Probing the effect of electron donation on CO2 absorbing 1,2,3-triazolide ionic liquids
Robert L. Thompson, Wei Shi, Erik Albenze, Victor A. Kusuma, David Hopkinson, Krishnan Damodaran, Anita S. Lee, John R. Kitchin, David R. Luebke, and Hunaid Nulwala, “Probing the effect of electron donation on CO2 absorbing 1,2,3-triazolide ionic liquids”, RSC Advances 4(25), 12748-12755 (2014). https://doi.org/10.1039/c3ra47097k
This summary was written by a language model from the paper and has not been checked line by line. Rely on the published paper for details.
Note on the source text: the extracted text is readable but two-column interleaved, and minus signs appear to have been lost from the computed interaction energies, so those are reported here as magnitudes. Figures are not available, so results are taken from the prose and tables.
Question
CO2-reactive ionic liquids (ILs) are candidate solvents for capturing CO2 from existing power plants, but ILs whose anions carry hydrogen-bonding protons tend to become very viscous on reaction. Aprotic heterocyclic anions (AHAs) avoid this, and 1,2,3-triazoles are attractive because their pKa (13.9 for the parent) suits carbon capture and substituted triazoles are easy to make by copper-catalyzed click chemistry. The paper asks how the steric bulk and electron-donating or electron-withdrawing character of the substituent at the 4-position of a triazolide anion control CO2 uptake, viscosity, thermal stability and interaction with water (Thompson et al. 2014).
Approach
Five tetrabutylphosphonium 4-substituted 1,2,3-triazolide ILs were prepared by neutralizing the triazoles with aqueous tetrabutylphosphonium hydroxide: phenyl (1, electron withdrawing), hexyl (2) and tert-butyl (3) (electron donating), a trimethylsilylmethyl group (4) and a fluorinated ether (5), the last two included as hydrophobic substituents. The ILs were characterized by density, Karl Fischer water content, TGA decomposition onset, viscosity (wet and dry, with and without CO2), 1H and 13C NMR (including 13CO2 exposure), LC-MS, and variable-pressure transmission FTIR. CO2 isotherms from 0 to 10 bar at 30 C were measured in a stirred Sievert’s apparatus. Gas-phase B3LYP/6-311++G(d,p) calculations in Gaussian 09 gave CHELPG atomic charges on the ring nitrogens and interaction energies of each anion with CO2 and H2O at the N1, N2 and N3 sites (Thompson et al. 2014).
Key findings
CO2 capacity at 30 C and about 1 bar spanned 0.07 to 0.40 mol CO2 per mol IL: 0.40 for hexyl 2, 0.30 for 4, 0.23 for 5, 0.10 for phenyl 1 and 0.07 for tert-butyl 3, compared with 0.032 for physisorption in [hmim][Tf2N]. Using phosphine cone angles as a steric measure (132 degrees for linear chains, 180 to 182 for the branched groups, 145 for phenyl), the authors argue that both sterics and electronics matter, with electronics dominant since bulky but electron-donating 4 outperformed phenyl 1. Electron-donating groups shifted the ring H5 proton upfield (7.12 to 7.23 ppm versus 7.36 to 7.59 ppm) and lowered the decomposition onset (151 to 227 C for 2, 3 and 4 versus 238 and 263 C for 1 and 5) (Thompson et al. 2014).
13C NMR after 13CO2 exposure showed two or three new peaks between 153 and 159 ppm, assigned to several carbamate isomers, and no free CO2 near 125 ppm, indicating chemisorption only. FTIR of 2 showed new bands at 1754, 1628 and 1280 cm-1 after 10 mbar CO2 that vanished on evacuation, consistent with reversible binding. Vacuum-dried ILs still held 0.9 to 2.4 wt% water (about 20 to 54 mol%). These wet ILs were viscous (1357 to 8275 cP) and unchanged by CO2, whereas the truly dry ILs were much less viscous (202 to 880 cP) but became markedly more viscous after CO2, and 2 and 5 even solidified. Water at up to 54 mol% did not change CO2 capacity of 2 or 3, which the authors explain with a parallel bicarbonate pathway. The ILs were hygroscopic but insoluble in water; 4 decomposed after two days in water (Thompson et al. 2014).
The calculations placed the largest negative charge at N1 for most anions (N3 for 2 and 5), but N1 was always the strongest binding site. For the unsubstituted anion the interaction energy magnitudes were 52 kJ/mol for CO2 and 67 kJ/mol for H2O; for phenyl 1 they were 36 and 61 kJ/mol, and for hexyl 2 51 and 66 kJ/mol. Electron donation strengthened both interactions, but more for CO2 than for water (Thompson et al. 2014).
Significance
The study shows that substituent choice on an aprotic triazolide anion tunes CO2 reactivity over a factor of about six, and gives a simple design rule: electron-donating, unbranched substituents favor uptake. It also warns that residual water strongly changes viscosity and physical state after CO2 absorption, so property comparisons between ILs require careful control of water content, even though water does not reduce capacity (Thompson et al. 2014).