This journal is© the Owner Societies 2017 Phys. Chem. Chem. Phys., 2017, 19, 15403--15411 | 15403

Cite this:Phys.Chem.Chem.Phys.,

2017, 19, 15403

New insights into the chemistry of ionicalkylorganic carbonates: a computational study†

Khaleel I. Assaf, *a Abdussalam K. Qaroush *b and Ala’a F. Eftaiha *c

A library of hydrogenated, perfluorinated aliphatic and aromatic (p-substituted) alcohols are selected

together with a combination of superbases (SBs) and metal hydrides (MHs) to understand the

thermodynamic parameters of the binary mixtures once serving as sorbents for the capture of CO2 via

ionic organic alkyl-carbonate (RCO3�) formation. Data are obtained using density functional theory (DFT)

calculations with the B3LYP/6-31+G* level of theory and compared with the experimental results

acquired from the literature using different spectroscopic techniques. It is found that the capturing

process has a favourable enthalpic contribution and an unfavourable entropic penalty regardless the

identity of the base, where the enthalpy values of alcohol/MH binary mixtures are almost two-fold

higher compared to their SB-based mixtures. The utilisation of perfluorinated aliphatic alcohols instead

of hydrogenated alcohols shows a negative impact on the formation of carbonate adducts, due to the

less reactive alkoxide anion along the carbon skeleton, which is attributed to the low charge density of

the nucleophilic oxygen atom. While perfluorinated phenol shows a higher reactivity than the parent

phenol. The calculations indicate that the reactivity of phenolic compounds is highly affected by the

electronic nature of the substituting groups, in which p-substituted phenols are more reactive towards

CO2 capturing when electron releasing groups are utilised. A pronounced solvent effect is observed,

in which the alkylcarbonate salts (RCO3� SBH+) are stabilized in solvents with high dielectric constant

(e.g., DMSO and MeCN). Simulated NMR and IR spectra of RCO3� are consistent with those reported for

the affiliated systems, which fortifies the results obtained for the unexplored substrate/MH mixtures,

filling a gap in the literature of CO2 sequestration using CO2 binding organic liquids (CO2BOLs) and

enabling a fair/quick prediction of potential substrates to be used as CO2 sorbents.

Introduction

Ionic alkyl-organic carbonates are a family of CO2 sequesteredzwitterionic compounds that were synthesized primarily byJessop and coworkers after their breakthrough work that waspublished in 2005.1 Its production came from the idea ofreacting alcohols with CO2 in the presence of a superbase(SB). Independently, two more research groups made it feasibleto synthesize the title molecules via two different routes, viz.,the 2016 noble prize laureate, Sir. Stoddart and coworkerswho were able to synthesize the molecules using a b-2 andg-cyclodextrin3 modified metal organic framework (MOF) in thesolid state via a dry scrubbing methodology. In contrast, awet-scrubbing based system was developed by Qaroush and

coworkers who were able to synthesize those carbonates inDMSO using a low molecular weight oligosaccharide, viz.,chitin.4,5 In this method, DMSO acts as an activator for thehydroxyl group, while the charged species formed are stabilizedvia a newly-discovered supramolecular chemisorption phenom-enon. Once again, CO2 binding organic liquids (CO2BOLs) wereintroduced by Jessop and co-workers,6 where chemi-/physisorptionmechanisms in dry neat alcohols with a high CO2 absorptioncapacity up to 19 wt% were observed via gravimetric methods,that is considered to be a high value among other similarsystems. The interaction with CO2 is reversible in terms ofcapturing and releasing temperatures. Moreover, they showedthat gas absorption is selective for CO2 in both concentrated anddilute gas streams (for more details, see the review of Y. Parket al.,7 2015). In comparison with currently employed aqueousalkanolamine-based scrubbing systems, CO2BOLs have thepotential to be much more energy efficient for CO2 release.The reason behind such inferiority is caused by the high specificheat of water (4.18 J g�1 deg�1).8 Park and co-workers were ableto demonstrate the thermal stabilities of the alkylcarbonate saltsof SBs in organic solutions in terms of manipulating the

a Department of Life Sciences and Chemistry, Jacobs University Bremen,

Campus Ring 1, 28759 Bremen, Germany. E-mail: k.assaf@jacobs-university.deb Department of Chemistry, Faculty of Science, The University of Jordan,

Amman 11942, Jordan. E-mail: a.qaroush@ju.edu.joc Department of Chemistry, The Hashemite University, P.O. Box 150459,

Zarqa 13115, Jordan. E-mail: alaa.eftaiha@hu.edu.jo

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c7cp02087b

Received 31st March 2017,Accepted 13th May 2017

DOI: 10.1039/c7cp02087b

rsc.li/pccp

PCCP

PAPER

15404 | Phys. Chem. Chem. Phys., 2017, 19, 15403--15411 This journal is© the Owner Societies 2017

composition of hydroxyl-based solvent and polar aprotic solventmixtures.9 In 2016, a new finding was reported by Heldebrantand coworkers, where they showed that CO2BOLs are consideredto be viscous which might hinder CO2 diffusion and render themimpractical for postcombustion capturing (PCC) into the neatsolutions.10 This was overcome using a single molecule CO2BOLthat was developed by their research group.10,11 As a result, fasterkinetic profiles arose from the close proximity of the amine andalcohol and the sites involved in CO2 binding via a concertedmechanism, resulting in a zwitterionic alkylcarbonate and aprotonated amine. The intramolecular hydrogen bonds betweenthe two functional groups determine the overall viscosity of thesolution.10 This observation of the proton shuttling phenomenon(it was discovered via ab initio molecular dynamics simulations)is novel and specific to a hydrogen-bonding network that can becontrolled by chemically tuning single molecule CO2 capturesolvents.10

