-free phenolic
As mentioned above in Section 3.1.2, according to the chemical resin was deliberately used in this study in order to eliminate
reactions between C and KOH during the chemical activation, the effects of any other heteroatoms in the precursors, a compar-
metallic K and quasi-chemical bonds of C–O–K as new surface ison of the K-ACs with the nitrogen-enriched carbons reported in
248 J. Liu et al. / CARBON 94 (2015) 243–255

Fig. 3. SEM images of (a) original PR0, (b) PR3_700 (KOH/AC mass ratio of 0.5:1), (c) PR4_700 (KOH/AC mass ratio of 1:1), and their corresponding cross-sectional images
(d–f).

As already been described that although various AC adsorbents

have been reported with high CO2 adsorption capacities, most of
the reported carbons were produced in powder forms via
processes that often require sophisticated post-treatments, such
as HNO3 oxidation and/or NH3 activation, leading to significantly
increased production costs and reduced carbon yields. Further,
carbons produced in fine powder forms cannot usually be used
in moving-bed or fluidized bed reactors and will have to be
further engineered via pelletisation or granulation to form
different types of shapes with required particle sizes, and this
can sharply reduce the adsorption capacity and kinetics of the
carbon adsorbents [42].
As indicated in Table 5, the K intercalation treatment as a means
to improve adsorption performance only has a relatively small
impact on the mechanical strength of the carbon beads. The addi-
tion of 0.1 and 0.3 KOH mass ratio to the carbon beads only reduces
the mechanical strength by ca. 10 and 20%, respectively. This
reduction is easily tolerated for fluidised bed operation given the
high mechanical strength of the initial beads.
One should also bear in mind that while the adsorption capacity
on a weight basis is important in terms of the scale-up of the
adsorption-based CO2 capture process, the form and density of
the activated carbons will also be critical, with the adsorption
capacity on a volumetric basis instead ultimately determining
the size and process efficiency of the adsorbent bed. Therefore, it
Fig. 4. High resolution TEM images of the activated carbon beads (sample is evident that the overall performance of the adsorbents will be
PR3_700). a compromise between adsorbent materials having high surface
area and micro-porosity while having sufficient density to maxi-
mize the volumetric adsorption capacity. To this end, the phenolic
the literature (Table 4) shows that the CO2 adsorption capacities of resin derived activated carbon beads from the present study are
our K-AC samples are overall comparable to the nitrogen-enriched highly advantageous over the reported carbons, given their signif-
carbons reported in literature, despite their significantly lower sur- icantly lower porosity, higher bulk density (0.39 g/cm3) and much
face areas (hence potentially higher capacities on a volumetric easier-to-handle spherical forms with practical diameters.
basis) and the absence of nitrogen functionalities (Tables 1, 2 and Scaled-up production of these novel carbon bead adsorbents is
4). More importantly, all the K-AC beads were measured as pro- underway in order to further evaluate their performance for CO2
duced with desirable spherical diameters typically varying capture on a volumetric basis using a kg-adsorbent scale fluidized
between 0.6 and 0.8 mm. bed reactor [71].
 J. Liu et al. / CARBON 94 (2015) 243–255 249

Fig. 5. SEM images of potassium mapping: (a) sample PR0, (b) sample PR3_700, (c) cross-section of sample PR3_700. (A color version of this figure can be viewed online.)

Fig. 6. TEM images of PR3_700, showing the distribution of different elements: (a) original image, (b) carbon mapping, (c) potassium mapping, (d) oxygen mapping. (A color
version of this figure can be viewed online.)

3.2.2. Heat of adsorption By fitting the obtained experimental isotherms at different temper-
Heat of adsorption is an important thermodynamic parameter atures with isotherms equations, one can acquire the pressure (P)
for describing the adsorption behaviour of an adsorbent and eval- needed to reach the same adsorption capacity (q) at different tem-
uating the energy performance of a process of adsorption and des- peratures (T), and then Qst can be calculated by fitting the Clausius–
orption, such as the pressure swing and temperature swing gas Clapeyron equation.
separations. Isosteric heat of adsorption (Qst) can be used to indi- In this work, the dual-site Langmuir (DSL) equation was used to
cate the surface affinity of an adsorbent towards an adsorbate. It fit the adsorption isotherms of all K-AC samples as the one-site
is usually estimated by using the Clausius–Clapeyron Eq. (1): Langmuir equation is no longer adoptable due to the surface
  heterogeneity of the carbon surface as a result of potassium inter-
@ ln P
Q st ¼ RT 2 ð1Þ calation [72,73]. Indeed, recent investigations have proposed that
@T q DSL equation is a better model to fit the adsorption isotherms of
where R represents specific gas constant; q stands for the adsorp- carbons with high surface heterogeneity [74,75]. The DSL Eq. (2)
tion capacity and T, P is the temperature and pressure, respectively. can be expressed as follows:
250 J. Liu et al. / CARBON 94 (2015) 243–255

(2) qS,A and qS,B should be kept constant at all temperatures to

fulfil their physical meanings;
(3) bX (X = A or B) usually decreases with increasing adsorption
temperature and is related to a temperature-independent
constant, b0,x by the Arrhenius Eq. (3).
 
