Enhanced CO2 capture by reducing cation-anion interactions in hydroxyl-pyridine anion-based ionic liquids

Article abstract of DOI:10.1039/c8dt04680h

In this work, an efficient strategy for improving CO₂ capture based on anion-functionalized ionic liquids (ILs) by reducing cation-anion interactions in ILs was reported. The influence of the cationic species on CO₂ absorption was investigated using 2-hydroxyl pyridium anions ([2-Op]) as a probe. CO₂ capture experiments indicated that the CO₂ absorption capacity in [2-Op] anion-based ILs varied from 0.94 to 1.69 mol CO₂ per mol IL at 30 °C and 1 atm. Spectroscopic analysis and quantum chemical calculations suggested that the increase of the CO₂ absorption capacity may be ascribed to the reduction of the strength of cation-anion interactions in ILs, and stronger cation-anion interactions would make one CO₂ site in the [2-Op] anion inactive. Furthermore, the effect of the cation unit on the anion was evidenced by FT-IR spectra, implying that strong interactions between ions may lead to the decrease of the IR absorption wavenumber of hydroxy pyridium and work against CO₂ capture. Following this strategy, it was finally found that [Ph-C₈eim][2-Op] (Ph-C₈eim = 1-N-ethyl-3-N-octyl-2-phenylimidazolium) with weaker cation-anion interactions exhibited a significant increase in the CO₂ uptake capacity, and extremely high capacities of 1.69 and 1.83 mol CO₂ per mol IL could be achieved at 30 and 20 °C, respectively. The study presented here would be helpful for further designing novel and effective ILs for advancing CO₂ capturing performance.

Full text of DOI:10.1039/c8dt04680h

Dalton
Transactions
View Article Online
View Journal
PAPER
Enhanced CO2 capture by reducing cation–anion
interactions in hydroxyl-pyridine anion-based
ionic liquids†
Cite this: DOI: 10.1039/c8dt04680h
Xiao-Yan Luo, a Xiao-Yan Chen,a Rong-Xing Qiu,a Bao-You Pei,a Yi Wei,a Min Hu,a
a,b
Jin-Qing Lin, *a Jing-Yang Zhanga and Geng-Geng Luo
*
In this work, an efficient strategy for improving CO2 capture based on anion-functionalized ionic liquids
(ILs) by reducing cation–anion interactions in ILs was reported. The influence of the cationic species on
CO2 absorption was investigated using 2-hydroxyl pyridium anions ([2-Op]) as a probe. CO2 capture
experiments indicated that the CO2 absorption capacity in [2-Op] anion-based ILs varied from 0.94 to
1.69 mol CO2 per mol IL at 30 °C and 1 atm. Spectroscopic analysis and quantum chemical calculations
suggested that the increase of the CO2 absorption capacity may be ascribed to the reduction of the strength
of cation–anion interactions in ILs, and stronger cation–anion interactions would make one CO2 site in the
[2-Op] anion inactive. Furthermore, the effect of the cation unit on the anion was evidenced by FT-IR
spectra, implying that strong interactions between ions may lead to the decrease of the IR absorption wave-
number of hydroxy pyridium and work against CO2 capture. Following this strategy, it was finally found that
[Ph-C8eim][2-Op] (Ph-C8eim = 1-N-ethyl-3-N-octyl-2-phenylimidazolium) with weaker cation–anion
interactions exhibited a significant increase in the CO2 uptake capacity, and extremely high capacities of
1.69 and 1.83 mol CO2 per mol IL could be achieved at 30 and 20 °C, respectively. The study presented here
would be helpful for further designing novel and effective ILs for advancing CO2 capturing performance.
Received 27th November 2018,
Accepted 2nd January 2019
DOI: 10.1039/c8dt04680h
rsc.li/dalton
which tend to acquire electrons.5,6 Functionalized ILs, such as
amine,7,8 azolium,9 phenolate10 and carbene-based ILs,11 were
1. Introduction
Rapid anthropogenic climate change caused by a large quan- reported as potential CO2 capturing solvents owing to their
tity of green-house gases is one of the most significant electron-rich properties, i.e., the basicity of ILs. Tuning the
environmental problems in the world today.1 Development of basicity of ILs is a great method to balance the CO2 absorption
sustainable and environmentally friendly technology to reduce capacity and energy consumption used for CO2 desorption.
the green-house gas emissions, especially carbon dioxide In this regard, Wang et al. found that the CO2 absorption
(CO2), is the focus of attention in many countries. Ionic capacity of azole-based ILs dominated by the pKa value of the
liquids (ILs) are typically low melting organic salts comprising anion had a rough linear relationship with the absorption
organic or inorganic cations and anions.2,3 Unique physico- enthalpy.12 Phenolate-based ILs and AHA-based ILs with
chemical properties such as high thermal stability, low vapour different substituents also exhibited a similar phenomenon.10
pressure and tunable properties make ILs suitable as CO2
Designing the steric hindrance of the functionalized moi-
absorbents.4 A CO2 molecule represents two delocalized Π eties was another strategy affecting the CO2 capture perform-
bonds from unhybridized p orbitals of the C atom and hybrid ance of ILs. In this respect, He and co-workers reported that
orbitals of the O atom in the electron-deficient ∏43 form, an amino group with steric hindrance prompted the formation
of carbamic acid.13 Brennecke et al. reported that the steric
hindrance in amino acid-based ILs would have lower viscosity
after absorbing CO2.14 The hydrogen bond formation or
aKey Laboratory of Environmental Friendly Function Materials, Ministry of
Education, College of Materials Science and Engineering, Huaqiao University,
Xiamen 361021, P.R. China. E-mail: linlab@hqu.edu.cn, ggluo@hqu.edu.cn
bState Key Laboratory of Structural Chemistry, Fujian Institute of Research on the
Structure of Matter, Chinese Academy of Sciences, Fuzhou 350002, P.R. China
†Electronic supplementary information (ESI) available: NMR and FT-IR spectral
data as well as Tables S1–S4 and Fig. S1–S4. See DOI: 10.1039/c8dt04680h
proton transfer has a great effect on CO2 capture, such as improv-
ing the CO2 absorption capacity,15,16 or decreasing the viscosity of
ILs.17,18 Recently, designing ILs with cooperative sites19–21 in
various forms22,23 was considered an attractive method to enhance
the CO2 gas absorption capacity.19–23 Overall, there is a great diver-
sity of functional design strategies to improve CO2 capture
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Scheme 1 Mechanism of CO2 capture by ILs with various functional
moieties and the strategies for improving CO2 capturing properties.