The interchange between theory and experiment is a power-ful tool to unravel ambiguities for the study of the mechanismof ionic alkyl-organic carbonate upon reacting alcohols withCO2 in the presence of SBs to have the premium sorbents thatcan work ideally for several sorption/desorption cycles. Such asystem does exist if a whole-in-one study that can set a library ofdifferent families is still a must, this is a good reason to startthe process by understanding the reactivity in terms of thermo-dynamics which can be well understood by the utilization ofcomputational methods. Examples for the use of quantum-chemical calculations to investigate the reactivity of amine-based ionic liquids are reported.12,13 Herein, the main objectiveof our study is to help in understanding the overall reactionupon changing the electronic as well as sterical effects towardssubstrates under different types of bases/stabilizers.

Computational methods

All calculations were performed with Gaussian 09.14 Geometryoptimizations and frequency calculations were carried outusing the B3LYP density functional with the 6-31+G* basisset.15,16 Minima were characterized by the absence of imaginaryfrequencies. A polarizable continuum model (PCM)17 was usedfor implicit solvent calculations (see the ESI,† for more details).

Results and discussion

Throughout this work, different bases (1–6, Fig. 1) have beenexplored to stabilize the ionic organic carbonates (RCO3

�)utilizing various alcohols (7–17, Fig. 1) that are used to effec-tively capture CO2. Those bases include metal hydrides (MHs;M: Li (1), Na (2), and K (3)) and organic superbases (SBs),namely, diazabicyclo[5.4.0]-undec-7-ene (DBU, 4), 1,1,3,3 tetra-methylguanidine (TMG, 5), and 2-(tert-butyl)-1,1,3,3-tetra-methylguanidine or more commonly known as Barton’s base(Bb, 6). The hydroxyl-based substrates comprised of hydroge-nated (acyclic/cyclic) and perfluorinated alcohols (PFAs), as wellas p-substituted phenols.

Generally, the reaction of alcohol with CO2 in the presenceof a SB results in the formation of an alkylcarbonate anion thatis stabilized by the cationic conjugate base (Scheme 1A).Instead, if MH is used, the reaction produces metal stabilizedcarbonate anions in addition to the evolution of hydrogen gas(Scheme 1B).

Table 1 shows the calculated pKa values of the studiedalcohols, their corresponding carbonic acids, and the suggestedbases in acetonitrile (MeCN) to mimic the reported experimentalconditions, with an ultimate goal to understand the ability of thealcohol/base binary mixtures to chemisorb CO2. Furthermore,gas-phase proton affinity (PA) for each substrate is calculated(Table S1, ESI†). Low pKa value reflects the alcohol ability to losea proton to form an alkoxide anion (RO�), which acts as anucleophile/base that attacks the electrophilic carbon of CO2 toform the ionic alkyl carbonate anion (RCO3

�). As shown inTable 1, aliphatic hydrogenated alcohols (7–11) have almostidentical pKa values in MeCN with a small variation (an averagevalue of 39.99 � 0.36), as expected from their similar electronicstructure.6 Phenols (12) show lower pKa values than their aliphatic

Fig. 1 Chemical structures of the investigated alcohols and bases: lithiumhydride (1), sodium hydride (2), potassium hydride (3), diazabicyclo[5.4.0]-undec-7-ene (DBU, 4), 1,1,3,3 tetramethylguanidine (TMG, 5), Barton’s base(Bb, 6), linear aliphatic hydrogenated alcohols; n r 5, (7); iso-propanol(i-PrOH, 8), tert-butanol (t-BuOH, 9), cyclohexanol (CyHexOH, 10), cyclo-hexylmethanol (CyHexMeOH, 11), p-substituted phenols (X: H, F, Cl, CH3,OCH3, NO2; 12), benzyl alcohol (BzOH, 13), linear perfluorinated alcohols(PFA; n r 5 (14)), perfluorinated tert-butanol (PF-t-BuOH, 15), pentafluoro-phenol (PF-PhOH, 16); perfluorobutylmethanol (PF-BuMeOH, 17).

Scheme 1 The proposed chemical reactions between alcohol and CO2 inthe presence of (A) SB or (B) MH.

Paper PCCP

This journal is© the Owner Societies 2017 Phys. Chem. Chem. Phys., 2017, 19, 15403--15411 | 15405

counterparts, which varied based on the p-substituent, forexample, p-cresol (X = CH3, 12) has a pKa value of 27.77, whileit is 16.01 for p-nitrophenol (X = NO2, 12). Acyclic PFAs (14, 15)exhibit much smaller pKa values than their hydrogenated counter-parts (7–11). Moreover, PF-PhOH (16) show a lower pKa valuecompared to the parent phenol (X = H, 12). pKa of PF-BuMeOH (17)is found to be closer to the hydrogenated alcohols more than tothe perfluorinated ones. It is anticipated that alcohols withsimilar pKa values will have comparable reactivity toward CO2

unless other factors such as substrate’s sterical demands, thedielectric constant of the reaction media and stability of theformed adduct play a crucial role in directing the reactionpathway. Based on the pKa values, the PFAs (14, 15) areexpected to have the highest reactivity towards capturing CO2,followed by phenols (12) and hydrogenated alcohols (7–11).Moreover, the formation/stabilization of the RCO3

