34
Ind. Eng. Chem. Res. 2008, 47, 34-38
Acceleration of CO2 Reaction with N,N-Diethylethanolamine in Aqueous Solutions by Piperazine Prakash D. Vaidya and Eugeny Y. Kenig* UniVersity of Dortmund, Department of Biochemical and Chemical Engineering, D-44227, Dortmund, Germany
N,N-Diethylethanolamine (DEEA), which can be prepared from renewable resources, represents a candidate alkanolamine for CO2 removal from gaseous streams. In this work, the reaction of CO2 with aqueous solutions containing N,N-diethylethanolamine (DEEA) was studied in a stirred cell reactor with a plane, horizontal gas-liquid interface, in the range of temperatures 298-308 K. The DEEA concentration in the aqueous solutions was varied in the range 2-3 kmol/m3. The liquid-side mass transfer coefficient and a combined parameter comprising the reaction rate constant and the diffusivity and solubility of CO2 in DEEA solutions were evaluated. The effect of the addition of piperazine (PIP) as a possible absorption activator was studied, and it was found that even with a small amount of PIP (0.1 kmol/m3) added, the CO2 absorption rate increased. 1. Introduction The reaction of CO2 with alkanolamines is of considerable industrial importance, for example, in treating natural/associated gas streams, in the manufacture of hydrogen via steam reforming of natural gas, in ammonia plants to make the gaseous streams free of CO2, and in thermal power stations to meet the discharge limits for CO2 in flue gas. Industrially important alkanolamines for CO2 removal are the primary amine monoethanolamine (MEA), the secondary amine diethanolamine (DEA), and the tertiary amine methyldiethanolamine (MDEA). Primary and secondary alkanolamines react rapidly with CO2 to form carbamates. However, the heat of absorption associated with carbamate formation is high. Consequently, this results in high solvent regeneration costs. Further, the CO2 loading capacity of such alkanolamines is limited to 0.5 mol of CO2/ mol of amine. In tertiary alkanolamines, which have a low reactivity with respect to CO2, there is no hydrogen atom attached to the nitrogen atom, as in case of MEA and DEA, and thus the carbamation reaction cannot take place. Instead, tertiary amines promote the CO2 hydrolysis reaction forming bicarbonates. The reaction heat released in bicarbonate formation is lower than that of carbamate formation, thus resulting in lower solvent regeneration costs. Moreover, tertiary amines have a high CO2 loading capacity of 1 mol of CO2/mol of amine. In this work, the reaction of CO2 with one such tertiary alkanolamine, namely, N,N-diethylethanolamine (DEEA), is investigated. It comprises two ethyl groups replacing the hydrogen atoms of the amino group in MEA. Diethylamine (obtained from ethanol) and ethylene oxide (obtained from the oxidation of ethylene which, in turn, can be prepared by the dehydration of ethanol) react to form DEEA. Ethanol, which is the major raw material for the manufacture of this alkanolamine, can be prepared from agricultural products and/or residues. Thus, DEEA is a potentially attractive absorbent for sour gas purification, as it can be prepared from renewable resources. The reaction between CO2 and DEEA in aqueous solutions was earlier studied by Kim and Savage1 at 323 K, by Spaninks et al.2 at 303 K, and by Benitez-Garcia et al.3 at 298 K. Kinetic data at various temperatures were reported by Sotelo et al.4 and * To whom correspondence should be addressed. Tel.: +49 231 755 2357. Fax: +49 231 755 3035. E-mail: e.kenig@ bci.uni-dortmund.de.
Littel et al.5 The agreement between the second-order reaction rate constants estimated in these works is poor. Besides, there exists no such data at high alkanolamine concentrations. In the present work, the rates of CO2 absorption into 2, 2.25, 2.5, and 3 kmol/m3 aqueous DEEA solutions are measured at 298, 303, and 308 K. The CO2 absorption rate in an aqueous DEEA solution can be enhanced by the addition of an activator. Piperazine (PIP), a cyclic diamine, was selected as a possible activator for this study, and the absorption of CO2 into formulated aqueous solutions of N,N-diethylethanolamine containing piperazine (DEEA + PIP + H2O) was investigated. 2. Theory There are a few comprehensive reviews on the reactions between CO2 and alkanolamines and the kinetic behavior of such reaction systems available.6,7 Donaldson and Nguyen8 were the first to propose that tertiary alkanolamines cannot react directly with CO2. Such amines have a base catalytic effect on the hydration of CO2. Some subsequent works on the reaction of CO2 with triethanolamine (TEA) and MDEA are in agreement with this mechanism.9-13 When CO2 is absorbed into an aqueous DEEA solution, base catalysis could be represented as follows (here DEEA is denoted as R2R′N, where R ) -C2H5 and R′ ) -CH2CH2OH): k′
R2R′N + H2O + CO2 98 R2R′N+H + HCO3-
(1)
The base catalysis reaction could also be explained by a zwitterion mechanism earlier proposed by Yu and Astarita:13
R2R′N + CO2 a R2R′NCOO-
(2)
R2R′NCOO- + H2O f R2R′N+H + HCO3-
(3)
Equation 2 represents a reaction of DEEA with CO2 to form an unstable complex. Equation 3 describes the homogeneous hydrolysis reaction. Water reacts with the zwitterion-type complex, which results in a bicarbonate formation. The following reactions may also take place simultaneously in an aqueous DEEA solution:
10.1021/ie070783y CCC: $40.75 © 2008 American Chemical Society Published on Web 12/05/2007
Ind. Eng. Chem. Res., Vol. 47, No. 1, 2008 35
R2R′N + H2O a R2R′N+H + OH-
(4)
H2O a H+ + OH-
(5)
kOH-
CO2 + OH- {\} HCO3kS
CO2 + H2O {\} HCO3- + H+
(6) (7)
