Journal of the Air & Waste Management Association

ISSN: 1096-2247 (Print) 2162-2906 (Online) Journal homepage: http://www.tandfonline.com/loi/uawm20

Carbon Dioxide Removal from Flue Gases

by Absorption/Desorption in Aqueous
Diethanolamine Solutions

Hanna Kierzkowska-Pawlak

To cite this article: Hanna Kierzkowska-Pawlak (2010) Carbon Dioxide Removal from Flue
Gases by Absorption/Desorption in Aqueous Diethanolamine Solutions, Journal of the Air & Waste
Management Association, 60:8, 925-931, DOI: 10.3155/1047-3289.60.8.925

To link to this article: https://doi.org/10.3155/1047-3289.60.8.925

Published online: 24 Jan 2012.

Submit your article to this journal

Article views: 2103

Citing articles: 9 View citing articles

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=uawm20
TECHNICAL PAPER
Carbon Dioxide Removal from Flue Gases by Absorption/
Desorption in Aqueous Diethanolamine Solutions
Hanna Kierzkowska-Pawlak and Andrzej Chacuk
Faculty of Process and Environmental Engineering, Technical University of Łódź, Łódź, Poland
ABSTRACT
The carbon dioxide (CO2) desorption rate from CO2-
loaded aqueous diethanolamine (DEA) solutions was
measured using a stirred cell with a flat gas-liquid inter-
face. The measurements were performed in the tempera-
ture range of 293.15–313.15 K and an amine concentra-
tion range of 10 –20% mass DEA. Measurements were
based on a semibatch isothermal absorption of the gas
until the equilibrium state was reached, followed by de-
sorption, which was initiated by the pressure release in
the system. A simplified mass transfer model based on the
film theory coupled with CO2 mass balance was devel-
oped to interpret the experimental data. On the basis of
the proposed model, the initial mass transfer rates were
calculated from the experimental results. The calculated
initial desorption rates enabled estimation of the en-
hancement factor for CO2 mass transfer from aqueous
DEA solutions. Analysis of the experimental data showed
that desorption took place in the diffusive mass transfer
regime.
INTRODUCTION
Absorption by aqueous alkanolamine solutions is the
dominant industrial process for separation of acid gases
such as carbon dioxide (CO2) and hydrogen sulfide
(H2S) from natural, refinery, and synthesis gas streams.
Another possible application of amine-based technolo-
gies is CO2 separation from the flue gases at fossil fuel
power plants. The chemical absorption of CO2 is gen-
erally recognized as the most efficient postcombustion
CO2 separation technology. The idea of CO2 sequestra-
tion, which includes its capture and storage in under-
ground rock formations, has progressed steadily over
the past 10 yr. It is claimed that this solution could play
an important role in solving the problem of increasing
greenhouse gas emissions. Postcombustion technology
can be fairly installed in the existing power plants,
IMPLICATIONS
The CO2 desorption rate from CO2-loaded aqueous DEA
solutions was measured. Information on kinetics of desorp-
tion in the CO2-loaded aqueous DEA system is of primary
importance for designing unit operations of acid gas re-
moval because it is directly related to the design require-
ments of the amine regeneration step. A better understand-
ing of the desorption process would be helpful in simulating
and improving the performance of CO2 strippers.
Volume 60 August 2010
ISSN:1047-3289 J. Air & Waste Manage. Assoc. 60:925–931
DOI:10.3155/1047-3289.60.8.925
Copyright 2010 Air & Waste Management Association
whereas other options such as the precombustion
method and oxy-fuel processes involve innovative ways
of power generation technology. Absorption with
amine-based solvents is the most promising technology
for CO2 removal regarding solvent capacity and regen-
erability.1 However, it is a process with considerable
inherent problems, particularly when used in a large
quantity of flue gases. Although several commercial
technologies of that type are offered, the cost of capture
is still too high and is the major barrier for the wide
application of CO2 capture in the power plant sector.2,3
The amines that have been proven to be of principal
commercial interest are primary amines monoethano-
lamine (MEA) and diglycolamine (DGA), the secondary
amines diethanolamine (DEA) and diisopropanolamine
(DIPA), and the tertiary amines methyldiethanolamine
(MDEA) and triethanolamine (TEA).4 Primary and second-
ary alkanolamines react rapidly with CO2 to form carbam-
ates. However, the heat of absorption associated with the
carbamate formation is high, which results in high sol-
vent regeneration costs. Tertiary alkanolamines possess
no hydrogen atom attached to the nitrogen atom and do
not react directly with CO2 to form carbamate. In aqueous
solutions, tertiary amines catalyze the CO2 hydrolysis re-
action, forming bicarbonate ions and the protonated
amine. Tertiary amines react more slowly with CO2 than
primary and secondary amines. The reaction heat released
in bicarbonate formation is lower than that of carbamate
formation, thus reducing solvent regeneration costs. Ter-
tiary amines have potentially higher CO2 absorbing ca-
pacity than primary and secondary amines because they
cannot form carbamates. However, they are not an ideal
choice for use in CO2 removal because their reaction with
CO2 is slow. The CO2 absorption rates in tertiary amines
can be enhanced by the addition of small amounts of
primary or secondary amines. In addition, other absorp-
tion activators such as piperazine (PZ), which is a cyclic
diamine, can also be used. Different amine blends of
conventional amines and novel individual amine sys-
tems with good potential for the removal of CO2 have
been extensively investigated in the past few years.4 –7
The requirements for excellent chemical absorbents in-