E
b ¼ b0 exp ð3Þ
RT
In Eq. (3), E can be used to indicate the average adsorption heat.
Since the fitted isosteric heat of adsorption was obtained as a func-
tion of adsorbed CO2 in which case Qst represents the average of all
adsorption sites that can be potentially occupied at a certain cov-
erage level [77,78], the calculated heat of adsorption corresponds
to the values at initial coverage (at a CO2 uptake level of
0.01 mmol/g). However, it is noteworthy that the calculated
adsorption heat may differ from the isosteric heat (Qst) of adsorp-
tion, but the difference should be quite small [78].
After taking account of these criteria, the experimental iso-
therms were fitted and some of the fitting curves are shown in
Fig. 7 while the others are provided in SI8 and SI9. Fig. 8a illustrates
the calculated Qst for both the parent and KOH-intercalated carbon
beads at a CO2 uptake level of 0.01 mmol/g. Generally, a consider-
able increase in isosteric adsorption heat was observed for all
K-intercalated carbon bead samples comparing with the parent
carbon beads (PR0). For instance, the calculated isosteric adsorp-
tion heat of K-intercalated PR1 and PR3 series carbons increased
by up to 60% to ca. 36–42 kJ/mol CO2 from 26 kJ/mol CO2 for the
parent carbon beads (PR0), respectively. This highlights the effec-
tiveness of K intercalation in improving the surface affinity of the
carbons for CO2 adsorption particularly at low CO2 partial
pressures.
Isosteric adsorption heat is also used to indicate the degree of
surface heterogeneity of the adsorbent or different interactions
between adsorbent and adsorbate. For energetically heterogeneous
surfaces, the isosteric heat of adsorption will decrease with
Fig. 7. CO2 adsorption isotherms of the carbon beads at 0 °C and 25 °C (symbols increasing quantities of adsorbed substances due to the occurrence
represent the measured values and lines are the fitted values with the dual-site
of different types of active adsorption sites. Fig. 8b shows the isos-
Langmuir isotherm equation). (A color version of this figure can be viewed online.)
teric adsorption heat both for the parent (PR0) and K-intercalated
carbons (PR3 series) at different CO2 uptake levels. The calculated
qs;A  bA  P qs;B  bB  P isosteric adsorption heats are by and large close to the upper range
q¼ þ ð2Þ of the reported adsorption heats for carbon-based adsorbents
1 þ bA  P 1 þ bB  P
[44,79,80]. While higher isosteric adsorption heats were obtained
where qS,A and qS,B are the mono-layer capacity of site A and B, bA for all K-ACs, the variation of isosteric heat of adsorption with
and bB are the adsorption equilibrium constants for site A and B. the quantities of CO2 adsorbed is much higher for the K-ACs than
During the fitting of isotherms, one must note [76]: for the initial carbon beads with the isosteric adsorption heat
remaining relatively constant. This demonstrates that the porous
(1) use absolute adsorption capacities;

Table 4
Table 3 Comparison on the adsorption capacities of PR3_700 and other carbon adsorbents.
CO2 adsorption capacities of the carbon beads.
Materials Precursor SBET Uptake at Ref.
Sample CO2 capacity at 0 °C CO2 capacity at 25 °C (m2/g) 25 °C*
(mmol/g) (mmol/g) (mmol/g)

0.15 bar 1 bar 0.15 bar 1 bar 0.15 bar 1 bar

PR0 1.52 4.43 0.79 2.78 PC-2 Agaricus 1479 0.88 3.46 [33]
PR1_600 1.92 4.45 1.14 3.02 KNC-A-K P-diaminobenzene 614 1.81 4.04 [36]
PR1_700 2.07 4.67 1.23 3.17 IBN9- P-diaminobenzene 890 1.62 4.50 [70]
PR1_800 2.06 4.74 1.26 3.27 NCI-A
PR2_600 2.06 3.95 1.38 2.97 VR-93 Vacuum residue 2895 1.02 4.83 [32]
PR2_700 2.19 4.58 1.40 3.28 CEM-750 N-doping carbon 3360 0.98 4.38 [34]
PR2_800 1.83 4.25 1.12 2.91 RFL-500 Resorcinol, formaldehyde 467 1.50 3.21 [46]
PR3_600 2.18 4.29 1.43 3.18 and lysine
PR3_700 2.33 4.59 1.51 3.35 CSA-700 Poly (acrylonitrile-co- 1231 2.11 3.80 [37]
PR3_800 1.95 4.25 1.23 3.06 acrylamide)
PR4_600 2.16 4.68 1.36 3.35 PR3_700 Phenolic resin 826 1.51 3.35 This
PR4_700 2.35 5.13 1.52 3.72 work
PR4_800 2.00 5.07 1.25 3.47 *
All results are measured by static volumetric method.
 J. Liu et al. / CARBON 94 (2015) 243–255 251