Besides, contrary to the proposal of Donaldson and Nguyen,8 a direct reaction between CO2 and DEEA may occur at extremely high pH. Jorgensen and Faurholt14 reported such a reaction between CO2 and TEA, another tertiary alkanolamine, thereby resulting in monoalkylcarbonate formation. However, at pH values lower than 12, the contribution of this reaction to the overall reaction can be neglected.3 The amine dissociation reaction (eq 4) is almost instantaneous. The reaction between CO2 and OH- (eq 6) is also fast. The total rate of all of the CO2 reactions in an aqueous solution containing DEEA is represented by the sum of the reaction rates given by eqs 1, 6, and 7:
rDEEA ) [kS + kOH- (OH-) + k′(R2R′N)](CO2) ) kobs(CO2)
(8) (9)
where kobs denotes the observed reaction rate constant, which can be measured and is expressed by
kobs ) [kS + kOH- (OH-) + k′(R2R′N)]
(10)
The solvent is in large excess and, therefore, the reaction order with respect to the solvent is assumed to be zero. Hence, eq 8 does not account for the water concentration. The apparent reaction rate constant (kap), which is used for the analysis of experimental data, is given by
kap ) [k′(R2R′N)]
(11)
kap can be obtained from kobs as follows:
kap ) kobs - [kS + kOH- (OH-)]
(12)
The reaction given by eq 7 is very slow compared to the other reactions. When the contribution of the reaction given by eq 6 to the overall rate is negligible (as suggested by BenitezGarcia et al.3 and Littel et al.5), kap becomes equal to kobs. In the context of the activated MDEA technology used by BASF, PIP is applied as an efficient activator. It is more effective than other absorption activators such as MEA and DEA.15 PIP reacts with CO2 rapidly, which results in carbamate formation. Besides, the dissociation reaction for PIP may occur.16,17 The reactions that may take place in presence of PIP (here denoted by R′′(NH)2) can be explained by the homogeneous activation mechanism:18
R′′(NH)2 + 2CO2 a R′′(NHCOO)2 R′′(NHCOO)2 + 2H2O f R′′(NH2)2+ + 2HCO3-
(13) (14)
The reaction described by eq 13 occurs simultaneously with that described by eq 2, and, CO2 could be transferred by R′′(NHCOO)2 to DEEA. This is reflected by the following reaction:
R′′(NHCOO)2 + 2R2R′N a R′′(NH)2 + 2R2R′NCOO- (15)
Figure 1. Experimental setup according to Kucka et al.19
Thus, PIP contributes to the absorption of CO2 and enhances the overall rate of absorption. The overall reaction can be regarded as a reaction between CO2 and PIP in parallel with the reaction of CO2 with DEEA. 3. Experimental Section 3.1. Materials. N,N-Diethylethanolamine, used in all of the experiments, was purchased from Riedel de Ha¨en. Anhydrous piperazine, with a minimum given purity of 99%, was purchased from Fluka Chemie. A carbon dioxide cylinder with a given purity of 99.995% was purchased from Messer Griesheim whereas nitrogen was obtained from Air Liquide. All of the solutions were prepared using deionized water. 3.2. Experimental Setup. A glass stirred cell reactor (Figure 1) with a plane, horizontal gas-liquid interface was used for the absorption rate measurements (see Kucka et al.19). This easyto-use experimental device (inner diameter 103.8 mm, height 130.5 mm) was operated batchwise. The total volume of the reactor was 1.1 dm3, and the interfacial surface area was 8.5 × 10-3 m2. The reactor was equipped with two flanges made of stainless steel. A pressure transducer (Afriso Euro-Index, 0-2.5 bar) mounted on the gas-side flange, coupled with a data acquisition system, enabled the measurement of the total pressure inside the reactor. Two temperature sensors were used to measure the gas and liquid temperatures. The reactor was also equipped with inlet and outlet ports for the gas and liquid phases. The entire assembly was proven to have no leak. The gas and liquid were stirred separately by two impellers. The speed of stirring could be adjusted to the desired value with an accuracy of (1 rpm. The impeller speed during kinetic measurements was limited to 70 rpm, to ensure that the gasliquid interface was undisturbed. Eight specially developed baffles inside the reactor prevented vortex formation in the liquid. Water at the desired temperature circulated through the reactor jacket to ensure isothermal conditions. The setup was supplied by two reservoirs (equipped with heat exchangers), one for the gas phase and one for the liquid phase. 3.3. Experimental Procedure. A series of experiments was conducted at different temperatures (298-308 K). The initial DEEA concentration in the aqueous solution was varied in the range of 2 to 3 kmol/m3. In each experiment, the reactor was charged with 0.52 dm3 of the alkanolamine solution from the liquid reservoir. It was then purged with nitrogen to ensure an inert atmosphere. All of the lines were closed, and the reactor content attained the desired temperature. CO2 from the gas reservoir was then charged inside the reactor, this being
36 Ind. Eng. Chem. Res., Vol. 47, No. 1, 2008
considered as the starting point for the reaction. The reactor content was stirred at the desired speed of agitation. The decrease in system pressure due to the reaction was monitored by the pressure transducer, and the PCO2 versus t data were recorded during 30 s using the data acquisition system. These data were plotted for the time interval between t ) 5 s and t ) 25 s and fitted to a third degree polynomial using the leastsquare regression. The absorption rates were calculated from the values of the slope -dPCO2/dt. This measurement method based on the fall-in-pressure technique enabled a simple and straightforward estimation of the absorption rates. Further, no analysis of the liquid phase was required, and the pressure decrease was the only factor necessary for the evaluation of the kinetic parameters. The reproducibility of the experiments was checked, and the error in all of the experimental measurements was found to be less than 3%. Figure 2. Rate of absorption as a function of CO2 partial pressure at 303 K.