clude high CO2 capacity, fast absorption and stripping
rates, low corrosivity and degradation rates, low vola-
tility, and low heat of absorption, which result in a
lower energy consumption for reversing the reaction
and stripping the CO2.8
The common feature of absorption processes is that
they are used in an absorber-stripper mode, requiring two
Journal of the Air & Waste Management Association 925
Kierzkowska-Pawlak and Chacuk
separate steps in CO2 separation. In an absorber the CO2
is absorbed into an amine solution, and in a stripper the
amine is regenerated and sent back to the absorber for
further absorption. Because regeneration includes heating
of the absorbent, the energy requirement significantly
increases when treating a large volume of flue gases and
thus has a major impact on the overall efficiency of the
power plant. Therefore, the reduction of heating energy
becomes a determining factor for optimizing the CO2
capture process.
To summarize, several commercially available tech-
nologies for CO2 capture exist, but the cost of removing
CO2 is still very high, in particular when considering the
energy requirements used for regeneration of the solvent.
Although in many practical situations the operational
and capital costs of the desorption column are greater
than those of the absorption column, studies devoted to
desorption are relatively scarce as compared with the ex-
tensive work done on the absorption. Although there are
many works dealing with absorption kinetics of acid gases
in various alkanoamine solutions, the direct measure-
ments of CO2 desorption rates from CO2-loaded aqueous
amine solutions are limited. Only a few papers deal with
CO2 desorption kinetics from loaded alkanolamine
solutions.9,10
In this paper, the CO2 desorption rate from CO2-
loaded aqueous DEA solutions was measured using a stirred
cell with a flat gas-liquid interface. DEA (C4H11NO2) is the
most popular secondary alkanolamine used for CO2
removal. It has several advantages, including high re-
activity, low solvent cost, and reasonable thermal sta-
bility.2 Adopting a small pressure drop into a gas-liquid
system that is in equilibrium initiated CO2 release from
the loaded DEA solution. The measurements were per-
formed in the temperature range of 293.15–313.15 K
and amine concentration range of 10 –20% mass DEA.
The pressure-driven desorption investigated in the
work presented here is the experimental technique re-
alized in isothermal conditions, which enabled deter-
mination of the initial CO2 desorption rate and the
initial desorption enhancement factor. It was used pre-
viously for studying the CO2 desorption kinetics from
loaded MDEA aqueous solutions by Cadours et al.9 It
should be stressed that no available literature has been
found regarding the kinetics of CO2 desorption from
aqueous DEA solutions in the range of operating pa-
rameters covered in the study presented here.
Because most of the kinetic and thermophysical
data of the system under consideration available in the
literature are collected in the limited temperature
range, the experimental desorption data can only be
interpreted and analyzed in terms of enhancement fac-
tor within this temperature range where relevant data
are available. Thus, the study presented here reports
preliminary results of low-temperature regeneration of
amine to check if the theory of chemical absorption can
be applied for a reversed process of desorption. How-
ever, it should be noted that the desorption tempera-
ture in an industrial stripper for flue gas application is
considerably higher than 313.15 K.
926 Journal of the Air & Waste Management Association
EXPERIMENTAL PROCEDURES
Apparatus
The measurements were performed in a heat flow reac-
tion calorimeter (CPA-Chemical Process Analyzer, Che-
miSens AB), which is a fully automated and computer-
controlled stirred reactor vessel with the possibility of
online measurement of thermal power developed by
the process. The reactor is a cylindrical, double-walled
glass vessel with an inner diameter of 70 mm. The total
volume available for gas and liquid is 250 ⫾ 0.5 cm3.
The reactor was equipped with an impeller stirrer and
four baffles, which helped in reducing vortex formation
in the reactor and promoting better mixing of the liq-
uid phase. It is capped, top and bottom, with 316 stain-
less steel. The reactor lid holds the shaft seal and the
necessary armature for charging and sampling from gas
and liquid phases. A schematic diagram of the experi-
mental setup is shown in Figure 1.
While in use the reactor is submerged in the thermo-
stating liquid bath. A Peltier element mounted inside of
the bottom of the reactor serves as an efficient heating
and cooling device and keeps temperature constant
within ⫾0.1 K. A pressure transducer mounted on the
reactor flange measured the absolute pressure. A separate
tube on the top flange allows evacuation of the cell or
introduction of a gas into the reactor. The measurement
accuracy of the digital pressure transducer was 0.1% of the
full range of 0 – 0.2 MPa.
Procedure
The experimental apparatus was developed to measure
acid gas absorption and desorption rates from alkano-
amine solutions under unsteady-state conditions. The
present procedure is specific for a batchwise operation
of a stirred tank reactor used in this work and is similar
to that applied by Cadours et al.,9 who investigated the
CO2 desorption from loaded MDEA solutions. MDEA is
a tertiary amine that reacts with CO2 according to a
distinct reaction mechanism than DEA.
Figure 1. Experimental setup: 1, storage tank; 2, downstream
pressure regulator; ns, stirring speed transducer; T, temperature
sensors; P, pressure transducers.
Volume 60 August 2010
 Kierzkowska-Pawlak and Chacuk

