Accelerated Aging and Qualitative Degradation Pathway Analysis of

Aug 1, 2012 - Mahendra Yadav , John C. Linehan , Abhijeet J. Karkamkar , Edwin van der Eide , and David. ... Eudes Eterno Fileti , Vitaly V. Chaban...
1 downloads 0 Views 226KB Size
Download PDF
Article pubs.acs.org/EF

Accelerated Aging and Qualitative Degradation Pathway Analysis of CO2 Capture Solvents Containing Ionic Liquids Andrew L. LaFrate,‡ Michael C. Huffman,‡ Nathan Brown,‡ Matthew S. Shannon,† Ken Belmore,§ Jason E. Bara,† and Alfred E. Brown*,‡ ‡

ION Engineering, Boulder, Colorado 80301, United States Department of Chemical and Biological Engineering, The University of Alabama, Tuscaloosa, Alabama 35487-0203, United States § Department of Chemistry, The University of Alabama, Tuscaloosa, Alabama 35487-0336, United States †

S Supporting Information *

ABSTRACT: A qualitative study was conducted to analyze degradation of CO2 capture solvents containing ionic liquids by GC/ MS under a variety of conditions representative of the carbon capture process. Aqueous monoethanolamine (MEA) was used as a benchmark for solvent degradation and compared to two advanced CO2 capture solvents containing ionic liquids. Accelerated aging experiments were conducted in the laboratory by exposing solvents to various gases (N2, air, CO2, or SO2) at elevated temperature in sealed cylinders. Degradation products were analyzed by GC/MS and key pathways leading to their formation were elucidated. For all solvents, degradation was most rapid and lead to a wider array of products in the presence of CO2 compared to the other studied gases. The ionic liquid solvents showed enhanced degradation and a different set of degradation products than those observed for MEA. In one case, the anion of the ionic liquid was found to be reactive and led to a unique primary degradation mechanism not seen in other solvents. The polar and charged nature of the ionic liquid solvents is also believed to enhance the nucleophilic substitution reaction rates that lead to amine oligomerization degradation products in these types of systems.

■

INTRODUCTION

A number of alternative systems have been proposed that reduce or eliminate water from the solvent in attempts to improve the energy efficiency of CO2 capture.7−9 Many of these efforts have focused on the use of very low-volatility ionic liquids (ILs), allowing significant reductions in total solvent regeneration energy.,10,11 However, most ILs are inefficient under postcombustion CO2 capture conditions as they lack sufficient capacity to absorb CO2 under low partial pressures.12 Including reactive sites covalently bound to ILs has thus been an area of focus to improve on this drawback inherent to early IL solvents.13,14 However, a more straightforward approach may be simply to mix ILs with amines. Camper and Bara first proposed the use of IL-amine mixtures for CO2 capture as an alternative to aqueous-based systems.15,16 ION Engineering has developed solvents that utilize an IL as the primary physical solvent, rather than water. This has resulted in lower solvent vapor pressure and an increase in process efficiency by reducing the heat of vaporization in the regeneration step. Degradation of solvents in CO2 capture processes has been studied for traditional aqueous amine systems and is a major process engineering, economic, and environmental consideration.17,18 Solvent degradation tends to decrease CO2 carrying capacity, cause precipitate formation, and increase solvent foaming and viscosity.19 All of these phenomena lead to decreased solvent performance. Emission of degradation impurities, such as nitrosamines, may also be of major