4. Results and Discussion 4.1. Estimation of kL. For the case of physical absorption in a stirred cell reactor, a mass balance for the solute for both the gas and the liquid-phase yields the following expression:20
ln
[
] [
]
P(t) - Pfinal (m′VL) + VG )kLA′t Pinitial - Pfinal VLVG
(17)
where xM is given by 2
(18)
kL
and Ei is the enhancement factor for an instantaneous reaction:
[
(DEEA)0 kmol/m3
PCO2 kPa
RCO2 × 106 kmol/(m2 s)
{HCO2 x(DCO2k2)} × 107 kmol1/2/(m1/2 s kPa)
298
2.25 2.50 3.0
15.62 14.10 14.26
1.86 2.49 2.79
0.79 1.12 1.13
303
2.0 2.0 2.0 2.0 2.25 2.25 2.25 2.25 2.50 2.50 2.50 2.50 3.0 3.0 3.0 3.0
4.54 6.36 7.58 12.74 5.30 7.73 10.46 15.62 2.27 4.99 8.94 13.19 3.94 6.82 10.16 14.41
0.77 0.90 1.30 1.84 0.98 1.47 1.80 2.58 0.48 1.22 2.19 2.98 0.98 1.91 2.66 3.52
1.20 1.0 1.21 1.02 1.23 1.27 1.15 1.10 1.34 1.55 1.55 1.43 1.44 1.62 1.51 1.41
308
2.0 2.25 2.50
16.38 15.62 14.86
3.19 3.47 4.17
1.38 1.48 1.77
When CO2 concentration in the bulk liquid is negligible, it can be shown, on the basis of the two-film theory of mass transfer, that the following relation holds:22
RCO2 ) PCO2HCO2 x{DCO2k2(DEEA)0}
xDCO kobs xM )
Ei ) 1 +
temperature K
(16)
To determine kL, the physical absorption of CO2 in water at 303 K was studied, and the fall-in-pressure due to absorption was recorded. Afterward, ln[(P(t) - Pfinal)/(Pinitial - Pfinal)] was plotted versus time t to estimate the value of kL. The slope of this graph equals -[((m′VL) + VG)/VLVG]kLA′. From the value of the slope (1.1 × 10-3) and from the knowledge on other relevant parameters (m′ ) 0.77 mol/mol, VG ) 0.58 dm3, VL ) 0.52 dm3, A′ ) 8.5 × 10-3 m2), the value of kL was found to be 4 × 10-5 m/s. The solubility of CO2 in water at 303 K was found to be 3.05 × 10-4 kmol/(m3 kPa), which agrees well with the published value.21 4.2. CO2-DEEA Reaction System. To determine the reaction rate constant, it is essential that the system belongs to the fast reaction regime, without depletion of the amine at the gas-liquid interface.22,23 The necessary conditions are
10 < xM < (Ei - 1)
Table 1. CO2 Absorption Rates into Aqueous DEEA Solution
]
(DEEA)0 DDEEA z(CO2) DCO2
(19)
In the fast pseudo-first-order reaction regime (E ) xM), the rate of absorption is independent of the liquid-side mass transfer coefficient kL, and hence it should not depend on the agitation speed. We studied this effect experimentally and found practically no change in the absorption rate, while varying the stirring speed in the range 50-100 rpm at 303 K. All further experiments were conducted at a speed of 70 rpm. The rate of absorption as a function of CO2 partial pressure at 303 K is shown in Figure 2.
(20)
The CO2 absorption rates into aqueous DEEA solutions at various temperatures and values of the parameter HCO2 x(DCO2k2) are represented in Table 1. Littel et al.5 reported correlations for the physical properties of aqueous solutions containing DEEA (