Chart 1 Ionic structures and the respective abbreviations for the
studied ILs.
properties, summarized in Scheme 1. However, the studies mainly
focused on functionalized moieties. Taking anion-functionalized
ILs for example, diverse strategies around the design of the anion
were developed for enhanced CO2 capture properties, while the
cation was usually considered as a counter-ion unless it was
unstable and participated in the reaction with CO2.24–27 As such,
understanding how the non-functionalized counter ions affect
CO2 capture is very important.
species on ILs was analysed through calculating the cation–
anion interaction energy and evidenced by vibration spec-
troscopy. It is proposed that weak cation–anion interactions
would contribute to the high CO2 absorption capacity, whereas
strong interactions would quench the activity of one site in the
anion. Accordingly, [Ph-C8eim][2-Op] (where [Ph-C8Eim]+
=
In recent years, the influence of non-functionalized moi-
eties of ILs on CO2 capture has also received attention. A repre-
sentative example is as follows: the stability of the cation
greatly affects the CO2 capture process. Thereinto, the imidazo-
lium and the phosphonium cations would compete to react
with CO2 in the carbene and ylide process, respectively.25–27
1-N-ethyl-3-N-octyl-2-phenylimidazolium) with weak cation–
anion interactions was designed as a promising CO2 capture
agent, and it showed extremely high capacities up to 1.69 and
1.83 mol CO2 per mol IL at 30 and 20 °C, respectively.
Riisager et al. reported that the [Lys] anion paired with [N66614
]
2. Results and discussion
2.1 Physical properties of [2-Op] anion-based ILs
and [P66614] displayed different CO2 capture routes, which
resulted in 1.5 and 2.0 mol of CO2 per mol of IL, respectively.28
Hatton et al. reported a series of [2-CNpyr] based ILs with Considering the fact that physical properties may affect the
various phosphonium cations, and the results showed that the stabilities and recycling properties of the ILs as well as the
CO2 capture process was less effected by the cationic species; mass transfer of CO2, some properties such as the viscosity
thus the gravimetric CO2 absorption capacity could be and thermal properties of ILs were measured and the data are
enhanced by decreasing the alkyl chain length from [P66614] to summarized in Table 1. For the phosphonium-based ILs, the
[P2228].29 However, Brennecke et al. reported that the CO2 viscosities of [P4442][2-Op], [P8884][2-Op], and [P66614][2-Op]
absorption capacity would decrease along with the increase of were 193.0, 525.4, and 310 cP, respectively, indicating that
the alkyl chain length of tetraalkylphosphonium due to the there was less effect of the phosphonium cation on their visc-
different entropies of reaction and the relative degrees of ioni- osities. However, the differences of viscosities between imida-
city.30 Thus, it is important to investigate the influence of the zolium-based ILs having the same [2-Op] anion and different
cation on CO2 capture by anion-functionalized ILs, and there imidazolium cation species could be quite large, spanning
is proposed guidance for enhancing the properties of ILs.
almost an order of magnitude. For instance, the viscosities of
The 2-hydroxypyridium anion ([2-Op]) with sharing of con- ILs [Me-C8mim][2-Op], [C4mim][2-Op], and [Ph-C8eim][2-op]
jugate Π electrons and two CO2 capture sites benefits CO2 were 137, 300 and 1112 cP, respectively, whereas ILs
capture with high capacity according to our previous results.31 [Me-C4mim][2-Op] and [Me-C6mim][2-Op] were solid at room
It was reported that the conjugated Π electron distribution of temperature. Thus, it was apparent that imidazolium ILs were
[2-Op] may be easily tuned by tuning the surroundings such as denser than phosphonium ILs. In addition, the analysis by
the cation.24 In this work, we report a series of [2-Op] anion- differential scanning calorimetry (DSC) implied that the
based ILs containing structurally diverse cations for effectively decomposition temperature of phosphonium- and imidazolium-
capturing CO2. The structures and abbreviations of different based ILs were in the range of 192–300 °C, which was suitable
types and sizes of cations for the studied ILs are depicted in for the application as CO2 capture agents, and phosphonium ILs
Chart 1 and Table S1 (ESI†). The influence of the cation were more thermally stable than the imidazolium ILs.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
Table 1 The viscosity and decomposition temperature of 2-hydroxy-
pyridine (2-Op)-based ILs
Viscositya
(cP)
Decomposition
Entry
IL
temperatureb (°C)
1
2
3
[P4442][2-Op]
[P8884][2-Op]
[P66614][2-Op]
193.0
525.4
—
288
270
—
4
5
6
7
8
9
10
11
[N4442][2-Op]
176
300
—
—
209
220
221
193
206
192
152
205
[C4mim][2-Op]
[Me-C4mim][2-Op]
[Me-C6mim][2-Op]
[Me-C8mim][2-Op]
[P4442OH][2-Op]
[Me-Pyr][2-Op]
[Ph-C8eim][2-Op]
137
213.6
165.0
1112
a Viscosity data were obtained using a Brookfield DV-II+Pro viscometer
at 25 °C. b Decomposition temperature was measured by DSC with the
temperature increase from 25 to 600 °C at a rate of 10 °C min−1 under
N2 gas flow.
