Energy 63 (2013) 252e259

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

HBGS (hydrate based gas separation) process for carbon dioxide

capture employing an unstirred reactor with cyclopentane
Leong Chuan Ho a,1, Ponnivalavan Babu a,1, Rajnish Kumar b, Praveen Linga a, *
a
Department of Chemical and Biomolecular Engineering, National University of Singapore, Singapore 117 576, Singapore
b
Chemical Engineering and Process Development Division, National Chemical Laboratory, Pune, India

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

Article history: The effect of CP (cyclopentane) as a promoter/additive, in the HBGS (hydrate based gas separation)
Received 20 March 2013 process for pre-combustion gas mixture was investigated by employing an unstirred reactor configu-
Received in revised form ration. Gas uptake measurements were performed at two different temperatures (275.7 K and 285.7 K)
24 August 2013
and at an experimental pressure of 6.0 MPa to determine the kinetics of hydrate formation. Experiments
Accepted 11 October 2013
Available online 9 November 2013
were conducted with three different volumes (7.5, 15 and 22 ml) of CP and based on induction time and
the rate of hydrate growth, 15 ml of CP was determined to be the optimal volume for carbon dioxide
capture at 6.0 MPa and 275.7 K. In addition, the effect of a kinetic promoter, SDS (sodium dodecyl sul-
Keywords:
Gas hydrates
fate), was investigated. Surprisingly, no improvement in kinetic performance was observed at 6.0 MPa
Gas separation and 275.7 K in the presence of SDS and CP. From the study, it was found that at the optimal 15 ml CP (CP
Pre-combustion layer thickness of 1.8 mm), the average composition of carbon dioxide in the hydrate phase was
Cyclopentane 90.36 mol% with a separation factor of 17.82. Furthermore, the unstirred reactor also yielded better ki-
Carbon dioxide capture netic performance over the stirred tank reactor with the unstirred reactor having a 2.28 times higher
Unstirred reactor average gas uptake.
Ó 2013 Elsevier Ltd. All rights reserved.

1. Introduction the process behind post-combustion capture and oxy-fuel com-

Increasing emissions of greenhouse gases, like carbon dioxide,
have been established as a cause for global warming, and there is
therefore a need to research into the technologies behind carbon
dioxide capture. With reference to a report by the IPCC (Intergov-
ernmental Panel on Climate Changes) [1], there exist three main
types of carbon dioxide capture from point sources, post-
combustion capture, pre-combustion capture, and oxy-fuel com-
bustion. Both post-combustion capture and oxy-fuel combustion
involve the capture of carbon dioxide from flue gases but with key
differences in the combustion process. In post-combustion capture,
the flue gas produced is a result of combustion between fuel and air
while in oxy-fuel combustion; oxygen with high purity is used in
place of air. This result in the carbon dioxide concentration in oxy-
fuel combustion to be significantly higher than in post-combustion
capture (>80 vol% in oxy-fuel combustion compared to 3 vol%e
15 vol% carbon dioxide in post-combustion capture). As the focus of
this paper will be on the pre-combustion capture of carbon dioxide,
* Corresponding author. Tel.: þ65 6601 1487; fax: þ65 6779 1936.
E-mail addresses: chepl@nus.edu.sg, praveen.linga@nus.edu.sg (P. Linga).
1
Both authors equally contributed to the work.
bustion will not be further detailed.
In pre-combustion, fossil fuels are partially oxidized, resulting in
the formation of “synthesis gas” (syngas), which consists of a
mixture of carbon monoxide and hydrogen. The syngas then un-
dergoes a wateregas shift reaction with steam to yield a carbon
dioxideehydrogen gas mixture, which is called as “Fuel gas”
mixture. Carbon dioxide can then be captured while hydrogen can
be used as the fuel for combustion. The presence of a relatively pure
emission stream with hydrogen as the output is one of the key
advantage of pre-combustion capture as the combustion process of
hydrogen is carbon-free and will not contribute to the emission of
greenhouse gases. Furthermore, the above processes produce gases
at high pressure that will be favorable for separation processes.
There exists a myriad of options available for carbon dioxide
capture [1e5] and the focus of this paper will be on HBGS (hydrate
based gas separation) process. Gas hydrates, also known as clath-
rate hydrates, are non-stoichiometric crystalline structures made
up of water and gas molecules. Water molecules act as a “cage” with
hydrogen bonds between the molecules while gas molecules are
held within it. Gas hydrate formation is an exothermic process and
it occurs above a specific temperature and pressure that are a
function of the type of gas involved, and the amount of gas and
water used. With the better understanding of gas hydrate