Increased atmospheric CO2 levels have been linked to climate change, and there has been significant interest in drastically reducing emissions of CO2 and other greenhouse gases (GHGs) in recent years.1 Capturing CO2 from point emissions sources, such as coal-fired power plants, is one strategy where a great deal of both applied and basic research is being conducted in both industry and academia.2 A major obstacle in capturing CO2 from power plant flue gas is isolating this gas from others components such as N2 while incurring a minimal energy penalty. CO2 typically exists in coal-fired power plant flue gas at a partial pressure of 2−3 psia (∼12% by volume).3 A number of methods have been proposed to separate CO2 from flue gas including membranes, solid sorbents, and liquid solvents. Aqueous amines have emerged as the top liquid solvent candidates due to their high level of development in an established process to remove acid gases (CO2, SO2, H2S, etc.) from natural gas.4 The Brønsted-basic and nucleophilic amine forms a covalent bond with CO2 via an exothermic substitution reaction, thus removing it from the gas stream. The reaction can be reversed by “stripping” the system at elevated temperature to release captured CO2 and regenerate the amine to capture more CO2 in a continuous process.5 Within the natural gas industry, the most common solvents employed utilize ∼30 wt % of an alkanolamine such as monoethanolamine (MEA) or N-methyldiethanolamine (MDEA).5 A major drawback of these aqueous solvents is that they inflict a high energy penalty during solvent regeneration, due largely in part to vaporization of water.6 © 2012 American Chemical Society

Received: April 30, 2012 Revised: June 27, 2012 Published: August 1, 2012 5345

dx.doi.org/10.1021/ef300731a | Energy Fuels 2012, 26, 5345−5349

Energy & Fuels

Article

environmental concern.20,21 Understanding how solvents degrade can simplify solvent purification and also offer insight into how to prevent or slow degradation by altering the solvent or process. Because the solvents investigated in this study incorporate ILs into the solvent, it is imperative to understand how these affect stability and compare this to other CO2 capture solvents. The goal of this study was to benchmark degradation of ION’s imidazolium-based IL solvents against the industry standard, aqueous MEA, and understand how structural changes to the IL affect degradation. The ILs [Emim][EtSO4] and [Emim][OTf] were chosen as carbon capture solvents in this study for a number of reasons including cost, commercial availability, and specific performance criteria such as viscosity and miscibility with amine and water. Previous studies examining solvent degradation in aqueous amine CO2 capture solvents have used a variety of chromatographic and spectroscopic methods to determine the identity of degradation products. Work done by Strazisar et al. looked at the breakdown of aqueous MEA in a carbon capture facility at a coal-fired power plant and found a number of new products formed that were not observed in laboratory studies.18 Because solvents containing ILs are not yet being used in a process with actual flue gas, this type of study is not possible. However, process conditions were simulated through accelerated aging in this study. This method has been used by others to mimic process conditions in the lab.22 Although this technique is not entirely representative of flue gas conditions, it does allow for comparison between solvents and for an understanding of relative solvent stability.