Fig. 1 13C NMR and FT-IR spectra of pure ILs compared with those of
their CO2 saturated ILs. 13C NMR spectroscopy change of [P4442][2-Op]
(a) and [Me-C4mim][2-Op] (b) along with CO2 capture; black line, pure
IL; red line, CO2 saturated IL. IR spectroscopy change of [P4442][2-Op] (c)
and [Me-C4mim][2-Op] (d) along with CO2 capture; black line, pure IL;
gray shade, CO2 saturated IL.
2.2 CO2 capture properties of [2-Op] anion-based ILs
These synthesized ILs were employed for CO2 absorption
capture at 30 °C and 1 atm ambient pressure. The data of
yielded CO2 absorption capacities are listed in Table 2.
Thereinto, the CO2 uptake capacity of IL [P4442][2-Op] was as
high as 1.40 CO2 per mol IL (entry 1, Table 2), which was
similar to the reported results.31 Entries 1–5 illustrate that the
CO2 uptake capacity by ILs ranges from 1.40 to 1.02 mol of
CO2 per mol of IL when changing the cation species, indicat-
ing that the CO2 absorption capacity by [2-Op]-based ILs
greatly depended on the cationic units.
To gain more insight into these interactions of the ILs with
CO2 molecules, [P4442][2-Op] was characterized during the
uptake of CO2 using FT-IR and NMR spectroscopic techniques
(Fig. 1). Compared with the IR spectrum of pure [P4442][2-Op],
two additional symmetric stretching bands of OvC–O− at
1670 and 1646 cm−1 along with two asymmetric stretching
vibrations of OvC–O− at 1373 and 1238 cm−1 were observed
after the CO2 reaction, indicating that the CO2 molecule was
fixed by [2-Op] in two forms.31 At the same time, the 13C solu-
tion NMR spectra of the CO2 saturated [P4442][2-Op] IL showed
two new signals at 177.6 and 159.9 ppm. The characteristics of
CO2 capture shown by [P4442][2-Op] were very similar to those
in our previously reported work in which the N and O sites of
the [2-Op] anion cooperatively bound a CO2 molecule.31
Furthermore, the 31P NMR spectra of [P4442][2-Op] before and
after the CO2 reaction are shown in Fig. S3 (ESI†), demonstrat-
ing that no phosphorus ylide (31.5 ppm) was present according
to ref. 24, 25 and 27. Therefore, the cation species in these ILs
did not bind CO2 molecules. In contrast, for the CO2 saturated
complex of the IL [Me-C4mim][2-Op], only two additional
stretching bands of the OvC–O− bond at 1602 and 1330 cm−1
were observed in the IR spectrum, and a carbon peak at
164.8 ppm in the 13C NMR spectrum was generated after the
CO2 reaction, indicating that just one form of CO2 was
captured by the N site in the anion.
Table 2 CO2 capture by ILs with different cation moieties and the cation–anion interaction energies of ILs
IR absorptiona
CO2 absorption
capacityc
CO2 absorption
Cation–anion
Entry
IL
Experimental
Calculationb
capacity of Tg mixturee
interaction energyf
1
2
3
4
5
6
7
[P4442][2-Op]
[N4442][2-Op]
[Bmim][2-Op]
[Me-C4mim][2-Op]
[Bmpr][2-Op]
[P4442OH][2-Op]
[Ph-C8eim][2-Op]
1434
1433
1425
1423
1428
1424
1434
1473
1469
1460
1420
1455
1423
1471
1.40
1.24
1.02
1.06
1.39
1.28
1.16
1.52
1.28
0.59
—
−322.3
−344.9
−366.8
−364.0
−370.4
−394.6
−317.5
1.17
0.94
1.69 (1.83)d
a cm−1; the wavenumber belonged to the vibrational absorption of hydroxyl pyridium from IR spectra. b cm−1; the data were calculated with the
Gaussian 03 program at the B3LYP/6-31++G(p,d) level with the correction factor as 0.961. c mol/mol IL; CO2 capture was operated at 30 °C and 1
atm CO2. d CO2 capture was operated at 20 °C and 1 atm CO2. e mol/mol IL; the mixture was obtained using the ratio of 1 : 1 (weight) of IL and Tg
solvent, CO2 capture was operated at 30 °C and 1 atm CO2 pressure. f kJ mol−1, the interaction energy was calculated using the Gaussian 03
program at the B3LYP/6-31++G(p,d) level with BSSE correction.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.

