Вы находитесь на странице: 1из 7

J. Chem.

Thermodynamics 71 (2014) 6470

Contents lists available at ScienceDirect

J. Chem. Thermodynamics
journal homepage: www.elsevier.com/locate/jct

Measurement and modelling of the solubility of carbon dioxide


in aqueous 1,8-p-menthane-diamine solution
Jie Li a,b,c, Xiao Lin a,b, Peng-Ge Ning a,b,, Hong-Bin Cao a,b,, Yi Zhang a,b
a

Beijing Engineering Research Centre for Process Pollution Control, Beijing 100190, China
Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, China
c
University of Chinese Academy of Sciences, Beijing 100049, China
b

a r t i c l e

i n f o

Article history:
Received 3 September 2013
Received in revised form 13 November 2013
Accepted 14 November 2013
Available online 22 November 2013
Keywords:
1,8-p-Menthane-diamine
Carbon dioxide
Chemical equilibrium constant
Gas solubility
KentEisenberg

a b s t r a c t
The solubility of CO2 in aqueous 1,8-p-menthane-diamine (MDA) solution with substance concentrations
of 0.625 and 1.25 mol  L1 was measured at temperatures (313.15, 333.15 and 353.15) K with CO2 partial
pressures ranging from (0.55 to 776.0) kPa and CO2 loading ranging from (0.120 to 1.97) mol CO2 per mol
MDA. The gas solubility results are expressed as the partial pressure of CO2 (PCO2 ) against its mole ratio,
i.e. aCO2 (mol CO2 per mol MDA). The chemical absorption reaction and thermodynamic model have been
proposed. The physicochemical KentEisenberg model was used to correlate all the experimental results
of the solubility of CO2 in the aqueous MDA solutions under investigation. The chemical equilibrium constants and model parameters were determined by tting the VLE data.
2013 Elsevier Ltd. All rights reserved.

1. Introduction
The greenhouse effect induced by CO2 emissions has been
receiving ever-increasing attention. The majority of industrial
CO2 emissions come from power plants using fossil fuels such as
coal and natural gas. Amine scrubbing is probably the only technology for the post-combustion capture of CO2 [1]. Monoethanolamine (MEA) solutions are widely used in CO2 capture because of
their fast reaction rate and low solvent cost. However, the loading
capacity of MEA reduces because of the formation of stable carbamate [2,3]. Methyldiethanolamine (MDEA) has a higher loading
capacity and lower absorption of heat, but it exhibits low reaction
rates [4]. Recently, piperazine (PZ) has received increasing attention due to its higher loading capacity and better physical properties [5,6]. It is worthwhile to search for amines with better physical
properties, higher loading capacities and faster reaction rates.
As shown in gure 1 1,8-p-menthane-diamine (MDA) comprises
an alicyclic ring with two primary NH2 groups. In the MDA molecule,
the steric hindrance introduced by the bulky substituent adjacent to
the amino group lowers the stability of the carbamate formation.
The steric hindrance enables the MDA to yield a higher thermodynamic capacity and a faster reaction rate with CO2 [7]. Furthermore,

MDA has several special physical and chemical properties, including


excellent antioxidant ability and high thermal stability. Kim et al. [8]
reported that the CO2 loading ratio of the MDA solution is higher
than that of 2-amino-2-methyl-1-propanol (AMP), MDEA and
MEA, and the regeneration heat of aqueous MDA is lower than that
of aqueous MEA. Lee et al. [9] measured the absorption rate of CO2
of an AMP solution with MDA as an additive; the obtained absorption rates were higher than those of MEA and AMP solutions. They
concluded that MDA can be used as an excellent absorbent or additive for AMP for CO2 capture. However, experimental data regarding
the solubility of CO2 in aqueous MDA solutions are insufcient and
the equilibrium constants for MDA-CO2H2O are not available. Measuring the fundamental (vapour + liquid) equilibrium (VLE) data and
modelling the data for applications in industrial design and operation purposes can prove worthwhile.
In this work, we measured the solubility of CO2 in aqueous MDA
solutions with concentrations of (0.625 and 1.25) mol  L1 at
T = (313.15, 333.15 and 353.15) K over a pressure range of (0.55
to 776.0) kPa. The VLE data were regressed using the well-known
KentEisenberg (KE) model [10]. Thereafter, the sterically hindering effect is discussed by comparing the equilibrium constants.
2. Materials and methods