Table 5 equilibrium CO2 adsorption capacity obtained by TGA-DSC in a

Mechanical strength of K-ACs (activated at 700 °C). calorimetric analysis.
Sample KOH/AC mass ratio Maximum load (N) Z n
1
PR0 0 3.08 Q int ¼ Q st dn ð4Þ
n 0
PR1_700 0.1:1 2.86
PR2_700 0.3:1 2.47 The calculated adsorption heat was then validated by the
PR3_700 0.5:1 2.21
adsorption heat experimentally measured by TGA-DSC. As shown
PR4_700 1:1 1.75
in Table 6, for the original steam-activated carbon bead sample,
the measured and calculated adsorption heats for the original
steam-activated carbon beads are almost identical while for the
surface of the K-intercalated carbons becomes energetically much
K-intercalated carbons, the calculated adsorption heat values
more heterogeneous, being indicative of the incorporation of new
appears to be slightly lower by 1–6 kJ/mol CO2, compared to the
active adsorption sites (Fig. 6) that led to higher interaction energy
values measured by TGA-DSC. The good agreement obtained
of CO2 with the surface.
between the calculated and measured adsorption heat for most
In order to validate the calculated isosteric heat of adsorption
of the carbon bead samples confirms the suitability of the DSL
with the dual-site Langmuir model, TGA-DSC has been used to
modelling for evaluating the adsorption behaviour of
experimentally measure the adsorption heat of the carbon beads.
surface-modified carbons.
Given that the measured value by TGA-DSC represents the aver-
aged adsorption heat integrated from the DSC heat flow corre-
3.3. The role of intercalated potassium in CO2 adsorption
sponding to the adsorption capacity of a carbon adsorbent, the
following Eq. (4) [81] was used to derive the integral adsorption
The results in the previous sections demonstrate that K interca-
heat values from the isosteric heat of adsorption. The integral
lation can significantly improve the CO2 adsorption performance of
adsorption values correspond to the defined ranges of
the phenolic resin derived carbon beads. It can be reasonably
sorption-phase concentration from 0 to n, where n equals to the
assumed that the increased CO2 adsorption capacity arose from
either increased micro-porosity or modified surface chemistries
or a combination of both. As Table 2 indicates, however, a consid-
erable decrease rather than an increase in surface area and micro-
porosity was observed for all the K intercalated carbon samples
under the conditions used, despite the washing with excessive
quantities of deionised water. This suggests that the improved
CO2 adsorption of the K-ACs cannot be accounted for by the surface
textual properties but by the modified surface chemistries. It has
been well established that KOH-activated carbons are usually
enriched in different surface oxygen functionalities [82–84], but
the presence of the surface functionalities alone is not sufficient
to explain the significant increase in CO2 capacities according to
the results reported elsewhere [37,85]. It is therefore considered
that the intercalated K is responsible for the enhanced CO2 adsorp-
tion performance of the K-ACs. To provide further insight, an
exhaustive Soxhlet extraction at 100 °C with deionised water as
the solvent was used to remove the intercalated potassium in the
K-AC samples to different levels. This was used to evaluate the
variation of CO2 adsorption with different levels of potassium
intercalation and hence to determine the minimal content of inter-
calated potassium that is required to achieve appreciable improve-
ment in CO2 adsorption. One of the K-AC samples, the PR3_700
which exhibited the highest CO2 capacity at 25 °C and a CO2 partial
pressure of 0.15 bar, was selected for the exhaustive Soxhlet

Table 6
Comparison between analytical integral heat of adsorption by DSL model and
measured DSC heat of adsorption.

Sample Analytical integral Experimental measured

heat of adsorption heat of adsorption
by DSL model (kJ/mol) by TG-DSC (kJ/mol)
PR0 26.0 26.7
PR1_600 34.1 33.9
PR1_700 36.9 39.7
PR1_800 30.6 33.7
PR2_600 35.6 38.2
PR2_700 31.4 35.1
PR2_800 29.6 33.0
PR3_600 33.3 38.8
PR3_700 33.3 35.5
PR3_800 32.4 34.2
Fig. 8. (a) Fitted isosteric adsorption heat at a CO2 uptake level of 0.01 mmol/g for PR4_600 33.2 34.9
all the carbon beads; (b) Fitted isosteric adsorption heat of PR0 and PR3 series as a PR4_700 31.6 34.6
function of absolute CO2 loading using dual-site Langmuir model. (A color version of PR4_800 25.3 31.3
this figure can be viewed online.)
252 J. Liu et al. / CARBON 94 (2015) 243–255