0360-5442/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.energy.2013.10.031
 L.C. Ho et al. / Energy 63 (2013) 252e259
formation over the years, there has been a growing interest in the
research of gas hydrate due to the large potential across many ap-
plications. Some of the possible applications include the use of gas
hydrate for transportation of natural gas/hydrogen [6e11], a source
of energy [12e18], and as a technology for carbon dioxide capture,
storage/sequestration [19e24].
The equilibrium condition for hydrate formation is a function of
the type of gas involved. At 280 K, the dissociation pressure of
hydrogen gas is 300 MPa while the dissociation pressure of carbon
dioxide is 2.91 MPa [25,26]. As such, the use of gas hydrate sepa-
ration as a means to capture carbon dioxide from pre-combustion
gas mixture is promising as carbon dioxide gas will be preferably
held in the water molecules “cage” at lower pressures [19,27,28].
Linga et al. [19] proposed a two stage HBGS process for capturing
carbon dioxide from a pre-combustion fuel gas mixture consisting
of 40% CO2 and rest H2 with the first stage requiring an operating
pressure of 7.5 MPa at 273.65 K. As the equilibrium pressure for this
gas mixture at 273.65 is 5.1 MPa [29], a higher operating pressure
provides the necessary driving force for gas hydrate formation to
occur at an appreciable rate, which also results in an increase in
compression costs for the HBGS process. There is therefore an
ongoing effort to search for additives that can lower the equilibrium
conditions required for gas hydrates.
In an attempt to lower the operating pressure required for the
HBGS process, the use of additives has been investigated by several
researchers by employing a STR (stirred tank reactor) [30e34]. In a
two-stage medium pressure hydrate process, Kumar et al. [30]
investigated the use of 2.5 mol% propane as an additive and re-
ported a significantly lower rate of hydrate formation and total
moles of gas consumed. As for the study on THF (tetrahydrofuran)
by Lee et al. [31], it also yielded similar observations with 1.0 mol%
THF reported as the optimal concentration. Kim et al. [35] and Li
et al. [32] investigated the use of TBAB (tetra-n-butyl ammonium
bromide) and the total number of moles of gas consumed on a
water free basis was lower compared to the THF experiments. From
the above investigations, it can be noted that in general, the use of
additives in the stirred tank reactor (STR) resulted in a lower rate of
hydrate formation and lower gas consumption when using the
clathrate process. Furthermore, it has also been established that the
use of stirred tank reactor leads to an agglomeration of hydrate
crystals, creating a barrier at the gas/liquid interface [22,36,37].
There is therefore a need to identify innovative reactor designs
in place of STR. One possible configuration was proposed by Linga
et al. [22] where a modular, mechanically agitated gas-inducing
crystallizer was used for both post- and pre-combustion carbon
dioxide capture. However, even though the gas uptake and sepa-
ration efficiency were higher than a STR, they reported that the cost
involved due to the mechanical agitation was significant and hence
prohibited industrial adoption. Other ways proposed to enhance
the performance of hydrate formation include the use of a fixed bed
reactor with silica gel or silica sand as a medium [38e40], a fluid-
ized bed reactor by employing a slurry bed or a gas bubble column
[41e43] or using an unstirred reactor configuration [44,45].
One of the possible approaches to overcome the above limita-
tions (higher operating pressure and effective gaseliquid contact)
is the use of CP (cyclopentane), as a promoter. CP has been iden-
tified as a possible candidate in the literature [26,33,45e48] and
has been shown to lower the equilibrium pressure and temperature
required for hydrate formation. Furthermore, CP can be used in an
unstirred reactor configuration [45], which is also associated with
an additional cost advantage in the practical implementation of the
HBGS process as highlighted by Linga et al. [22]. Zhang et al. [45]
reported faster kinetics for pure CO2 hydrates in an unstirred
reactor configuration. Li et al. [33] investigated the CO2 separation
efficiencies of a fuel gas stream with CP in a stirred tank reactor and
253
reported a highest gas uptake of 0.003 mol of gas/mole of water for
the 36.0 ml CP at 4.5 MPa and 273.15 K. Recently, Lim et al. [48]
investigated the morphology of CO2/H2/CP mixed hydrates and
presented the mechanism of hydrate formation in quiescent con-
ditions. They proposed that for experiments with a CP layer thick-
ness of 1.8 mm, the gas diffuses into the CP layer and reaches the
water layer and nucleation begins at the CPewater interface, hy-
drates then begin to grow upward and radially inward towards the
center of the reactor. The formation of hydrate layer then further
resulted in a faster rate of gas consumption for hydrate formation.
The objective of this paper is to evaluate the macroscopic ki-
netics of cyclopentane (CP) as a promoter for effective separation of
CO2 from CO2/H2 mixture at moderate pressures in an unstirred
reactor configuration. Effect of the use of CP with SDS (sodium
dodecyl sulfate), a kinetic promoter, is also investigated with
varying temperature as a driving force for both sets of experiments.
Finally, the kinetic comparison between STR and unstirred reactor
is also presented.
2. Experimental
2.1. Materials
A 40 mol% CO2/60 mol% H2 gas mixture, provided by Soxal Pte., Ltd.,
was used in the experiment. Cyclopentane with purity 99.98% and SDS
with purity 99% were supplied by Alfa Aesar and Amresco respec-
tively. Distilled and deionized water was used in all experiments.
2.2. Apparatus
2.2.1. Unstirred reactor
The apparatus for the hydrate kinetic experiments has been
described elsewhere in the literature [40]. The adapted schematic of
the experimental setup is shown in Fig.1. Briefly, the setup consists of a
crystallizer immersed in a temperature controlled water bath coupled
with an external refrigerator. Thermocouples with uncertainty of 0.1 K
are used to measure the temperature of the gas phase and liquid phase
in the crystallizer. Pressure transmitter and pressure gauge are
employed to measure the pressure of the crystallizer. The temperature
and pressure data are recorded using a data acquisition system
coupled with a computer. Hydrate formation experiments are carried
out in a batch manner (constant volume) at a constant temperature.
2.2.2. Stirred tank reactor (STR)
Detailed description of stirred tank reactor (STR) is available in
our previous work [49]. Briefly, the setup consists of a crystallizer
with a pair of circular sapphire glass viewing windows at the side of
the reactor to observe the hydrate formation process. The in-built
cooling arrangement in the crystallizer is connected to an
external refrigerator to create a temperature-controlled environ-
ment. A thermocouple with uncertainty of 0.1 K is located at the
fixed head of the crystallizer to measure the temperature of the
liquid phase in the crystallizer. Pressure transmitter and pressure
gauge are employed to measure the pressure of the crystallizer. A
mechanical stirrer connected to a controller is also employed to
control the stirring speed (rpm). The temperature and pressure
data are recorded using a data acquisition system coupled with a
computer. Hydrate formation experiments are carried out in a
batch manner (constant volume) at a constant temperature.
2.3. Procedures
2.3.1. Preparation of SDS solution
SDS is a white powdered solid at room temperature and pres-
sure. To prepare the SDS solution, the required amount of SDS (0.1,
254 L.C. Ho et al. / Energy 63 (2013) 252e259
Fig. 1. Schematic of the experimental setup.
0.3 and 1 g) was weighed on a mass balance and transferred into a
volumetric flask with 1 L of distilled water. The solution was then
thoroughly mixed to obtain the required concentration (100, 300
and 1000 ppm) of SDS solution.
2.3.2. Hydrate formation procedure
140 ml of distilled water or SDS solution was transferred to the
crystallizer, followed by the addition of required quantity of CP (7.5
or 15.0 or 22.5 ml). CP formed a clear second phase above the
distilled water or SDS solution. Thermocouples were connected and
the crystallizer was closed, and placed inside the temperature
controlled water bath. The crystallizer was cooled to the experi-
mental temperature (275.7 or 285.7 K). The crystallizer was pres-
surized and depressurized to atmospheric pressure with the feed
gas thrice in order to remove any presence of air bubble in the
system. Subsequently, the crystallizer was then pressurized to
experimental pressure of 6.0 MPa and temperature was allowed to
reach the desired value. This time was recorded as time zero for
formation experiments, and temperature and pressure data was
recorded for every 20 s. As hydrate formation proceeds, gas gets
consumed and hence the pressure in the crystallizer decreases. The
experiment was allowed to continue until no significant change in
the crystallizer pressure was observed.
2.3.2.1. Procedure to determine the amount of gas consumed.
Equation (1) is used to determine the number of moles of gas
consumed at any time during the experiment.
   