Corresponding authors. Address: No.1, Beiertiao, Zhongguancun, Beijing


100190, China. Tel./fax: +86 1082544845 (P.-G. Ning). Tel./fax: +86 1082544879
(H.-B. Cao).
E-mail addresses: pgning@home.ipe.ac.cn (P.-G. Ning), hbcao@home.ipe.ac.cn
(H.-B. Cao).
0021-9614/$ - see front matter 2013 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.jct.2013.11.018

2.1. Reagents
The MDA was purchased from Rohm and Haas Chemicals LLC;
by means of chromatographic analyses (Agilent 7890A, FID

65

J. Li et al. / J. Chem. Thermodynamics 71 (2014) 6470

H3C

NH2

9DFXXP
SXPS

0DVVIORZ
PHWHU

H3C C CH3
NH2
FIGURE 1. Molecular structure of MDA.

detector, HP-5 column), it was determined that the purity was


more than 97%. The solution was prepared with deionised water.
The gases and other reagent used in this experiment were listed
in table 1.

7
1

:DWHU
EDWK

2.2. Solubility measurement

2.2.2. Absorption procedure


The CO2 absorption experiment was carried out as follows. First,
the concentration of the studied aqueous solution of MDA were
prepared by mass using an analytical balance (Mettler ML104,
0.0001 g). 100 mL (298.15 K) aqueous MDA solution was placed
in the reactor. A vacuum pump was used to remove the remaining
air; then, N2 gas was introduced into the reactor. The pressure in
the reactor was maintained to be stable after stirring at a speed
of 150 rpm for approximately 1 h. Then, the VLE in the reactor
was achieved, and the pressure was recorded as P N2 . The CO2 gas
was introduced into the gas container via the mass ow meter,
and the volume of the standard state introduced into the reactor
was recorded. The temperature in the reactor was then adjusted
to the desired level. The VLE was achieved again after approximately 10 h, and the equilibrium pressure in the reactor was recorded as PT.
The PengRobinson cubic equation of state (PR EOS) [11] was
used to represent the non-ideal gas. The original equation can be
expressed as

RT

aT

v  b v v b bv  b

5HDFWRU

FIGURE 2. Schematic diagram of the experimental equipment.

2.2.1. Apparatus
The schematic representation of the apparatus used to measure
the CO2 solubility is shown in gure 2. The reactor (GCF-0.5, Dalian
Automatic Control Equipment Factory, China) consisted of a cylindrical stainless steel tank (595 cm3 by calibration) that was tted
with a magnetically coupled stirrer. A thermometer (TM-301, AS
ONE Co., Japan) with an uncertainty of 0.1 K was used to measure
the temperature in the reactor. A pressure transducer (2600T, ABB
Ltd., Switzerland) with an uncertainty of 0.075% was used to measure the pressure in the reactor. A water bath (0.1 K) was used to
control the temperature of the reactor tank. A mass ow meter
(CS200, SEVENSTAR Ltd., China) was used to control the total mass
of CO2 introduced into the reactor.

&2

Here, v represents the molar volume of a gas at a particular


pressure and temperature, which can be calculated using the PR
EOS if the volume, pressure and temperature of the gas phase of

TABLE 1
Source and purity of chemicals used in this work.

the system are already known. The number of moles of gas in


the system (n) can be calculated as

n V=v ;

where V is the system volume occupied by the gas phase. The moles
of the gas introduced into the reactor measured by the mass ow
meter are dened as nin.With the assumption that the gas phase
obeys Daltons law, when the VLE is achieved in the reactor, the partial pressure of CO2 (P CO2 ) can be determined using

PCO2 PT  PN2 :

According to the PR EOS, the number of moles of CO2 in the


reactor is nCO2 . The moles of CO2 absorbed in 100 mL of MDA
(nab) at the pressure P CO2 can be calculated as

nab nin  nCO2 :