extraction for different extraction times. The samples produced

from different periods of extraction were then subjected to CO2
adsorption tests and other characterisations.
Table 7 summaries the textural properties of the extracted sam-
ples and their CO2 adsorption capacities measured at both 0 and
25 °C and a CO2 partial pressure of 0.15 bar. As expected, the
amount of K present decreased with increasing extraction time
and after 20 h extraction, the potassium content was reduced from
9.25% to only 0.57% on a weight basis. As can be seen from Table 7,
the BET surface area of the carbon sample increased considerably
with increasing removal of the intercalated potassium from the
carbons, particularly at the early stages of Soxhlet extraction.
This confirms that the lower surface areas obtained for the K-AC
samples (Table 2) was due to the potassium intercalation which
led to partial or even complete occlusion of micropores.
Measurements of CO2 adsorption capacity before and after the
Soxhlet extraction, presented in Table 7 and Fig. 9, demonstrate
that the capacity shows no significant change with decreasing
residual potassium content from ca. 9.2 to 2.8 wt.%, but a sharp
decrease was observed when potassium content was further
Fig. 9. The influence of K concentration (determined by ICP) on the CO2 capacities
reduced to below 1 wt.%. It appears that there is a critical concen-
for deionised water extracted samples from PR3_700. (A color version of this figure
tration of potassium above which the excess residual potassium can be viewed online.)
contributes little to CO2 adsorption capacity. It can be reasonably
assumed that the hard-to extract residual quantities of K are dee-
ply intercalated into the fine carbon structure in the form of temperature of 25 °C and a CO2 partial pressure of 0.15 bar (N2 bal-
extra-framework K+ cations. Previous investigations [36,86] ance). As shown in Fig. 10a, a final equilibrium CO2 adsorption
showed that stronger CO2 adsorption sites of zeolite, MOF adsor- capacity of 1.9 mmol/g is obtained with approximately 90% of
bents reside predominantly around unsaturated framework metal the equilibrium capacity achieved in 5 min under the TGA test con-
cations. Therefore, it is believed that the significantly improved ditions. This highlights the high capacity and fast adsorption kinet-
CO2 adsorption as a result of the intercalation of K cations is due ics of the K-AC beads, although it takes a slightly longer time for
to the involvement of stronger electrostatic forces created by the the K-ACs to reach equilibrium adsorption compared to the parent
exposed framework K+ ions on the surface, with the strength of ACs. Fig. 10b shows the cyclic adsorption/desorption testing results
the electrostatic interaction being determined by the surface den- for the K-AC carbon bead sample. The cyclic life-time performance
sity and nature of the extra-framework K+ cations, an adsorption testing demonstrates the excellent reversible adsorption perfor-
mechanism similar to that observed for the MOF-based CO2 adsor- mance of the K-intercalated carbon beads, despite the slight
bents [36,86]. Indeed, strong electrostatic properties were decrease in CO2 capacity in the first 5 cycles, which appears to be
observed in handing the K-AC samples which were found to be attributable to the irreversible CO2 adsorption on
easily attached to the interior surface of the sample cell and non-intercalated potassium species (as shown by bright red spots
showed appreciable repulsion behaviour between different K-AC in Fig. 5b) in the dry flue gas conditions.
carbon beads while similar phenomena were not observed for Apart from overall stability in multi-cycle tests, the regenera-
the parent carbon beads. tion energy needs to be considered in order to evaluate the effi-
It is very likely that the strong electrostatic interaction may lead ciency and cost-effectiveness of the K-AC samples as CO2
to the formation of labile carbonate-like complexes upon CO2 adsorbents. The required regeneration energy of a solid adsorbent
adsorption around the extra-framework cations [87]. As high- is the sum of the sensible heat that is required to heat the adsor-
lighted in the following cyclic adsorption/desorption tests in dry bent from the adsorption temperature to the regeneration temper-
gas conditions, the enhanced CO2 adsorption of the K-ACs, does ature and the latent heat (heat of adsorption) that is required to
not require the participation of water or moisture in the adsorption overcome the bonding energy to remove CO2 from the adsorbent
process unlike the K2CO3-based adsorbents where the presence of in the desorption process [89]. It has been reported [90] that
significant quantities of moisture in the gas stream is essential to amine-functionalised solid adsorbents such as PEI-silica generally
the CO2 adsorption [88]. require less regeneration energy (in a cyclic temperature swing
adsorption/regeneration CO2 capture process) compared with the
benchmark MEA scrubbing technology. The facts that the K-AC
3.4. Cyclic adsorption/desorption and stability testing of the K-AC
beads developed in this study have much lower heat of adsorption
adsorbents
and specific heat capacity (e.g. 36 kJ/mol, 1.1 kJ/kg.K for PR3_700)
than PEI/silica adsorbents (60–90 kJ/mol, 1.7 kJ/kg.K [71,91,92]),
Fig. 10 presents the adsorption/desorption characteristics of one
comparable CO2 capacities and superior regenerability indicate
of the K/AC samples (PR3_700) obtained from TGA at a
that K-AC adsorbents are potentially better alternatives for
Table 7 PEI-silica for PCC in coal-fired PF coal-fired power plants.
Characterization results for the extracted samples from PR3_700 (K concentrations
determined by ICP).

Sample SBET (m2 g1) K (wt.%) 4. Conclusions

PR3_700_1.5 h 892 9.25
PR3_700_2.5 h 984 2.79 The use of potassium intercalation via relatively mild KOH acti-
PR3_700_4.0 h 1053 0.96 vation has been investigated as a novel effective means to boost
PR3_700_20 h 1074 0.57 the adsorption performance of activated carbons for
PR0 1128 <0.01
post-combustion carbon capture. A nitrogen-free phenolic resin
 J. Liu et al. / CARBON 94 (2015) 243–255 253

Fig. 10. (a) The CO2 adsorption curve of PR0 and PR3_700; (b) 50 cycles of the adsorption and desorption for PR3_700; both under an atmosphere of 15% CO2 and 85% N2 at
25 °C. (A color version of this figure can be viewed online.)

derived spherical carbon material with desirable spherical diame- Acknowledgements

