Novel Non-aqueous Amine Solvents for Biogas Upgrading - Energy

Jul 21, 2014 - ... which allowed us to identify and quantify the carbonated species in solution originated from both amine and alcohol carbonatation. ...
0 downloads 0 Views 1MB Size
Download PDF
Article pubs.acs.org/EF

Novel Non-aqueous Amine Solvents for Biogas Upgrading Francesco Barzagli,†,‡ Sarah Lai,‡ Fabrizio Mani,*,† and Piero Stoppioni‡ †

Institute of Organometallic Chemistry (ICCOM), National Research Council (CNR), via Madonna del Piano 10, 50019 Sesto Fiorentino, Florence, Italy ‡ Department of Chemistry, University of Florence, via della Lastruccia 3, 50019 Sesto Fiorentino, Florence, Italy ABSTRACT: Reducing the CO2 and H2S contents of biogas is a prerequisite to raise its quality to that of natural gas. The chemical capture of carbon dioxide was accomplished with non-aqueous single 2-amino-2-methyl-1-propanol (AMP), 2-(tertbutylamino)ethanol (TBMEA), 2-(isopropylamino)ethanol (IPMEA), and N-methyl-2,2′-iminodiethanol (MDEA), with their 1:1 blends dissolved in either an ethylene glycol/1-propanol mixture or single diethylene glycol monomethyl ether. The gas mixtures used contain either 15 or 40% CO2 in air, sometimes added with 50 ppm H2S. We designed two different experimental procedures: (1) separate experiments of CO2 absorption and desorption aimed at selecting the most efficient amines and (2) continuous cycles of CO2 absorption (20 °C) and desorption (90−95 °C) featuring CO2 removal efficiency in the range of 89− 96%. The CO2/amine/alcohol equilibria were analyzed by 13C nuclear magnetic resonance (NMR) spectroscopy, which allowed us to identify and quantify the carbonated species in solution originated from both amine and alcohol carbonatation. In some continuous cycles of CO2 absorption, H2S was selectively captured by aqueous H2O2 and separated as either elemental sulfur or CaSO4.

■

INTRODUCTION Methane production from municipal biomass waste is a topic of increasing interest as an environmentally benign and available energy option. Besides CH4, biogas from landfill wastes contains approximately 15−50% CO2, small amounts of H2S and N2, and trace amounts of other compounds. Biogas cleaning and upgrading to natural gas quality is the prerequisite for most of its use; afterward, it can be mixed with natural gas and injected into an existing gas grid or used in power plants or as transportation fuel. The cleaning of the biogas requires H2S, water, and trace compound removal. In particular, H2S must be removed because of its harmfulness and corrosive nature in the presence of water. To raise its energy content, it is mandatory to upgrade the biogas by means of CO2 removal. The most common techniques for CO2 removal are the high-pressure water scrubbing, pressure and thermal swing absorption, physical and chemical capture by organic solvents or amines, and membrane or cryogenic separation.1−10 More recently, metal− organic frameworks (MOFs) are emerging as promising alternatives for CO2 separation.11 All of these techniques display both advantages and disadvantages, and their cost to benefit assessment helps to choose the technique for the implementation in a large-scale plant. Amine scrubbing is based on the chemical capture of CO2 by aqueous alkanolamines, in particular, 2-aminoethanol (MEA), and the absorbent regeneration at high temperature.12−18 However, the high energy consumption associated with the amine regeneration19−22 and the amine degradation23,24 (up to 30% every year) because of the high stripping temperature (110−140 °C at a pressure of over 1 bar) are the major obstacles to the large-scale application of this technology. In an effort to take advantage of the high efficiency of aqueous alkanolamines while reducing their disadvantages, we have devised the strategy to replace water with organic solvents © 2014 American Chemical Society