Using the moles of CO2 dissolved in MDA (nab) and the total moles of MDA introduced into the reactor (namine), the CO2 loading
capacity (aCO2 ) in the liquid phase can be dened as

aCO2

nab
:
namine

The uncertainty in the measured temperature, pressure and volume were 0.1 K, 0.075% and 0.5%, respectively. The uncertainty
in determining the MDA concentration was estimated as 1%. From
Eqs. (3)(5), the uncertainty of CO2 loading can be determined
using the following equation:

aCO2

nin  nCO2
:
namine

The contributions of Dnin and DnCO2 account for the uncertainties in


the measured temperature, pressure and volume. The contribution
of Dnamine can be attributed to the MDA concentration. The uncertainty of a determined from the above uncertainties was estimated
as 3.5%.
To verify the uncertainty of the experiment, the solubility of
CO2 in 30 wt% (4.95 mol  L1) MEA solution was measured at
T = 313.15 K and the obtained results were compared with the literature data [12,13] (gure 3).
The experimental VLE data used for correlation are listed in tables 2 and 3.
2.3. Thermodynamic framework

Chemicals

Source

Mass fraction purity

MDA
MEA
N2
CO2

Rohm and Haas Co.


SigmaAldrich
Qianxi Gas Co.
Qianxi Gas Co.

>0.97
P0.99
P0.999
P0.999

Numerous ionic species exist in the liquid phase and certain


chemical equilibrium constants are unknown; therefore, the physicochemical KE model [10] was used in this study to correlate the
VLE data. This model was developed by Hu and Chakma [14], who

66

J. Li et al. / J. Chem. Thermodynamics 71 (2014) 6470

Shen was used to correlate the solubility data for CO2 in an aqueous MDA solution.
Although the two NH2 groups connecting to a large alicyclic ring
are inherently unsymmetrical, we suppose that their chemical
equilibrium constants are the same in order to simplify our
calculations. In our material balance equations, the ionic species
[H+], [OH], [CO32] and [OOCHNRNHCOO] are neglected.
K1

CO2 H2 O H2 NRNH2 HCO3 H2 NRNH3 ;


K2

H2 NRNH2 HCO3 H2 NRNHCOO H2 O;


K3

H2 NRNH3 HCO3 H3 N RNHCOO ;


K4

H2 NRNH3 CO2 H2 O H3 N RNH3 HCO3 :


FIGURE 3. Solubility of CO2 in 30 wt% MEA solution at T = 313.15 K.

T = 333.15 K

H3C

9
10

C(CH3)2

T = 353.15 K

aCO2

P CO2 /kPa

aCO2

P CO2 /kPa

aCO2

P CO2 /kPa

0.182
0.417
0.612
0.794
1.20
1.38
1.55
1.71
1.82

0.550
1.12
1.46
2.13
4.74
7.47
11.2
25.5
58.0

0.148
0.297
0.441
0.608
0.726
0.886
1.13
1.32
1.50
1.63
1.72
1.78

1.23
2.07
3.08
4.17
6.47
10.7
13.3
23.0
42.7
70.1
109.2
152.4

0.260
0.366
0.472
0.599
0.749
0.831
0.922
1.19
1.39
1.55
1.66
1.75
1.82
1.90

6.80
10.7
12.7
19.9
30.2
37.0
49.5
62.1
112.1
188.4
265.7
352.1
508.0
690.6

Standard uncertainties are: u(T) = 0.1 K, u(P) = (0.02P) kPa, u(a) = 0.035a.

In the KE model, the non-idealities of the solution were not considered. So the K values cannot be kept constants and were treated
as functions of amine concentration, gas loading and temperature
as follows:

K i expai bi T1 ci aCO2 di a1


CO2 ei ln C 0

T = 333.15 K

11

Here, ai, bi, ci, di and ei are the parameters that need to be regressed.
When the CO2 loading is less than 1, H3 N RNH
3  is neglected to
simplify the regression and the material balance equations are
expressed as follows:

H2 NRNH3  HCO3  H2 NRNHCOO ;

12

H2 NRNH2 0 H2 NRNH3  H2 NRNH2  H2 NRNHCOO 


H3 N RNHCOO ;

