Absorption of carbon dioxide into a mixed aqueous solution of diethanolamine and piperazine

Transcription

1 Indian Journal of Chemical Technology Vol. 17, November 2010, pp Absorption of carbon dioxide into a mixed aqueous solution of diethanolamine and piperazine M K Mondal Department of Chemical Engineering and Technology, Institute of Technology, Banaras Hindu University, Varanasi , India mkmondal13@yahoo.com Received 4 August 2009; revised 17 August 2010 The CO 2 loading in aqueous mixtures of diethanolamine (DEA) and piperazine(pz), from a mixture of CO 2 and N 2, has been measured for total amine concentrations and mole ratios of PZ to total amine ranging from 2.0 to 3.0 M and 0.01 to 0.20, respectively, at K and kpa CO 2 partial pressure. Measurements were made by a saturation method using a laboratory scale bubble column. The results of CO 2 loading are expressed as X CO2 (mole CO 2 /mole of total amine) for all experimental runs. A model is given to predict the CO 2 loading in aqueous mixture of DEA and PZ. The model predictions have been in good agreement with the experimental data of CO 2 loading in aqueous mixture of DEA and PZ with the average deviation of Keywords: DEA, PZ, CO 2 Loading, Total amine, Bubble column The removal of acidic gases (in particular carbon dioxide and hydrogen sulphide) from a gas stream is an area of industrial importance. Examples of such streams include natural gases, synthesis gases from the gasification of coal and heavy oils, and tail gases from sulphur plants and petroleum chemical plants. The removal of acid gases has a goal to increase the industrial and commercial utility of the hydrocarbon streams, reducing contaminant emissions to the environment during the combustion of such streams, to reduce the corrosion problems in equipment and pipelines due to the presence of such acid gases, and additionally to take advantage of these gases for applications in other industrial processes, such as in the sulphur production in the case of hydrogen sulphide 1. Due to their chemically active nature, these acidic gases may be absorbed by a number of different chemical and physical absorbents 2. The removal of acid gases from a gas stream is carried out mainly by means of a chemical reaction rather than only physical absorption in which different chemical reactions occur between the acid gases and the constituents of the aqueous solution 3. Particularly, the removal of carbon dioxide by using chemical absorbents has been of great importance, since it was found that the global warming effect is primarily due to excessive discharge of carbon dioxide and methane 4. Aqueous alkanolamine solutions have been extensively used for the removal of CO 2 from gas streams 5. A wide variety of alkanolamines, such as monoethanolamine (MEA), diethanolamine (DEA), di-2-propanolamine (DIPA) and N- methyldiethanolamine (MDEA), has been used for industrial gas treating processes 6. Aqueous MEA solutions are the most frequently used absorbents because of high reactivity to such chemicals as CO 2. However, these solutions can also react with materials in the reaction vessels, tubing lines, and several process compartments. For this reason, highly MEAconcentrated aqueous solutions should be avoided for the CO 2 removal process. The use of amine blends may have the potential of solving this problem. A blended amine solvent, which is an aqueous blend of a primary or a secondary amine with a tertiary or a hindered amine, combines the higher equilibrium capacity of the tertiary or hindered amine, with the higher reaction rate of the primary or secondary amine. Thus the use of blended amine solvents, requiring lower circulation rates and lower regeneration energy, can bring about considerable improvement and great savings in individual gas-treating processes 7. The low vapour pressure of DEA makes it suitable for low pressure operations, as vaporization losses are quite negligible. Besides DEA solution are in general, less corrosive than MEA solutions. In view of this, DEA based