� ions isdetermined by the difference in the pKa values between thecarbonic acid and the conjugate acid of the used base. As such,it is expected that the efficiency to capture CO2 increases whenthe difference between the pKa values of the carbonic acid andthe base increases.10 It is worth mentioning that due to theinstability of these acids, the experimental measurement oftheir pKa values is still a challenge.18 The calculated PA values(see the ESI†) in the gas phase agree with the calculated pKa

values in MeCN. For example, aliphatic hydrogenated alcohols(7–11) showed average PA values of 369 � 2 kcal mol�1,followed by PFAs (14, 15) (310 � 5 kcal mol�1). Furthermore,the pKa and PA values of the organic SBs indicated that Bb (6)acts as the strongest base followed by DBU (4) and TMG (5).

Thermodynamic parameters for CO2 capturing

The thermodynamic parameters for the reaction of CO2 withalcohols in the presence of different bases (as shown inScheme 1) in MeCN and in the gas phase are given in Table 2and Table S2 (ESI†), respectively.23 In general, the capturing

process with all alcohols has a negative (favourable) enthalpiccontribution (DH) and a negative entropic (unfavourable) penaltywhatsoever the base used. In the case of MHs, the enthalpy valuesare almost two-fold higher compared to SBs, which points to thestrong ion–ion interactions between the carbonate anions and themetal cations (Li+, Na+, and K+). In addition, the smaller size ofthe metal cations compared to SBs provides much lower entropicpenalties (ca. two-fold less), which results in more negative freeenergy (DG) values, this is accompanied by the compensationof entropic penalty as a result of hydrogen gas evolution as aby-product. In MeCN, DH values descending in the followingorder: LiH B NaH o KH, where DH for KH is more negative by4 kcal mol�1. In contrast, almost identical entropic contribu-tions are obtained for all investigated MHs (B�11 kcal mol�1).

In the case of SBs, the DG values for most reactions arepositive. The discrepancy between the calculated results andthe experimental data of simple alcohols (e.g., linear aliphatichydrogenated alcohols (7); n = 3–5, i-PrOH (8)) reported in theliterature6 is attributed to our methodology excluding the bulkproperties. However, the calculation predicts very well therelative reactivity trends of the linear hydrogenated alcoholswith respect to the used SB, and descending in the followingorder: Bb (6) 4 DBU (4) d TMG (5), as a result of decreasingthe exothermicity of the reactions. The same order was reportedexperimentally by Jessop and co-workers6 as shown in Fig. 2A.This behaviour is attributed to the relative strength (pKa) of theinvestigated SBs as shown in Table 1, which follows the sameorder. PhOH (X = H, 12) was less reactive compared to thealiphatic hydrogenated alcohols (7–11, Table 2). The DG valuesas a function of the number of carbon atoms for both fullylinear hydrogenated alcohols (7) and PFAs (14) are shown inFig. 2B and C, respectively. Individually, DG values for bothalcohol categories are almost independent of the number ofcarbons in the chain, which is expected since these alcoholshave similar pKa values in MeCN (Table 1, vide supra). Further-more, the carbonate moiety is the only part which is involved inthe interaction with the base, while the organic tail does notcontribute to the interaction and therefore it is expected to havea negligible effect (if any) on the overall interaction energy.6

Less favourable DG values for secondary and tertiary alcoholsare obtained, which is explained by both electronic and stericeffects.6 In contrast to the calculation outcome, i-PrOH is anoutlier among the other alcohols with the three SBs oncecompared to the experimental data (Fig. 2A). Experimentally,t-BuOH was found to be unable to capture CO2 in the presenceof SBs, this fact is clearly indicated by the large unfavourableDG value (Table 2). In this context, KH is expected to be themost promising MH to facilitate CO2 capturing. The goodagreement between the calculated and experimental findingsvalidates the results for the newly suggested alcohol basedcandidates in the presence of MHs instead of their organicSB counterpart.

Fig. 3 shows the optimised geometries for the HexO–CO2�countercation adducts in both gas phase and MeCN. The gasphase was selected as an extremely low dielectric constantmedium (e = 1) compared to MeCN (e = 37.5). The interaction

Table 1 Calculated pKa values for water, selected bases (4–6), alcohols(7–17), and the carbonic acid counterparts formed in MeCN

Substrate

pKa

Substrate

pKa

ROH RCO3H ROH RCO3H

H2Oa 39.84 13.28 12 (X = F) 25.28 9.074 23.99 [24.3]b n.a. 12 (X = Cl) 24.12 [25.44]d 8.825 22.90 [23.3]b n.a. 12 (X = CH3) 27.77 [27.45]d 10.206 26.20 n.a. 12 (X = OCH3) 27.31 10.447 (n = 0) 39.57 15.21 12 (X = NO2) 16.01 [20.7]d 8.807 (n = 1) 39.67 15.39 13 37.40 13.247 (n = 2) 39.78 15.31 14 (n = 0) 5.31 —e

7 (n = 3) 39.98 10.72 14 (n = 1) 5.68 —e

7 (n = 4) 39.89 10.61 14 (n = 2) 6.09 —e

7 (n = 5) 39.98 11.11 14 (n = 3) 5.81 —e

8 40.04 15.57 14 (n = 4) 5.36 —e

9 40.81 17.12 14 (n = 5) 5.33 —e

10 40.31 11.43 15 12.88 —e

11 39.83 14.68 16 15.10 —e

12 (X = H) 26.00 [29.14]c 8.93 17 31.10 10.90

a Water was taken as a model for comparison reasons. b Data obtainedfrom ref. 19 and 20. c Data obtained from ref. 21. d Data obtained fromref. 22. e The carbonate anions were not stable, therefore no pKa valueswere calculated.