TABLE 3
Experimental solubility data of CO2 in 1.25 mol  L1 MDA.
T = 313.15 K

Here, R represents

TABLE 2
Experimental solubility data of CO2 in 0.625 mol  L1 MDA.
T = 313.15 K

13

H2 NRNH2 0 a H2 NRNHCOO  HCO3 


T = 353.15 K

H3 N RNHCOO :

aCO2

P CO2 /kPa

aCO2

P CO2 /kPa

aCO2

P CO2 /kPa

0.227
0.355
0.507
0.659
0.807
1.15
1.30
1.45
1.61
1.77
1.88
1.92

0.78
1.00
1.61
2.18
3.01
3.82
7.12
10.8
19.7
43.9
89.4
116.4

0.12
0.263
0.509
0.705
0.898
1.24
1.42
1.57
1.72
1.83
1.93
1.97

1.27
3.00
5.00
8.40
11.8
30.8
50.1
79.5
127.5
210.4
378.5
547.5

0.248
0.447
0.609
0.773
1.25
1.31
1.49
1.62
1.71
1.83
1.92

8.90
13.0
22.4
40.7
130.2
169.5
245.5
354.4
474.3
628.4
776.0

Standard uncertainties are: u(T)=0.1 K, u(P)=(0.02P) kPa, u(a)=0.035a.

correlated the equilibrium constants with temperature, partial


pressure of the acid gas and amine concentration. In the study of
Li and Shen [15], the equilibrium constants were treated as a
function of temperature, amine concentration and gas loading. Later, Dong et al. [3] and Trejo et al. [16] used this model, while Kim
et al. [17] developed a modied KE model to simplify their computation. However, the KE model that deals with diamines has not
been reported yet. In this work, the model developed by Li and

14

The equilibrium constants are expressed as follows (the nomenclature and calculation equations are listed in table 4):

K1

H2 NRNH3 HCO3  C 2  C 3

CO2 H2 NRNH2 


C5  C4

C 0 a  C 6 C 0 a  C 6  C 1
;
C 5 C 0 1  a  C 1 C 5 

K2

H2 NRNHCOO 
C1

HCO3 H2 NRNH2  C 3  C 4

C1
;
C 0 a  C 6  C 1 C 0 1  a  C 1 C 5 

K3

H3 N RNHCOO 
C6

H2 NRNH3 HCO3  C 2  C 3

C6
:
C 0 a  C 6 C 0 a  C 6  C 1

15

16

17

For the MDA-CO2H2O system with a known initial MDA concentration C0 and a, the concentrations of six components, i.e.
C1C6, can be solved using the six equations, i.e. Eqs. (12)(17)

67

J. Li et al. / J. Chem. Thermodynamics 71 (2014) 6470


TABLE 4
Nomenclature and calculation equations of each species in MDA-CO2-H2O system.
Constituent

Nomenclature

[H2NRNHCOO]
[H2NRNH3+]
[HCO
3]
[H2NRNH2]
[CO2]
[H3N+RNHCOO]
[H3N+RNH3+]
[H2NRNH2]0

C1
C2
C3
C4
C5
C6
C7
C0

Calculation

(see Appendix). Then, the partial pressure of the CO2 in relation to


the CO2 concentration physically dissolved in the water can be calculated by means of Henrys law using Eq. (18):

PCO2 HCO2  HC 5

18
1 1

Here, H/{kPa(mol  L ) } is the Henrys constant for CO2,


which can be expressed as follows [15]:

H exp20:2669  1:38306  104 =T 0:06913  108 =T2


 0:015589  1011 =T3 0:012  1013 =T4

19

A simplex program is then used to solve the non-linear equations with the objective function in Eq. (20):

F obj 1=n

X
jPex  Pca =Pex j:

20

Hence, the parameters for K1, K2 and K3 can be regressed in the a < 1
region.
When the CO2 loading is greater than 1, [H2NRNHCOO] and
[H2NRNH2] are neglected and the material balance equations can
be expressed as follows:

a<1

a>1

C1
C2 = C0a  C6
C3 = C0a  C6  C1
C4 = C0(1  a)  C1 + C5
C5
C6
0
Initial MDA concentration

0
C2 = C0  C7  C6
C3 = C0 + C7  C6
0
C5 = C0(a  1)  C7
C6
C7

3. Results and discussion


Table 5 contains the tted values for K1, K2, K3 and K4. Figures 4
and 5 show the plots of the experimental data and the correlated
results for different MDA concentrations. The total average absolute deviation (AAD) for the correlation is 13.45%. The AAD for each
temperature and C0 value is listed in table 6. Figures 6 and 7 show
the plots of the K3 values vs. a at MDA concentrations of (0.625 and
1.25) mol  L1, respectively. Evidently, a few peaks near the a  1
can be seen for each temperature, which demonstrates that the
maximum concentration for C6 ([H3N+RNHCOO]) exists in the system (see gure 8). Two sets of K3 parameters (in the a < 1and a > 1
regions) coincide with each other, which illustrates that our
regression is reliable. Figure 8 shows the concentration changes
in C1C7 with CO2 loading. From gure 8, it is evident that the concentrations of C1 and C6 are very low when compared with those in
PZ-CO2H2O system [18]. This phenomenon can be attributed to
the sterically hindered structure of the MDA.
In this work, we calculated the chemical equilibrium constants
of K1, K2 and K3 for the MDA-CO2H2O system by using the tted
parameters. We can compare pKa, K2 and K3 values with those for