P P
Dnt ¼ V V (1)
zRT 0 zRT t
where Dnt is the number of moles of gas consumed after t minutes,
V is the volume of the gas phase in the crystallizer, P is the pressure
of the crystallizer, T is the temperature of the liquid phase, R is the
ideal gas constant and z is the compressibility factor, which is
calculated by Pitzer’s correlation [50].
2.3.2.2. Calculation of average rate of hydrate formation. To deter-
mine the rate of hydrate formation, the rate of gas consumed from
induction time is first determined using the forward difference
 L.C. Ho et al. / Energy 63 (2013) 252e259 255

Table 1
Summary of experimental conditions and results.

Exp. no. System Pressure Temperature Driving Induction Gas consumption, 5 h CO2 composition in S.F.
(MPa) (K) force (K)a time, IT (h) after IT hydrate phase
(mol of gas/mol of water)

1 15 ml CP 6.0 285.7 5.1 5.49 0.0133 e e

2 15 ml CP 6.0 285.7 5.1 4.64 0.0133 77.79 16.33
3 15 ml CP 6.0 275.7 15.1 0.38 0.0338 89.02 16.98
4 15 ml CP 6.0 275.7 15.1 0.15 0.0355 91.97 18.26
5 15 ml CP 6.0 275.7 15.1 0.27 0.0321 90.09 18.21
6 22.5 ml CP 6.0 275.7 15.1 1.53 0.0237 90.59 18.17
7 22.5 ml CP 6.0 275.7 15.1 0.62 0.0350 91.02 18.86
8 7.5 ml CP 6.0 275.7 15.1 0.56 0.0295 89.36 15.98
9 7.5 ml CP 6.0 275.7 15.1 0.60 0.0313 88.96 17.46
10 15 ml CP/300 ppm SDS 6.0 275.7 15.1 0.76 0.0282 87.67 14.99
11 15 ml CP/300 ppm SDS 6.0 275.7 15.1 0.54 0.0294 87.76 19.86
12 15 ml CP/1000 ppm SDS 6.0 275.7 15.1 0.77 0.0240 78.08 13.74
13 15 ml CP/1000 ppm SDS 6.0 275.7 15.1 0.52 0.0281 e e
14 15 ml CP/100 ppm SDS 6.0 275.7 15.1 0.59 0.0261 89.44 15.89
15 15 ml CP/100 ppm SDS 6.0 275.7 15.1 0.72 0.0280 88.38 15.78
16 15 ml CP/100 ppm SDS 6.0 275.7 15.1 1.14 0.0269 88.67 15.61
17 15 ml CP/300 ppm SDS 6.0 285.7 5.1 13.67 0.0174 e e
18 15 ml CP/300 ppm SDS 6.0 285.7 5.1 11.62 0.0171 83.99 19.01
19 STR 15 ml CP 6.0 275.7 15.1 0.41 0.0166 85.93 e
20 STR 15 ml CP 6.0 275.7 15.1 0.32 0.0134 91.06 e
21 STR 15 ml CP 6.0 275.7 15.1 0.33 0.0135 88.03 e
22 15 ml CP 4.5 275.7 13.8 0.15 0.0188 e e
a
Equilibrium temperature of system at 4.5 MPa and 6.0 MPa is 289.5 K and 290.8 K respectively [46].