ters (0.6–0.8 mm) which is suitable for direct practical applications
without further palletisation or granulation treatment has been The authors wish to acknowledge the financial support from the
used so that the enhanced CO2 adsorption achieved by potassium UK EPSRC (Grant nos: EP/G063176/1 and EP/J020745/1) and the
intercalation could be delineated from any other effects. The National Natural Science Foundation of China (Grant nos:
results demonstrated that the intercalation of K significantly 51061130536, 51172251). The authors also wish to thank Dr.
increased the CO2 capacity of the AC beads by a factor of up to 2 Wenbin Zhang from the University of Nottingham for assisting in
at 0.15 bar CO2 partial pressure while the effects of potassium the measurements of adsorption heat.
intercalation on the mechanical strength and morphological fea-
tures of the carbon beads were negligible at KOH/AC mass ratios Appendix A. Supplementary data
of 60.3.
The results suggest that the enhanced CO2 adsorption of Supplementary data associated with this article can be found, in
potassium-intercalated carbons, as highlighted by their modestly the online version, at http://dx.doi.org/10.1016/j.carbon.2015.06.
increased heat of adsorption, is closely related to the formation 036.
of carbon framework or extra-framework K+ cations, which can
lead to the potential formation of Kd+  Od zwitterion-like struc- References
tures that can boost the electrostatic interaction of the carbon sur-
face with CO2. In addition, there appears to be a low critical K [1] M.M. Halmann, M. Steinberg, Greenhouse Gas Carbon Dioxide Mitigation:
concentration, which is not much more than ca. 1 wt.% in the case Science and Technology, CRC Press, 1998.
[2] D.B. Dell’Amico, F. Calderazzo, L. Labella, F. Marchetti, G. Pampaloni,
of the carbon beads examined, below which the beneficial effect of Converting carbon dioxide into carbamato derivatives, Chem. Rev. 103 (10)
K intercalation on CO2 adsorption becomes negligible. The results (2003) 3857–3898.
also demonstrate that the dual-site Langmuir model can be used [3] V. Scott, S. Gilfillan, N. Markusson, H. Chalmers, R.S. Haszeldine, Last chance for
carbon capture and storage, Nat. Clim. Change 3 (2) (2013) 105–111.
to effectively characterise the adsorption behaviour of the
[4] J.T. Houghton, Y. Ding, D.J. Griggs, M. Noguer, P.J. van der Linden, X. Dai, et al.,
K-intercalated carbons where high surface heterogeneity exists. Climate Change 2001: The Scientific Basis, Cambridge University Press,
The significantly increased CO2 adsorption capacity, modest Cambridge, 2001.
adsorption heat, fast kinetics and good mechanical strength augurs [5] P.H. Stauffer, G.N. Keating, R.S. Middleton, H.S. Viswanathan, K.A. Berchtold,
R.P. Singh, et al., Greening coal: breakthroughs and challenges in carbon
extremely well that the K-AC adsorbents could provide a sound capture and storage, Environ. Sci. Technol. 45 (20) (2011) 8597–8604.
alternative to the most studied polyethyleneimine-based sorbents [6] B. Metz, O. Davidson, H. De Coninck, M. Loos, L. Meyer, Carbon dioxide capture
for improved energy penalty of post-combustion capture using low and storage, 2005.
[7] E.S. Rubin, C. Chen, A.B. Rao, Cost and performance of fossil fuel power plants
temperature solid adsorbent looping technology. with CO2 capture and storage, Energy Policy 35 (9) (2007) 4444–4454.
254 J. Liu et al. / CARBON 94 (2015) 243–255