to redirect the reaction of CO2 capture toward less stable carbonated species, which thereby require lower regeneration temperature.25,26 Additional benefits come from the much lower heat capacity of the organic solvents compared to water, lower ammine degradation and evaporation because of the reduced stripping temperature, and lower corrosion of the equipment. In the experimental study reported here, we have investigated an efficient method for the selective separation of CO2 and H2S from a simulated biogas containing either 15 or 40% CO2 in air and, in some experiments, 50 ppm H2S. It was mandatory to perform the absorption experiments with the CO2/air mixture instead of CO2/methane, because the use of methane is not allowed in our laboratories because of safety rules. However, air and methane display the same features in the absorption experiments because of the methane inertness toward amines. The absorbents were either single or blended amines (1:1, on a molar scale), overall 3.0 mol dm−3, dissolved in a 1:1 (volume scale) mixture of ethylene glycol (EG) and 1propanol (PrOH) or in single diethylene glycol monomethyl ether (DEGMME). The selected amines were 2-amino-2methyl-1-propanol (AMP), 2-(tert-butylamino)ethanol (TBMEA), 2-(isopropylamino)ethanol (IPMEA), and Nmethyl-2,2′-iminodiethanol (MDEA). We have designed two sets of experiments: (1) CO2 absorption and, separately, thermal amine regeneration aimed at selecting the most efficient absorbents and (2) continuous cycles of CO 2 absorption−desorption carried out in packed columns with the purpose of verify the efficiency of CO2 and H2S capture. The carbonated species in solution were qualitatively and quantitatively analyzed by 13C nuclear magnetic resonance Received: February 6, 2014 Revised: July 18, 2014 Published: July 21, 2014 5252

dx.doi.org/10.1021/ef501170d | Energy Fuels 2014, 28, 5252−5258

Energy & Fuels

Article

(NMR) spectroscopy27,2 and were found to be originated from the amine and alcohol carbonatation. Until now, non-aqueous solutions of alkanolamines (MEA, DEA, and MDEA) were used for kinetic studies of CO2 capture.29,30 Additionally, alkanolamines combined with room-temperature ionic liquids31 and properly designed organic liquids32,33 were investigated for reversible CO2 capture. However, the very high viscosity of the carbonated species and the procedures often required by an efficient CO2 capture or absorbent regeneration prevented their application to a continuous CO2 absorption−desorption cycle.

■

RESULTS AND DISCUSSION Batch Experiments of CO2 Absorption and Desorption. The choice of the AMP-based absorbents was due to the high thermal stability of the amine and, on the contrary, to the relatively low stability of the corresponding carbamate, as compared, for example, to those of 2-aminoethanol (MEA), Nmethyl-2-aminoethanol (MMEA), and 2,2′-iminodiethanol (DEA). The choice of the alcohols was based on four main constraints: (i) solubility of the carbonated compounds, (ii) high boiling temperatures, (iii) low costs, and (iv) avoidance of foaming problems. We found that ethylene glycol (hereafter indicated as EG) prevented the formation of precipitates upon CO2 absorption as well as immiscible liquid phases and foaming. To the purpose of lowering the high absolute viscosity (μ = 19.9 mPa s, at 20 °C) of pure EG that would reduce the CO2 absorption efficiency and would be an hindrance to the absorbent flow, a 1:1 (v/v) mixture of EG with 1-propanol (μ = 1.07 mPa s, at 20 °C) was used and a boiling temperature up to 150 °C was still attained. Finally, the amine solutions in pure 2(2-methoxyethoxy)ethanol [diethylene glycol monomethyl ether (DEGMME)] were used for comparison purposes. The absorption experiments at 25 °C were carried out into a 2.0 dm3 flask containing the absorbent (0.040 dm3) and filled with the appropriate gas mixture (see the Experimental Section) at room pressure. The airtight flask is equipped with an electronic pressure gauge and a magnetic stirrer. From the pressure decrease during the CO2 absorption (15 and 40%, v/v, CO2 in air at 25 °C), the amount of absorbed CO2 was measured as a function of time. When the pressure did not change with time (20−45 min), the steady state was reached and the experiment was stopped. The percentage of CO2 absorbed by single or 1:1 (molar scale) amine blends, overall 3.0 mol dm−3, in 1:1 (v/v) EG/PrOH or DEGMME is summarized in Figures 1 and 2 as a function of the absorption time. Overall, the percentage (in other words, the absorption efficiency) decreases in the order: (1) AMP > IPMEA > TBMEA > MDEA, (2) AMP/IPMEA ≥ AMP/TBMEA > AMP/MDEA, (3) amine blends > single amines (with the exception of AMP), (4) EG/PrOH > DEGMME, and (5) 15% CO2 > 40% CO2. As far as the single amines are concerned, the absorption efficiency decreases with increasing steric hindrance at amine function. The absorption efficiency is comprised between 13% (MDEA−DEGMME and 40% CO2) and 100% (several amines, EG/PrOH, and 15% CO2). The greater amine efficiency with 15% CO2 compared to 40% CO2 is due to the greater amine/CO2 ratio. For the same experimental conditions, the mixture EG/PrOH has an appreciable advantage over DEGMME because of the greater reactivity toward CO2 (reaction 2). In particular, the MDEA efficiency increases from 13% in DEGMME to 52% in EG/PrOH (15% CO2). The tertiary amine MDEA is unable to react with CO2 in

Figure 1. Variation of the CO2 (A, 15%; B, 40%) absorption efficiency of the different amines in DEGMME solutions as a function of time.