Each experiment consisted of two steps, including

the absorption and subsequent desorption initialized
by the small pressure release. All parts of the setup were
carefully examined for gas leaking. The absorption was
operated batchwise with respect to the liquid phase δ

while the gas was continuously fed to the reactor. The

gas was charged from the storage tank of a known
volume (150 ⫾ 0.5 cm3) to the reactor through the
downstream pressure controller (Brooks, 5866), which
∆
allowed the constant total pressure (P) to be maintained
in the reactor during the whole absorption experiment.
After the reactor was filled with the desired solution,
the liquid was degassed and heated to a given temper-
ature. When the thermal equilibrium was established
(i.e., the temperature in the reactor became constant to
⫾0.01 K), the stirring was stopped for a moment. Then,
under the solution vapor pressure, pure CO2 was started
to feed to the upper part of the cell up to a fixed Figure 2. Treatment of desorption experimental data (experiment
pressure in the range of 0.06 – 0.12 MPa. After a few no. 14).
seconds, when the pressure in the reactor reached the
set value, the absorption process was initiated by RESULTS AND DISCUSSION
switching on the stirrer. The pressure in the storage Theoretical Background—Reaction Mechanism
tank started to decrease as the gas was being absorbed in The following reactions may take place when CO2 is ab-
the solution, whereas the pressure in the reactor re- sorbed into an aqueous solution of DEA11:
mained constant. The subsequent decline of the gas
pressure in the storage tank was recorded as a function K1a,k1,k⫺1
of time until the gas-liquid equilibrium state was CO 2 ⫹ RR⬘NH ¢
O ¡ RR⬘NH⫹COO ⫺ (1)
reached. The pressure drop in the storage tank from the
moment of switching on the stirrer until the final equi-
librium state was necessary for calculating the amount K2,k2,k⫺2
RR⬘NH ⫹ COO ⫺ ⫹ RR⬘NH ¢
O ¡ RR⬘NH2 ⫹ ⫹ RR⬘NCOO ⫺
of gas absorbed in the liquid. The gas solubility was
expressed by a loading, ␣, which is defined as the ratio (2)
of moles of absorbed CO2 per moles of alkanolamine in
the solution. The CO2 partial pressure in the reactor, pA,
K3,k3,k⫺3
was calculated according to the total pressure in the RR⬘NH ⫹ COO ⫺ ⫹ H2O ¢
O ¡ H3O ⫹ ⫹ RR⬘NCOO ⫺ (3)
system, P, corrected for solution vapor pressure by use
of Raoults’s law. In these calculations, the amine vapor
pressure was neglected, because it was very small as K4,k4,k⫺4

compared with the solvent vapor pressure. RR⬘NH ⫹ COO ⫺ ⫹ OH ⫺ ¢

O ¡ H2O ⫹ RR⬘NCOO ⫺ (4)
The desorption run was conducted in a similar man-
ner. The stirring was stopped for a moment while the K5,k5
system was rapidly decompressed via a release valve to RR⬘NCOO ⫺ ⫹ H2O ¢
O¡ RR⬘NH ⫹ HCO3⫺ (5)
impose the required supersaturation of the solution. The
resulting supersaturation of the system subsequently led K6
to CO2 release from the liquid. The corresponding pres- RR⬘NH 2⫹ ⫹ OH⫺ ¢
¡ RR⬘NH ⫹ H2O
O (6)
sure increase was recorded up to another gas-liquid equi-
librium pressure (Figure 2). Small pressure drops were
K7,k8
applied to avoid bubble desorption and to keep the liquid CO 2 ⫹ OH⫺ ¢
O¡ HCO3 ⫺ (7)
composition approximately constant during the whole
experiment.
K8
The experiments were carried out at the temperature
HCO 3⫺ ⫹ OH⫺ ¢
¡ CO32⫺ ⫹ H2O
O (8)
of 293.15–313.15 K. The chemicals used (CO2 [L’Air Liq-
uide, 99.995 vol % pure] and DEA [Fluka, 99 mass %
pure]) were used without any further purification. Alkano- K9