2 432 INDIAN J. CHEM. TECHNOL., NOVEMBER 2010 blends appear to be potential solvents for gas treating processes. Xu et al. 8 measured the solubility data for CO 2 in 4.28 kmol/m 3 MDEA with the Piperazine (PZ) concentration ranging from 0 to kmol/m 3 and CO 2 partial pressure ranging from 3.83 to kpa. Bishnoi and Rochelle 9 showed that PZ has a large effect on solubility when the ratio of total CO 2 to PZ is less than unity. Piperazine is an effective activator for an industrial CO 2 removal process. However the absorption data of CO 2 in the aqueous blends of DEA with PZ is very scarce. In this study, the absorption data of CO 2 in DEA-PZ-H 2 O solution was systematically determined and represented by a simple model. Experimental Procedure Materials Reagent grade DEA was obtained from Sisco Research Laboratory Pvt. Ltd., Mumbai with a purity of 98%. PZ was obtained from SD Fine Chemical Ltd., Mumbai with a purity of better than 99% and was used without further purification. All solutions were prepared with distilled water. The carbon dioxide and nitrogen gases provided were of a commercial grade with a purity of 99.5 mole%. Apparatus and procedure For the absorption of CO 2 into an aqueous blend of DEA and PZ, a borosilicate glass bubble column was used. The schematic diagram of experimental set-up is shown in Fig. 1 to determine the experimental data for this work. Carbon dioxide with a partial pressure of kpa from a gas cylinder is passed through the gas rotameter at fixed flow rate and sent to the bubble column. The bubble column containing an aqueous blend of DEA and PZ was submerged in a constant temperature water bath. The temperature in the bubble column was controlled within ± 0.1 K of the desired level with a thermometer, and all measurements were done at atmospheric pressure. The dimensions of various main units including material of construction were reported elsewhere 10 and the experimental conditions used in the presents work are shown in Table 1. Before starting the experiment, a constant temperature was attended inside the bath to maintain the desired temperature inside the content of the bubble column. After that the main gas stream is slowly turned on and maintained the minimum possible gas flow rate so that it was bubbled through the liquid. Determination of the CO 2 concentration in the gas phase was made for each run with the help of microprocessor based CO 2 analyzer (UNIPHS 225 PM, United Phosphorous Limited, Mumbai). The equilibrium state was assumed when the outlet composition becomes equal to the inlet gas composition. The equilibrium CO 2 loading in the liquid phase was determined by acidulating known volume of the loaded liquid sample with 6 M HCl and Table 1 Experimental conditions used in the present work CO 2 Partial pressure in inlet gas stream (kpa) Temperature of the liquid bed (K) Volume of the liquid (cm 3 ) 500 Height of the liquid bed (cm) 28.2 ph of the distilled water used 6.95 Mole ratio of PZ in total amine Total amine concentration (M) Fig. 1 Experimental set-up for bubble column: 1- Gas cylinder (CO 2 and N 2 ); 2-2-Stage S.S. pressure regulator; 3- S.S. Valve; 4-Gas rotameter; 5-Gas mixing and pressure release chamber; 6- Bubble column; 7- Constant temperature water bath; 8- Glass Tee; 9- Moisture trap column; 10- CO 2 analyzer; 11- Wet gas flow meter; 12- Flat bottom flask with absorbing solution