Figure 2. Variation of the CO2 (A, 15%; B, 40%) absorption efficiency of the different amines in EG/PrOH solutions as a function of time.

the absence of water, and DEGMME is poorly reactive toward CO2, contrary to EG and PrOH. The single amines AMP and IPMEA and the mixed AMP/IPMEA display a greater initial reaction rate, irrespective of the solvent employed and the CO2 percentage. The amine regeneration occurred in a conical flask heated at 90 °C, and the CO2 release was measured with two 100 cm3 gas burets filled with CO2 saturated water, as previously 5253

dx.doi.org/10.1021/ef501170d | Energy Fuels 2014, 28, 5252−5258

Energy & Fuels

Article

described.34 The CO2 desorption required a maximum of 60 min. Figures 3 and 4 summarize the average values of three

Figure 4. CO2 desorption efficiency of the different amines carbonated by (A) 15% CO2 and (B) 40% CO2 in EG/PrOH. The desorption efficiency is compared to the average absorption efficiency in the same solutions.

Figure 3. CO2 desorption efficiency of the different amines carbonated by (A) 15% CO2 and (B) 40% CO2 in DEGMME. The desorption efficiency is compared to the average absorption efficiency in the same solutions.

CO2 + 2AmH ⇄ AmCO2− + AmH 2+

(1)

In the absence of water, tertiary amines, such as MDEA, cannot react with CO2. However, in a mixture with primary or secondary amines, MDEA takes part in the CO2 capture by deprotonating the carbamic acid. Besides primary and secondary amines, the alcohol too could react with CO2 in the presence of the amine, yielding the alkyl carbonate (eq 2; R denotes an alkyl group).

measurements of CO2 absorption and desorption efficiency (the latter is the percentage of desorbed CO2 with respect to that absorbed), after the steady state was reached, according to the different amines and experimental conditions. In general, the desorption and absorption efficiencies display an opposite trend; this feature is particularly evident for MDEA. By considering both absorption and desorption efficiencies, the rate of CO2 absorption, and the absence of solid compounds during the CO2 capture, AMP, AMP/IPMEA, and AMP/ TBMEA dissolved in EG/PrOH were selected for the experiments in a closed cycle of absorption−desorption, where the kinetic constraints, presumably, prevail over the thermodynamic properties. Analogous experiments carried out for comparison purposes with the amines MEA, DEA, and MMEA gave high absorption efficiency (over 95%) but were discarded for cyclic experiments because of the low desorption efficiency (15−30% at 90 °C). Chemistry of CO 2 Absorption in Non-aqueous Solvents and Analysis of the Carbonated Species Based on 13C NMR Spectroscopy. In anhydrous conditions, CO2 reacts with an excess of both primary and secondary amines, yielding the amine carbamates (eq 1; AmH denotes either primary or secondary amine).

CO2 + AmH + ROH ⇄ ROCO2− + AmH 2+

(2)

In EG/PrOH and DEGMME, R is CH 3 (CH 2 ) 2 −, HOCH2CH2−, and to a much less extent, CH3O(CH2)2O(CH2)2−. The fast CO2 alkylation that we have found to easily occur in the presence of common amines at room conditions is noteworthy. As a matter of fact, the carbonatation reaction of saturated alcohols already reported has required specific promoters, such as alkyl bromide and Cs2CO3,35 dibutyl tin(IV)oxide,36 and amidine or guanidine,32 whereas the carbonatation of unsaturated alcohols occurred in the presence of tert-butyl ipoiodide.37 However, the alkyl carbonates are less stable than the carbamates of MEA, MMEA, and DEA, and consequently, reaction 2 occurred with the sterically hindered amines that give relatively unstable carbamates. The thermal decomposition of stable carbamates is slower than that of alkyl carbonates and, consequently, requires higher temperatures. On the contrary, the amines AMP, TBMEA, and IPMEA in EG/ 5254

dx.doi.org/10.1021/ef501170d | Energy Fuels 2014, 28, 5252−5258

Energy & Fuels

Article

Figure 5. 13C NMR spectrum of IPMEA−EG/PrOH carbonated by 40% CO2. The inset reports the chemical shifts of carbonyl atoms of IPMEA carbamate (δ = 164.7 ppm) and alcohol carbonates (δ = 160.1−159.8 ppm). Asterisks denote the chemical shifts of carbon backbones of IPMEA carbamate and both free IPMEA and protonated IPMEA fast exchanging in the NMR scale.