PCCP Paper

15406 | Phys. Chem. Chem. Phys., 2017, 19, 15403--15411 This journal is© the Owner Societies 2017

within the formed adducts is dominated by the electrostaticinteraction between the negatively charged carbonate anionand the positively charged countercation (vide supra). It isnoteworthy that the distances between the oxygen atom andthe metal cation are shorter in the gas phase than in MeCN,

where the distances become longer with increasing size of thecation. When DBU is used, it is found to form two short contactpoints with the carbonate anion, in which the first oxygen isinvolved in hydrogen bonding (H-bonding) with the amidiniumproton, and another short contact exists between the second

Table 2 Thermodynamic parameters (kcal mol�1) for the capture of CO2 using different alcohol/base binary mixtures in MeCN (charges are omitted forclarity)

R(Ar)O–CO2�countercation DH TDS DGcapture R(Ar)O–CO2�countercation DH TDS DGcapture

MeO–CO2�Li �47.88 �11.85 �36.03 PhO–CO2�Li �42.40 �12.29 �30.11MeO–CO2�Na �47.88 �11.61 �36.27 PhO–CO2�Na �42.90 �11.18 �31.72MeO–CO2�K �51.53 �11.13 �40.40 PhO–CO2�K �47.50 �12.51 �34.99MeO–CO2�DBUH �20.62 �20.45 �0.17 PhO–CO2�DBUH �15.95 �19.35 3.40MeO–CO2�TMGH �18.05 �19.7 1.65 PhO–CO2�TMGH �13.45 �18.55 5.10MeO–CO2�BbH �21.72 �20.69 �1.03 PhO–CO2�BbH �17.25 �21.24 3.99EtO–CO2�Li �46.90 �11.93 �34.97 BzO–CO2�Li �46.41 �12.54 �33.87EtO–CO2�Na �46.74 �11.75 �34.99 BzO–CO2�Na �46.34 �12.06 �34.28EtO–CO2�K �50.33 �11.66 �38.67 BzO–CO2�K �50.11 �11.66 �38.45EtO–CO2�DBUH �20.12 �20.06 �0.06 BzO–CO2�DBUH �19.01 �19.8 0.79EtO–CO2�TMGH �16.89 �19.48 2.59 BzO–CO2�TMGH �17.04 �22.38 5.34EtO–CO2�BbH �20.49 �21.61 1.12 BzO–CO2�BbH �20.21 �21.29 1.08PrO–CO2�Li �46.97 �11.73 �35.24 PF-MeO–CO2�Li �42.90 �12.92 �29.98PrO–CO2�Na �46.78 �11.87 �34.91 PF-MeO–CO2�Na �44.13 �11.55 �32.58PrO–CO2�K �50.37 �11.38 �38.99 PF-MeO–CO2�K �49.36 �10.83 �38.53PrO–CO2�DBUH �20.02 �20.21 0.19 PF-MeO–CO2�DBUH �18.63 �19.97 1.34PrO–CO2�TMGH �16.86 �19.86 3.00 PF-MeO–CO2�TMGH �16.16 �19.3 3.14PrO–CO2�BbH �20.53 �21.37 0.84 PF-MeO–CO2�BbH �20.72 �20.33 �0.39i-PrO–CO2�Li �43.26 �12.59 �30.67 PF-EtO–CO2�Li �43.88 �11.51 �32.37i-PrO–CO2�Na �42.99 �11.04 �31.95 PF-EtO–CO2�Na �44.10 �11.26 �32.84i-PrO–CO2�K �46.37 �11.29 �35.08 PF-EtO–CO2�K �49.27 �10.95 �38.32i-PrO–CO2�DBUH �15.44 �20.82 5.38 PF-EtO–CO2�DBUH �18.56 �19.85 1.29i-PrO–CO2�TMGH �12.89 �20.05 7.16 PF-EtO–CO2�TMGH �16.14 �18.65 2.51i-PrO–CO2�BbH �16.40 �20.95 4.55 PF-EtO–CO2�BbH �20.70 �19.49 �1.21BuO–CO2�Li �46.93 �11.92 �35.01 PF-PrO–CO2�Li �42.09 �12.84 �29.25BuO–CO2�Na �46.79 �11.7 �35.09 PF-PrO–CO2�Na �44.15 �11.34 �32.81BuO–CO2�K �50.34 �11.25 �39.09 PF-PrO–CO2�K �49.36 �11.31 �38.05BuO–CO2�DBUH �19.99 �20.8 0.81 PF-PrO–CO2�DBUH �18.71 �20.45 1.74BuO–CO2�TMGH �16.81 �20.07 3.26 PF-PrO–CO2�TMGH �16.21 �19.75 3.54BuO–CO2�BbH �20.61 �20.38 �0.23 PF-PrO–CO2�BbH �20.81 �20.72 �0.09t-BuO–CO2�Li �42.77 �12.43 �30.34 PF-BuO–CO2�Li �42.63 �13.94 �28.69t-BuO–CO2�Na �42.38 �12.06 �30.32 PF-BuO–CO2�Na �44.19 �12.11 �32.08t-BuO–CO2�K �45.73 �11.43 �34.30 PF-BuO–CO2�K �49.36 �11.8 �37.56t-BuO–CO2�DBUH �14.82 �20.86 6.04 PF-BuO–CO2�DBUH �19.35 �22.33 2.98t-BuO–CO2�TMGH �7.49 �20.13 12.64 PF-BuO–CO2�TMGH �16.19 �23.22 7.03t-BuO–CO2�BbH �15.74 �21.51 5.77 PF-BuO–CO2�BbH �20.85 �21.07 0.22PentO–CO2�Li �46.94 �11.76 �35.18 PF-t-BuO–CO2�Li �39.74 �12.2 �27.54PentO–CO2�Na �46.79 �11.79 �35.00 PF-t-BuO–CO2�Na �40.87 �10.35 �30.52PentO–CO2�K �50.38 �11.41 �38.97 PF-t-BuO–CO2�K �45.64 �10.31 �35.33PentO–CO2�DBUH �20.00 �20.77 0.77 PF-t-BuO–CO2�DBUH �14.79 �20 5.21PentO–CO2�TMGH �16.88 �19.85 2.97 PF-t-BuO–CO2�TMGH �12.21 �19.15 6.94PentO–CO2�BbH �20.49 �20.42 �0.07 PF-t-BuO–CO2�BbH �16.66 �20.52 3.86HexO–CO2�Li �46.98 �11.8 �35.18 PF-PentO–CO2�Li �42.12 �12.73 �29.39HexO–CO2�Na �46.81 �11.84 �34.97 PF-PentO–CO2�Na �44.23 �11.82 �32.41HexO–CO2�K �50.36 �11.18 �39.18 PF-PentO–CO2�K �49.41 �10.99 �38.42HexO–CO2�DBUH �19.99 �21.01 1.02 PF-PentO–CO2�DBUH �19.29 �22.49 3.20HexO–CO2�TMGH �16.82 �20.1 3.28 PF-PentO–CO2�TMGH �16.18 �19.6 3.42HexO–CO2�BbH �20.52 �20.49 �0.03 PF-PentO–CO2�BbH �20.87 �20.39 �0.48CyHexO–CO2�Li �46.54 �12.03 �34.51 PF-HexO–CO2�Li �42.07 �11.78 �30.29CyHexO–CO2�Na �46.39 �11.78 �34.61 PF-HexO–CO2�Na �44.22 �11.82 �32.40CyHexO–CO2�K �49.79 �11.33 �38.46 PF-HexO–CO2�K �49.38 �10.93 �38.45CyHexO–CO2�DBUH �18.87 �20.4 1.53 PF-HexO–CO2�DBUH �18.60 �19.7 1.10CyHexO–CO2�TMGH �16.27 �19.79 3.52 PF-HexO–CO2�TMGH �16.80 �21.42 4.62CyHexO–CO2�BbH �19.98 �21.64 1.66 PF-HexO–CO2�BbH �20.82 �20.57 �0.25CyHexMeO–CO2�Li �47.50 �12.44 �35.06 PF-PhO–CO2�Li �44.13 �13.12 �31.01CyHexMeO–CO2�Na �47.47 �12.1 �35.37 PF-PhO–CO2�Na �44.91 �11.16 �33.75CyHexMeO–CO2�K �50.93 �11.21 �39.72 PF-PhO–CO2�K �50.15 �12.81 �37.34CyHexMeO–CO2�DBUH �19.82 �21.22 1.40 PF-PhO–CO2�DBUH �18.64 �18.94 0.30CyHexMeO–CO2�TMGH �17.45 �19.87 2.42 PF-PhO–CO2�TMGH �16.10 �18.68 2.58CyHexMeO–CO2�BbH �20.97 �21.49 0.52 PF-PhO–CO2�BbH �20.43 �20.54 0.11