method from the 5 min data of the number of moles of gas occupation of CO2 in hydrate phase than that of H2. With the
consumed per mole of water [22]. Subsequently, the average of 7 composition of the carbon dioxide and hydrogen gas determined,
points between each 30 min interval is calculated to obtain the plot equation (3) is used to determine the separation factor [19]:
of rate (mole of gas consumed/mole of water/h) against time [40].
The forward difference method is shown in equation (2). nH Gas
CO2 ,nH2
Separation factorðS:F:Þ ¼ (3)
H2 ,nCO2
nH Gas
 
dDnH ðDnH ÞtþDt  ðDnH Þt
¼ ; Dt ¼ 5 min (2)
dt t Dt where nGas
CO2 is the number of moles of carbon dioxide in the gas
phase at the end of the formation experiment, nGas
H2 is the number of
After completion of the formation experiment, three samples of
moles of hydrogen in the gas phase at the end of the experiment,
the gas mixture were collected to determine the composition using
nH
CO2 is the number of moles of carbon dioxide gas in the hydrate
the GC (Gas Chromatography). The hydrates were then allowed to
phase, and nH H2 is the number of moles of hydrogen gas in the hy-
decompose by rapid depressurization to atmospheric pressure.
drate phase.
Temperature of the water bath was then increased to 35.0  C. Once
the pressure profile was steady, another three samples of the gas
mixture were collected and the composition of the dissociated gas 3. Results and discussion
from hydrates was determined using GC.
Table 1 summarizes the results for all kinetic experiments in the
2.3.2.3. Calculation of separation factor. Linga et al. [19] proposed a presence of various concentrations of CP and SDS at the experi-
metric called separation factor that indicates the preferential mental temperature of 275.7 K and 285.7 K, at 6.0 MPa, for both
unstirred and STR configurations. Pressure of 6.0 MPa was chosen
as a typical fuel gas mixture coming out of IGCC power plant is in
the range of 2.0e7.0 MPa [1].

3.1. Typical gas consumption curve

A typical gas consumption curve at 285.7 K, 6.0 MPa, and 15 ml

CP, is shown in Fig. 2 for a period of 16 h. The general shape of the
gas consumption curve is in line with the detailed description in the
literature [25,51] and three distinct phases were observed. Prior to
first nucleation event (A), dissolution phase can be observed, fol-
lowed by the hydrate growth phase. The second nucleation (B)
resulted in a second growth phase, as can be seen from the sharp
increase in gas uptake. For Experiment 1, the induction time was
5.49 h (A) and a second nucleation (B) was observed at approxi-
mately 12 h. The gas uptake during hydrate formation reached
0.0185 mol/mol of water after 16 h. A temperature spike was
Fig. 2. Gas uptake measurement curve at 285.7 K, 6.0 MPa, and 15 ml CP (Experiment observed at nucleation since hydrate formation is an exothermic
1). process. The temperature of the crystallizer was then restored to
256 L.C. Ho et al. / Energy 63 (2013) 252e259
Fig. 3. Gas uptake measurement curve at 275.7 K, 6.0 MPa, and 15 ml CP (Experiment
3).
the experimental temperature of 285.7 K with the help of the
temperature controlled water bath. As seen in Fig. 2, heat transfer is
effective for this system even in the unstirred configuration as
temperature was restored to 285.7 K in a short period of time.
Finally, the system reached steady state when there is no further
hydrate formation.
The gas uptake curve for an experiment conducted at the lower
temperature of 275.7 and 6.0 MPa with 15 ml CP is shown in Fig. 3.
It is interesting to note in Fig. 3 that apart from 5 multiple nucle-
ation events (AeE), the induction time was shorter as can also be
seen in Table 1. It can also be observed in Fig. 3 that the experiment
reached steady state in much shorter time period when the
experiment is conducted at a lower temperature of 275.7 K
compared to 285.7 K. The induction time was at 0.38 h and gas
uptake during hydrate formation increased to 0.0338 mol/mol of
water after 5 h.
3.2. Effect of driving force
Driving force is defined as the difference in operating temper-
ature or pressure from the equilibrium temperature or pressure.
Driving forces can be quantified as either the degree of sub-cooling
(DT ¼ Tequilibrium  Texperimental) or the pressure difference
(DP ¼ Pexperimental  Pequilibrium). The equilibrium temperature for
Fig. 4. Effect of driving force on hydrate growth at 6.0 MPa, 15 ml CP. [Time zero
corresponds to induction time or first nucleation point].
Fig. 5. Effect of driving force on average rate of hydrate growth at 6.0 MPa, 15 ml CP.
[Time zero corresponds to induction time or first nucleation point].
40 mol% CO2/60 mol% H2/CP at 6.0 MPa is 290.8 K [46]. To deter-
mine the effect of driving force, experiments were conducted at
275.7 and 285.7 K (driving force of 15.1 K and 5.1 K). The higher
driving force resulted in a shorter induction time. With reference to
Table 1, comparing Experiments 1e2 (DT ¼ 5.1 K) and Experiments
3e5 (DT ¼ 15.1 K), the average induction time decreased by a factor
of 19, from 5.07 h to 0.27 h. Fig. 4 shows the effect of driving force
on the gas consumption for hydrate growth. As expected, the gas
consumed for hydrate growth increased when the driving force was
increased. The average gas consumption after 5 h from first in-
duction point increased by 2.92 times for DT ¼ 15.1 K (0.0338 mol of
gas/mol of water) than that experiments at DT ¼ 5.1 K (0.0133 mol
of gas/mol of water). Fig. 5 shows the rate of hydrate formation at
two different driving forces employed in this study. The initial rate
of hydrate formation was significantly higher at higher driving
force for first few hours. From a practical point of view, shorter
induction time and faster hydrate growth rates are desirable for
implementation of the HBGS process for pre-combustion capture of
CO2. The experimental condition of 6.0 MPa and 275.7 K is favorable
compared to 285.7 K in terms of achieving higher gas consumption
for hydrate formation and shorter induction time.
Fig. 6. Effect of volume of CP on average induction time at 6.0 MPa, 275.7 K.
 L.C. Ho et al. / Energy 63 (2013) 252e259 257