is the prevailing species in the IPMEA and AMP/IPMEA solutions carbonated by 15% CO2 (70−75% in AMP/IPMEA; AMP−CO2−, PrOH > DEGMME and the carbamate stability decreases in the order IPMEA > AMP > TBMEA. The tertiary amine MDEA is unable to form the corresponding carbamate, and consequently, MDEA− DEGMME is the least efficient absorbent; the small absorption capacity (13−16%) measured is, presumably, due to a physical absorption. The 40% CO2 absorption by AMP, IPMEA, and TBMEA and their blends in DEGMME gave rise to the amine carbamates [AmCO2−][AmH2+] in the solid state (AmH denotes the amine). The lattice energy because of the ionic interactions stabilizes the carbamates of AMP and TBMEA, despite their relative instability in solution.25 Continuous Cycles of CO2 Absorption−Desorption and H2S Capture. The CO2 absorption efficiency was measured in continuous cycles of CO2 capture and solvent regeneration carried out, separately, in two glass columns packed with glass rings and set at the appropriate temperatures.25,26 In such a cyclic process, the carbonated and regenerated solutions are continuously transferred from the absorber to the desorber and vice versa. The hot lean solution exiting from the desorber became cold at room temperature, while it was recycled to the absorber. The gas mixture of either 15 or 40% (v/v) CO2 in air came in contact with the absorbent in the countercurrent mode. The gas exited from the top of the absorber was dried and purified before being gas chromatography (GC)-analyzed. Each complete experiment lasted 24−36 h before the reactions of CO2 capture and amine regeneration reached a steady state. On the basis of the aforesaid batch results, the amines AMP, AMP/IPMEA, and AMP/TBMEA in

PrOH give less stable carbamates and promote the formation of alkyl carbonates of EG and PrOH. Therefore, these absorbents may be efficient in both absorption and desorption steps. To evaluate the distribution of the species in solution at the end of the absorption experiments, we analyzed the CO2loaded solutions by 13C NMR spectroscopy, as already reported for aqueous solutions.27,28 The peak assignment was performed according to literature data.38 The most intense signals are due to the carbon backbones of the alcohols and both protonated and free amines that are fast exchanging on the NMR time scale, whereas the lower intensity resonances were easily assigned to amine carbamates and alkyl carbonates. The two distinct resonances in the ranges of 67.96−66.98 and 61.75− 60.81 ppm are assigned to carbon atoms of the monoalkyl carbonate derivative of ethylene glycol (HO−CH2−CH2− OCO2−), and the resonance at 160.11−159.10 ppm is ascribed to alkyl carbonate quaternary carbon (R−OCO2−). The resonances of 1-propanol carbonate are in the ranges of 160.11−159.32, 67.16−66.12, 22.71−21.45, and 10.88−10.26 ppm. Finally, the resonances of the carbamate derivatives of AMP were identified at 165.10−164.05, 72.26−71.61, 53.54− 51.62, and of IPMEA 22.71−21.07 ppm and at 164.76−164.14, 61.52−60.36, 52.42−51.62, 49.65−48.55, and 21.68−21.07 ppm. The 13C resonances of TBMEA carbamate are very weak. As an example, the 13C NMR spectrum of carbonated IPMEA (40% CO2) in EG/PrOH is reported in Figure 5. The careful integration of the carbon resonances of carbamate R−CO2− and carbonate R′−OCO2− falling in the range of 164−159 ppm provides a reliable estimation (estimated error 5%) of the relative amounts of carbamate and alkyl carbonate, despite the lower intensity resonances of the quaternary 13C atoms of R−OCO2− and R′−CO2−, because of the much longer relaxation time, as compared to the −CH2− groups. The CO2 capture as either alkyl (EG, PrOH, and DEGMME) carbonate or amine (AMP, IPMEA, and TBMEA) carbamate is the result of the competition between the relative stability of the alkyl carbonates and amine carbamates and the CO2/amine ratio. The IPMEA carbamate 5255

dx.doi.org/10.1021/ef501170d | Energy Fuels 2014, 28, 5252−5258

Energy & Fuels

Article

1:1 (v/v) EG/PrOH were employed in the continuous cycle experiments. A summary of the operating conditions is reported in Table 1, whereas the results of the cyclic experiments are reported in Table 2.

higher than that of 15% is most likely due to the increased CO2 partial pressure, which enhances the reaction rate. The selective capture of H2S and its conversion into a harmless compound was performed with aqueous H2O2 in two absorption−desorption cycles of AMP/IPMEA/DEGMME solution (0.400 dm3, 3.00 mol dm−3) sequentially to the CO2 and H2S stripping. In one experiment, a gas mixture of 15% CO2 in air also containing 50 ppm H2S was used. The oxidation reaction was

Table 1. Operating Conditions Employed in the Continuous Absorption−Desorption Experiments solution volume absorption temperature desorption temperature/pressure overall amine concentration liquid flow rate gas flow rate gas mixture

0.400 and 0.450 dm3 of 50% carbonated solution 20 °C 90 and 95 °C/1 bar

H 2S + H 2O2 → S + 2H 2O

To verify whether this system successfully works with a greater amount of sulfide as well, the absorbent solution was charged with 5.0 g of Na2S·9H2O (0.021 mol) dissolved in the minimum amount of water. With 0.65 mol dm−3 H2O2 and a molar ratio 3:1 H2O2/H2S, the oxidation reaction was

3.00 mol dm−3 (single or 1:1 molar ratio) 0.330, 0.650, and 0.950 dm3/h 12.2 dm3/h 15 and 40% (v/v) CO2 in air

Table 2. CO2 Absorption Efficiency of Amines in EG/PrOH in the Continuous Absorption−Regeneration Cycles as a Function of the Operational Parameters amine

liquid flow rate (dm3/h)

Tabsorb (°C)

Tdesorb (°C)

H 2S + 4H 2O2 → SO4 2 − + 2H+ + 4H 2O

AMP−IPMEA

AMP−TBMEA

0.33 0.65 0.65 0.95 0.33 0.65 0.65 0.65 0.95 0.65

efficiency (%)

SO4 2 − + 2H+ + CaO → CaSO4 + H 2O 20 20 20 20 20 20 20 20 20 20

95 90 95 95 90 90 95 95 90 95

95.0 93.5a 96.5a 92.0a 89.1 92.1 94.0 94.7a 89.4 95.7a

AMP−IPMEA

AMP−TBMEA

0.33 0.33 0.33 0.33 0.33 0.33 0.65 0.33 0.65

20 20 20 20 20 20 20 20 20

90 95 95 90 95 90 95 95 95

91.3 93.3 95.4a 88.9 91.3 89.4a 86.7 91.0a 93.3a

(5)

The oxidation reactions 3 and 4 were complete, and no detectable amount of H2S exited from the system (see the Experimental Section).

■

CONCLUSION

The study reported here describes the chemical capture of CO2 by non-aqueous solvents containing single AMP, IPMEA, TBMEA, and MDEA or their blends and ethylene glycol/ propanol mixture or diethylene glycol monomethyl ether. The 13 C NMR analysis indicates that CO2 is captured as monoalkyl carbonate, R−OCO2− (R = CH2CH2OH and n-C3H7), and amine (AMP and IPMEA) carbamate. Absorption efficiencies up to 96% are due to the fast reactions at room temperature, whereas the stripping temperatures of 90−95 °C at room pressure are due to the lower stability of both monoalkyl carbonates and amine carbamates with respect to HCO3− and stable amine carbamates that are formed in aqueous solutions. The potential advantages of the non-aqueous absorbents with respect to the conventional aqueous amines are (i) the lower heat capacity and evaporation enthapy of the organic solvents (about half) compared to water and (ii) the lower stripping temperature, as compared to that of aqueous alkanolamines (110−140 °C at a pressure over 1 bar), that should reduce the amine evaporation and degradation and the equipment corrosion. In two continuous absorption−desorption experiments, H2S was selectively oxidized with aqueous H2O2 and separated from CO2 by either harmless elemental sulfur or CaSO4 with the addition of CaO. The selective capture of H2S reported here is more simple and efficient, in our opinion, than the traditional capture protocols based on Fe(II) precipitation, oxidation with Fe(III) complexes, activated carbon, etc. As a conclusion, the possible replacement in a commercial plant of the conventional aqueous amines by the non-aqueous absorbents reported here requires an accurate assessment of the operating and investment costs and benefits of the entire process, and tests performed in a pilot plant are necessary to select the most efficient amine/alcohol blend and the best operational conditions.

