CC BY
105
УДК 536.7+544.3
Вестник СПбГУ. Сер. 4. 2013. Вып. 1
D. Schäfer, A. P.-S. Kamps, B.Rumpf, G. Maurer
EXPERIMENTAL INVESTIGATION ON THE INFLUENCE OF BORIC ACID ON THE SOLUBILITY OF CARBON DIOXIDE IN AQUEOUS SOLUTIONS OF POTASSIUM HYDROXIDE*
Introduction. Aqueous solutions of potassium carbonate are often used to remove carbon dioxide from gas streams. One important example for such an application is the "hot-potash" process [1]. In the "hot-potash" process carbon dioxide is typically removed by absorption into an aqueous solution of K2CO3 at low temperatures (between about 310 K and 350 K) and the loaded solvent solution is regenerated in a desorption step at higher temperatures (around 400 K). As long as the amount of dissolved CO2 is considerably below the amount of potassium ions, CO2 is predominantly dissolved chemically as bicarbonate HCOg~ (as carbonate ions CO^- and neutral solute molecules CO2 are converted to HCOg"). When nearly all carbonate is converted to bicarbonate more carbon dioxide can only be dissolved physically, i. e., more carbon dioxide has to be dissolved as neutral species in an aqueous solution of potassium bicarbonate. Increasing the temperature shifts the chemical reaction equilibrium in favor of CO2 (i. e., decreases the "chemical" solubility of CO2) which is essential for the regeneration of the CO2-loaded solvent. In previous work, we performed experimental work on the equilibrium solubility of CO2 in aqueous solutions of K2CO3 and developed a thermodynamic model to describe the combined "chemical" and "physical" solubility of CO2 in such aqueous solutions [2]. However, in industrial applications it is common practice to add some "activators" (to speed up the conversion of CO2 to HCOg") and "corrosion inhibitors" (which allow the use of cheaper construction materials) to such solutions. Various additives — activators and corrosion inhibitors — have been described in the literature, but only a limited number has been used in industrial applications. For example, in the so-called Lurgi-process the additive is boric acid [3], in the Benfield-process it is diethanolamine [3] or vanadium pentoxide [4] and in the Giammarco—Vetrocoko-process the additive is either arsenic trioxide or glycine [1]. One of the most commonly applied activated hot-potash processes for the removal of CO2 is the Catacarb-process [1]. However, for the Catacarb-process the additives have not been disclosed. The favorable influence of such additives on the absorption/desorption of CO2 in aqueous solutions of K2CO3 processes is well assessed, although there are no openly accessible publications on the influence of these additives neither on the equilibrium solubility of CO2 nor on the reaction paths and reaction kinetics.
In this investigation we report experimental data for the influence of boric acid on the solubility of CO2 in an aqueous solution of K2CO3 at 343 K and at 383 K, i. e., at temperatures that are typical for CO2 removal and solvent regeneration, respectively, in the hot-potash process. The potassium content and the boric acid content of the solvent are expressed by the stoichiometric molality of potassium hydroxide токон and of boric acid mH3во3 in water. To discuss the influence of boric acid on the equilibrium solubility of CO2
Dirk Schäfer — University of Kaiserslautern, Germany.
Álvaro Pérez-Salado Kamps — University of Kaiserslautern, Germany.
Bernd Rumpf — University of Kaiserslautern, Germany.
Gerd Maurer — prof. dr.-ing., professor, University of Kaiserslautern, Germany; e-mail: maurer@mv.uni-kl.de
* The authors appreciate financial support of this investigation by BASF SE.
© D. Schäfer, A. P.-S. Kamps, B. Rumpf, G. Maurer, 2013
in aqueous solutions of K2CO3 the new results are compared to calculation results from the previously published model [2] for the solubility of CO2 in additive-free aqueous solutions of K2CO3.
Apparatus. The equipment used in the experimental work has been employed before, for example, to determine the simultaneous solubility of ammonia and a sour gas like carbon dioxide or sulfur dioxide in water, in methanol, and in aqueous solutions of methanol as well as to determine the influence of a single salt on the solubility of those gases in aqueous solutions [5-13]. Therefore, only a short description is given here. A schematic of the equipment is shown in Fig. 1.