■

temperature, cylinders containing CO2 were still under high pressure), and a plastic syringe (Norm-Ject from Henke Sass Wolf) with a 12 in. stainless steel needle was used to extract 1 mL of solvent through the septum. The same 12 in. needle was then used to sparge the solvent with the gas used to age the sample. The needle was then removed quickly, the valve was sealed, and the cylinder was placed back in the oven until another sample was acquired. Samples of aqueous MEA were diluted 10-fold with high purity anhydrous methanol (MeOH, Honeywell B&J, 99.99% pure). Samples of solvents A and B containing IL-based solvents were treated differently because the low volatility ionic components may foul GC columns. For these samples, the organics were extracted into a 1:1 mixture of ethyl acetate (EtOAc) and hexanes (both HPLC-UV grade from Pharmco-AAPER) which phase separated from the aged solvent. These extractions were done in a 5 mL plastic syringe by pulling 1 mL of solvent from the aging cylinder, changing to a clean needle, then pulling 2 mL of the 1:1 EtOAc-hexanes mixture and approximately 1 mL of air. The syringe was shaken vigorously and the two phases were allowed to separate, with the denser IL-containing solvent on the bottom layer. The bottom layer was discarded along with 0.5 mL of the EtOAc−hexanes mixture to ensure the entire IL layer was washed away. All samples were filtered through Millex 0.45 μm PTFE syringe filters (Millipore) while transferring to GC autosampler vials to remove any particulates as a precaution. GC/MS Analysis and Data Workup. GC/MS analysis was conducted using a Hewlett-Packard 6890 Series GC with a 6890 Series autoinjector coupled to a Hewlett-Packard 5973 mass selective detector. The carrier gas was high purity helium (99.999% pure), and a RESTEK Rtx-5 Amine column (30 m long, 0.32 mm diameter, 1 μm film thickness) was used in this study. Separations were performed with a temperature profile beginning at 80 °C and ramping 15 °C/min to 220 °C with a helium flow rate of 1.5 mL/min. Data analysis was conducted using Chem Station (Version D.00.01.27) and AMDIS (Version 2.66) software packages to analyze chromatograms. The 50 largest area peaks (excluding carrier gas and extraction/dilution solvents) were identified using the NIST MS Search program (Version 2.0f). NMR Experimental Details. NMR experiments were conducted on a Bruker 500 MHz instrument in DMSO-d6 (some spectra contain TMS as an internal standard). All initial mixtures contained a 60:40 (vol:vol) mixture of [EMIM][EtSO4] and NMEA . Samples for NMR were taken from the bulk solvent (∼50 mL) after exposing the mixture to heat (80−120 °C) and adding enough CO2 (on a theoretical 1:2 basis) to saturate all amine molecules. The accelerated aging experiments were carried out in stainless steel chambers similar to the method described above.

EXPERIMENTAL SECTION

Reagents. Reagents for this study were purchased from commercial suppliers and used without further purification. Solvents were analyzed by GC/MS using the standard protocol outlined below prior to this study to determine baseline impurities so they would not be mistaken for degradation products. Distilled water was used for all experiments. Reagents were procured from the following suppliers: monoethanolamine (MEA, 99% pure) and N-methylethanolamine (NMEA, 99% pure) from Univar USA; 1-ethyl-3-methylimidazolium ethylsulfate ([EMIM][EtSO4]) from Evonik Industries; and 1-ethyl-3methylimidazolium trifluoromethanesulfonate ([EMIM][OTf], 98% pure) from IoLiTec USA. Industrial grade N2 (99.00% pure), CO2 (98.00% pure), and 1000 ppm SO2 in N2 (confirmed by gas analysis performed by vendor) cylinders were all obtained from General Air. Air was obtained from the atmosphere using an aquarium pump. Solvent Aging. Solvents for aging experiments were mixed in the following proportions by weight to prepare 60 g total for each. Aqueous MEA: 30% MEA, 70% water. Solvent A: 33% NMEA, 57% [EMIM][EtSO4], 10% water. Solvent B: 33% NMEA, 57% [EMIM][OTf], 10% water. The solvent was then sparged with N2 and stirred for 15 min to desorb any dissolved gases. Solvent (20 mL) was added to a 50 mL cylinder (Swagelok SS-4CS-TW-50) fitted with a ball valve (Swagelok SS-43G86) and a high-temperature silicone septum. All Swagelok vessels and fittings are made of 316 stainless steel. The gas of interest (N2, air, CO2, or SO2) was bubbled through the solvent for 5 min via 12 in. needle while venting through the septum with a disposable, single-use needle. The cylinder was then sealed and placed in a laboratory oven set to 120 °C (Boekel Scientific) and allowed to age for a period of time. Tables outlining sampling time points and degradation products observed in this study can be found in the Supporting Information. Sampling and Dilution/Extraction. Samples of aged solvent were acquired at predetermined intervals to monitor degradation. The cylinders were removed from the oven and allowed to cool to room temperature for approximately one hour. The valve was opened carefully in a ventilated fume hood (CAUTION: even at room