[8] J. Gibbins, H. Chalmers, Carbon capture and storage, Energy Policy 36 (12) [39] V. Presser, J. McDonough, S.-H. Yeon, Y. Gogotsi, Effect of pore size on carbon
(2008) 4317–4322. dioxide sorption by carbide derived carbon, Energy Environ. Sci. 4 (8) (2011)
[9] M.E. Boot-Handford, J.C. Abanades, E.J. Anthony, M.J. Blunt, S. Brandani, N. Mac 3059–3066.
Dowell, et al., Carbon capture and storage update, Energy Environ. Sci. 7 (1) [40] M. Sevilla, P. Valle-Vigón, A.B. Fuertes, N-doped polypyrrole-based porous
(2014) 130–189. carbons for CO2 capture, Adv. Funct. Mater. 21 (14) (2011) 2781–2787.
[10] S. Choi, J.H. Drese, C.W. Jones, Adsorbent materials for carbon dioxide capture [41] S.-H. Yoon, S. Lim, Y. Song, Y. Ota, W. Qiao, A. Tanaka, et al., KOH activation of
from large anthropogenic point sources, ChemSusChem. 2 (9) (2009) 796–854. carbon nanofibers, Carbon 42 (8) (2004) 1723–1729.
[11] T.C. Drage, C.E. Snape, L.A. Stevens, J. Wood, J. Wang, A.I. Cooper, et al., [42] N. Sun, C. Sun, H. Liu, J. Liu, L. Stevens, T. Drage, et al., Synthesis,
Materials challenges for the development of solid sorbents for post- characterization and evaluation of activated spherical carbon materials for
combustion carbon capture, J. Mater. Chem. 22 (7) (2012) 2815–2823. CO2 capture, Fuel 113 (2013) 854–862.
[12] A.S. Bhown, B.C. Freeman, Analysis and status of post-combustion carbon [43] E.S. Kikkinides, R.T. Yang, S.H. Cho, Concentration and recovery of carbon
dioxide capture technologies, Environ. Sci. Technol. 45 (20) (2011) 8624–8632. dioxide from flue gas by pressure swing adsorption, Ind. Eng. Chem. Res. 32
[13] A. Goeppert, M. Czaun, R.B. May, G.S. Prakash, G.A. Olah, S. Narayanan, Carbon (11) (1993) 2714–2720.
dioxide capture from the air using a polyamine based regenerable solid [44] S. Himeno, T. Komatsu, S. Fujita, High-pressure adsorption equilibria of
adsorbent, J. Am. Chem. Soc. 133 (50) (2011) 20164–20167. methane and carbon dioxide on several activated carbons, J. Chem. Eng. Data
[14] J.-R. Li, Y. Ma, M.C. McCarthy, J. Sculley, J. Yu, H.-K. Jeong, et al., Carbon dioxide 50 (2) (2005) 369–376.
capture-related gas adsorption and separation in metal-organic frameworks, [45] W.R. Alesi, M. Gray, J.R. Kitchin, CO2 adsorption on supported molecular
Coord. Chem. Rev. 255 (15) (2011) 1791–1823. amidine systems on activated carbon, ChemSusChem. 3 (8) (2010) 948–956.
[15] N.R. Stuckert, R.T. Yang, CO2 capture from the atmosphere and simultaneous [46] G.P. Hao, W.C. Li, D. Qian, A.H. Lu, Rapid synthesis of nitrogen-doped porous
concentration using zeolites and amine-grafted SBA-15, Environ. Sci. Technol. carbon monolith for CO2 capture, Adv. Mater. 22 (7) (2010) 853–857.
45 (23) (2011) 10257–10264. [47] M. Plaza, C. Pevida, A. Arenillas, F. Rubiera, J. Pis, CO2 capture by adsorption
[16] K. Sumida, D.L. Rogow, J.A. Mason, T.M. McDonald, E.D. Bloch, Z.R. Herm, et al., with nitrogen enriched carbons, Fuel 86 (14) (2007) 2204–2212.
Carbon dioxide capture in metal-organic frameworks, Chem. Rev. 112 (2) [48] M.S. Shafeeyan, W.M.A.W. Daud, A. Houshmand, A. Arami-Niya, Ammonia
(2011) 724–781. modification of activated carbon to enhance carbon dioxide adsorption: effect
[17] L.K. de Souza, N.P. Wickramaratne, A.S. Ello, M.J. Costa, C.E. da Costa, M. of pre-oxidation, Appl. Surf. Sci. 257 (9) (2011) 3936–3942.
Jaroniec, Enhancement of CO2 adsorption on phenolic resin-based mesoporous [49] M.-M. Titirici, Sustainable Carbon Materials from Hydrothermal Processes,
carbons by KOH activation, Carbon 65 (2013) 334–340. Wiley Online Library, 2013.
[18] Y.C. Park, S.-H. Jo, C.K. Ryu, C.-K. Yi, Demonstration of pilot scale carbon [50] L.C. Dolores, P.M.L. Juan, C. Falco, M.M. Titirici, C.A. Diego. Porous Biomass-
dioxide capture system using dry regenerable sorbents to the real coal-fired Derived Carbons: Activated Carbons. Sustainable Carbon Materials from
power plant in Korea, Energy Procedia 4 (2011) 1508–1512. Hydrothermal Processes, 2013, pp. 75–100.
[19] J. Wang, S. Kaskel, KOH activation of carbon-based materials for energy [51] V. Verheyen, R. Rathbone, M. Jagtoyen, F. Derbyshire, Activated extrudates by
storage, J. Mater. Chem. 22 (45) (2012) 23710–23725. oxidation and KOH activation of bituminous coal, Carbon 33 (6) (1995) 763–
[20] C. Pevida, T. Drage, C.E. Snape, Silica-templated melamine-formaldehyde resin 772.
derived adsorbents for CO2 capture, Carbon 46 (11) (2008) 1464–1474. [52] J.B. Parra Soto, J. De Sousa, R.C. Bansal, J. Pis Martínez, Characterization of
[21] Y. Juan, Q. Ke-qiang, Preparation of activated carbon by chemical activation activated carbons by the BET equation: an alternative approach, Adsorpt. Sci.
under vacuum, Environ. Sci. Technol. 43 (9) (2009) 3385–3390. Technol. 12 (1) (1995) 51–66.
[22] C. Pevida, M. Plaza, B. Arias, J. Fermoso, F. Rubiera, J. Pis, Surface modification [53] K.S. Sing, Reporting physisorption data for gas/solid systems with special
of activated carbons for CO2 capture, Appl. Surf. Sci. 254 (22) (2008) 7165– reference to the determination of surface area and porosity
7172. (Recommendations 1984), Pure Appl. Chem. 57 (4) (1985) 603–619.
[23] J. Figueiredo, M. Pereira, M. Freitas, J. Orfao, Modification of the surface [54] A. Oya, S. Yoshida, J. Alcaniz-Monge, A. Linares-Solano, Formation of
chemistry of activated carbons, Carbon 37 (9) (1999) 1379–1389. mesopores in phenolic resin-derived carbon fiber by catalytic activation
[24] P. Chingombe, B. Saha, R. Wakeman, Surface modification and characterisation using cobalt, Carbon 33 (8) (1995) 1085–1090.