Paper PCCP

This journal is© the Owner Societies 2017 Phys. Chem. Chem. Phys., 2017, 19, 15403--15411 | 15407

oxygen and hydrogen atom on the carbon next to the protona-tion site24 (on the six-membered ring). The geometry ofHexO–CO2�DBUH adduct shows no significant changes goingfrom the gas phase to MeCN, where the ion-pair adductcomprised of BbH+ (conjugate acid of Barton’s base) tends tobe more compact in the gas phase. The optimised geometryof the HexO–CO2�TMGH adduct does not show ion-pair for-mation in the gas phase, viz., HexO–CO2H�TMG. This might beexplained by the weak binding of the proton and the nitrogenatoms at TMG, which results in the formation of carbonic acid.This is not the case when MeCN is used, where the ion-pairis formed due to the higher pKa value of TMG at a higherdielectric constant medium. The optimised structure of

MeO–CO2�DBUH agrees very well with the reported X-Raydiffraction (XRD) structure.25

Hydrogenated versus perfluorinated alcohols

We have investigated the ability of PFAs to capture CO2 andcompared the results with the corresponding fully hydrogenatedcounterpart. The differences in DDG (DGcapture (F) � DGcapture (H))values between both fluorinated and hydrogenated alcohols aresummarized in Table 3. Hydrogenated alcohols show morefavourable DG values than PFAs in MeCN with MHs, exceptt-BuOH (9) once mixed with NaH or KH. Similarly, DDG valuesindicate a more favourable CO2 binding in the presence of DBUand TMG for the hydrogenated alcohols over the perfluorinatedones. Using Bb, most of the PFAs are more capable of capturingCO2 than the hydrogenated alcohols. PF-PhOH showed a morefavourable DG compared to PhOH (X = H, 12) with allinvestigated bases.