■

RESULTS AND DISCUSSION Aqueous MEA. Aging experiments were conducted with 30% aqueous MEA at 120 °C in sealed cylinders under the studied gases. Degradation was negligible under N2 after four weeks and was slightly more prevalent in the presence of air. As with all other solvents in this study, degradation of MEA was observed to increase with CO2 relative to the other gases. Samples aged in the presence of CO2 contained hydroxylamine and ammonia after one day in the oven at 120 °C. No further degradation was observed with these samples until after three weeks, at which point oxazolidinone (1) and amine dimers (2 and 3) were detected (Scheme 1). The carbamate oligomerizaScheme 1

5346

dx.doi.org/10.1021/ef300731a | Energy Fuels 2012, 26, 5345−5349

Energy & Fuels

Article

Additional characterization of amine alkylation by the IL [EMIM][EtSO4] was performed via 1H NMR spectroscopy. Spectra of [EMIM][EtSO4] and NMEA were acquired for virgin and aged samples and can be found in the Supporting Information. The chemical shifts of the methylene protons on NMEA change after aging the solvent, as indicated by the disappearance of the signals at 2.55 and 3.32 ppm in the spectrum. This is attributed to alkylation of the amine by the ethylsulfate anion, in addition to formation of amine carbamates. The chemical shifts of the protons on the imidazolium cation of the IL also change from the virgin to the aged sample. The proton at the C(2)-position of the cation has shifted from 9.13 to 9.26 after aging. Smaller changes are also observed for the protons at the 4- and 5-positions. This is likely due to chemical changes to the anion from ethylsulfate to hydrogen sulfate after alkylation of the amine (Scheme 3).

tion pathway outlined in Scheme 1 is common for these types of alkanolamine systems.22 When the amine is exposed to CO2 it forms the semistable oxazolidinone 1 which is electrophilic and can react with the nucleophilic amine via a SN2 reaction mechanism. Depending on which component of the amine acts as the nucleophile in this reaction, different degradation products (i.e., 2 or 3) are possible. The amine oligomers will continue to react with 1 to build larger oligomers and polymers. However, detection of these higher molecular weight compounds was limited in this study because GC/MS can only detect volatile compounds and such polymers have high boiling points. This is one limitation of using GC/MS to study degradation because only volatile products can be detected. A more comprehensive method would employ liquid chromatography and ion chromatography to detect nonvolatile organic and inorganic species, respectively. The carbamate oligomerization amine degradation pathway observed for aqueous MEA in this study has also been reported by other authors studying degradation of alkanolamines.18,22 Solvent A. Solvent A is a mixture of water, NMEA, and [EMIM][EtSO4] and was aged under the same conditions as aqueous MEA. From previous, unpublished work, this solvent was known to degrade faster than other solvents containing ILs, so samples were taken more frequently, and the duration of the experiment was shorter than for MEA or solvent B. The solvent underwent similar carbamate oligomerization of NMEA to that observed for MEA (Scheme 2); however, this was not the

Scheme 3

Changes to the composition of the anion are known to affect the chemical shift of the imidazolium protons in the 1H NMR spectrum.25 These NMR spectra further support alkylation of the amine by the ethylsulfate anion as the primary degradation mechanism of solvent A. Solvent B. Degradation was most prominent in solvent B which contains the IL [EMIM][OTf]. The primary mechanism was amine-carbamate oligomerization, which led to a greater variety of degradation products than those observed in aqueous MEA or solvent A. Not only were more degradation products detected in this solvent, but they also formed earlier than in the other solvents, with many appearing within one week in the oven and several after only 24 h. As with the other solvents, degradation was more rapid and produced a wider array of compounds under CO2 compared to air or N2. The initial degradation steps for solvent B were identical to those shown in Scheme 2 for solvent A; NMEA reacts with CO2 to form oxazolidinone 4, which then reacts with NMEA to form dimers 5 and 6. However, unlike with solvent A, dimers 5 and 6 continued to react in solvent B forming multiple piperazine derivatives (Scheme 4). For example, dimer 6 can eliminate