View Article Online
Paper
Dalton Transactions
2.3 Theoretical calculation analysis
peak appeared at 1434 cm−1 when the [2-Op] anion was paired
with the [P4442] cation, while the wavenumber of νpy was
shifted to 1423 cm−1 upon coupling with the cation [Me-
C4mim]. The wave number of νpy obtained from the IR spectra
was enhanced with the increase of the interaction energy
between the cation and anion, which was consistent with the
results from DFT calculations as discussed above. It is worth
noting that the CO2 absorption capacity has a good linear
relationship with the cation–anion interaction energy and νpy,
respectively, as shown in Fig. 2b. Accordingly, weakening the
cation–anion interactions may be an efficient approach to
improve the CO2 absorption capacity, and the ability of CO2
capture by [2-Op]-based ILs can be reflected by IR spectroscopy.
Tetraglyme (Tg) was considered as a solvent that could
weaken the strength of cation–anion interactions in ILs
because Tg preferentially solvates cations.33,34 Accordingly, ILs
mixed with additive Tg were applied in the exploration of the
influence of Tg on ILs. The effect of Tg on different ILs was
further investigated using IR spectra because the characterized
vibrational absorption of the anion might reflect the real situ-
ation. A comparison of the νpy between pure ILs and their Tg
mixtures and the extent of the effect of Tg on CO2 capture are
listed in Table S3.† The results indicated that the trend of the
increase of the νpy effect of Tg agreed with the trend of the CO2
capture experiment. Taking ILs [P4442][2-Op] and [Me-C4mim]
[2-Op] as examples, the characteristic absorption peak (νpy) of
[Me-C4mim][2-Op] was observed at 1423 cm−1 but it blue-
shifted to 1435 cm−1 after adding Tg, while there was just a
tiny shift of the νpy from 1434 to 1435 cm−1 for [P4442][2-Op], as
shown in Fig. 3b and d, respectively. It could be inferred
that strong interactions between [Me-C4mim]+ and [2-Op]−
were destroyed by the Tg solvent, while [P4442][2-Op] was
insensitive to the addition of Tg due to the weak ion inter-
actions. Thus in order to improve the CO2 capture capacity of
ILs, the mixtures of ILs and Tg were applied for CO2 capture,
In order to better explain the relationship between the cation–
anion interactions in these ILs with the CO2 absorption
capacity, the Mulliken atomic charge density of the nitrogen
and oxygen atoms in the anion was calculated using the
Gaussian 03 suite of programs with DFT, the B3LYP method,
and 6-31G** as the basis set. From the DFT calculations
(Fig. 2a), it was observed that the CO2 absorption capacity has
a good linear correlation with the atomic charge density of N
and O atoms. The greater negative charge on the O atom and
the less negative charge on the N atom would result in the
higher CO2 absorption capacity. As seen from the optimized
structures of ILs (Table S2, ESI†), the formation of a hydrogen
bond between the anion [2-Op] and the cation group in
addition to electrostatic interactions would make the charge
more aggregated to interact with the cation, and thus weaken
the ability of ILs to capture CO2. The interaction energies and
the vibrational absorption of the hydroxyl pyridium (νpy) of the
optimized ILs were also calculated. As can be seen in Table 2,
CO2 absorption capacities by ILs increased along with a
reduction of the energies of interactions between the cationic
and anionic groups as well as an increase of the wavenumber of
the νpy of the [2-Op]-based IL. For example, [P4442][2-Op] with a
high interaction energy (−322.3 kJ mol−1) and a large νpy value
(1473 cm−1) possesses a high CO2 absorption capacity up to
1.40 mol CO2 per mol IL at 30 °C, while [Me-C4mim][2-Op]
having a low interaction energy (−364.0 kJ mol−1) and a small νpy
value (1420 cm−1) approaches a low CO2 absorption capacity
(1.06 mol CO2 per mol IL). DFT calculations corroborate the
experimental results and it can thus be inferred that the presence
of strong cation–anion interactions may affect the properties of
the [2-Op] anion and sequester one CO2 absorption site in the
anion, resulting in the reduction of the CO2 uptake capacity.
2.4 The effect of an additive and the cationic chain on CO2
capture
It is well-known that IR vibrational spectroscopy has found
continued use as a fundamental tool to characterize and
detect the changes of functional groups of ILs,32 and it was
observed that the vibrational absorption of hydroxyl pyridium
νpy appeared at different wavenumbers. For example, an IR
Fig. 3 Comparison of the CO2 absorption capacity and IR spectra of ILs
and their mixtures with Tg. (a) CO2 absorption capacity of [P4442][2-Op]
compared with that of its Tg solution; (b) IR spectra of [P4442][2-Op] (red
dot) compared with those of its Tg solution (black line); (c) CO2 absorption
capacity of [Me-C4mim][2-Op] compared with that of its Tg solution; (d) IR
spectra of [Me-C4mim][2-Op] (red dot) compared with those of its Tg solu-
tion (black line); the blue dotted-line indicates the IR spectra of tetraglyme.
Fig. 2 (a) Correlation of the CO2 absorption capacity with the com-
puted atomic charge density of O (green dots) and N (red dots) in the
anion. (b) Correlation of the CO2 absorption capacity with the inter-
action energy between the cation and [2-Op] (green dots) and the
change of the IR absorption of the pyridine ring of [2-Op] (pink dots).
Dalton Trans.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
and the CO2 absorption capacity is showed in Fig. 3a and c.
