(733d) Revealing the Mechanism of Intriguing Predominant Urea Formation from Thermal Degradation of CO2-Loaded Aqueous Ethylenediamine

Amine scrubbing using aqueous amine solutions remains to be the most efficient near-term solution for post-combustion CO₂ capture from flue gas at coal-fired power plants.^1,2 Monoethanolamine (MEA) is one of the most extensively tested alkanolamines, but its commercial application tends to be limited because of degradation and corrosion issues and enormous energy requirement for solvent regeneration.^3–6 About (~10 %) of total capital cost of CO₂ capture is attributed to makeup of amine solvent loss.^7–10 Moreover, some degradation products are considered to be dangerous if exposed to the environment.

Ethylenediamine (EDA) has been reported to have several advantages over MEA for its higher CO₂ capacity, faster CO₂ absorption rate, easier producible implementation from renewable resources, and slightly lower degradation rates at moderate CO₂ loading.^11–16 EDA and MEA have a similar molecular structure as a primary amine; the former contains two primary amine groups and the latter has one primary amine and one hydroxyl group, while the end groups are separated by two -CH₂- groups in both molecules. Despite their structural similarity, the major thermal degradation products of CO₂-loaded EDA and MEA solutions have been found to be distinctly different.^11,17

Thermal degradation of amines is generally understood to occur primarily in the stripper due to high temperature (~ 100 - 120 ^oC) and is known to increase with CO₂ loading.¹⁸ Previous experimental studies proposed that thermal degradation of primary and secondary amines would be initiated by polymerization of carbamates, which can ultimately degrade to urea and imidazolidinones.^19,20 The MEA degradation is thought to proceed via 2-oxazolidinone (OZD) formation that may undergo further reaction with amines to form more stable products.²¹ Our previous study predicted isocyanate formation from dehydration of carbamic acid to be highly probable, and also showed that cyclization of intermediate isocyanate to OZD would be kinetically more facile than reaction with MEA to form urea in aqueous MEA solution. The formation of isocyanate, albeit very small quantity, has also been experimentally reported.^22,23 In contrast to the case of MEA, the major thermal degradation products of EDA have been identified to be 1,3-Bis(2-aminoethyl)urea and 2-imidazolidone from mass spectrometry and high performance liquid chromatography measurements, while urea formation appears to be preferred.^11,17

In this talk, we will present molecular mechanisms underlying thermal degradation of EDA in CO₂-loaded solution, especially the underlying reason for the preferential formation of urea, rather than cyclic compounds, in contrast to the case of MEA. The thermodynamic favorability of previously proposed reaction pathways, including the relative stabilities of intermediates and products, were evaluated using quantum mechanical calculations with an implicit solvent model. The free-energy barriers for key elementary reactions involved in the EDA degradation were determined using ab initio molecular dynamics simulations combined with metadynamics. Besides the simulation results, further analysis of the interactions between short-lived intermediates and surrounding solvent molecules provide an explanation as to why EDA and MEA show different degradation behaviors. This work highlights the importance of kinetic effects associated with the local structure and dynamics of solvent molecules around the intermediates that may significantly influence the degradation process of amine solvents.

References

[1] G. T. Rochelle, Science, 2009, 325, 1652–4.

[2] P. Luis, Desalination, 2016, 380, 93–99.

[3] M. R. M. Abu-Zahra, L. H. J. Schneiders, J. P. M. Niederer, P. H. M. Feron and G. F. Versteeg, Int. J. Greenh. Gas Control, 2007, 1, 37–46.

[4] M. R. M. Abu-Zahra, J. P. M. Niederer, P. H. M. Feron and G. F. Versteeg, Int. J. Greenh. Gas Control, 2007, 1, 135–142.

[5] D. M. D’Alessandro, B. Smit and J. R. Long, Angew. Chem. Int. Ed. Engl., 2010, 49, 6058–82.

[6] H. M. Stowe and G. S. Hwang, Ind. Eng. Chem. Res., 2017, 56, 6887–6899.

[7] A. B. Rao and E. S. Rubin, Environ. Sci. Technol., 2002, 36, 4467–4475.

[8] A. J. Reynolds, T. V. Verheyen, S. B. Adeloju, A. L. Chaffee and E. Meuleman, Int. J. Greenh. Gas Control, 2015.

[9] G. S. Goff and G. T. Rochelle, Ind. Eng. Chem. Res., 2006, 45, 2513–2521.

[10] S. A. Mazari, B. Si Ali, B. M. Jan, I. M. Saeed and S. Nizamuddin, Int. J. Greenh. Gas Control, 2015.

[11] S. Zhou, X. Chen, T. Nguyen, A. K. Voice and G. T. Rochelle, ChemSusChem, 2010, 3, 913–918.

[12] M. Rabensteiner, G. Kinger, M. Koller, G. Gronald and C. Hochenauer, Int. J. Greenh. Gas Control, 2014, 27, 1–14.

[13] S. Kadiwala, A. V. Rayer and A. Henni, Chem. Eng. J., 2012, 179, 262–271.

[14] J. Li, A. Henni and P. Tontiwachwuthikul, Ind. Eng. Chem. Res., 2007, 46, 4426–4434.

[15] Z. Wang, M. Fang, Y. Pan, S. Yan and Z. Luo, Chem. Eng. Sci., 2013.

[16] A. P. Salvi, P. D. Vaidya and E. Y. Kenig, Can. J. Chem. Eng., 2014, 92, 2021–2028.

[17] J. Thompson, H. Richburg and K. Liu, in Energy Procedia, 2017.

[18] D. Hatchell, O. Namjoshi, K. Fischer and G. T. Rochelle, Energy Procedia, 2014, 63, 1558–1568.

[19] H. Lepaumier, D. Picq and P. L. Carrette, in Energy Procedia, 2009.

[20] H. Lepaumier, E. F. Da Silva, A. Einbu, A. Grimstvedt, J. N. Knudsen, K. Zahlsen and H. F. Svendsen, in Energy Procedia, 2011.

[21] J. Davis and G. Rochelle, Energy Procedia, 2009, 1, 327–333.

[22] T. Sakakura, J. C. Choi and H. Yasuda, Chem. Rev., 2007, 107, 2365–2387.

[23] T. E. Waldman and W. D. McGhee, J. Chem. Soc. Chem. Commun., 1994, 957.