3.3. Effect of volume of CP used average hydrate growth for the experiments conducted for the
three CP volumes. From the figure it can be seen that for experi-
With the improved kinetics achieved at operating conditions of ments with 15 ml CP the gas consumption rate due to hydrate
6.0 MPa and 275.7 K, the effect of volume of CP (7.5 ml, and growth were slightly higher than that of experiments with 7.5 ml
22.5 ml CP) on hydrate formation kinetics was investigated. Li CP and significantly higher than the experiments conducted with
et al. [33] reported an optimum 5 vol% of CP/TBAB solution ratio at 22.5 ml CP. Based on our results, it can be concluded that 15 ml of CP
4.0 MPa and 274.65 K in a STR, which translates to a volume of (layer thickness of 1.8 mm) yields the optimal kinetics with the
7 ml of CP for 140 ml of distilled water used. A more realistic and lowest induction time (Fig. 6), and higher initial hydrate growth
comparable way of presenting the presence of CP is by the rate (Fig. 7).
thickness of the CP layer for an unstirred reactor configuration
owing to difference in dimension of the reactors and also the fact
3.4. Effect of SDS on hydrate formation for CP experiments
that CP is immiscible in water. According to our reactor di-
mensions 7.5, 15.0 and 22.5 ml of CP will translate to a thickness of
Sodium dodecyl sulfate (SDS), an anionic surfactant has been
0.9, 1.8 and 2.7 mm respectively. Fig. 6 shows the average induc-
reported to be the best among surfactants in the literature for
tion time for the different CP volumes or CP layer thickness
different hydrate forming systems [52e54]. The effect of SDS on
employed in this work. As can be seen, 15 ml of CP (thickness of
hydrate formation in the presence of CP was investigated. Wata-
1.8 mm) resulted in the shortest induction time compared to 7.5
nabe et al. [44] studied the effect of SDS using CH2F2 gas in an
and 22.5 ml of CP. Quantitatively, it can be seen from the figure
unstirred system and reported that the use of SDS beyond an op-
that changing the volume of CP to 7.5 ml and 22.5 ml has led to a
timum concentration can result in a lower rate of hydrate formation
longer induction time from an average of 0.27 h for 15 ml CP to
[44,55]. Hence the optimum concentration of SDS that will yield the
0.58 h for 7.5 ml CP, and 1.08 h for 22.5 ml CP.
best kinetic performance has to be determined.
In theory, the difference in induction time could be attributed to
Using the optimal experimental condition of 6.0 MPa, 275.7 K
the differences in thickness of CP layer above the distilled water in
and 15 ml CP, the concentration of SDS was varied to determine the
the two-phase mixture. As CP is immiscible and has lower density
optimal concentration. Fig. 8 shows the effect of SDS concentration
than water, it forms a clear layer above water. A larger volume of CP
on induction time for the experiments conducted at 275.7 K,
used will result in a larger thickness of the CP layer, which can
6.0 MPa, and 15 ml CP solution. Increasing the SDS concentration
potentially increase the diffusion time required for the gas to come
from 100, 300 and 1000 ppm increased the average induction time
in contact with water for hydrate formation. Hence a longer in-
from 0.82, 0.65 and 0.64 h respectively while for the experiment
duction time for experiments with 22.5 ml of CP was observed than
without SDS, the average induction time was 0.27 h.
that of experiments conducted with 15 ml of CP.
The gas consumption data for the SDS experiments are given in
However, it has to be noted that when the volume of CP is
Table 1. The addition of 100, 300 and 1000 ppm SDS decreased the
reduced to 7.5 ml (0.9 mm CP layer thickness) from 15 ml, the in-
average gas consumption for hydrate formation to 0.0270
duction time does not decrease despite the expected lower thick-
(0.0010), 0.0288 (0.0008) and 0.0261 (0.0029) mol of gas/mol
ness of the CP layer. In this case, the limiting factor for hydrate
of water respectively while the average gas consumption was
formation may not be the diffusion time. One possible explanation
0.0338 mol of gas/mol of water in the absence of SDS. Based on our
could be that the concentration of gas available in the CP layer is not
results, it can be concluded that in an unstirred configuration, SDS
sufficient for nucleation compared to case of when 15 ml of CP is
has an inhibitory effect on the kinetics of hydrate formation for CP
used, thereby increasing the induction time. Recently, Lim et al. [48]
experiments both based on induction time (Fig. 8) and gas con-
observed that with a CP layer thickness of 0.9 mm, the nucleation
sumption (Table 1).
did not happen at the CPewater interface (as was the case for
Similar to CP experiments, the use of higher driving force for CP
1.8 mm CP layer thickness) instead crystals were found to form in
and SDS experiments was also associated with a significantly
the bulk liquid water resulting in longer induction times.
shorter induction time, higher gas uptake during hydrate formation
Effect of the volume of CP or thickness of the CP layer on hydrate
(Fig. S1 given in supplementary information).
growth is shown in Fig. 7. It is noted that the data in the figure is the