3 MONDAL: ABSORPTION OF CARBON DIOXIDE 433 measuring the volume of the evolved gas by a precisely graduated gas burette. At a given temperature and pressure, at least two liquid equilibrium samples were taken in order to check reproducibility and the estimated error in the measured data is about ± 0.5%. The temperature of the liquid inside the bubble column was controlled within ± 0.1 K upto K. The total pressure was measured for each run with an uncertainty of ± 0.5 kpa upto kpa. Model development The absorption of CO 2 into an aqueous blend of PZ and DEA can be explained by a homogeneous activation mechanism. The reaction of CO 2 with PZ can be regarded as the rapid pseudo-first order reaction in parallel with that of CO 2 with DEA. Piperazine contains two basic nitrogens and can theoretically react with 2 mole of CO 2. PZ could also be protonated. Therefore, the effective free PZ, although its concentration is very low, can transfer CO 2 to DEA as a homogeneous activator and obviously promote the CO 2 absorption rate of activation DEA aqueous solutions. In an aqueous phase for the CO 2 -DEA-PZ-H 2 O system, the following chemical equilibria are involved 11 : PZH + PZ + H + (1) DEAH + DEA + H + (2) H 2 O+CO 2 H + + HCO - 3 (3) HCO - 3 H + + CO 2-3 (4) H 2 O H + + OH - (5) Experimental work in the present study is limited to low partial pressure range, the gas phase is considered to be ideal. In the low range of CO 2 partial pressure the fugacity of CO 2 becomes equal to its partial pressure and gas-liquid phase equilibria for CO 2 may be described by Henry s law. In solution, the equilibrium constants of the above mentioned equilibrium reactions are: K 1 = C PZ C H+ / C PZH+ (6) K 2 = C DEA C H+ /C DEAH+ (7) K 3 = C H+ C HCO3 - /C CO2 C H2O (8) K 4 = C H+ C CO3 2- /C HCO3 - (9) K 5 = C H+ C OH- /C H2O (10) The concentration of H + and OH - are rather low, so it is reasonable to neglect their effects on the mass and charge balance equations. Another assumption is that all forms of the absorbed carbon dioxide are regarded as bicarbonate since the contents of CO 3 2- and CO 2 are also very low. Thus only three main dissociation reactions are taken in consideration and are also expressed as given by Eqs (6-8). Mass and charge balances for the reacting species can be written as: C PZ, total = C PZ + C PZH + C DEA, total = C DEA + C DEAH + (11) (12) X CO2 = C HCO3 _ /C PZ, total + C DEA, total (13) C PZH + + C DEAH+ = C HCO3 _ (14) Gas-liquid equilibrium of CO 2 may be expressed as: p CO2 = HC CO2 (15) The CO 2 loading can be expressed as its partial pressure by solving Eqs (6-8) and Eqs (11-15) and given below: X CO2 = 2K 3 p CO2 /[2K 3 p CO2 + H {-b + (b 2 4ac)^(1/2)}] (16) where a = (C PZ, total + C DEA, total ) (1 X CO2 ) (17) b = (C PZ, total K 2 + C DEA, total K 1 ) (1 X CO2 ) (C PZ, total K 1 + C DEA, total K 2 ) X CO2 (18) c = - (C PZ, total + C DEA, total ) K 1 K 2 X CO2 (19) The thermodynamic equilibrium constants and Henry s law constant are taken from literature and presented in Table 2. Table 2 Temperature dependence of the equilibrium constants and Henry s constant Parameter Expression Reference K 1 ln K 1 = /T Pagano et al. 12 K 2 ln K 2 = /T ln T Austgen et al. 13 K 3 ln K 3 = /T /T /T /T 4 Kent and Eisenberg 14 H ln H = /T /T / T / T 4 Kent and Eisenberg 14

4 434 INDIAN J. CHEM. TECHNOL., NOVEMBER 2010 Results and Discussion Comparison of CO 2 loading of present study with literature data In literature, a lot of amine blends had been worked out for measuring CO 2 loading. Among all of previous works, some of amine blends are compared with the experimental data of present study for same experimental conditions such as partial pressure of CO 2 in inlet gas stream, total concentration of amine blend and mole ratio of an amine constituent in blend. For comparison purpose, blends consisting either DEA or PZ as one of the component has been used. The CO 2 loading comparisons of experimental data of present study with the other blends are taken at total amine concentrations of 2.0, 2.5 and 3.0 M, respectively. The CO 2 loading comparisons are shown in Figs 2 to 4. From Figs 2-4, it is seen that CO 2 loading is highest for DEA-PZ blend of present study at all total concentration range M with experimental conditions of K and kpa CO 2 partial pressure in comparison with other blends available in literature. Therefore, DEA-PZ blend for present study is assumed to be superior to other blends available in literature viz. DEA-AMP, MDEA-PZ, DEA-MDEA, TIPA-PZ regarding CO 2 loading at experimental conditions of low temperature and low CO 2 partial pressure. Comparison between the experimental and model values The model has been solved directly without iterative calculation. In the model, CO 2 loading is expressed interms of partial pressure of CO 2 in inlet gas stream, Henry s law constant, equilibrium constants of the reactions, and total concentrations of DEA and PZ. Henry s constant and equilibrium constants are function of temperature only, and Fig. 2 Comparison of CO 2 loading in 2 M DEA-PZ blend with other 2 M blends available in literature Fig. 4 Comparison of CO 2 loading in 3 M DEA-PZ blend with other 3 M blends available in literature Fig. 3 Comparison of CO 2 loading in 2.5 M DEA-PZ blend with other 2.5 M blends available in literature Fig. 5 Comparison of the experimental solubility of CO 2 (X CO2) and model values

5 MONDAL: ABSORPTION OF CARBON DIOXIDE 435 can be calculated using equations as given in Table 2. From Eqs (17)-(19), the terms a, b and c can be dependent on X CO2 only after knowing the values of K 1, K 2 and total concentrations of DEA and PZ. Thereby, finally using Eq. (16) and with the help of Henry s constant and equilibrium constant K 3 and p CO2, the values of X CO2 can be obtained. A comparison of the experimental value X CO2, exp and the model value X CO2, mod for the CO 2 loading is shown in Fig. 5. As shown in Fig. 5, the model value is in good agreement with the experimental value with a mean deviation of 8.67%. Conclusion In this work, the CO 2 loading has been measured into mixed solutions of DEA and PZ within total concentration range M and with mole ratios of PZ to total amine of at K and kpa CO 2 partial pressure. CO 2 loading in the concentration range M of DEA-PZ blends of present study has been found to be superior to other blends such as DEA-AMP, MDEA-PZ, DEA-MDEA, TIPA PZ available in literature. The data obtained from model is agreed well with the experimental data of CO 2 loading. The average deviation between the calculated and experimental CO 2 loading data is 8.67%. Acknowledgement Author gratefully acknowledges Banaras Hindu University for providing the necessary help for carrying out the present work. Nomenclature C i = concentration of solute i, kmol/m 3 H = Henry s law constant, kpa.m 3 /kmol K 1 K 5 = equilibrium constants p CO2 = CO 2 partial pressure, kpa X CO2 = CO 2 loading, mole of CO 2 /mole of total amine Subscripts i = species or component exp = experimental mod = model Amine abbreviations AMP = aminomethylpropanol DEA = diethanolamine MDEA = methyldiethanolamine PZ = piperazine TIPA = triisopropanolamine References 1 Rebolledo-Libreros M E & Trejo A, Fluid Phase Equilibr, 218 (2004) Carson J K & Marsh K N, J Chem Thermo, 32 (2000) Maddox R N, Diers J, Bhaini A M, Thomas-Looper P A & Elizondo E M, Plant Oper Prog, 6 (1987) J H, Park S B, Yoon J H & Lee H, J Chem Eng Data, 42 (1997) Jenab M H, Abdi M A, Najibi S H, Vahidi M & Matin N S, J Chem Eng Data, 50 (2005) Kundu M & Bandyopadhyay S S, Fluid Phase Equilibr, 248 (2006) Kundu M & Bandyopadhyay S S, J Chem Eng Data, 51 (2006) Xu G W, Zhang C F, Qin S J, Cao W H & Liu H B, Ind Eng Chem Res, 37 (1998) Bishnoi S & Rochelle G T, Ind Eng Chem Res, 41 (2002) Mondal M K, Fluid Phase Equilibr, 262 (2007) Mondal M K, J Chem Eng Data, 54 (2009) Pagano J M, Godeberg D E & Fernelius W C, J Phys Chem, 65 (1961) Austgen D M, Rochelle G T & Chen C C, Ind Eng Chem Res, 30 (1991) Kent R L & Eisenberg B, Hydrocarbon Process, 55 (1976) Ali B S & Arora M K, Int J Thermophys, 25 (2004) Daneshvar N, Moattar M T Z, Abdi M A & Aber S, Sep Purif Technol, 37 (2004) Murrieta-Guevara F, Rebolledo-Libreros M E, Romero- Martinez A & Trejo A, Fluid Phase Equilibr, 150 (1998) Seo D & Hong W, J Chem Eng Data, 41 (1996) 258.