15% CO2 AMP

(4)

SO42− was separated from the solution with the stoichiometric amount of CaO that neutralized the acidity of reaction 4.

40% CO2 AMP

(3)

a Volume of the absorbent is 0.450 dm3, and in all of the other experiments, the volume of the absorbent is 0.400 dm3.

As expected, the absorption efficiency increases with the desorption temperature and with the increased volume of amine solution. The dependence upon the liquid flux deserves a comment. The lower flux, 0.330 dm3/h, increases the residence time of the liquid into both absorber and desorber, thereby enhancing both the absorption and desorption reactions. The higher flux, 0.950 dm3/h, makes a greater amount of the sorbent to come from the desorber to the absorber and vice versa. The higher absorption efficiency was the result of the best compromise between the two opposite features that allowed us to transfer the greatest amount of both loaded and regenerated absorbent between the desorber and the absorber. Anyway, an absorption efficiency up to 95−96% with 40% CO2 and 91− 95% with 15% CO2 was obtained by adopting the best operational conditions. The absorption efficiency of 40% CO2 5256

dx.doi.org/10.1021/ef501170d | Energy Fuels 2014, 28, 5252−5258

Energy & Fuels

■

Article

the solutions at room temperature with a procedure that has already been described.25,26 Chemical shifts are high-frequency relative to tetramethylsilane as an external standard at 0.00 ppm. CH3CN was used as an internal reference (CH3, δ = 1.47). To obtain a quantitative analysis of the 13C{1H} spectra, the pulse sequence with proton decoupling and nuclear Overhauser effect (NOE) suppression was used. The acquisition parameters were pulse angle of 90.0°, acquisition time of 1.3632 s, delay time of 2−30 s, data points of 65 000, and number of scans of 250−500. The relative integration of signals of the −CH2− carbon atoms containing the same number of attached protons was not affected by an increased acquisition time and/or relaxation delay (up to 60 s). The carbon atoms of R−CO3− and R′− CO2−, with no attached hydrogen, have longer relaxation times and lower intensity resonances than those of −CH2−.39,40 Consequently, there is some uncertainty in the integration of the carbon atoms of R− CO3− and R′−CO2−, and we estimated a 5% deviation of the relative amounts of the two species.