of a coal-based activated carbon, Carbon 43 (15) (2005) 3132–3143. [55] S.R. Tennison, Phenolic-resin-derived activated carbons, Appl. Catal. A 173 (2)
[25] M.S. Shafeeyan, W.M.A.W. Daud, A. Houshmand, A. Shamiri, A review on (1998) 289–311.
surface modification of activated carbon for carbon dioxide adsorption, J. Anal. [56] M. Lillo-Ródenas, D. Cazorla-Amorós, A. Linares-Solano, Understanding
Appl. Pyrol. 89 (2) (2010) 143–151. chemical reactions between carbons and NaOH and KOH: an insight into the
[26] F. Rodriguez-Reinoso, M. Molina-Sabio, Activated carbons from lignocellulosic chemical activation mechanism, Carbon 41 (2) (2003) 267–275.
materials by chemical and/or physical activation: an overview, Carbon 30 (7) [57] E. Raymundo-Piñero, P. Azaïs, T. Cacciaguerra, D. Cazorla-Amorós, A. Linares-
(1992) 1111–1118. Solano, F. Béguin, KOH and NaOH activation mechanisms of multiwalled
[27] D. Vargas, L. Giraldo, A. Erto, J. Moreno-Piraján, Chemical modification of carbon nanotubes with different structural organisation, Carbon 43 (4) (2005)
activated carbon monoliths for CO2 adsorption, J. Therm. Anal. Calorim. 114 (3) 786–795.
(2013) 1039–1047. [58] J. Romanos, M. Beckner, T. Rash, L. Firlej, B. Kuchta, P. Yu, et al., Nanospace
[28] A. Erto, A. Silvestre-Albero, J. Silvestre-Albero, F. Rodríguez-Reinoso, M. engineering of KOH activated carbon, Nanotechnology 23 (1) (2012)
Balsamo, A. Lancia, et al., Carbon-supported ionic liquids as innovative 015401.
adsorbents for CO2 separation from synthetic flue-gas, J. Colloid Interface [59] P.I. Neel, B. Viswanathan, T. Varadarajan, Methods of activation and specific
Sci. 448 (2015) 41–50. applications of carbon materials, 2009.
[29] F. Montagnaro, A. Silvestre-Albero, J. Silvestre-Albero, F. Rodríguez-Reinoso, A. [60] D. Lozano-Castello, J. Calo, D. Cazorla-Amoros, A. Linares-Solano, Carbon
Erto, A. Lancia, et al., Post-combustion CO2 adsorption on activated carbons activation with KOH as explored by temperature programmed techniques, and
with different textural properties, Microporous Mesoporous Mater. (2014). the effects of hydrogen, Carbon 45 (13) (2007) 2529–2536.
[30] A. Wahby, J. Silvestre-Albero, A. Sepúlveda-Escribano, F. Rodríguez-Reinoso, [61] T. Otowa, R. Tanibata, M. Itoh, Production and adsorption characteristics of
CO2 adsorption on carbon molecular sieves, Microporous Mesoporous Mater. MAXSORB: high-surface-area active carbon, Gas Sep. Purif. 7 (4) (1993) 241–
164 (2012) 280–287. 245.
[31] Y. Liu, J. Wilcox, Effects of surface heterogeneity on the adsorption of CO2 in [62] A. Ahmadpour, D. Do, The preparation of activated carbon from macadamia
microporous carbons, Environ. Sci. Technol. 46 (3) (2012) 1940–1947. nutshell by chemical activation, Carbon 35 (12) (1997) 1723–1732.
[32] J. Silvestre-Albero, A. Wahby, A. Sepúlveda-Escribano, M. Martínez-Escandell, [63] K. Kierzek, E. Frackowiak, G. Lota, G. Gryglewicz, J. Machnikowski,
K. Kaneko, F. Rodríguez-Reinoso, Ultrahigh CO2 adsorption capacity on carbon Electrochemical capacitors based on highly porous carbons prepared by KOH
molecular sieves at room temperature, Chem. Commun. 47 (24) (2011) 6840– activation, Electrochim. Acta 49 (4) (2004) 515–523.
6842. [64] N.P. Wickramaratne, M. Jaroniec, Importance of small micropores in CO2
[33] J. Wang, A. Heerwig, M.R. Lohe, M. Oschatz, L. Borchardt, S. Kaskel, Fungi-based capture by phenolic resin-based activated carbon spheres, J. Mater. Chem. A 1
porous carbons for CO2 adsorption and separation, J. Mater. Chem. 22 (28) (1) (2013) 112–116.
(2012) 13911–13913. [65] Y. Zhu, S. Murali, M.D. Stoller, K. Ganesh, W. Cai, P.J. Ferreira, et al., Carbon-
[34] Y. Xia, R. Mokaya, G.S. Walker, Y. Zhu, Superior CO2 adsorption capacity on N- based supercapacitors produced by activation of graphene, Science 332 (6037)
doped, high-surface-area, microporous carbons templated from zeolite, Adv. (2011) 1537–1541.
Energy Mater. 1 (4) (2011) 678–683. [66] R. Krishna, Adsorptive separation of CO2/CH4/CO gas mixtures at high
[35] L. Zhao, L.Z. Fan, M.Q. Zhou, H. Guan, S. Qiao, M. Antonietti, et al., Nitrogen- pressures, Microporous Mesoporous Mater. 156 (2012) 217–223.
containing hydrothermal carbons with superior performance in [67] A. Myers, P. Monson, Adsorption in porous materials at high pressure: theory
supercapacitors, Adv. Mater. 22 (45) (2010) 5202–5206. and experiment, Langmuir 18 (26) (2002) 10261–10273.
[36] Y. Zhao, X. Liu, K.X. Yao, L. Zhao, Y. Han, Superior capture of CO2 achieved by [68] D.B. Robinson, D.-Y. Peng, S.Y. Chung, The development of the Peng–Robinson
introducing extra-framework cations into N-doped microporous carbon, equation and its application to phase equilibrium in a system containing
Chem. Mater. 24 (24) (2012) 4725–4734. methanol, Fluid Phase Equilib. 24 (1) (1985) 25–41.
[37] B. Zhu, K. Li, J. Liu, H. Liu, C. Sun, C.E. Snape, et al., Nitrogen-enriched and [69] D.B. Robinson, D.-Y. Peng, The Characterization of the Heptanes and Heavier
hierarchically porous carbon macro-spheres–ideal for large-scale CO2 capture, Fractions for the GPA Peng-Robinson Programs, Gas Processors Association,
J. Mater. Chem. A 2 (15) (2014) 5481–5489. 1978.
[38] D. Lozano-Castello, M. Lillo-Rodenas, D. Cazorla-Amoros, A. Linares-Solano, [70] Y. Zhao, L. Zhao, K.X. Yao, Y. Yang, Q. Zhang, Y. Han, Novel porous carbon
Preparation of activated carbons from Spanish anthracite: I. Activation by materials with ultrahigh nitrogen contents for selective CO2 capture, J. Mater.
KOH, Carbon 39 (5) (2001) 741–749. Chem. 22 (37) (2012) 19726–19731.
 J. Liu et al. / CARBON 94 (2015) 243–255 255