lamine aqueous solutions were prepared from distilled, ¡ OH ⫺ ⫹ H3O ⫹

deionized water.
The solution volume was maintained at 100 ⫾ 0.1
cm3 in each experiment. The respective mass of the solu-
tions was determined within ⫾0.01 g. The stirring speed
was in the range of 100 –180 min⫺1.
2H 2 O ¢
O (9)
The proposed mechanism for the reaction between CO2
and DEA (RR’NH) involves the formation of zwitterion
(RR’NH⫹COO⫺) followed by the deprotonation of zwitte-
rion by a base, to produce carbamate (RR’NCOO⫺) and

Volume 60 August 2010 Journal of the Air & Waste Management Association 927
Kierzkowska-Pawlak and Chacuk

protonated base. Any base present in the solution such as pA

water (H2O), hydroxyl ion (OH⫺), or amine may contrib- c AL * ⫽ (15)
He
ute to the deprotonation of zwitterion. The contribution
of each base would depend on its concentration and its
where pA (Pa) is the CO2 partial pressure and He (Pa 䡠
basicity. In the case of aqueous DEA solutions, the reac-
m3/mol) is Henry’s law constant. The CO2 desorption rate
tions associated with the deprotonation of zwitterions to
can be calculated by the mass balance in the reactor gas
H2O and OH⫺ can be ignored and the main contribution
phase according to
to the deprotonation of the zwitterions comes from the
amine itself.11 The overall reaction, which accounts for
carbamate formation, is given by the sum of reactions V G dpA
⫽ ⫺NAL* (16)
represented by eqs 1 and 2. R 䡠 T 䡠 A dt

K1 where VG (m3) is the gas-phase volume in the reactor, T

CO 2 ⫹ 2RR⬘NH ¢
¡ RR⬘NH2 ⫹ RR⬘NCOO
O ⫹ ⫺
(10) (K) is the temperature, R ( J 䡠 mol⫺1 䡠 K⫺1) is a gas
constant, and A (m2) is the gas-liquid interfacial area.
Very small amounts of CO2 are transferred to the gas
Deprotonation of the zwitterions is often the rate control- phase, leaving quasi-unchanged the CO2 and free
ling step. Applying steady-state principle to the interme- amine liquid bulk concentrations during experiments
diate zwitterions and assuming that eq 2 is irreversible (CBL␦ ⫽ const).5 Then, the CO2 bulk concentration is
according to Zhang et al.12, the rate of eq 10 can be assumed constant and calculated from Henry’s law at
expressed as equilibrium pressure, pA␦. This assumption leads to the
equation
k1关CO2兴 关RR⬘ NH兴
r CO 2 ⫽ (11) dp A R 䡠 T 䡠 A kAL
k⫺1 ⫽⫺ 共p ⫺ pA␦兲 䡠 EA (17)
1⫹ dt VG He A
k2 关RR⬘ NH兴

Furthermore, U was denoted by the following expression

The experimental desorption data obtained in this study
according to Cadours et al.9:
are interpreted by the use of an apparent second-order
rate constant.
RTA
U ⫽ kAL E (18)
r CO 2 ⫽ kapp关CO2兴 关RR⬘ NH兴 (12) VGHe A

The apparent second-order rate constant of the overall and eq 17 can be simplified to
reaction between CO2 and DEA was determined by Zhang
et al.12 as dp A
⫽ ⫺U共pA ⫺ pA␦兲 (19)
dt

k app 冉 m3
kmol 䡠 sec 冊
⫽ 24.515 ⫺ 5411.3/T (13) Integration of eq 19 from the initial conditions at t ⫽ tdep
to t yields
Calculation of Desorption Rate and
Enhancement Factor
The general expression for CO2 desorption rate N AL * has Y ⫽ ln 冉 pA ⫺ pA␦
pA,dep ⫺ pA␦ 冊
⫽ ⫺U共t ⫺ tdep兲 (20)
the following form:
where pA,dep and pA are the partial pressures at time t ⫽
N AL * ⫽ ⫺kAL共cAL* ⫺ cAL␦兲 䡠 EA (14) tdep and at some arbitrary time t. Linear regression of Y
over t shows that eq 20 is fulfilled for all experiments. It is
where NAL* is expressed in (mol 䡠 m⫺2 䡠 sec⫺1), kAL (m/sec) then possible to calculate the initial CO2 desorption rates
is the mass transfer coefficient of CO2 in the liquid phase, and the experimental enhancement factor EAexp as
and cAL* and c AL␦(mol/m3) are the interfacial and bulk
CO2 concentrations, respectively. The enhancement fac- VG
tor, EA, describes the influence of a chemical reaction on N AL * ⫽ ⫺U 䡠 共p ⫺ pA␦兲 (21)
RTA A
the mass transfer rate. The enhancement factor is defined
as the ratio of the absorption rate of a gas in the liquid in
the presence of a chemical reaction to the absorption rate VG He
E A exp ⫽ U (22)
in the absence of a reaction at identical concentration kAL R 䡠 T 䡠 A
differences of the absorbing gas between the interface and
the liquid bulk. Henry’s law constants necessary for calculating the exper-
The interfacial concentration of CO2 can be obtained imental enhancement factor were obtained with the cor-
from Henry’s law. relation proposed by Abu-Arabi et al.13 The values of the