EXPERIMENTAL SECTION

General Information. All reagents were reagent-grade. The alkanolamines and alcohols (Sigma-Aldrich) were used as received without further purification. Gas mixtures of 15 and 40% CO2, 50 ppm H2S, and air (Rivoira Spa) were used to simulate flue gas. Flow rates were measured with gas mass flow meters (Aalborg) equipped with gas controllers (Cole Parmer). The inlet and outlet CO2 concentrations in the flue gas mixture were measured with a Varian CP-4900 gas chromatograph calibrated with 15 and 40% (v/v) CO2/air reference mixtures (Rivoira Spa) and 100% CO2 reference gas (Rivoira Spa). The starting amine solutions were fixed at overall 3.00 mol dm−3; amine blends were a 1:1 molar ratio. Mixed EG/PrOH was 1:1 (v/v). Gas mixtures of 15 and 40% CO2 in air were used to simulate the biogas. In some experiments, the gas mixture contains 50 ppm H2S. The CO2 absorption in the batch experiments was measured with a gastight apparatus, which comprised of a 2.0 dm3 flask (actual volume of 2.20 dm3) equipped with a digital pressure gauge, magnetic stirrer, and pressure-equalizing dropping funnel containing 40.0 cm3 of the appropriate ammine (0.12 mol) solution. After the air was removed with a vacuum pump, the flask was filled with the appropriate CO2 mixture at room pressure. This operation was repeated 5 times before the final one. After the amine solution was quickly introduced from the funnel into the flask, the stirring was started and the pressure decrease shown by the pressure gauge allowed us to measure the CO2 absorption as a function of time. The experiment was stopped when the pressure did not change with time (20−45 min). The thermal release of pure CO2 was accomplished with two 100 cm3 gas burets (Volac), as already described.34 The desorption rate at 90 °C rapidly decreased with time, and a steady state was reached within 60 min. The apparatus and operational procedures used for continuous absorption−desorption cycles have already been described.26 The absorber and desorber are glass columns packed with glass rings and heated by the liquid circulating through the jackets and maintained at the appropriate temperature by means of a thermostated bath (Julabo model F33-MC bath). The solutions circulate continuously between the absorber and desorber at the flow rate of 0.330, 0.650, and 0.950 dm3 h−1. Because of the endothermic reaction of CO2 release, the temperature of the desorption column was 1.5−2.0 °C below that of the heating jacket (90 and 95 °C). The absorbent and gas mixture come into contact in the countercurrent mode; the regenerated absorbent dropped from the top of the absorber, while the gas mixture was continuously fed into the bottom of the column with a flow rate of 12.2 dm3 h−1 (15% CO2, 0.0750 mol h−1; 40% CO2, 0.200 mol h−1, at 24 °C). The carbonated absorbent was sent to the top of the desorber column for regeneration and CO2 release. A water-cooled condenser refluxed the possible overhead vapor to the desorber. The equipment was charged with 0.400−0.450 dm3 of amine− alcohol solution that had been previously 50% saturated with CO2. These starting solutions were prepared by mixing 0.200−0.225 dm3 of the CO2-saturated amine solution with the volume of the amine solution necessary to obtain the overall volume. The gas exited from the absorber was dried and purified before being analyzed by the gas chromatograph. A complete cyclic experiment lasted 24−36 h and was stopped when the efficiency of CO2 capture remained constant with time. In the experiments aimed at separating H2S from CO2, a gas stream containing 50 ppm H2S was employed. Another set of H2S capture experiments were carried out with the AMP/IPMEA/ DEGMME solution charged with 5.0 g of Na2S·9H2O (0.021 mol) dissolved in the minimum amount of water. The selective removal of H2S was accomplished subsequently to the CO2−H2S stripping by flowing the gas mixture through a H2O2 aqueous solution of either 0.10 or 0.65 mol dm−3. The precipitation of CaSO4 was accomplished in a separate vessel with the stoichiometric amount of CaO. The unreacted H2O2 solution was recycled to the H2S absorber. To check the efficiency of H2S removal, the CO2 stream that exited from the H2S absorber was bubbled through an aqueous solution of Cu(II). No CuS precipitation was detected in all of the experiments. 13 C NMR Spectroscopy. The Bruker AvanceIII 400 spectrometer operating at 100.613 MHz was used to obtain the 13C NMR spectra of

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This study has been accomplished with the financial support of Agenzia Nazionale per le Nuove Tecnologie, l’Energia e lo Sviluppo Economico Sostenibile (ENEA, Rome, Italy). Thanks are expressed to Maurizio Passaponti for technical assistance.

■ ■

DEDICATION This paper is dedicated to the memory of our colleague and friend Piero Stoppioni. REFERENCES

(1) Heyne, S.; Harvey, S. Int. J. Energy Res. 2014, 38, 299−318. (2) Privalova, E. I.; Mäki-Arvela, P.; Eränen, K.; Avetisov, A. K.; Mikkola, J. P.; Murzin, D. Y. Chem. Eng. Technol. 2013, 36, 740−748. (3) Lopez, M. E.; Rene, E. R.; Veiga, M. C.; Kennes, K. In Environmental Chemistry for a Sustainable World; Lichtfouse, E., Schwarzbauer, J., Robert, D., Eds.; Springer: Dordrecht, Netherlands, 2012; Vol. 2, pp 347−378. (4) Xue, Q.; Liu, X. Sep. Sci. Technol. 2011, 46, 679−686. (5) Götz, M.; Koppel, W.; Reinert, R.; Graf, F. Chem. Ing. Tech. 2011, 83, 1−10. (6) Lopes, F. V. S.; Grande, C. A.; Rodrigues, A. E. Chem. Eng. Sci. 2011, 66, 303−317. (7) Chaffee, A. L.; Knowles, G. P.; Liang, Z.; Zhang, J.; Xiao, P.; Webley, P. A. Int. J. Greenhouse Gas Control 2007, 1, 11−18. (8) Yang, H.; Xu, Z.; Fan, M.; Gupta, R.; Slimane, R. B.; Bland, A. E. J. Environ. Sci. 2008, 20, 14−27. (9) Sridhar, S.; Smitha, B.; Aminabhavi, T. M. Sep. Purif. Rev. 2007, 36, 113−174. (10) Bucklin, R. W.; Schendel, R. L. Energy Prog. 1984, 4, 137−142. (11) Chaemchuen, S.; Kabir, N. A.; Zhou, K.; Verpoort, F. Chem. Soc. Rev. 2013, 42, 9304−9332. (12) Bottoms, R. R. Ind. Eng. Chem. 1931, 23, 501−504. (13) Teller, A. J.; Ford, H. E. Ind. Eng. Chem. 1958, 50, 1201−1206. (14) Caplow, M. J. J. Am. Chem. Soc. 1968, 90, 6795−6803. (15) Oyenekan, B. A.; Rochelle, G. T. Ind. Eng. Chem. Res. 2006, 45, 2457−2464. (16) Idem, R.; Wilson, M.; Tontiwachwuthikul, P.; Chakma, A.; Veawab, A.; Aroonwilas, A.; Gelowitz, D. Ind. Eng. Chem. Res. 2006, 45, 2414−2420. (17) Bonenfant, D.; Mimeault, M.; Hausler, R. Ind. Eng. Chem. Res. 2003, 42, 3179−3184. 5257