H2 NRNH2 o H2 NRNH3  H3 N RNHCOO 


H3 N RNH3 ;

21

H2 NRNH2 o a H3 N RNHCOO  HCO3  CO2 ;

22

HCO3  2H3 N RNH3  H2 NRNH3 ;

23

K3

K4

H3 N RNHCOO 
C6

H2 NRNH3 HCO3  C 2  C 3
C6
;
C 0  C 7  C 6 C 0 C 7  C 6

24

H3 N RNH3 HCO3  C 3  C 7 C 0 C 7  C 6 C 7

:
C 2  C 5 C 0  C 7  C 6 C 5
CO2 H2 NRNH3 

25

The concentrations of ve components, i.e. C2, C3, C5, C6 and C7


can be solved using the ve equations, i.e. Eqs. (21)(25) (see
Appendix). Then, the CO2 pressure and parameters for K3 and K4
can be regressed in the a > 1 region using the simplex program.

FIGURE 4. Correlation results for the MDA-CO2H2O system in 0.625 mol  L1 MDA
solution (points: experimental data; lines: correlated results).

TABLE 5
Fitted values from the experimental data for MDA-CO2-H2O system.
Fitted parameters

a<1

K1
K2/L  mol1
K3/L  mol1

10.969
3.7500
1.7851

4973.4
1641.7
1104.7

a>1

K3/L  mol1
K4

6.3905
15.251

1349.3
6170.79

4.5702
0.0201
0.46967

0.03305
0.01052
1.0954

0.56008
0.18629
1.33401

0.50932
1.0109

9.1894
0.06003

0.95847
0.58172

68

J. Li et al. / J. Chem. Thermodynamics 71 (2014) 6470

FIGURE 7. K3 values vs. a in 1.25 mol  L1 MDA solution.


1

FIGURE 5. Correlation results for the MDA-CO2H2O system in1.25 mol  L


solution (points: experimental data; lines: correlated results).

MDA

pK a1 log10 K a1 log10 K 1 =K CO2 ;


TABLE 6
AAD values for correlation of VLE data in MDA-CO2-H2O system.

K4

C0(MDA)/mol  L1

T/K

Correlation data points

AADa/%

0.625
0.625
0.625
1.25
1.25
1.25

313.15
333.15
353.15
313.15
333.15
353.15
Total

9
12
14
12
12
11
70

11.94
16.35
11.91
9.354
13.40
17.74
13.45

AAD% 1n

n jP exp

 P cal j  100=P exp .

29

H3 N RNH3 HCO3  H HCO3  H3 N RNH3 


CO2 
CO2 H2 NRNH3 
H H2 NRNH3 
K CO2
;
K a4

30

pK a4 log10 K a4 log10 K 4 =K CO2 :

31

Here, K CO2 is the chemical equilibrium constant for the following


reaction:
K CO

CO2 H2 O HCO3 H

32

K CO2 exp241:828 29:8253  104 =T  1:48528


 108 =T 2 0:332647  1011 =T 3  0:282393
 1013 =T 4

33
1

The expression of K CO2 (mol  L ) and its parameters can be


obtained
from
the
literature
[15].
At
T = 313.1 K,
K CO2 = 5.386  107 (mol  L1). For MDEA, MEA and PZ, the pKa1,
K2, K3 and pKa4 values were calculated. Table 7 lists their values
at 313.15 K for comparison. For pKa1, K2 and pKa4 of MDA, the

FIGURE 6. K3 values vs. a in 0.625 mol  L1 MDA solution.

conventional amines, such as MDEA, MEA and PZ. The denition of


pKa is as follows:

Ka

RNH2 H 
;
RNH3 

pK a log10 K a :

26
27

For the MDA-CO2-H2O system,


K1

K CO2
H2 NRNH3 HCO3  H HCO3  H2 NRNH3 
;

CO2 H2 NRNH2 


CO2 
H H2 NRNH2  K a1

28

FIGURE 8. Predicted C1C7 values vs. a at T = 313.15 K in 1.25 mol  L1 MDA
solution.