928 Journal of the Air & Waste Management Association Volume 60 August 2010
 Kierzkowska-Pawlak and Chacuk

Table 1. Experimental data of CO2 desorption from loaded aqueous DEA solutions.

Experiment DEA Mass Loading ␣ ⌬P ns kAL ⴛ 105 U ⴛ 104 N AL* ⴛ 105 Experimental EA
No. T (K) (%) (mol CO2/ mol DEA) (Pa) (minⴚ1) (m/sec) (secⴚ1) (mol 䡠 mⴚ2 䡠 secⴚ1) Experimental

1 313.15 20 0.759 8051 120 1.32 7.61 9.17 4.25

2 313.15 20 0.753 7679 150 1.53 7.72 8.75 3.67
3 313.15 10 0.890 4447 150 2.01 13.30 8.96 4.32
4 313.15 10 0.881 8712 120 1.74 8.38 10.96 3.12
5 313.15 10 0.883 8499 140 1.92 10.07 12.71 3.35
6 313.15 10 0.841 6417 160 2.10 8.31 7.84 2.51
7 313.15 10 0.837 6245 180 2.27 10.58 9.63 2.93
8 293.15 20 0.867 9168 120 0.98 3.08 4.55 1.74
9 293.15 20 0.864 6432 180 1.28 5.18 5.22 2.17
10 293.15 20 0.858 9897 150 1.13 5.11 8.08 2.46
11 293.15 20 0.851 8020 100 0.86 4.03 5.29 2.60
12 293.15 10 0.979 6149 140 1.31 10.17 10.17 3.60
13 293.15 10 0.973 8137 120 1.18 5.80 7.60 2.25
14 293.15 10 0.968 5999 140 1.31 6.45 6.16 2.24
15 293.15 10 0.947 5959 120 1.18 7.44 7.10 2.88
16 293.15 10 0.942 5584 160 1.43 7.21 6.33 2.26
17 293.15 10 0.929 5544 160 1.43 5.59 4.89 1.76
18 293.15 10 0.925 5787 180 1.66 6.10 5.52 1.64
19 293.15 10 0.923 5227 140 1.31 4.47 3.68 1.53
20 293.15 10 0.919 5099 140 1.31 4.11 3.30 1.41
21 313.15 10 0.831 5036 160 2.10 7.94 5.90 2.40
22 313.15 10 0.828 4718 140 1.92 7.21 5.02 2.39

physical mass transfer coefficient of CO2 in aqueous DEA a given loading, desorption rate values for various mea-
solutions were obtained from a mass transfer correlation surements may be different because of different values of
between dimensionless numbers (ShL, ReL, ScL) that has the remaining operating parameters, which are presented
been established for this apparatus from CO2 absorption in Figure 3.
experiments into the pure H2O and nitrous oxide (N2O) Nevertheless, some general conclusions can be drawn
absorption into aqueous DEA solutions in the following from the data presented here. At a given temperature and
form: total DEA concentration, increasing the CO2 loading
leads to an increase of the CO2 desorption rate. This
observation agrees with the conclusions of Cadours et al.9
Sh L ⫽ 0.2258ReL0.6632 Sc L0.33 (23)
and Jamal et al.10 The CO2 desorption rate increases with
temperature and stirring speed.
All needed physical properties including solvent viscosity
and solvent density were estimated using correlations Theoretical Predictions
summarized by Rinker et al.,14 whereas CO2 diffusivity For interpreting the experimental desorption data, the
was calculated using the relation from Snijder et al.15 The Hatta numbers and infinite enhancement factors (E⬁)
raw experimental data of kAL established for this appara-
tus are reported by Bzoma.16
The different experimental conditions and the corre-
sponding initial desorption rates and enhancement fac-
tors obtained from eqs 21 and 22 are listed in Table 1.
Generally, in all experiments the increase in stirring
speed and pressure resulted in increasing the mass transfer
rate, indicating the diffusive mass transfer regime. The
selected results of experimental desorption rates for a 10%
mass DEA solution are presented in Figure 3 as a function
of CO2 loading.
On the basis of the determined quantities, the accu-
rate assessment of the influence of the operational condi-
tions on the desorption rate is impossible. The initial
desorption rate is a function of the physical mass transfer
coefficient and the driving force of the process, whereas
the latter one depends on the pressure difference, ⌬P, and
Henry’s constant (see eq 17). The physical mass transfer
coefficient depends on the stirring speed, the concentra- Figure 3. Influence of the loading on the experimental desorption
tion of the DEA solution, and the temperature. Hence, for rates from 10% mass DEA solutions at different stirring speeds.