Fig. 7. Effect of volume of CP on average gas consumption for hydrate growth at

6.0 MPa, 275.7 K. [Time zero corresponds to induction time or first nucleation point]. Fig. 8. Effect on SDS concentration on average induction time at 6.0 MPa and 275.7 K.
258 L.C. Ho et al. / Energy 63 (2013) 252e259
3.5. Hydrate phase composition
The hydrate phase composition was determined using GC after
decomposing the hydrate crystals by depressurization and thermal
stimulation. The hydrate phase composition for all experiments is
given in Table 1. The average CO2 composition in hydrate phase
increased from 77.79 mol% to 90.36 mol% when the degree of sub-
cooling increased from 5.1 K to 15.1 K in an unstirred reactor
configuration. This can be attributed to the increase in solubility of
carbon dioxide with decreasing temperature. Furthermore, at
higher driving force the gas consumed for hydrate formation is also
high. The average CO2 composition in hydrate phase for experi-
ments in the presence of SDS is lower than that in the absence of
SDS. In STR experiments, the average CO2 composition in the hy-
drate phase was 88.34 mol% while that in an unstirred reactor it
was 90.36 mol% at the same experimental conditions (6.0 MPa,
275.7 K and 1.8 mm thick CP layer).
Separation factor was calculated based on equation (3) and
summarized in Table 1. At the optimal experimental condition of
6.0 MPa, 275.7 K, and 15 ml CP, the average separation factor was
17.82, which means that carbon dioxide is 17.82 times more
favorable than hydrogen in hydrate formation. It is noted that Li
et al. [56] reported 84 mol% CO2 composition in the hydrate phase
and a separation factor of 6.8 at 4.0 MPa, 274.65 K and 9 ml of CP in
a stirred tank reactor.
3.6. Comparison between STR and unstirred reactor for optimal CP
system
In order to evaluate the advantage of employing an unstirred
reactor over stirred tank reactor (STR), experiments were carried
out in a STR at the same experimental conditions of 1.8 mm thick CP
layer, 6.0 MPa and 275.7 K. In order to achieve thickness of 1.8 mm
of CP layer, 5.8 ml of CP was used in the STR. A typical gas con-
sumption curve at 6.0 MPa and 275.7 K in presence of 1.8 mm thick
CP layer in a STR is shown in Fig. S2 (given in supplementary in-
formation). The general characteristics of the gas consumption
curve in STR resemble as that of unstirred reactor (Fig. S2 and
Fig. 3). Similar to unstirred reactor experiments, multiple nucle-
ation events were observed in all STR experiments.
Fig. 9 compares the performance of unstirred reactor with that
of STR at the same experimental conditions. The average gas
consumed for hydrate formation after 5 h from nucleation in an
unstirred reactor (0.0228 mol of gas/mol of water) is 7 times higher
than that in a STR (0.0031 mol of gas/mol of water) at 6.0 MPa,
Fig. 9. Gas consumed for hydrate formation at 6.0 MPa, 275.7 K, 1.8 mm thick CP layer
in STR and unstirred systems. [Time zero corresponds to induction time].
275.7 K and 1.8 mm thick CP layer. There is no significant change in
induction time irrespective of the reactor configuration (unstirred
or STR) at the same experimental conditions (Table 1). The effect of
reactor configuration on the rate of hydrate formation is shown in
Fig. S3 (given in supplementary information). In an unstirred
reactor, the initial average rate of hydrate formation (0.0171 mol/
mol h) is 3.8 times higher than that in STR (0.0045 mol of gas/mol of
water h). Li et al. [33] reported a gas consumption of 0.003 mol of
gas/mol of water for hydrate formation in a STR at experimental
conditions of 4.5 MPa and 273.15 K in the presence of 15 ml of CP.
For a realistic comparison with STR results in literature, we per-
formed an experiment at 275.7 K and 4.5 MPa (experiment 22 in
Table 1) in an unstirred rector configuration. A gas consumption of
0.0188 mol of gas/mol of water was obtained which is 6.3 times
higher than that reported by Li et al. [33] employing an STR.
The superior performance of unstirred reactor over STR is due to
the presence of immiscible CP layer above the water/SDS solution.
CP layer aids in diffusion of guest gas to the bulk phase in an
unstirred reactor as reported recently by Lim et al. [48] while in a
STR, CP is dispersed in the bulk phase due to stirring. Linga et al.
[22] reported that power consumption associated with stirring will
be significant when a STR is employed for HBGS. In case of
employing unstirred reactor for HBGS, apart from enhanced ki-
netics of hydrate formation, the equipment and operation cost
associated with stirring can be completely avoided.
4. Conclusion
Experiments were conducted at two different temperatures of
275.7 and 285.7 K and at 6.0 MPa to determine the effect of CP as a
promoter. For CP systems, 15 ml of CP yielded the optimal kinetic
performance with the average gas uptake of 0.0338 mol of gas/mol
of water at 6.0 MPa and 275.7 K. Subsequently, using the optimal
volume of CP, the effect of different concentration of SDS was
investigated. Surprisingly, at a higher driving force of 15.1 K, the
addition of SDS resulted in a lower rate of hydrate formation and
gas consumption for hydrate formation. At the optimal conditions,
the average composition of carbon dioxide analyzed in the hydrate