Fig. 1. Schematic of the experimental equipment:
(A) thermostat; (B) equilibrium cell; (C) solvent tank; (D) container for CO2; (E) magnetically coupled stirrer; (F) liquid level indicator; (G) pressure transducing device; (H) pressure gauge; (I) sampling valve; (J) gas chromatograph; (K) platinum resistance thermometer; (L) automatic AC-bridge; (M) display for liquid level indicator; (N) stirrer motor
The heart of the apparatus is a thermostated high-pressure cell. Its internal volume is about 2.0 dm3. About one kilogram of the solvent mixture (an aqueous solution of KOH and H3BO3) was filled from a burette into the previously evacuated cell. The composition of the solvent was known from its gravimetrical preparation. The CO2-free solvent mixture was prepared by dissolving known amounts of KOH (typically about 250 g) and H3BO3 (typically about 20 g) in degassed water (typically about 1.5 kg). The amounts of both solutes were determined with an uncertainty of less than 0.005 g. The amount of water was determined with an uncertainty of less than 0.01 g. The amount of solvent which was filled into the cell was determined by weighing the burette before and after the filling with an uncertainty of less than 0.2 g. Then, CO2 was added stepwise from a small tank. The amount of CO2 was determined by weighing that tank before and after the filling procedure. In the first step the cell was charged with about 50 g of CO2, whereas in each of the additional steps about 20 g of CO2 were charged. The mass of CO2 charged in a single step was determined with an experimental uncertainty of less than 0.05 g. After each addition, the phases were equilibrated before temperature, pressure, and vapor-phase volume were measured, and small vapor phase samples were taken and analyzed by on-line gas chromatography.
The temperature was determined with a coated, calibrated platinum-resistance thermometer placed in the thermostat. The absolute uncertainty of the measured temperature amounts to 0.1 K. The pressure is transferred by a thin foil (that serves as the bottom of the cell) to silicon oil and measured with a calibrated pressure transducer. The uncertainty for the solubility pressure is 0.5 % of the actual pressure reading. The vapor phase volume is determined using a calibrated device consisting of a rod through the ceiling of the cell that is electrically isolated from the cell as long as its tip is not in contact with the liquid phase. There is a small electrical voltage between the rod and the cell. A small electric
current indicates that the rod is in contact with the liquid. That current disappears when there is no contact. The position of the rod is read from a dial meter. A correlation between the rod position and the vapor phase volume was determined from (isothermal) calibration measurements. The uncertainty of the experimental results for the vapor phase volume is estimated to be less than 5 cm3. The vapor phase samples were analyzed on-line in a gas chromatograph (HP Agilent, type 6890) equipped with a capillary column (Alltech, type Heliflex AT-Q 30 m 0.32 mm I.D.) and a thermal conductivity detector. A correlation between the primary data collected in the chromatographic measurements (i. e., the peak areas of carbon dioxide and water) and the (small) vapor phase mole fraction of water yH2O (or of CO2 ycO2) was determined in calibration measurements with binary gaseous mixtures of (carbon dioxide + water). The relative uncertainty of the experimental results for the mole fraction of water is estimated to 7 %.
For more details on the experimental procedure etc. see Schafer et. al [13], Schafer [14] and Kurz [15]. The experimental results for the composition of the vapor phase, temperature, pressure, and volume of the vapor phase were used to calculate (from the truncated virial equation of state) the amounts of carbon dioxide and water in that vapor phase. These (typically small) amounts were used together with the known amounts of the feed to determine the stoichiometric composition of the liquid phase. Some additional (also small) corrections were applied to account for the small amounts of the volatile components previously withdrawn from the cell for analysis (for more details see [13-15]).
Materials and Sample Pretreatment. Carbon dioxide (mole fraction ^ 0.99999) was purchased from Messer—Griesheim, Ludwigshafen, Germany and used as supplied. Potassium hydroxide (p.a., mass fraction > 0.85, residue — water) and boric acid (p.a., mass fraction 0.998) were bought from Riedel-de Haen, Seelze, Germany and Merck KGaA, Darmstadt, Germany, respectively. The samples were only degassed before use. The amount of water in (degassed) KOH was determined by potentiometric titration and taken into account in the calculations for the composition of the aqueous solvent feed. Water was double-distilled and degassed before use.