Volume 60 August 2010 Journal of the Air & Waste Management Association 929
Kierzkowska-Pawlak and Chacuk

were estimated according to the equations reported by Table 2. Results of equilibrium calculations and estimation of infinite
Zarzycki and Chacuk.17 The infinite enhancement factor enhancement factor and Hatta number.
for a single, second-order, irreversible reaction and the
Hatta number are expressed by the following expressions: Experiment 关DEA兴tot pA,dep cDEA␦ Calculated
No. (mol/m3) (kPa) (mol/m3) Ha Eⴥ

DDEAcDEA␦
E⬁ ⫽ 1 ⫹ (24) 1 1929 102.7 459 72.9 6.07
␯DEADALcAL* 2 1929 91.3 481 64.4 6.77
3 954 114.2 138 29.4 2.28
冑 kIIc DEA␦DAL 4 954 99.8 145 35.0 2.37
Ha ⫽ (25) 5 954 82.5 158 32.9 2.60
kAL 6 954 66.8 193 33.3 3.84
7 954 64.4 198 31.2 4.03
where DDEA (m2/sec) is the amine diffusion coefficient in 8 1944 99.5 208 33.3 2.22
the liquid phase, cDEA␦ is the free amine bulk concentra- 9 1944 99.7 212 25.7 2.32
10 1944 78.0 238 30.7 2.61
tion, DAL is the CO2 diffusion coefficient in the liquid
11 1944 81.6 234 39.9 2.62
phase, kII is a second-order rate constant, and ␯DEA is the
12 961 119.8 48 12.4 1.24
stoichiometric coefficient of DEA in the reaction with 13 961 103.1 50 14.1 1.28
CO2. Henry’s law relationship is used to find the concen- 14 961 100.1 52 12.9 1.31
tration of dissolved but unreacted CO2 at the gas-liquid 15 961 79.9 63 15.8 1.46
interface. The equation derived by Snijder et al.15 was 16 961 75.7 66 13.4 1.50
used to estimate the DEA diffusion coefficient in the liq- 17 961 65.8 74 14.1 1.65
uid phase. Because eq 11 (which describes the kinetics of 18 961 62.5 76 12.4 1.70
the overall reaction of the carbamate formation ) has the 19 961 67.4 76 15.7 1.74
form of a second-order irreversible chemical reaction rate 20 961 64.7 81 16.2 2.24
expression, in further considerations it was assumed that 21 954 67.2 212 34.9 4.53
22 954 65.4 218 38.7 4.73
the rate constant kII corresponds to the apparent rate
constant kapp.
To determine the free amine bulk concentration
(cDEA␦) and bulk concentrations of all chemical species, a with experimental results.18 It is essential to interpret the
thermodynamic model was developed based on eqs 1–9. experimental desorption rate data obtained in this work
The input data of the model included the initial DEA according to a realistic kinetic mechanism in which all
concentration, the CO2 loading of the solution, the equi- possible chemical reactions are taken into account. For
librium constants of the reactions, and the equilibrium complicated kinetics such as that of CO2 with aqueous
CO2 partial pressure (pA␦). The concentrations of all spe- DEA, the only reliable method for predicting desorption
cies were calculated by solving the amine and carbon rate is to use such a rigorous model.
balance, the Henry’s law correlation for the CO2 equilib-
rium concentration, and expressions for five independent
CONCLUSIONS
equilibrium constants: K1, K6–K9. The corresponding
chemical equilibrium constants for the needed reactions In the paper presented here, the CO2 desorption rate from
in terms of molar concentration were taken from Rinker CO2-loaded aqueous DEA solutions was measured using a
et al.14 The eight nonlinear algebraic equations were stirred cell with a flat gas-liquid interface. The measure-
solved for eight unknowns of the bulk concentrations: ments were performed in the temperature range of
CO2, RR’NH, RR’NH⫹, RR’NCOO⫺, HCO3⫺, CO32⫺, 293.15–313.15 K and amine concentration range of 10 –
H3O⫹, and OH⫺. 20% mass DEA. Measurements were based on a batch
The calculated values of free amine bulk concentra- isothermal absorption of the gas until the equilibrium
tions are listed in Table 2. As can be noticed, increasing state was reached, followed by desorption, which was
the CO2 loading leads to decreasing cDEA␦ and then the initiated by the pressure release in the system. The subse-
Hatta number. As can be seen from these data, the Hatta quent pressure increase was accurately recorded during
numbers are significantly greater than 1 for all experi- batch desorption until another equilibrium state was
ments, indicating the fast reaction regime.17 Also, the reached. A simplified mass transfer model was developed
requirement Ha ⬎ 10E⬁ is satisfied irrespective of the based on the film theory coupled with CO2 mass balance.
operating parameters. The fulfilling of both requirements On the basis of the proposed model, the initial mass
means that desorption occurs in the instantaneous or transfer rates were calculated from the experimental re-
close-to-instantaneous reaction regime with respect to sults. The calculated initial desorption rates enabled esti-
mass transfer. In this regime, the desorption rate is con- mation of the enhancement factor of gas-liquid mass
trolled by the rate of mass transfer. transfer.
Any approximate analytical solution of mass transfer Analysis of the experimental data showed that de-
equations for multiple chemical reactions published in sorption took place in the diffusive mass transfer regime
the literature does not adequately represent the set of (i.e., the desorption rate is controlled by the rate of mass
multiple reactions that are involved in the CO2-DEA sys- transfer). The experimental values of the enhancement
tem and thus cannot be used to theoretically predict the factor are close to the values corresponding to the infi-
desorption rate and the enhancement factor to compare nitely fast reaction. Although the values of the desorption