phase was 90.36 mol% with a separation factor of 17.82. Further-
more, the unstirred reactor also yielded better kinetic performance
over the stirred tank reactor with the unstirred reactor having a
2.28 times higher gas consumption. The advantages of employing
cyclopentane and an unstirred reactor configuration are that both
capital costs (compression costs) and operating costs can be
reduced for the hydrate based gas separation process.
Acknowledgment
The financial support from the Ministry of Education’s AcRF Tier
1 (R-279-000-317-133) is greatly appreciated. Rajnish Kumar wants
to thank the Council of Scientific and Industrial Research (CSIR) for
the financial support. In addition, the authors would like to thank
Hari Prakash Veluswamy for his help with the STR experiments.
Appendix A. Supplementary information
Supplementary information related to this article can be found
at http://dx.doi.org/10.1016/j.energy.2013.10.031.
References
[1] Metz B, Davidson O, De Coninck H, Loos M, Meyer L. IPCC, 2005: IPCC special
report on carbon dioxide capture and storage. Prepared by working group III
of the Intergovernmental Panel on Climate Change. Cambridge, United
Kingdom and New York, NY, USA: Cambridge University Press; 2005.
 L.C. Ho et al. / Energy 63 (2013) 252e259
[2] Mondal MK, Balsora HK, Varshney P. Progress and trends in CO2 capture/
separation technologies: a review. Energy 2012;46(1):431e41.
[3] Aaron D, Tsouris C. Separation of CO2 from flue gas: a review. Sep Sci Technol
2005;40(1e3):321e48.
[4] D’Alessandro DM, Smit B, Long JR. Carbon dioxide capture: prospects for new
materials. Angew Chem Int Ed 2010;49(35):6058e82.
[5] Herzog HJ, Drake EM. Carbon dioxide recovery and disposal from large energy
systems. Annu Rev Energy Environ 1996;21(1):145e66.
[6] Gudmundsson JS, Parlaktuna M, Khokhar AA. Storing natural-gas as frozen
hydrate. SPE Prod Facil 1994;9(1):69e73.
[7] Kerr RA. Energy e gas hydrate resource: smaller but sooner. Science
2004;303(5660):946e7.
[8] Nakayama T, Tomura S, Ozaki M, Ohmura R, Mori YH. Engineering investi-
gation of hydrogen storage in the form of clathrate hydrates: conceptual
design of hydrate production plants. Energy Fuels 2010;24:2576e88.
[9] Lee H, Lee JW, Kim DY, Park J, Seo YT, Zeng H, et al. Tuning clathrate hydrates
for hydrogen storage. Nature 2005;434(7034):743e6.
[10] Carter BO, Wang W, Adams DJ, Cooper AI. Gas storage in “dry water” and “dry
gel” clathrates. Langmuir 2010;26(5):3186e93.
[11] Florusse LJ, Peters CJ, Schoonman J, Hester KC, Koh CA, Dec SF, et al. Stable
low-pressure hydrogen clusters stored in a binary clathrate hydrate. Science
2004;306(5695):469e71.
[12] Kvenvolden K. Potential effects of gas hydrate on human welfare. Proc Natl
Acad Sci U S A 1999;96:3420.
[13] Makogon YF, Trebin FA, Trofimuk AA, Tsarev VP, Cherskii N. Detection of a
pool of natural gas in a solid (hydrated gas) state. Dokl Akad Nauk SSSR
1971;196:203e6.
[14] Klauda JB, Sandler SI. Global distribution of methane hydrate in ocean sedi-
ment. Energy Fuels 2005;19(2):459e70.
[15] Makogon YF, Holditch SA, Makogon TY. Natural gas-hydrates e a potential
energy source for the 21st century. J Pet Sci Eng 2007;56(1e3):14e31.
[16] Linga P, Haligva C, Nam SC, Ripmeester JA, Englezos P. Recovery of methane
from hydrate formed in a variable volume bed of silica sand particles. Energy
Fuels 2009;23(11):5508e16.
[17] Haligva C, Linga P, Ripmeester JA, Englezos P. Recovery of methane from a
variable-volume bed of silica sand/hydrate by depressurization. Energy Fuels
2010;24(5):2947e55.
[18] Koh CA, Sum AK, Sloan ED. State of the art: natural gas hydrates as a natural
resource. J Nat Gas Sci Eng 2012;8:132e8.
[19] Linga P, Kumar R, Englezos P. The clathrate hydrate process for post and pre-
combustion capture of carbon dioxide. J Hazard Mater 2007;149(3):625e9.
[20] Klara SM, Srivastava RD, McIlvried HG. Integrated collaborative technology,
development program for CO2 sequestration in geologic formations e
United States Department of Energy R&D. Energy Convers Manag
2003;44(17):2699e712.
[21] Kang SP, Lee H. Recovery of CO2 from flue gas using gas hydrate: thermo-
dynamic verification through phase equilibrium measurements. Environ Sci
Technol 2000;34(20):4397e400.
[22] Linga P, Kumar R, Lee JD, Ripmeester J, Englezos P. A new apparatus to
enhance the rate of gas hydrate formation: application to capture of carbon
dioxide. Int J Greenh Gas Control 2010;4(4):630e7.
[23] Rehder G, Kirby SH, Durham WB, Stern LA, Peltzer ET, Pinkston J, et al.
Dissolution rates of pure methane hydrate and carbon-dioxide hydrate in
undersaturated seawater at 1000-m depth. Geochim Cosmochim Acta
2004;68(2):285e92.
[24] Lee S, Liang L, Riestenberg D, West OR, Tsouris C, Adams E. CO2 hydrate
composite for ocean carbon sequestration. Environ Sci Technol 2003;37(16):
3701e8.
[25] Sloan ED, Koh C. Clathrate hydrates of natural gases. CRC; 2007.
[26] Zhang JS, Lee JW. Equilibrium of hydrogen þ cyclopentane and carbon dioxide þ
cyclopentane binary hydrates. J Chem Eng Data 2008;54(2):659e61.
[27] Deppe G, Tam SS, Currier RP, Young JS, Anderson GK, Le L, et al. Developments
in the SIMTECHE process e separation of CO2 from coal syngas by formation