Aiming at further exploring the influence of the strength of
The CO2 absorption capacity of the [P4442][2-Op]–Tg mixture cation–anion interactions on the CO2 uptake capacity, another
was 1.39 mol/mol IL similar to that of the pure IL (1.40 mol/ phosphonium-based IL, [P4442OH][2-Op], with an additional
mol IL), which might be ascribed to the weak interaction of [2- hydroxyl group in the cationic unit was investigated. Generally,
OP] with [P4442]. However, the addition of Tg to [Me-C4mim][2- the hydroxyl group in the cation may be in favour of hydrogen-
Op] IL dramatically increased the capacity of the IL for CO2, bond formation with the anion, which will contribute to
leading to an excess of 0.46 mol CO2 being captured. These greater intermolecular interactions. As seen in Table 2, the
experiments demonstrated that strong interactions between computed interaction energy of [2-Op]− with [P4442OH]+
the cation and the [2-Op] anion went against CO2 capture, and (−394.6 kJ mol−1) was obviously more negative than that of [2-
the CO2 capture ability of [2-Op] would be enhanced through Op]− with [P4442]+ (−322.3 kJ mol−1), and the calculated νpy of
the isolation of the cation from the anion after adding Tg.
[P4442OH][2-Op] (calculated νpy: 1424 cm−1) was less than that
We then focused our attention on increasing the alkyl chain of [P4442OH][2-Op] (calculated νpy: 1473 cm−1). The vibrational
lengths of the cation, which was considered as one of the spectrum of [P4442OH][2-Op] showed a characteristic peak at
effective methods to weaken interactions between ions. Two 1424 cm−1 assigned to the vibration of the pyridine ring skel-
kinds of [2-Op]-based ILs including phosphonium-based eton, which was clearly lower than that of [P4442][2-Op]
([P4442][2-Op], [P8884][2-Op] and [P66614][2-Op]) and imidazo- (1434 cm−1) (see Fig. S2 of the ESI†) and coincided with the
lium-based ILs ([Me-C4mim][2-Op], [Me-C6mim][2-Op] and theoretical results. As expected, [P4442OH][2-Op] could only
[Me-C8mim][2-Op]) (Chart 1) with different alkyl chain lengths capture 0.94 mol CO2 per mol of IL, demonstrating that the
were synthesized to investigate how the cationic size influences influence of the cation on CO2 capture is to a large extent
the CO2 capture properties. The IR vibrational spectra of these through the strength of cation–anion interactions. On adding
ILs are displayed in Fig. 4b. The wavenumber νpy of phos- Tg into [P4442OH][2-Op], we found that the νpy shifted to
phonium-based ILs ([Pmmmn][2-Op]) always remains at 1423 cm−1, because the Tg solvent might not be sufficient to
∼1434 cm−1 despite the fact that the alkyl chain lengths on the break the hydrogen bond with the hydroxyl functional group
phosphonium cation varies from butyl to octyl, which was an according to ref. 34 and 35.
indication that the lengths of the alkyl chain in the phos-
On the basis of the above results, it seemed reasonable to
phonium cation have no bearing on the anion. By comparison, assume that the reduction of the cation–anion interactions in
the νpy of imidazolium-based ILs ([Cn-mim][2-Op]) ranged from ILs might be helpful in improving the CO2 capture capacity. In
1424 to 1434 cm−1 with an increase in the lengths of the alkyl addition, the effect of the cation species on the CO2 capturing
chain from butyl to octyl on the imidazolium cation. When performance in the [2-Op]-based ILs may be predicted by ana-
these ILs were applied to CO2 capture under the conditions of lysing the changes of the calculated interaction energy and
30 °C and 1 atm ambient pressure, the CO2 absorption measured vibrational spectra. Following this point, we further
capacities achieved (Fig. 4a) indicated that the CO2 absorption designed another IL, [Ph-C8eim][2-Op]. From the view of struc-
capacity still remained steady when changing the size of the ture, this IL has an additional phenyl group located at the C-2
phosphonium cation. However, a pronounced influence was site of the cation, which may contribute towards weakening
noticed on the imidazolium-based ILs ([Me-Cnmim][2-Op]): the the strength of cation–anion interactions due to the weak
CO2 absorption capacity increased from 1.06 to 1.41 mol CO2 polarity of the aromatic ring. The DFT calculations showed
per mol IL as the lengths of the alkyl chain increased on the that the interaction energy between the ion pair of [Ph-C8eim]+
imidazolium cation. All the results presented here combined and [2-Op]− was −317.5 kJ mol−1, which was weaker than that
with the data of vibrational spectra implied that both CO2 of the ion pair of [P4442]+ and [2-Op]− (−322.3 kJ mol−1). In the
absorption capacity and wavenumber in imidazolium-based FT-IR spectrum of the IL [Ph-C8eim][2-Op], the vibration of the
ILs increased with the reduction of cation–anion interactions.
hydroxyl pyridium ring skeleton appears at a high wavenumber
of 1433 cm−1 (Fig. 4b), suggestive of its potential as an
efficient CO2 absorbent. It was visible from Fig. 5a that the
extremely high capacities of 1.69 and 1.83 mol of CO2 per mol
of [Ph-C8eim][2-Op] IL were determined at 30 °C and 20 °C,
respectively. It should be noteworthy that the value of the
capacity exhibited by [Ph-C8eim][2-Op] (1.83 mol/mol IL at
20 °C) was much higher than that of the previously reported IL
[P66614][2-Op] (1.58 mol CO2 per mol IL at 20 °C; see Table S4,
ESI†).31 The effect of the CO2 partial pressure shown in Fig. 5b
indicates that [Ph-C8eim][2-Op] has a high CO2 absorption
capacity of 1.40 mol/mol IL at a low CO2 pressure of 0.1 atm.