Experimental Investigations and Results. The solubility of CO2 was investigated at 343 K and at 383 K in an aqueous solution of KOH and H3BO3. The experimental results are summarized in Table. There, the experimental results for temperature T, composition of the liquid phase, i. e., stoichiometric molalities Wj of KOH, H3B03 and C02, pressure p, and vapor phase composition (partial pressures of CO2 (pCO2 = yCO2p) and water (pH2O = yH2Op)) are reported. The stoichiometric molality of component i is defined as the amount-of-substance (the number of moles) of that component per kilogram of water. The mass fractions of KOH and H3BO3 in the liquid feed were 0.1443 and 0.0124, respectively, corresponding to stoichiometric molalities tokoh = 3.05 mol/(kg H20) and toh3bo3 = 0.237 mol/(kg H20). The corresponding estimated uncertainties are Atokoh = 0.06 mol/(kg H20) — resulting essentially from the potentiometric titration — and Atoh3bo3 = 0.001 mol/(kg H2O). For the gas solubility measurements at 343 K (383 K), the experimental total pressures ranged between 0.03 MPa and 0.7 MPa (0.13 MPa and 0.7 MPa). The stoichiometric molality of CO2 in the liquid phase varied between 1.18 and 3.1 mol/(kg H2O) at 343 K and between 1.35 and 2.96 mol/(kg H2O) at 383 K. The experimental uncertainty of that stoichiometric molality is estimated from the uncertainties given above for the amounts of charged CO2 and water in the cell. These uncertainties as well as the estimates for the uncertainties of the partial pressures of CO2 and water are also given in Table.
The experimental results for the total pressure and the partial pressure of water are plotted versus the stoichiometric molality of CO2 in Fig. 2 and 3.
Experimental results for the solubility of carbon dioxide (2) in aqueous solutions of KOH (mkoh = 3.05 mol/(kg H20)) and H3BO3 (toh3bo3 = 0.237 mol/(kg H20))*
T, K mCo2, mol/(kg H20) p, MPa J3co2, MPa Ph2o, MPa
343.2 ±0.1 1.180 ±0.001 0.0289 ± 0.0001 < 0.0001 0.0289 ± 0.0001
1.565 ± 0.004 0.0293 ± 0.0001 0.0001 ± 0.0001 0.0292 ± 0.0001
1.970 ±0.007 0.0311 ±0.0002 0.0019 ± 0.0002 0.0292 ± 0.0002
2.324 ±0.011 0.0385 ± 0.0002 0.0117 ± 0.0009 0.0269 ± 0.0009
2.515 ± 0.014 0.0496 ± 0.0002 0.0241 ± 0.0013 0.0255 ±0.0013
2.702 ± 0.018 0.0737 ± 0.0004 0.0499 ± 0.0017 0.0238 ±0.0017
2.860 ±0.023 0.1290 ±0.0006 0.107 ±0.002 0.023 ± 0.002
2.942 ± 0.027 0.199 ±0.001 0.176 ± 0.002 0.023 ± 0.002
3.035 ± 0.031 0.387 ± 0.002 0.364 ± 0.002 0.024 ± 0.002
3.086 ± 0.035 0.601 ± 0.003 0.576 ± 0.002 0.024 ± 0.002
3.101 ± 0.039 0.691 ±0.003 0.667 ±0.002 0.024 ± 0.002
383.2 ±0.1 1.351 ± 0.002 0.1329 ±0.0007 < 0.0001 0.1329 ±0.0007
1.795 ± 0.004 0.1355 ±0.0007 0.0021 ± 0.0002 0.1334 ± 0.0002
2.188 ±0.007 0.1486 ±0.0007 0.021 ± 0.002 0.128 ±0.002
2.540 ±0.012 0.195 ±0.001 0.077 ± 0.005 0.118 ±0.005
2.766 ± 0.016 0.300 ± 0.002 0.186 ± 0.008 0.114 ±0.008
2.852 ± 0.019 0.395 ± 0.002 0.282 ± 0.009 0.113 ±0.009
2.917 ±0.023 0.519 ±0.003 0.406 ± 0.009 0.113 ±0.009
2.965 ± 0.027 0.661 ±0.003 0.547 ±0.010 0.115 ±0.010
* T — temperature; m,KOH> mH3B03 respectively; p — pressure; pco2, PH2O —
> mCo2 — stoichiometric molality of KOH, H3BO3 and CO2, partial pressure of CO2 and water, respectively.
Adding CO2 to the particular aqueous solution of (KOH + H3BO3) does not result in an increase of the total pressure beyond 0.05 MPa as long as the molar ratio of CO2 to KOH (i. e., the loading of KOH by CO2) is lower than about 0.85 because in that range CO2 is nearly completely dissolved as bicarbonate. Calculations for the speciation of the liquid phase for the boric acid-free aqueous solution with the model by Perez-Salado Kamps et al. [2] show that at a = 0.85 and 343 K more than 81 % of the totally dissolved CO2 is present as bicarbonate, less than 18 % as carbonate, and only about 0.1 % as CO2. At 383 K the amount of neutrally dissolved CO2 increases to about 0.2 %. At the upper limit of the investigated composition range, i. e., at a = 1 at 343 K (at 383 K) more than 96 % (95 %) of dissolved CO2 is present as bicarbonate, only about 1.5 % (2.1 %) as carbonate, and about 1.7 % (2.3 %) as CO2.