The optimised geometries for BuOH (7; n = 3) and PF-BuOH(14; n = 3) with the different bases are shown in Fig. 4. Thenature of the interaction between the carbonate anion andthe counter cation is similar for both alcohol categories. Theshorter interaction distances obtained in the case of BuOHmight indicate stronger interactions and explain its higherreactivity over the perfluorinated counterpart. Indeed, the highelectronegativity of the fluorine atom makes the formed RO�

less reactive towards nucleophilic attack to form the carbonatespecies. This is demonstrated by the low charge density of the

Fig. 3 Optimised geometries of the HexO–CO2�countercation adducts in both (A) the gas phase and (B) MeCN. Distances between the hydrogen atomon the base and carbonate oxygens are given in Å.

Table 3 DDG (DGcapture (F) � DGcapture (H), kcal mol�1)a for CO2 bindingbetween perfluorinated and fully hydrogenated alcohols in MeCN

Li Na K DBU TMG Bb

7 (n = 0) 6.05 3.69 1.87 1.51 1.49 0.647 (n = 1) 2.6 2.15 0.35 1.35 �0.08 �2.337 (n = 2) 5.99 2.1 0.94 1.55 0.54 �0.937 (n = 3) 6.32 3.01 1.53 2.17 0.54 0.457 (n = 4) 5.79 2.59 0.55 2.43 0.45 �0.417 (n = 5) 4.89 2.57 0.73 0.08 1.34 �0.229 2.80 �0.19 �1.02 �0.83 �5.70 �1.9112 (X = H) �0.90 �2.04 �2.35 �3.11 �2.52 �3.89

a Positive values indicate that hydrogenated alcohols are more capableof capturing CO2 than PFAs.

Fig. 2 (A) Experimental versus calculated DG for CO2 capture; the R2 value in black is obtained after omitting the i-PrOH data points, while the R2 valuein red is obtained for all data points. Free energy values for CO2 capturing (DGcapture) in MeCN for linear alcohols versus number of carbon atoms found inmolecules: (B) hydrogenated alcohols (7); n = 0–5. (C) PFAs (14); n = 0–5. (Li: J, Na: &, K: D, DBU: ", TMG:B, Bb: 2).

PCCP Paper

15408 | Phys. Chem. Chem. Phys., 2017, 19, 15403--15411 This journal is© the Owner Societies 2017

oxygen atom of the PFAs, together with longer O� � �CO2 bonds,even in the presence of the counter base which is supposed tostabilize the carbonate anion formation (Fig. S3, ESI†).

Interestingly, PF-PhOH (16) is found to have a favourable DDGvalue over PhOH (X = H; 12) with all used MHs and SBs, particularly,when Bb is utilized, it gets stabilized by ca. 4 kcal mol�1 (Table 3).The optimised structures of PhCO3

� are found to be stable overPF-PhCO3

� (Fig. 5A and B, respectively) in the absence of thecounter cation (1.56 and 2.52 Å, respectively). This is attributedto the electronic repulsion between the oxygen atoms on theCO2 and the fluorine atoms. Herein, no significant difference inthe charge density of the oxygen atom of ArO� of 12 and 16 isfound (Fig. S4, ESI†). In the presence of any counter cation, e.g.,BbH+ (Fig. 5C and D), both ArCO3

� are stabilized throughelectrostatic attractive forces together with H-bonding. Thethermodynamic data show that DH for the CO2 capture byPF-PhOH using BbH+ is favourable by B3 kcal mol�1 over Ph-OH.This might be attributed to additional electrostatic interactionsbetween the negatively charged fluorine atoms and the positivelycharged hydrogen atoms on the superbase core.

para-Substituted phenols (p-PhOHs) (12)

Substituents, particularly those located at the para position onthe benzene ring to the hydroxyl group, can dramaticallyinfluence the acidity of the phenols and thus their ability to

act as CO2 sorbents due to the resonance and/or inductiveeffects. On one hand, it is anticipated that electron withdrawinggroups enhance the acidity of p-PhOHs (decrease the pKa). Onthe other hand, electron donating substituents decrease their

Fig. 4 Optimised structure of (A) the BuO–CO2 based adducts. (B) PF-BuO–CO2 based adducts in MeCN. Distances are given in Å. Longer O� � �CO2

bonds are obtained in the case of perfluorinated based adducts versus the hydrogenated ones (1.55 versus 1.39 Å).

Fig. 5 The optimised structures of ArCO3� in MeCN of: (A) PhCO3

�,(B) PF-PhO�CO2

� (in the absence of a counter cation). (C) PhCO3�BbH,(D) PF-PhCO3�BbH (distances are given in Å).

Fig. 6 The optimised structures of p-(X)-PhCO3�DBUH in MeCN, where Xis: (A) H–, (B) Me–, (C) MeO–, (D) F–, (E) Cl–, (F) NO2–. Distances betweenthe N–hydrogen of the DBUH+ and carbonate oxygens are given in Å.Relative DG values with respect to (A) are shown below the structures.(G) Hammett plot for DG values of p-PhOHs.