Scheme 2

primary method of solvent degradation. The EtSO4 anion is electrophilic, and nucleophilic amine alkylation was the most significant form of solvent degradation. Like MEA, NMEA is capable of forming a reactive oxazolidinone (4, Scheme 2) when exposed to CO2 . Compound 4 was detected after one day of aging in the presence of CO2. After one week, amine dimers (5 and 6, Scheme 2) began forming due to NMEA reacting with 4. These degradation products were not detected for samples aged under N2 or air, indicating CO2 is required to form 4. As mentioned previously, formation of amine-carbamate oligomers was a secondary degradation mechanism of solvent A, which only occurs in the presence of CO2. The primary degradation pathway of solvent A is alkylation of the amine by the [EMIM][EtSO4] anion. After one week of aging, all samples contained alkylated NMEA compounds. The ethylsulfate anion is a strong electrophile and is known to act as an alkylating agent in aqueous environments.23 Therefore, it is not surprising that this behavior is observed in this solvent environment with elevated temperatures. This degradation path may even be expected since ILs are known to enhance the rate of these types of nucleophilic substitution reactions.24 The presence of N-ethyl-N-methylethanolamine (7) in the aged samples was a strong indicator that this alkylation reaction between NMEA and the ethylsulfate anion was occurring as a primary degradation mechanism independent of the CO2promoted carbamate oligomerization observed for MEA.

Scheme 4

water to form dimethylpiperazine (8). Piperazine derivatives 9 and 10 (Scheme 4) were also detected in these samples and are formed by more complex oligomerization and cyclization reactions. The rapid formation and greater abundance of carbamate oligomerization products in solvent B can likely be attributed to 5347

dx.doi.org/10.1021/ef300731a | Energy Fuels 2012, 26, 5345−5349

Energy & Fuels

Article

as salts and would not be volatile or would partition into the IL layer during extraction, and thus be undetectable by GC/MS. It is also possible that the amines formed heat stable salts in the presence of SO2 that were nonvolatile. Aside from the issue of detecting the SO2 degradation products by GC/MS, the concentration of the gas is also important to consider. The samples were exposed to pure CO2 gas in the aging cylinders, while the SO2 was 1000 ppm in N2. The reduced partial pressure of SO2 will result in a lower concentration of this gas in the solvent than for CO2, which will have a dramatic impact on the reaction rates. Assuming the solubility of CO2 and SO2 is similar in the solvent mixture, there would be 1,000 times as much CO2 in solution as SO2, which will likely have a marked effect on rates of degradation. Conclusions cannot be drawn from this study about how these solvents behave at a high concentration of SO2. However, the concentration of SO2 in coal-fired flue gas will be orders of magnitude lower than that of CO2, and the results from this study offer valuable information on SO2-driven amine degradation under simulated flue gas conditions. Obviously, CO2 increases degradation when compared to relatively inert gases, such as air and N2. This must be considered when designing processes and solvents for CO2 capture.