dx.doi.org/10.1021/ef501170d | Energy Fuels 2014, 28, 5252−5258

Energy & Fuels

Article

(18) Mandal, B. P.; Biswas, A. K.; Bandyopadhyay, S. S. Sep. Purif. Technol. 2004, 35, 191−202. (19) Reynolds, A. J.; Verheyen, T. V.; Adeloju, S. B.; Meuleman, E.; Feron, P. Environ. Sci. Technol. 2012, 46, 3643−3654. (20) Oexmann, J.; Kather, A. Int. J. Greenhouse Gas Control 2010, 4, 36−43. (21) Rao, A. B.; Rubin, E. S. Ind. Eng. Chem. Res. 2006, 45, 2421− 2429. (22) Jassim, M. S.; Rochelle, G. T. Ind. Eng. Chem. Res. 2006, 45, 2465−2472. (23) Lepaumier, H.; Picq, D.; Carrette, P. L. Ind. Eng. Chem. Res. 2009, 48, 9061−9067. (24) Goff, G. S.; Rochelle, G. T. Ind. Eng. Chem. Res. 2006, 45, 2513− 2521. (25) Barzagli, F.; Mani, F.; Peruzzini, M. Int. J. Greenhouse Gas Control. 2013, 16, 217−223. (26) Barbarossa, V.; Barzagli, F.; Mani, F.; Lai, S.; Stoppioni, P.; Vanga, G. RSC Adv. 2013, 3, 12349−12355. (27) Barzagli, F.; Mani, F.; Peruzzini, M. Int. J. Greenhouse Gas Control 2011, 5, 448−456. (28) Barzagli, F.; Mani, F.; Peruzzini, M. Energy Environ. Sci. 2009, 2, 322−330. (29) Park, S. W.; Lee, J. W.; Choi, B. S.; Lee, J. W. J. Ind. Eng. Chem. 2005, 11, 202−209. (30) Xu, H. J.; Zhang, C. F.; Zheng, Z. S. Ind. Eng. Chem. Res. 2002, 41, 6175−6180. (31) Camper, D.; Bara, J. E.; Gin, D. L.; Noble, R. D. Ind. Eng. Chem. Res. 2008, 47, 8496−8498. (32) Heldebrant, D. J.; Yonker, C. R.; Jessop, P. G.; Phan, L. Energy Environ. Sci. 2008, 1, 487−493. (33) Liu, Y. X.; Jessop, P. G.; Cunningham, M.; Eckert, C. A.; Liotta, C. L. Science 2006, 313, 958−960. (34) Mani, F.; Peruzzini, M.; Stoppioni, P. Energy Fuels 2008, 22, 1714−1719. (35) Kim, S. I.; Chu, F.; Dueno, E. E.; Jung, K. W. J. Org. Chem. 1999, 64, 4578−4579. (36) Jimil, G.; Yogesh, P.; Muthukumaru, P. S.; Pradip, M. J. J. Mol. Catal. A: Chem. 2009, 304, 1−7. (37) Minakata, S.; Sasaki, I.; Ide, T. Angew. Chem., Int. Ed. 2010, 49, 1309−1311. (38) Ma’mun, S.; Jakobsen, J. P.; Svendsen, H. F.; Juliussen, O. Ind. Eng. Chem. Res. 2006, 45, 2505−2512. (39) Hook, R. J. Ind. Eng. Chem. Res. 1997, 36, 1779−1790. (40) Breitmaier, E.; Voelter, W. Carbon-13 NMR Spectroscopy, 3rd ed.; Wiley-VCH: Weinheim, Germany, 1990.

5258

dx.doi.org/10.1021/ef501170d | Energy Fuels 2014, 28, 5252−5258