Paper PCCP

This journal is© the Owner Societies 2017 Phys. Chem. Chem. Phys., 2017, 19, 15403--15411 | 15409

acidity (Table 1). Calculations show that in the presence of DBUas an activator (Fig. 6A–F), the MeO-substituent has the mostfavourable relative DG value (�0.42 kcal mol�1), while theNO2– group has the least value (0.53 kcal mol�1). This is furtherfortified by the obtained distances between ArCO3

� and theproton of DBUH+. In the case of MeO–, they are 1.73 and 2.85 Å,compared to NO2– which are 1.77 and 2.89 Å. These results arein a good agreement with Schneider’s26 findings that indicateda similar trend when p-PhOHs are used to capture CO2 in thepresence of ionic liquids. DG values of p-PhOHs are plotted as afunction of Hammett constant27 (s) for different substituentsas shown in Fig. 6G. A linear correlation was obtained, exceptfor CH3– which is the only outlier.

Solvent effect (dielectric constant) on CO2 binding

It is anticipated that the reaction energetics depend strongly onthe choice of the solvent. We have selected the HexOH (7, n = 5)/DBU binary mixture as a model to investigate the solvent effecton the DG values for CO2 capturing. The Hex–CO3�DBUHadduct is stabilized by the high dielectric constant mediumdue to its ability to separate charges upon solvation. As shownin Fig. 7, the values of DG are descending as a function of e asknown in the literature which is further verified by our compu-tational findings. Herein, water was excluded to be used as asolvent due to its reactivity toward CO2 to form the bicarbonateion rather than the formation of the organic Hex–CO3

� adduct.

Spectroscopic predictions (NMR and IR)

Experimentally the CO2 capturing process using an alcohol/SBbinary system is investigated using Fourier transform infrared(FT-IR) spectroscopy, in which the newly formed RCO3

� showsthe emergence of new peaks ascribed to the chemisorptionprocess. The appearance of new peaks at B1670 cm�1 and1630 cm�1 corresponds to the stretching frequencies of CQO andO� � �C� � �O within the ionic organic carbonate anion, respectively.28

The calculated IR spectra of Hex–CO3�BH (B: MHs (1–3), and SBs(4–6)) and the asymmetric stretching frequencies of O� � �C� � �O areshown in Fig. 8 and Table 4, respectively. The calculated frequencyvalues vary as a function of the counter cation, ranging within1617–1644 cm�1 for M+ and 1647–1659 cm�1 for SBH+ which is inaccordance with the experimentally reported values.

The nuclear magnetic resonance (NMR) chemical shifts arecalculated with DFT using a 6-31+G* basis set for Hex–CO3�SBH(Fig. 9). The obtained chemical shifts show a good agreement withthe experimentally reported values.25 The most up-fielded chemicalshift observed in 1H NMR is attributed to the hydrogen bondingbetween the protonated SB and anionic carbonate species (H1, Fig. 9).This agrees with the experimental results obtained for similarsystems.29 The calculated carbon chemical shift of the formedcarbonate are 147, 147, 146 ppm with DBU, TMG, and Bb, respec-tively. Experimentally, the formation of R-CO3�SBH is confirmedby 13C NMR through the emergence of a peak at B157–158 ppm,which is distinguishable from the physisorbed CO2 (at 125 ppm).

Conclusions

In summary, quantum chemical calculations are used to inves-tigate the ability of different alcohols to capture CO2 in thepresence of inorganic and organic bases. General conclusionscan be made: (i) the electronic and sterical effects play a keyrole in the stability of the formed carbonate species, in whichprimary alcohols are found to be more reactive compared tosecondary and tertiary alcohols. (ii) Similarly, hydrogenatedalcohols are more reactive than the fluorinated counterparttoward CO2 capturing. (iii) Base selection is an importantmodulator in the CO2 capture with alcohols. The calculationsindicate that metal hydrides act as promising bases that canefficiently stabilize the formed carbonate anions. The reactivityof alcohols in the presence of organic SBs depends on the pKa

Fig. 7 Calculated DG values for CO2 capture as a function of the solventdielectric constant (e) for the HexOH/DBU binary mixture.

Fig. 8 Partially calculated (B3LYP/6-31+G*) IR spectra of Hex–CO3�BHin MeCN.

Table 4 The calculated (B3LYP/6-31+G*) asymmetric stretching frequencyof OQCQO of Hex–CO3�BH in MeCN

Carbonate adduct Freq./cm�1

HexO–CO2�Li 1617.11HexO–CO2�Na 1630.30HexO–CO2�K 1643.97HexO–CO2�DBUH 1646.64HexO–CO2�TMGH 1653.66HexO–CO2�BbH 1658.90

PCCP Paper

15410 | Phys. Chem. Chem. Phys., 2017, 19, 15403--15411 This journal is© the Owner Societies 2017

values of the used SBs. (iv) Solvents with high dielectric con-stants are found to stabilize the ionic carbonate adducts.

Acknowledgements

KIA acknowledges the Computational Laboratory for Analysis,Modelling, and Visualization (CLAMV) at the Jacobs UniversityBremen. AFE acknowledges the Deanship of Scientific Researchat the Hashemite University.

References

1 P. G. Jessop, D. J. Heldebrant, X. Li, C. A. Eckert andC. L. Liotta, Nature, 2005, 436, 1102.

2 J. J. Gassensmith, H. Furukawa, R. A. Smaldone, R. S.Forgan, Y. Y. Botros, O. M. Yaghi and J. F. Stoddart, J. Am.Chem. Soc., 2011, 133, 15312–15315.

3 R. A. Smaldone, R. S. Forgan, H. Furukawa, J. J. Gassensmith,A. M. Z. Slawin, O. M. Yaghi and J. F. Stoddart, Angew. Chem.,Int. Ed., 2010, 49, 8630–8634.