930 Journal of the Air & Waste Management Association Volume 60 August 2010

enhancement factor calculated in such a way approxi-
mate the experimental ones, the real value of the en-
hancement factor may be obtained only by way of the
numerical solution of CO2 diffusion equations with a
complex reaction in a liquid phase.
Desorption rate data should be interpreted accord-
ing to a realistic kinetic mechanism in which all possi-
ble chemical reactions are taken into account. For com-
plicated kinetics such as that of CO2 with aqueous DEA,
the only reliable method for predicting the desorption
rate is to use such a rigorous mass transfer model. The
study presented here has provided a better understand-
ing of the desorption process and would be helpful in
simulating and improving the performance of CO2
strippers. Information on the kinetics of desorption in a
CO2-DEA-H2O system is of primary importance for de-
signing unit operations of acid gas removal because it is
directly related to the design requirements of the amine
regeneration steps.
ACKNOWLEDGMENTS
This work was funded by the Ministry of Science and
Higher Education of Poland (project no. 1 T09C 018 30).
NOMENCLATURE
A ⫽ gas-liquid interfacial area, m2
c ⫽ concentration, mol 䡠 m⫺3
DiL ⫽ diffusion coefficient of species i in liquid phase, m2
䡠 sec⫺1
ds ⫽ turbine diameter, m
E⬁ ⫽ infinite enhancement factor
EA ⫽ enhancement factor
Ha ⫽ Hatta number
He ⫽ Henry’s law constant for CO2, Pa 䡠 m3 䡠 mol⫺1
kAL ⫽ liquid side mass transfer coefficient of dissolved
CO2, m 䡠 sec⫺1
kapp⫽ the apparent second-order rate constant, m3 䡠
kmol⫺1 䡠 sec⫺1
K ⫽ equilibrium constant
ns ⫽ stirring speed, sec⫺1
p ⫽ partial pressure, Pa
NAL*⫽ desorption rate of CO2 at the interface, mol 䡠 m⫺2 䡠
sec⫺1
ReL ⫽ Reynolds number, ReL ⫽ nsds2␳L/␩L
r ⫽ chemical reaction rate, kmol 䡠 m⫺3 䡠 sec⫺1
ScL ⫽ Schmidt number, ScL ⫽ ␩L/␳LDAL
ShL ⫽ Sherwood number, ShL ⫽ kALds/DAL
t ⫽ time, sec
T ⫽ temperature, K
R ⫽ universal gas constant, 8.3143 J/(mol 䡠 K)
U ⫽ defined by eq 10
V ⫽ volume, m3
Greek Letters
␦ ⫽ refers to a liquid bulk
␣ ⫽ mole ratio in the liquid phase, mol CO2/mol DEA
␳ ⫽ density, kg 䡠 m⫺3
␩ ⫽ viscosity, Pa 䡠 sec
␯DEA ⫽ stoichiometric coefficient of DEA in the reaction
with CO2
Volume 60 August 2010
Kierzkowska-Pawlak and Chacuk
Subscripts
A ⫽ CO2
G ⫽ gas phase
L ⫽ liquid phase
DEA ⫽ diethanoloamine, C4H11NO2
REFERENCES
1. Ma’mun, S.; Svendsen, H.F.; Hoff, K.A.; Juliussen, O. Selection of New
Absorbents for Carbon Dioxide Capture; Energy Convers. Manage.
2006, 48, 251-258.
2. Aaron, D.; Tsouris, C. Separation of CO2 from Flue Gas: a Review; Sep.
Sci. Tech. 2005, 40, 321-348.
3. Notz, R.; Asprion, N.; Clausen, I.; Hasse, H. Selection and Pilot Plant
Tests of New Absorbents for Post-Combustion Carbon Dioxide Cap-
ture; Chem Eng. Res. Des. 2007, 85, 510-515.
4. Vaidya, P.D.; Kenig, E.Y. CO2-Alkanolamine Reaction Kinetics: a Re-
view of Recent Studies; Chem. Eng. Tech. 2007, 30, 1467-1474.