The IR spectra in Fig. 5c demonstrate 1673 and 1241 cm−1
attributed to CO2 captured by the oxygen site while 1647 and
1379 cm−1 are attributed to another CO2 captured by the nitro-
gen site.31 Notably, the absorbed CO2 gas at 30 °C and 1 atm
Fig. 4 (a) CO2 absorption capacity in [2-Op]-based ILs with different
alkyl carbon chain lengths on phosphonium and the 1,2-dimethyl-3-
alkyl-imidazolium cation at 30 °C and 1 atm pressure. (b) The FT-IR
absorption spectra of these ILs.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.

View Article Online
Paper
Dalton Transactions
Scheme 2 The proposed reversible CO2 absorption/desorption
mechanism shown by [2-Op] anion-based ILs.
Fig. 5 (a) The CO2 absorption curve of [Ph-C8eim][2-Op] with time at
30 °C (green circle) and 20 °C (red cricle) under 1 atm CO2 pressure; (b)
IR spectroscopy change of [Ph-C8eim][2-Op] along with CO2 capture
(black line: pure IL; gray shade: CO2 saturated IL); (c) CO2 capture by
[Ph-C8eim][2-Op] varied with pressure at 30 °C, and (d) six cycles of
CO2 absorption by [Ph-C8eim][2-Op] at 30 °C.
amount of water vapour in the flue gas did not significantly
reduce the CO2 capture capability of [Ph-C8eim][2-Op] com-
pared with a dry CO2 flow (1.69 mol CO2 per mol IL).
Drawing on previous studies for the CO2 capturing mecha-
nisms by the [2-Op]-based ILs,31 a proposed CO2 absorption/
desorption mechanism in the present system is depicted in
Scheme 2. Briefly, the [2-Op] anion structurally exhibits two
potential binding sites including electronegative N and O
atoms for CO2 capture. When CO2 entered into the [2-Op]-based
ILs, a high CO2 uptake capacity would be realized in the case of
weak cation–anion interactions, while strong interactions
between the cation and the anion only led to a low CO2 absorp-
tion capacity because the N site of the anion was inactive. That
is to say the ability of CO2 capture by [2-Op]-based ILs would be
better used in the absence of the interference of the cation.
Finally, the regeneration of the CO2-saturated IL solution was
observed at 80 °C with a vacuum pressure of 1 kPa.
can be readily desorbed at 80 °C with a vacuum pressure of
1
kPa. Six consecutive absorption cycles of CO2 by IL
[Ph-C8Eim][2-Op] are displayed in Fig. 5d and exhibit reversibility.
To a greater or lesser extent, water may be one of the stron-
gest competitors to CO2 in the absorption on ILs,36 affecting
their performances as CO2 capturing agents in carbon capture
utilization. Therefore, the CO2 capturing performances of [Ph-
C8eim][2-Op] were then tested in a wet gas mixture containing
4.2% water vapour at 30 °C, aiming to assess the effect of
water vapour on CO2 capture (Fig. 6). The results indicated
that the CO2 absorption capacity of [Ph-C8eim][2-Op] with the
copresence of 4.2% water vapour still remained at 1.56 mol
CO2 per mol IL. Thus, it was demonstrated that a small
3. Experimental section
3.1 Materials and instrumentation
An anion exchange resin in its chloride form (95%), and tri-
butylamine (98%) were obtained from Sinopharm Group
Chemical Company. Tributylethylphosphonium bromide
(98%) was purchased from Nanjing Chemlin Biomedical
Technology Company. 1,2-Dimethylimidazole (98%) was pur-
chased from J&K Chemical Company Limited. The following
chemical reagents were purchased from Aladdin Industrial
Corporation: tetraglyme (99%), 1-chlorohexane (98%), 1-bro-
modecane (98%), 1-methylimidazole (99%), 2-bromoethanol
(95%), bromoethane (98%), dimethyltetragol (99%), 2-hydroxy-
pyridine (98%), 1-bromooctane (98%), 1-bromobutane (98%),
N-octylphosphine (98%), tributylphosphine (98%), 1-chloro-
hexane (98%), and 2-phenylimidazole (98%). All the chemical
reagents mentioned above were used without further
purification.
Fig. 6 The effect of water vapour on CO2 capture by [Ph-C8eim][2-Op].
Black filled square, weight increase of [Ph-C8eim][2-Op] by bubbling
60 ml min−1 N2 with 4.2% water vapour; red filled circle, weight increase
of [Ph-C8eim][2-Op] by bubbling 60 ml min−1 CO2 with 4.2% water
vapour; green open circle, CO2 absorption capacity by [Ph-C8eim][2-Op]
with the deduction of captured water (see the Experimental section).
Dalton Trans.
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
1H and 13C NMR spectra were measured at room tempera- cation–anion interactions in the [2-Op] anion-functionalized
ture using a Bruker spectrometer (500 MHz) with tetramethyl- ILs. A series of [2-Op] anion-based ILs containing various
silane (TMS, 0.00 ppm) as an internal standard. FT-IR spectra phosphonium, quaternary ammonium, and imidazolium
were recorded using a Nicolet IS 50 FT-IR spectrometer. The cations were synthesized and applied for CO2 capture.
viscosity of ILs was detected using
a Brookfield DVII Spectroscopic analysis and theoretical calculations suggest
Pro+viscometer at 25 °C. DSC was performed from 25 °C to that the increase of the CO2 absorption capacity may be
600 °C on a Shimadzu DTG-60H instrument at a heating rate assigned to the reduction of the cation–anion interaction
of 10 °C min−1 under N2 flow. The water content of ILs was strengths in ILs, and favorable cation–anion interactions
determined by Karl Fischer titration, and was found to be would make the N site of the [2-Op] anion inactive to CO2
lower than 0.5 wt%.