The experimental results for the total pressure remain practically unchanged when about 0.24 mol of boric acid are added to the aqueous solution that consists of one kilogram of water and 3.05 mol of KOH. The differences between the experimental results for the pressure above the CO2 loaded aqueous KOH-solution (with boric acid) and the calculation results from the model of Perez-Salado Kamps et al. [2] (without boric acid) are of the same order of magnitude as the experimental uncertainties. According to the model by Perez-Salado Kamps et al. [2] adding another neutral and nonreacting solute has no influence on the partial pressure of CO2 as long as interactions between the newly added solute and the other solutes (in the solvent, i. e., water) can be neglected. Adding boric acid can result in the formation of B(OH)J as well as other reaction products, i. e., in ionic solute species. As more ionic species increase the ionic strength of the solution, that might have an influence on the speciation in the liquid phases as well as on the volatility of CO2 even when all model parameters for specific interactions between the new species and the other solutes are
mCO, mol/(kg H2O)
Fig. 2. Experimental results for the pressure above aqueous solutions of KOH (rôKOH = 3.05 mol/ (kg H20)), H3BO3 (tohjBOî = 0.237 mol/(kg H2O)) and CO2 at different temperatures compared to prediction results [2] for the boric acid-free aqueous solutions (curves)
0.030
0.022
0.020
n (15(1 _
b
mCO, mol/(kg H2O)
mCO, mol/(kg H2O)
Fig. 3. Experimental results for the partial pressure of water pH2O above aqueous solutions of KOH (mKOH = 3.05 mol/(kg H20)), H3BO3 (mH3B03 = 0.237 mol/(kg H20)) and CO2 at different temperatures compared to prediction results [2] for the boric acid-free aqueous solutions (curves): the experimental uncertainty is indicated by vertical bars
a
neglected. However, at least within the investigated ranges of composition and temperature that influence of H3BO3 on the partial pressure of CO2 is small and it cannot be detected in the experiments as it is hidden behind the experimental uncertainties.
Fig. 3 shows a comparison between the experimental data and calculation results (for the boric acid-free solution) for the partial pressure of water. The experimental results agree with the calculation results within experimental uncertainty as long as the gas loading a < 2/3. At higher loadings the experimental results for the partial pressure of water (which in that region have a rather large experimental uncertainty) are lower than the calculation results (for the boric acid-free solution). The deviations reach between 10 % and 20 % at high gas loadings. However, the deviations reduce by a factor of 2 when the estimated experimental uncertainty for the partial pressure of water is taken into account, i. e., when
the deviations are calculated using the upper limit of the experimental results for the partial pressure of water. According to the model adding another solute results in a decrease of the partial pressure of the solvent. However, a crude estimate (which neglects any interaction parameters between the new solute and the other solute species) shows that a decrease of 5 % in the partial pressure of water requires about 2.8 moles of the new solute in one kilogram of water, whereas in the experiments the molality of H3BO3 was only 0.24 mol/(kg H2O). Therefore, the reason for the observed differences remains unknown. Nevertheless, the new experimental data show that at least in the investigated range of state, boric acid has only a very small influence on the volatility of CO2 in aqueous solutions of KOH.
Conclusions. In a "hot-potash" process carbon dioxide is removed from a gas stream by absorption in aqueous solutions of K2CO3 at temperatures around 70 °C. The loaded aqueous solutions are regenerated at around 110 C. Typical absorption media of such processes contain additional compounds which are described as "corrosion inhibitors" (allowing the use of cheaper construction material) and "activators" (favorably influencing either — or even both — the equilibrium solubility and the solubility kinetics). The mode of action of those additives is often unknown. One of the commonly applied additives is boric acid.
In order to develop a model for the computer-assisted design of the "hot-potash" process, in the first step a model for the equilibrium solubility of carbon dioxide is required that in a second step has to be extended to include the kinetics of the process. Based on previous work on the solubility of carbon dioxide in aqueous solutions of potassium hydroxide, the present investigation dealt with the influence of small amounts of boric acid on the equilibrium solubility of carbon dioxide under typical absorption/desorption conditions. The experimental results reveal that the influence of boric acid on the equilibrium solubility of CO2 is very small. As boric acid does not increase the solubility of CO2 in such solutions, the reason for its use is assumed in the area of reaction kinetics where it might be able to accelerate the formation of bicarbonate from carbon dioxide (maybe together with a second additive). However, that has to be proven in further investigations.