of hydrates. In: 2nd annual conference on carbon sequestration. Alexandria,
VA: 2003.
[28] Klara SM, Srivastava RD. U.S. DOE integrated collaborative technology
development program for CO2 separation and capture. Environ Prog
2002;21(4):247e53.
[29] Kumar R, Wu HJ, Englezos P. Incipient hydrate phase equilibrium for gas
mixtures containing hydrogen, carbon dioxide and propane. Fluid Phase
Equilib 2006;244(2):167e71.
[30] Kumar R, Englezos P, Moudrakovski I, Ripmeester JA. Structure and compo-
sition of CO2/H2 and CO2/H2/C3H8 hydrate in relation to simultaneous
CO2capture and H2 production. AlChE J 2009;55(6):1584e94.
259
[31] Lee HJ, Lee JD, Linga P, Englezos P, Kim YS, Lee MS, et al. Gas hydrate formation
process for pre-combustion capture of carbon dioxide. Energy 2010;35(6):
2729e33.
[32] Li XS, Xia ZM, Chen ZY, Yan KF, Li G, Wu HJ. Gas hydrate formation process for
capture of carbon dioxide from fuel gas mixture. Ind Eng Chem Res
2010;49(22):11614.
[33] Li X-S, Xu C-G, Chen Z-Y, Wu H-J. Hydrate-based pre-combustion carbon di-
oxide capture process in the system with tetra-n-butyl ammonium bromide
solution in the presence of cyclopentane. Energy 2011;36(3):1394e403.
[34] Sabil KM, Duarte ARC, Zevenbergen J, Ahmad MM, Yusup S, Omar AA, et al.
Kinetic of formation for single carbon dioxide and mixed carbon dioxide and
tetrahydrofuran hydrates in water and sodium chloride aqueous solution. Int J
Greenh Gas Control 2010;4(5):798e805.
[35] Kim SM, Lee JD, Lee HJ, Lee EK, Kim Y. Gas hydrate formation method to
capture the carbon dioxide for pre-combustion process in IGCC plant. Int J
Hydrogen Energy 2011;36(1):1115e21.
[36] Englezos P. Nucleation and growth of gas hydrate crystals in relation to kinetic
inhibition. Oil Gas Sci Technol 1996;51(6):789e95.
[37] Mori YH. Recent advances in hydrate-based technologies for natural gas
storageea review. J Chem Ind Eng (China) 2003;54:1e17.
[38] Adeyemo A, Kumar R, Linga P, Ripmeester J, Englezos P. Capture of carbon
dioxide from flue or fuel gas mixtures by clathrate crystallization in a silica gel
column. Int J Greenh Gas Control 2010;4(3):478e85.
[39] Seo Y-T, Moudrakovski IL, Ripmeester JA, Lee J-w, Lee H. Efficient recovery of
CO2 from flue gas by clathrate hydrate formation in porous silica gels. Environ
Sci Technol 2005;39(7):2315e9.
[40] Babu P, Kumar R, Linga P. Pre-combustion capture of carbon dioxide in a fixed
bed reactor using the clathrate hydrate process. Energy 2012;50(1):364e73.
[41] Hashemi S. Carbon dioxide hydrate formation in a three-phase slurry bubble
column; 2009.
[42] Xu CG, Li XS, Lv QN, Chen ZY, Cai J. Hydrate-based CO2 (carbon dioxide)
capture from IGCC (integrated gasification combined cycle) synthesis gas
using bubble method with a set of visual equipment. Energy 2012;44(1):
358e66.
[43] Xu CG, Cai J, Li XS, Lv QN, Chen ZY, Deng HW. Integrated process study on
hydrate-based carbon dioxide separation from integrated gasification com-
bined cycle (IGCC) synthesis gas in scaled-up equipment. Energy Fuels
2012;26(10):6442e8.
[44] Watanabe K, Imai S, Mori YH. Surfactant effects on hydrate formation in an
unstirred gas/liquid system: an experimental study using HFC-32 and sodium
dodecyl sulfate. Chem Eng Sci 2005;60(17):4846e57.
[45] Zhang J, Lee JW. Enhanced kinetics of CO2 hydrate formation under static
conditions. Ind Eng Chem Res 2008;48(13):5934e42.
[46] Zhang J, Yedlapalli P, Lee JW. Thermodynamic analysis of hydrate-based pre-
combustion capture of CO2. Chem Eng Sci 2009;64(22):4732e6.
[47] Zhong DL, Daraboina N, Englezos P. Recovery of CH4 from coal mine model gas
mixture (CH4/N2) by hydrate crystallization in the presence of cyclopentane.
Fuel 2013;106:425e30.
[48] Lim YA, Babu P, Kumar R, Linga P. Morphology of carbon dioxide-hydrogen-
cyclopentane hydrates with or without sodium dodecyl sulfate. Cryst
Growth Des 2013;13(5):2047e59.
[49] Veluswamy HP, Linga P. Macroscopic kinetics of hydrate formation of mixed
hydrates of hydrogen/tetrahydrofuran for hydrogen storage. Int J Hydrogen
Energy 2013;38(11):4587e96.
[50] Smith JM, Van Ness HC, Abbot MM. Introduction to chemical engineering
thermodynamics. 6th ed. McGraw Hill; 2001.
[51] Bishnoi PR, Natarajan V. Formation and decomposition of gas hydrates. Fluid
Phase Equilib 1996;117(1):168e77.
[52] Kumar A, Sakpal T, Linga P, Kumar R. Influence of contact medium and
surfactants on carbon dioxide clathrate hydrate kinetics. Fuel 2013;105(0):
664e71.
[53] Yoslim J, Linga P, Englezos P. Enhanced growth of methaneepropane clathrate
hydrate crystals with sodium dodecyl sulfate, sodium tetradecyl sulfate, and
sodium hexadecyl sulfate surfactants. J Cryst Growth 2010;313(1):68e80.
[54] Zhong Y, Rogers RE. Surfactant effects on gas hydrate formation. Chem Eng Sci
2000;55(19):4175e87.
[55] Okutani K, Kuwabara Y, Mori YH. Surfactant effects on hydrate formation in an
unstirred gas/liquid system: amendments to the previous study using HFC-32
and sodium dodecyl sulfate. Chem Eng Sci 2007;62(14):3858e60.
[56] Li X-S, Xu C-G, Chen Z-Y, Cai J. Synergic effect of cyclopentane and tetra-n-
butyl ammonium bromide on hydrate-based carbon dioxide separation
from fuel gas mixture by measurements of gas uptake and X-ray diffraction
patterns. Int J Hydrogen Energy 2012;37(1):720e7.