capture. Furthermore, the effect of the cation on the anion can
be detected using FT-IR spectra, implying that strong inter-
actions between ions may decrease the IR absorption wave-
number of the vibration of the hydroxy pyridinium ring skel-
eton and work against CO2 capture. Thus the [2-Op] anion
paired with cation [Pmmmn] showed weak cation–anion inter-
actions and no obvious change of the CO2 capture capacity
which varied with the carbon length of the cation. In contrast,
the increase of the CO2 absorption capacity is particularly
evident for imidazolium-based ILs when the cation alkyl chain
length increases from [Me-C4mim][2-Op] to [Me-C8mim][2-Op]
with the decrease of the interaction between the [Me-Cnmim]+
cation and the [2-Op]− anion. Following this strategy, the
newly synthesized IL [Ph-C8Eim][2-Op] exhibited the highest
CO2 capture capacity, up to 1.69 and 1.83 mol of CO2 per mol
of IL at 30 and 20 °C, respectively. Moreover, the CO2 absorp-
tion capacity remained steady during six cycles of CO2 absorp-
tion/desorption, implying that this CO2 capture process by
[Ph-C8Eim][2-Op] was highly reversible. The results presented
here would be of value to design novel and effective ILs for
enhancing CO2 absorption capturing performance in the future.
3.2 General method of preparation of [2-Op] based ILs
The preparation of the ILs was achieved by the stepwise syn-
thetic procedures including first the synthesis of IL halides, a
further anion exchange reaction, and final neutralization with
2-hydroxypyridine. The detailed structural characterization of
all ILs is given in the ESI.† In a typical experiment, tributyl-
ethylammonium 2-hydroxypyridine ([N4442][2-OP]) was syn-
thesized according to the procedures given below: (1) tributyl-
ethylammonium bromide ([N4442][Br]) was synthesized using
literature methods;37 (2) then
a
solution of tributyl-
ethylammonium hydroxide ([N4442][OH]) in ethanol was
obtained from [N4442][Br] from the anion-exchange method;5
(3) equimolar 2-hydroxypridine was added to the ethanol solu-
tion of [N4442][OH], and the mixture was subjected to a
vacuum-rotary evaporation procedure to remove most of the
ethanol after stirring at room temperature for 24 h. The
obtained sample was dried under reduced pressure at 60 °C
for 1 d to remove the possible residual moisture. CO2 absorp-
tion experiments with the IL were carried out according to our
previous report.31 CO2 desorption was operated under a
vacuum pressure of 1 kPa for 2 hours to recover the IL. The
mixture of the IL and Tg was prepared with the weight ratio as
1 : 1, and the CO2 absorption capacity of the IL was obtained
from deducting CO2 dissolution by Tg.
Conflicts of interest
There are no conflicts to declare.
3.3 Computational methods
Acknowledgements
All calculations were performed in this study using the
Gaussian 03 programs package. For each set of calculations,
we calculated the geometry optimization for the free anion and
the cation–anion complex, and the interaction energy of the
cation–anion at the B3LYP/6-31G++(d,p) level. The basis set
superposition error (BSSE)38 was corrected for all interaction
calculations using the counterpoise method.38–40
A gas flow with the copresence of 4.2% water vapour was
achieved by bubbling pure N2 or CO2 into 30 °C H2O (saturated
vapor pressure is 4.2455 kPa), the CO2 absorption capacity using
moist CO2 was calculated through the increase of weight by bub-
bling the CO2 gas flow with 4.2% water vapour minus the increase
of weight by bubbling the N2 gas flow with 4.2% water vapour.
We gratefully acknowledge the financial support for this work
from the National Natural Science Foundation of China
(21803021, 21641011, and 21246008), the Natural Science
Foundation of Fujian Province (JZ160407, JAT170032,
2016J01060, and 2015H0024), and the Education and
Scientific Research Project of Middle and Young Teachers in
Fujian (JAT170032). G. Luo also thanks the Provincial
Innovative
Foundation
Project
for
undergraduates
(201810385072).
Notes and references
1 M. R. Raupach, G. Marland, P. Ciais, C. Le Quere,
J. G. Canadell, G. Klepper and C. B. Field, Proc. Natl. Acad.
Sci. U. S. A., 2007, 104, 10288–10293.
4. Conclusions
In summary, we have developed an efficient controlled strategy
for enhancing the CO2 capture capacity via reducing the
2 X. P. Zhang, X. C. Zhang, H. F. Dong, Z. J. Zhao, S. J. Zhang
and Y. Huang, Energy Environ. Sci., 2012, 5, 6668–6681.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans.

View Article Online
Paper
Dalton Transactions
3 G. K. Cui, J. J. Wang and S. J. Zhang, Chem. Soc. Rev., 2016, 21 R. Vaidhyanathan, S. S. Iremonger, G. K. H. Shimizu,
45, 4307–4339.
P. G. Boyd, S. Alavi and T. K. Woo, Science, 2010, 330, 650–
653.
4 J. H. Huang and T. Ruther, Aust. J. Chem., 2009, 62, 298–
308.
22 L. Xia, Q. Cui, X. Suo, Y. H. Li, X. L. Cui, Q. W. Yang,
J. H. Xu, Y. W. Yang and H. B. Xing, Adv. Funct. Mater.,
2018, 28, 1–10.
5 Q. W. Song, Z. H. Zhou and L. N. He, Green Chem., 2017,
19, 3707–3728.