J. Li et al. / J. Chem. Thermodynamics 71 (2014) 6470

69

average values in the low ionic concentration were taken. For K3 of


MDA, the value was taken at the maximum concentration of zwitterions in order to compare with other amines.
High pKa values represent the high basicity of amines. High K2
value indicates the high stability of carbamates formed in the
amine. The high K3 value implies the high stability of zwitterions
formed in the amine.
Then, it is straightforward to perform a quantitative comparison. For solutions with the same concentration of NH2 groups
and at the same temperature (313.15 K), three plots of P CO2 vs. a
for MDEA, MEA and PZ solutions when compared with the corresponding values of the MDA solution are shown in gures 911,
respectively. With regard to the MDEA solution, the basicity is
marginally lesser than the MDA solution. A slightly higher solubilTABLE 7
Thermodynamic constants for amines at T = 313.15 K.

PZ
MEA
DEA
MDEA
MDA

pKa1

K2/(L  mol1)

K3/(L  mol1)

pKa4

Data source

9.37
9.64
9.06
8.25
8.43

18.4
26.77
2.297
0
0.2

12.43
0
0
0
0.11

5.07

[18]
[17]
[17]
[20]
This work

8.68

FIGURE 11. Comparison of the CO2 solubility in 0.625 mol  L1 MDA solution (this
work) and 0.6 mol  L1 PZ solution [19] at T = 313.15 K.

ity of CO2 for the MDA than the MDEA solution can be observed
(gure 9). When compared with the MEA and PZ solutions (see gures 10 and 11), the solubility in MDA is lower in the low-a region
due to the low basicity, while in the high-a region, the solubility in
MDA is higher: this can be attributed to the steric hindrance. From
Eqs. (16) and (17), it can be observed that low K2 and K3 values
would prevent the formation of carbamates and zwitterions, and
increase the concentration of [HCO
3 ], which consequently increases the loading capacity of CO2 in the MDA solution. Compared
with PZ, the high second protonation constant (pKa4) for MDA
would increase the concentration of [H3N+RNH3+] and increase
the solubility at high-a region (see gure 11). All the solubility
comparisons show that MDA can act as a good absorbent of CO2.
4. Conclusions

FIGURE 9. Comparison of the CO2 solubility in 1.25 molL1 MDA solution (this
work) and 30 wt.% (2.596 molL1) MDEA solution [12] at T = 313.15 K.

In this work, the solubility of CO2 in aqueous MDA solutions


with concentrations of (0.625 and 1.25) mol  L1 was measured
at different temperatures and pressures. The absorption process
was categorised into a < 1 and a > 1 regions. For each region, the
chemical reactions and material balance equations were determined. The KE model was used to regress the VLE data, and satisfactory results were obtained. Each equilibrium constant was
calculated and compared with those obtained for conventional
amines. The comparison results show that MDA has a higher
capacity to absorb CO2 due to steric hindrance.
Acknowledgement
The authors appreciate the nancial support from National
Natural Science Foundation of China (21206168, 51178446).
Appendix A
In the a < 1 region, using Eq. (15),

C5

C 0 a  C 6 2  C 1 C 0 a  C 6
:
K 1 C 0 1  a  C 1 

A:1

From Eqs. (15) and (17), we have

C 6 K 1 K 3 C 5 C 0 1  a  C 1 C 5   K 1 K 3 C 5 C 0 1  a:

A:2

From Eqs. (15) and (16), we have


FIGURE 10. Comparison of the CO2 solubility in 1.25 mol  L1 MDA solution (this
work), (2 mol  L1 MEA + 0.25 mol  L1 MDA) solution (this work) and 2.5 mol  L1
MEA solution [13] at T = 313.15 K.