5. Goto, K.; Okabe, H.; Shimizu, S.; Onoda, M.; Fujioka, Y. Evaluation
Method of Novel Absorbents for CO2 Capture; Energy Procedia 2009, 1,
1083-1089.
6. Shim, J.-G.; Kim, J.-H.; Lee J.H.; Jang, K.-R. Highly Efficient Absorbents
for Post-Combustion CO2 Capture; Energy Procedia 2009, 1, 779-782.
7. Chowdhury, F.A.; Okabe, H.; Shimizu, S.; Onoda, M; Fujioka, Y. De-
velopment of Novel Tertiary Amine Absorbents for CO2 Capture;
Energy Procedia 2009, 1, 1241-1248.
8. Blomena, E.; Hendriksa, C.; Neeleb, F. Capture Technologies: Improve-
ments and Promising Developments; Energy Procedia 2009, 1, 1505-1512.
9. Cadours, R.; Bouallou, H.; Gaunand, A. Kinetics of CO2 Desorption
from Highly Concentrated and CO2-Loaded Methyldiethanolamine
Aqueous Solutions in the Range 312–383 K; Ind. Eng. Chem. Res. 1997,
36, 5384-5391.
10. Jamal, A.; Meisen, A.; Jim Lim, C. Kinetics of Carbon Dioxide Absorp-
tion and Desorption in Aqueous Alkanolamine Solutions Using a
Novel Hemispherical Contactor. I. Experimental Apparatus and Math-
ematical Modeling; Chem. Eng. Sci. 2006, 61, 6571-6589.
11. Rinker, E.B.; Ashour, S.S.; Sandall, O.C. Absorption of Carbon Dioxide
into Aqueous Blends of Diethanolamine and Methyldiethanolamine;
Ind. Eng. Chem. Res. 2000, 39, 4346-4356.
12. Zhang, X.; Zhang, C.F.A. Kinetics of Absorption of Carbon Dioxide
into Aqueous Solution of MDEA Blended with DEA; Ind. Eng. Chem.
Res. 2002, 41, 1135-1141.
13. Abu-Arabi, M.K.; Al-Jarrah, A.M.; El-Eideh, M.; Tamimi, A. Physical
Solubility and Diffusivity of CO2 in Aqueous Diethanolamine Solu-
tions; J. Chem. Eng. Data 2001, 46, 516-522.
14. Rinker, E.B.; Oelschlager, D.W.; Colussi, A.T.; Henry, K.R.; Sandall,
O.C. Viscosity, Density, and Surface Tension of Binary Mixtures of
Water and n-Methyldiethanolamine and Water and Diethanolamine
and Tertiary Mixtures of These Amines with Water over the Temper-
ature Range 20 –100 °C; J. Chem. Eng. Data 1994, 39, 392-395.
15. Snijder, E.D.; Riele, M.J.M.; Versteeg, G.F.; van Swaaij, W.P.M. Diffu-
sion Coefficients of Several Aqueous Alkanolamine Solutions; J. Chem.
Eng. Data 1993, 38, 475-480.
16. Bzoma, B. M.Sci. Thesis, Technical University of Łódź, Łódź, Poland,
2009.
17. Zarzycki R., Chacuk A. Absorption: Fundamentals and Applications; Per-
gamon: Oxford, United Kingdom, 1993.
18. Swaaij W.P.M:; Verteeg, G.F. Mass Transfer Accompanied with Com-
plex Reversible Chemical Reactions in Gas-Liquid Systems: an Over-
view; Chem. Eng. Sci. 1992, 47, 3181-3195.
About the Authors
Hanna Kierzkowska-Pawlak and Andrzej Chacuk are re-
searchers at the Faculty of Process and Environmental
Engineering at the Technical University of Łódź. Please
address correspondence to: Hanna Kierzkowska-Pawlak,
Faculty of Process and Environmental Engineering, Tech-
nical University of Łódź, 90-924 Łódź, ul. Wólczańska 213,
Poland; phone: ⫹48-42-6313774; fax: ⫹48-42-6313746;
e-mail: kierzkow@p.lodz.pl.
Journal of the Air & Waste Management Association 931