4 A. F. Eftaiha, F. Alsoubani, K. I. Assaf, W. M. Nau, C. Trolland A. K. Qaroush, RSC Adv., 2016, 6, 22090–22093.

5 A. F. Eftaiha, F. Alsoubani, K. I. Assaf, C. Troll, B. Rieger,A. H. Khaled and A. K. Qaroush, Carbohydr. Polym., 2016,152, 163–169.

6 D. J. Heldebrant, C. R. Yonker, P. G. Jessop and L. Phan,Energy Environ. Sci., 2008, 1, 487–493.

7 Y. Park, K.-Y. A. Lin, A.-H. A. Park and C. Petit, Front. EnergyRes., 2015, 3, 42.

8 D. R. Lide, CRC Handbook of Chemistry and Physics,CRC Press, Boca Raton, Florida, 84th edn, 2003.

9 R. Rajamanickam, H. Kim and J.-W. Park, Sci. Rep., 2015,5, 10688.

10 D. J. Heldebrant, P. K. Koech, J. E. Rainbolt, F. Zheng,T. Smurthwaite, C. J. Freeman, M. Oss and I. Leito,Spec. Sect. Symp. Post-Combust. Carbon Dioxide Capture, 2011,171, 794–800.

11 D. C. Cantu, J. Lee, M.-S. Lee, D. J. Heldebrant, P. K. Koech,C. J. Freeman, R. Rousseau and V.-A. Glezakou, J. Phys.Chem. Lett., 2016, 7, 1646–1652.

12 B. Cao, J. Du, S. Liu, X. Zhu, X. Sun, H. Sun and H. Fu,RSC Adv., 2016, 6, 10462–10470.

13 B. Seyedhosseini, M. Izadyar and M. R. Housaindokht,RSC Adv., 2016, 6, 85924–85932.

14 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone,B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato,X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng,J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda,J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao,H. Nakai, T. Vreven, J. A. Montgomery, Jr., J. E. Peralta,F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin,V. N. Staroverov, T. Keith, R. Kobayashi, J. Normand,K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar,J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene,J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo,R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin,R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin,K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,J. J. Dannenberg, S. Dapprich, A. D. Daniels, O. Farkas,J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian09, Revision B.01, Gaussian, Inc., Wallingford CT, 2010.

15 J. E. Del Bene, W. B. Person and K. Szczepaniak, J. Phys.Chem., 1995, 99, 10705–10707.

16 A. Rudnitskaya, B. Torok and T. Dransfield, Comput. Theor.Chem., 2011, 963, 191–199.

17 S. Miertus, E. Scrocco and J. Tomasi, Chem. Phys., 1981, 55,117–129.

18 J. L. Gohres, A. T. Marin, J. Lu, C. L. Liotta and C. A. Eckert,Ind. Eng. Chem. Res., 2009, 48, 1302–1306.

Fig. 9 Calculated (B3LYP/6-31+G*) 1H NMR spectra of Hex–CO3�SBHin MeCN.

Paper PCCP

This journal is© the Owner Societies 2017 Phys. Chem. Chem. Phys., 2017, 19, 15403--15411 | 15411

19 I. Kaljurand, A. Kutt, L. Soovali, T. Rodima, V. Maemets,I. Leito and I. A. Koppel, J. Org. Chem., 2005, 70, 1019–1028.

20 R. A. Kunetskiy, S. M. Polyakova, J. Vavrık, I. Cısarova, J. Saame,E. R. Nerut, I. Koppel, I. A. Koppel, A. Kutt, I. Leito andI. M. Lyapkalo, Chem. – Eur. J., 2012, 18, 3621–3630.

21 A. Kutt, V. Movchun, T. Rodima, T. Dansauer, E. B. Rusanov,I. Leito, I. Kaljurand, J. Koppel, V. Pihl, I. Koppel, G. Ovsjannikov,L. Toom, M. Mishima, M. Medebielle, E. Lork, G.-V.Roschenthaler, I. A. Koppel and A. A. Kolomeitsev, J. Org.Chem., 2008, 73, 2607–2620.

22 F. Eckert, I. Leito, I. Kaljurand, A. Kutt, A. Klamt andM. Diedenhofen, J. Comput. Chem., 2009, 30, 799–810.

23 DG values for selected systems were also calculated at amore reliable but also more computationally demandingquantum mechanical method (second order Møller–Plesset:MP2). The MP2 calculations resulted in identical trends,

but with more favorable energy values compared to theB3LYP at the same basis-set.

24 S. Scheiner, Phys. Chem. Chem. Phys., 2011, 13, 13860–13872.25 L. Phan, D. Chiu, D. J. Heldebrant, H. Huttenhower, E. John,

X. Li, P. Pollet, R. Wang, C. A. Eckert, C. L. Liotta andP. G. Jessop, Ind. Eng. Chem. Res., 2008, 47, 539–545.

26 T. B. Lee, S. Oh, T. R. Gohndrone, O. Morales-Collazo,S. Seo, J. F. Brennecke and W. F. Schneider, J. Phys.Chem. B, 2016, 120, 1509–1517.

27 C. Hansch, A. Leo and R. W. Taft, Chem. Rev., 1991, 91,165–195.

28 R. M. Silverstein, G. C. Bassler and T. C. Morrill, SpectrometricIdentification of Organic Compounds, Wiley, New York,5th edn, 1991, vol. 26.

29 Y. Yang, H. Xie and E. Liu, Green Chem., 2014, 16,3018–3023.

PCCP Paper