C 1 C 5 K 1 K 2 C 0 1  a  C 1 C 5 2 =C 0 a  C 6
 C 5 K 1 K 2 C 0 1  a2 =a:

A:3

70

J. Li et al. / J. Chem. Thermodynamics 71 (2014) 6470

Substituting Eqs. (A.2) and (A.3) into Eq. (A.1), we obtain a quadratic
equation of C5:C5 can be solved using the following equation.
C5

2C 0 aK 3 K 2 C 0 1  a 1 

q
K 2 C 0 1  a 12 4C 0 aK 3  K 2

21  aK 1 K 23 C 0 K 2 K 3 C 0 1  a=a K 2 =a

A:4

A:5

From Eq. (25), we have

C 6 C 0 C 7

C 27

 K 4 C 5 C 0  C 7 =C 7  K 4 C 5 :

A:6

4K 4 C 5  1=K 3 C 37  C 0 C 27 =K 3 2K 4 C 5 C 0 K 24 C 25 C 7 =K 3
A:7

Using the expression of C7 (Eq. (A.8)), a quartic equation of C5


(Eq. (A.8)) can be established to solve C5 using the Newton iteration
method with C5 = C0(a  1) as the initial value:

C 7 C 0 a  1  C 5 ;

[7]
[8]
[9]

Using the fact that Eq. (A.5) equals Eq. (A.6), an equation comprising C7 and C5 is obtained as follows:

 K 24 C 25 C 0 =K 3 0:

[1]
[2]
[3]
[4]
[5]
[6]

In the a > 1 region, from Eq. (24), we have

q
C 6 2C 0 1=K 3  2C 0 1=K 3 2  4C 20  C 27 =2:

References

[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]

G.T. Rochelle, Science 325 (2009) 16521654.


Y.G. Li, A.E. Mather, Ind. Eng. Chem. Res. 33 (1994) 20062015.
L.H. Dong, J. Chen, G.H. Gao, J. Chem. Eng. Data 55 (2010) 10301034.
Y.G. Li, A.E. Mather, Ind. Eng. Chem. Res. 36 (1997) 27602765.
R. Dugas, G. Rochelle, Greenhouse Gas Control Technol. 9 (1) (2009) 1163
1169.
S.A. Freeman, J. Davis, G.T. Rochelle, Int. J. Greenhouse Gas Control 4 (2010)
756761.
G. Sartori, D.W. Savage, Ind. Eng. Chem. Fundam. 22 (1983) 239249.
S.T. Kim, J.W. Kang, J.S. Lee, B.M. Min, Korean J. Chem. Eng. 28 (2011) 2275
2281.
S.S. Lee, S.M. Mun, W.J. Choi, B.M. Min, S.W. Cho, K.J. Oh, J. Environ. Sci. 24
(2012) 897902.
R.L. Kent, B. Eisenberg, Hydrocarb. Process 55 (1976) 8790.
D. Peng, D.B. Robinson, Ind. Eng. Chem. Fundam. 15 (1976) 5964.
K.P. Shen, M.H. Li, J. Chem. Eng. Data 37 (1992) 96100.
J.I. Lee, F.D. Otto, A.E. Mather, J. Appl. Chem. Biotech. 26 (1976) 541549.
W. Hu, A. Chakma, Chem. Eng. Commun. 94 (1990) 5361.
M.H. Li, K.P. Shen, Fluid Phase Equilib. 85 (1993) 129140.
M.A. Rebolledo-Morales, M.E. Rebolledo-Libreros, A. Trejo, J. Chem.
Thermodyn. 43 (2011) 690695.
S.H. Park, K.B. Lee, J.C. Hyun, S.H. Kim, Ind. Eng. Chem. Res. 41 (2002) 1658
1665.
V. Ermatchkov, A.P.S. Kamps, G. Maurer, J. Chem. Thermodyn. 35 (2003) 1277
1289.
S. Bishnoi, G.T. Rochelle, Chem. Eng. Sci. 55 (2000) 55315543.
D. Barth, C. Tondre, J.J. Delpuech, Chem. Eng. Sci. 39 (1984) 17531757.

A:8

4K 4 C 45  12C 0 a  1K 4 1=K 3  K 24 =K 3 C 35 12C 20 a  12 K 4


3C 0 a  1=K 3 C 0 =K 3 2K 4 C 0 =K 3 K 24 C 0 =K 3
 K 24 C 0 a  1=K 3 C 25  4C 30 a  13 K 4 3C 20 a  12 =K 3
2C 20 a  1=K 3 2K 4 C 20 a  1=K 3 C 5 C 30 a  12 a=K 3  0:
A:9

JCT 13-510