This paper dedicated to Profs. N. A. Smirnova and A. Morashevsky. References
1. EickmeyerA. G., Johnson R. F., Goar B. G. Carbon Dioxide Removal // Encyclopedia of Chemical Processing and Design / eds J. J. McKetta, W. A. Cunningham. New York; Basel: Marcel Dekker Inc., 1978. Vol. 6. P. 292-310.
2. Kamps A. P.-S., Meyer E, Rumpf B., Maurer G. Solubility of CO2 in aqueous solutions of KCl and in aqueous solutions of K2CO3 // J. Chem. Eng. Data. 2007. Vol. 52. P. 917-932.
3. Mundo K., Weber W. Anorganische Stickstoffverbindungen // Chemische Technologie, Anorganische Technologie I / eds H. Harnisch, R. Steiner, K. Winnacker, L. Kuchler. München; Wien: Carl Hanser Verlag, 1982. Bd. 2, N 4. S. 93-203.
4. Bauer G., Gunther V., Hess H. et al. Vanadium and Vanadium Compounds // Ullmann's Encyclopedia of Industrial Chemistry: 5th rev. edition / eds B. Elvers, S. Hawkins. VCH Weinheim, Germany, 1996. Vol. A27. 628 p.
5. Göppert U., Maurer G. Vapor—liquid equilibria in aqueous solutions of ammonia and carbon dioxide at temperatures between 333 and 393 K and pressures up to 10 MPa // Fluid Phase Equilib. 1988. Vol. 41. P. 153-185.
6. Kurz F., Rumpf B., Maurer G. Vapor—liquid—solid equilibria in the system NH3—CO2—H2O from around 310 to 470 K: New experimental data and modeling // Fluid Phase Equilib. 1995. Vol. 104. P. 261-275.
7. Bieling V., Kurz F., Rumpf B., Maurer G. Simultaneous solubility of ammonia and carbon dioxide in aqueous solutions of sodium sulfate in the temperature range 313-393 K and pressures up to 3 MPa // Ind. Eng. Chem. Res. 1995. Vol. 34. P. 1449-1460.
8. Kurz F., Rumpf B., SingR., Maurer G. Vapor—liquid and vapor—liquid—solid equilibria in the system ammonia — carbon dioxide — sodium chloride — water at temperatures from 313 to 393 K and pressures up to 3 MPa // Ind. Eng. Chem. Res. 1996. Vol. 35. P. 3795-3802.
9. Sing R., Rumpf B., Maurer G. Solubility of ammonia in aqueous solutions of single electrolytes sodium chloride, sodium nitrate, sodium acetate, and sodium hydroxide // Ind. Eng. Chem. Res. 1999. Vol. 38. P. 2098-2109.
10. Meier E., Ermatchkov V., Kamps A. P.-S., Maurer G. Simultaneous solubility of SO2 and NH3 in (salt + H2O) and enthalpy change upon dilution of (SO2 + NH3 + salt + H2 O) in (salt + H2O) {salt = (NH4)2SO4 or Na2SO4}: Experimental results and model predictions // Fluid Phase Equilib. 2006. Vol. 244. P. 137-152.
11. Schäfer D., VogtM., Kamps A. P.-S., Maurer G. Solubility of ammonia in liquid mixtures of (water + methanol) // Fluid Phase Equilib. 2007. Vol. 261. P. 306-312.
12. Schäfer D., VogtM., Kamps A. P.-S., Maurer G. Influence of a single salt (NaCl/Na2 SO4) on the solubility of ammonia in liquid mixtures of (water + methanol) // Ind. Eng. Chem. Res. 2008. Vol. 47. P. 5112-5118.
13. Schäfer D., Kamps A. P.-S., Maurer G. Experimental investigation of the simultaneous solubility of chemical reacting gases NH3 and CO2 in liquid mixtures of (H2O + CH3OH) // J. Chem. Eng. Data 2011. Vol. 56. P. 1084-1095.
14. SchaferD. Untersuchungen zur Löslichkeit von Ammoniak und Kohlendioxid in organisch-wassrigen Losungsmittelgemischen. Doctoral Dissertation. Kaiserslautern, Germany, 2003.
15. KurzF. Untersuchungen zur simultanen Lösung von Ammoniak und Kohlendioxid in Wasser und salzhaltigen wassrigen Lösungen. Doctoral Dissertation. Kaiserslautern, Germany, 1994.
Статья поступила в редакцию 24 июля 2012 г.