[71] W. Zhang, H. Liu, C. Sun, T.C. Drage, C.E. Snape, Performance of [82] M.J. Bleda-Martínez, J.A. Maciá-Agulló, D. Lozano-Castelló, E. Morallón, D.
polyethyleneimine–silica adsorbent for post-combustion CO2 capture in a Cazorla-Amorós, A. Linares-Solano, Role of surface chemistry on electric
bubbling fluidized bed, Chem. Eng. J. 251 (2014) 293–303. double layer capacitance of carbon materials, Carbon 43 (13) (2005) 2677–
[72] J.A. Mason, K. Sumida, Z.R. Herm, R. Krishna, J.R. Long, Evaluating metal– 2684.
organic frameworks for post-combustion carbon dioxide capture via [83] A. Pandolfo, A. Hollenkamp, Carbon properties and their role in
temperature swing adsorption, Energy Environ. Sci. 4 (8) (2011) 3030–3040. supercapacitors, J. Power Sources 157 (1) (2006) 11–27.
[73] Z.R. Herm, R. Krishna, J.R. Long, CO2/CH4, CH4/H2 and CO2/CH4/H2 separations [84] E. Frackowiak, Carbon materials for supercapacitor application, Phys. Chem.
at high pressures using Mg2(dobdc), Microporous Mesoporous Mater. 151 Chem. Phys. 9 (15) (2007) 1774–1785.
(2012) 481–487. [85] T.C. Drage, J.M. Blackman, C. Pevida, C.E. Snape, Evaluation of activated carbon
[74] T. Ben, Y. Li, L. Zhu, D. Zhang, D. Cao, Z. Xiang, et al., Selective adsorption of adsorbents for CO2 capture in gasification, Energy Fuels 23 (5) (2009) 2790–
carbon dioxide by carbonized porous aromatic framework (PAF), Energy 2796.
Environ. Sci. 5 (8) (2012) 8370–8376. [86] J. Liu, P.K. Thallapally, B.P. McGrail, D.R. Brown, J. Liu, Progress in adsorption-
[75] S.M. Mahurin, J. Gorka, K.M. Nelson, R.T. Mayes, S. Dai, Enhanced CO2/N2 based CO2 capture by metal-organic frameworks, Chem. Soc. Rev. 41 (6) (2012)
selectivity in amidoxime-modified porous carbon, Carbon 67 (2014) 457–464. 2308–2322.
[76] W. Lu, J.P. Sculley, D. Yuan, R. Krishna, H.-C. Zhou, Carbon dioxide capture from [87] A. Martin-Calvo, J.B. Parra, C. Ania, S. Calero, Insights on the anomalous
air using amine-grafted porous polymer networks, J. Phys. Chem. C 117 (8) adsorption of carbon dioxide in LTA zeolites, J. Phys. Chem. C 118 (44) (2014)
(2013) 4057–4061. 25460–25467.
[77] T.M. McDonald, D.M. D’Alessandro, R. Krishna, J.R. Long, Enhanced carbon [88] S.C. Lee, H.J. Chae, S.J. Lee, B.Y. Choi, C.K. Yi, J.B. Lee, et al., Development of
dioxide capture upon incorporation of N, N0 -dimethylethylenediamine in the regenerable MgO-based sorbent promoted with K2CO3 for CO2 capture at low
metal-organic framework CuBTTri, Chem. Sci. 2 (10) (2011) 2022–2028. temperatures, Environ. Sci. Technol. 42 (8) (2008) 2736–2741.
[78] P.M. Mathias, R. Kumar, J.D. Moyer, J.M. Schork, S.R. Srinivasan, S.R. Auvil, et al., [89] W. Zhang, H. Liu, C. Sun, T.C. Drage, C.E. Snape, Capturing CO2 from ambient air
Correlation of multicomponent gas adsorption by the dual-site Langmuir using a polyethyleneimine–silica adsorbent in fluidized beds, Chem. Eng. Sci.
model. Application to nitrogen/oxygen adsorption on 5A-zeolite, Ind. Eng. 116 (2014) 306–316.
Chem. Res. 35 (7) (1996) 2477–2483. [90] S. Sjostrom, H. Krutka, Evaluation of solid sorbents as a retrofit technology for
[79] R.C. Bansal, M. Goyal, Activated carbon adsorption, CRC Press, 2010. CO2 capture, Fuel 89 (6) (2010) 1298–1306.
[80] C. Shen, C.A. Grande, P. Li, J. Yu, A.E. Rodrigues, Adsorption equilibria and [91] M. Gray, J. Hoffman, D. Hreha, D. Fauth, S. Hedges, K. Champagne, et al.,
kinetics of CO2 and N2 on activated carbon beads, Chem. Eng. J. 160 (2) (2010) Parametric study of solid amine sorbents for the capture of carbon dioxide,
398–407. Energy Fuels 23 (10) (2009) 4840–4844.
[81] M. Bülow, D. Shen, S. Jale, Measurement of sorption equilibria under isosteric [92] A. Ebner, M. Gray, N. Chisholm, Q. Black, D. Mumford, M. Nicholson, et al.,
conditions: the principles, advantages and limitations, Appl. Surf. Sci. 196 (1) Suitability of a solid amine sorbent for CO2 capture by pressure swing
(2002) 157–172. adsorption, Ind. Eng. Chem. Res. 50 (9) (2011) 5634–5641.