the IL component of the solvent. The primary mechanism of degradation in solvent A was amine alkylation by the ethylsulfate anion, which obscured the secondary degradation mechanism of amine carbamate polymerization. Because this process formed a tertiary amine (7) which cannot oligomerize in the same manner as NMEA or MEA (secondary and primary amines, respectively), these products were less abundant in solvent A. The triflate anion of the IL in solvent B is not electrophilic and cannot undergo an amine alkylation degradation mechanism as observed in solvent A (i.e., no “shift” of the −CF3 group is possible). Thus NMEA is available in solution and readily undergoes carbamate oligomerization and cyclization reactions to produce piperazine derivatives. As mentioned previously, ILs are known to increase the rate of substitution reactions such as these, which explains why more of these products are seen in solvent B than with aqueous MEA.24 Another aspect to consider for solvent B is purity of the IL used in the study and its degradation. Traces of imidazoles (1methyl- and 1-ethylimidazole) were present in the [EMIM][OTf] used in this study and also in all of the degradation samples analyzed. These compounds are themselves reactive, so it is possible they play a role in degradation reaction pathways or react to form degradation products. Because this study was not quantitative, it is not possible to say whether the concentration of imidazoles increased during the study. Imidazoles are also known to form as degradation products of NMEA in aqueous systems. Lepaumier et al. report the presence of multiple imidazole degradation products in aqueous NMEA, which they attribute to reaction of the amine with formic acid.22 Decoupling the formation of imidazoles as NMEA degradation products from their presence as impurities in the IL is a difficult task and was beyond the scope of this study, but it could be done through isotopic labeling. SO2 Degradation. Samples of all three solvents were aged under SO2, and GC/MS was used to monitor formation of degradation products. In the case of all three solvents, degradation was slower and fewer degradation products were observed in the presence of SO2 than with CO2. SO2 is much more toxic to humans than CO2, is a strong oxidizer, and is generally thought of as a problematic impurity in flue gas and in the atmosphere. Given the highly reactive/acidic nature of SO2, we anticipated much more degradation compared to CO2 and conducted the SO2 experiment for a shorter amount of time and took samples more frequently than for solvent B or aqueous MEA. In the case of solvent A, amine alkylation was still the primary degradation mechanism with significant amounts of alkylated amines appearing in less than a week. For aqueous MEA and solvent B, amine oligomerization was the primary degradation pathway and identical products to those seen with CO2 were observed in the presence of SO2. In the case of all solvents studied here, none of the degradation products observed by GC/MS contained any sulfur atoms. It was not expected that samples exposed to SO2 would degrade slower and produce fewer degradation products than those exposed to CO2. Because SO2 is an acidic (in the presence of water) and oxidizing gas, it was expected to enhance amine degradation and lead to a different set of degradation products. No sulfur-containing products were observed by GC/MS, nor were oxidation products, such as carboxylic acids. However, this does not mean these types of compounds did not form. Any carboxylic acids formed by degradation would be neutralized by the Brønsted-basic amine

■

CONCLUSIONS The degradation of carbon capture solvents containing ILs was studied and compared to that of the industry standard, aqueous MEA. In all cases, degradation was faster in the presence of CO2 than other studied gases (air, N2, or SO2). This was attributed to the formation of an electrophilic oxazolidinone (1 or 4) that forms in the presence of CO2 and reacts with the amine via a nucleophilic substitution reaction leading to the formation of amine oligomers. This degradation pathway was enhanced in solvent B, which contained the IL [EMIM][OTf], and was attributed to the ionic character of the solvent and its well-known ability to accelerate these types of reactions. In the case of solvent A, which contains the electrophilic ethylsulfate anion, the primary mechanism of solvent degradation was amine alkylation by the anion. This study was qualitative and aimed to examine relative rates of degradation and identify degradation products formed in the presence of ILs. This study provides no information on the absolute amount of degradation. Future work on solvent degradation in CO2 capture solvents will focus on quantitative analysis of amine and solvent degradation using GC/MS and other methods, such as NMR spectroscopy. The use of methods that can detect nonvolatile degradation products, such as salts and high molecular weight amine oligomers, will also be investigated.

■

ASSOCIATED CONTENT

S Supporting Information *

Tables listing degradation products, NMR spectra, and a figure of the accelerated aging apparatus. This material is available free of charge via the Internet at http://pubs.acs.org.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Phone: (303) 997-7097. Notes

The authors declare no competing financial interest. 5348

dx.doi.org/10.1021/ef300731a | Energy Fuels 2012, 26, 5345−5349

Energy & Fuels

■

Article

ACKNOWLEDGMENTS Funding from the United States Department of Energy − National Energy Technology Laboratory (DE-FE00005799) is gratefully acknowledged.

■

REFERENCES

(1) Metz, B., Davidson, O. R., Bosch, P. R., Dave, R., Meyer, L. A., Eds. In Climate Change; Cambridge University Press: New York, 2007. (2) Bhown, A. S.; Freeman, B. C. Environ. Sci. Technol. 2011, 45, 8624−8632. (3) Astarita, G.; Savage, D. W.; Bisio, A. Gas Treating with Chemical Solvents; John Wiley & Sons: New York, 1983. (4) Mangalapally, H. P.; Hasse, H. Chem. Eng. Sci. 2011, 66, 5512− 5522. (5) Rochelle, G. T. Science 2009, 325, 1652−1654. (6) Van Straelen, J.; Geuzebroek, F. Energy Procedia 2011, 4, 1500− 1507. (7) Heldebrant, D. J.; Koech, P. K.; Rainbolt, J. E.; Zheng, F. Energy Procedia 2011, 4, 216−223. (8) McDonald, T. M.; Lee, W. R.; Mason, J. A.; Wiers, B. M.; Hong, C. S.; Long, J. R. J. Am. Chem. Soc. 2012, 134, 7056−7065. (9) Goeppert, A.; Czaun, M.; May, R. B.; Prakash, G. K. S.; Olah, G. A.; Narayanan, S. R. J. Am. Chem. Soc. 2011, 133, 20164−20167. (10) Bara, J. E.; Carlisle, T. K.; Gabriel, C. J.; Camper, D.; Finotello, A.; Gin, D. L.; Noble, R. D. Ind. Eng. Chem. Res. 2009, 48, 2739−2751. (11) Heldebrant, D. J.; Koech, P. K.; Ang, T.; Liang, C.; Rainbolt, J. E.; Yonker, C. R.; Jessop, P. G. Green Chem. 2010, 12, 713−721. (12) Cadena, C.; Anthony, J. L.; Shah, J. K.; Morrow, T. I.; Brennecke, J. F.; Maginn, E. J. J. Am. Chem. Soc. 2004, 126, 5300− 5308. (13) Shannon, M. S.; Bara, J. E. Sep. Sci. Technol. 2012, 47, 178−188. (14) Bates, E. D.; Mayton, R. D.; Ntai, I.; Davis, J. H. J. Am. Chem. Soc. 2002, 124, 926−927. (15) Finotello, A.; Bara, J. E.; Narayan, S.; Camper, D.; Noble, R. D. J. Phys. Chem. B 2008, 112, 2335−2339. (16) Camper, D.; Bara, J.; Koval, C.; Noble, R. Ind. Eng. Chem. Res. 2006, 45, 6279−6283. (17) Goff, G. S.; Rochelle, G. T. Ind. Eng. Chem. Res. 2004, 43, 6400− 6408. (18) Strazisar, B. R.; Anderson, R. R.; White, C. M. Energy Fuels 2003, 17, 1034−1039. (19) Kohl, A. L.; Nielsen, R. Gas Purification, 5th ed.; Gulf Professional Publishing: Houston, TX, 1997. (20) Nguyen, T.; Hilliard, M.; Rochelle, G. T. Int. J. Greenhouse Gas Contol 2010, 4, 707−715. (21) Karl, M.; Wright, R. F.; Berglen, T. F.; Denby, B. Int. J. Greenhouse Gas Control 2011, 5, 439−447. (22) Lepaumier, H.; Grimstvedt, A.; Vernstad, K.; Zahlsen, K.; Svendsen, H. F. Chem. Eng. Sci. 2011, 66, 3491−3498. (23) Wolfenden, R.; Yuan, Y. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 83−86. (24) Kim, D. W.; Song, C. E.; Chi, D. Y. J. Org. Chem. 2003, 68, 4281−4285. (25) LaFrate, A. L.; Gin, D. L.; Noble, R. D. Ind. Eng. Chem. Res. 2010, 49, 11914−11919.

5349

dx.doi.org/10.1021/ef300731a | Energy Fuels 2012, 26, 5345−5349