6 Z. H. Pan, G. G. Luo, J. W. Zhou, J. X. Xia, K. Fang and 23 W. Fam, J. Mansouri, H. Y. Li, J. W. Hou and V. Chen, ACS
R. B. Wu, Dalton Trans., 2014, 43, 8499–8507. Appl. Mater. Interfaces, 2018, 10, 7389–7400.
7 B. E. Gurkan, J. C. de la Fuente, E. M. Mindrup, L. E. Ficke, 24 X. W. An, X. Du, D. H. Duan, L. J. Shi, X. G. Hao, H. F. Lu,
B. F. Goodrich, E. A. Price, W. F. Schneider and
J. F. Brennecke, J. Am. Chem. Soc., 2010, 132, 2116–2117.
G. Q. Guan and C. J. Peng, Phys. Chem. Chem. Phys., 2017,
19, 1134–1142.
8 E. D. Bates, R. D. Mayton, I. Ntai and J. H. Davis, J. Am. 25 T. R. Gohndrone, T. B. Lee, M. A. DeSilva, M. Quiroz-
Chem. Soc., 2002, 124, 926–927.
Guzman, W. F. Schneider and J. F. Brennecke,
9 B. Gurkan, B. F. Goodrich, E. M. Mindrup, L. E. Ficke,
ChemSusChem, 2014, 7, 1970–1975.
M. Massel, S. Seo, T. P. Senftle, H. Wu, M. F. Glaser, 26 S. Seo, M. A. DeSilva and J. F. Brennecke, J. Phys. Chem. B,
J. K. Shah, E. J. Maginn, J. F. Brennecke and
W. F. Schneider, J. Phys. Chem. Lett., 2010, 1, 3494–3499.
10 C. M. Wang, H. M. Luo, H. R. Li, X. Zhu, B. Yu and S. Dai,
Chem. – Eur. J., 2012, 18, 2153–2160.
2014, 118, 14870–14879.
27 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.
11 G. Gurau, H. Rodriguez, S. P. Kelley, P. Janiczek, R. S. Kalb 28 S. Saravanamurugan, A. J. Kunov-Kruse, R. Fehrmann and
and R. D. Rogers, Angew. Chem., Int. Ed., 2011, 50, 12024–
12026.
12 C. M. Wang, X. Y. Luo, H. M. Luo, D. E. Jiang, H. R. Li and
S. Dai, Angew. Chem., Int. Ed., 2011, 50, 4918–4922.
13 A. H. Liu, R. Ma, C. Song, Z. Z. Yang, A. Yu, Y. Cai,
A. Riisager, ChemSusChem, 2014, 7, 897–902.
29 P. Brown, B. E. Gurkan and T. A. Hatton, AIChE J., 2015, 61,
2280–2285.
30 S. Seo, M. A. DeSilva, H. Xia and J. F. Brennecke, J. Phys.
Chem. B, 2015, 119, 11807–11814.
L. N. He, Y. N. Zhao, B. Yu and Q. W. Song, Angew. Chem., 31 X. Y. Luo, Y. Guo, F. Ding, H. Q. Zhao, G. K. Cui, H. R. Li
Int. Ed., 2012, 51, 11306–11310.
and C. M. Wang, Angew. Chem., Int. Ed., 2014, 53, 7053–
14 B. F. Goodrich, J. C. de la Fuente, B. E. Gurkan,
7057.
Z. K. Lopez, E. A. Price, Y. Huang and J. F. Brennecke, 32 V. H. Paschoal, L. F. O. Faria and M. C. C. Ribeiro, Chem.
J. Phys. Chem. B, 2011, 115, 9140–9150. Rev., 2017, 117, 7053–7112.
15 B. Hong, L. D. Simoni, J. E. Bennett, J. F. Brennecke and 33 A. Sharma, Y. Zhang, T. Gohndrone, S. Oh, J. F. Brennecke,
M. A. Stadtherr, Ind. Eng. Chem. Res., 2016, 55, 8432–8449.
16 F. Ding, X. He, X. Y. Luo, W. J. Lin, K. H. Chen, H. R. Li and
C. M. Wang, Chem. Commun., 2014, 50, 15041–15044.
M. J. McCready and E. J. Maginn, Chem. Eng. Sci., 2017,
159, 43–57.
34 X. Y. Luo, F. Ding, W. J. Lin, Y. Q. Qi, H. R. Li and
C. M. Wang, J. Phys. Chem. Lett., 2014, 5, 381–386.
17 M. G. Pan, R. Vijayaraghavan, F. L. Zhou, M. Kar, H. R. Li,
C. M. Wang and D. R. MacFarlane, Chem. Commun., 2017, 35 E. S. J. Reid, C. E. S. Bernardes, F. Agapito, F. Martins,
53, 5950–5953.
S. Shimizu, M. E. M. da Piedade and A. J. Walker, Phys.
Chem. Chem. Phys., 2017, 19, 28133–28138.
36 O. Y. Fajardo, F. Bresme, A. A. Kornyshev and M. urbakh,
ACS Nano, 2017, 11, 6825–6831.
18 X. Y. Luo, X. Fan, G. L. Shi, H. R. Li and C. M. Wang,
J. Phys. Chem. B, 2016, 120, 2807–2813.
19 Y. J. Huang, G. K. Cui, Y. L. Zhao, H. Y. Wang, Z. Y. Li,
S. Dai and J. J. Wang, Angew. Chem., Int. Ed., 2017, 56, 37 Y. Y. Jiang, G. N. Wang, Z. Zhou, Y. T. Wu,
13293–13297.
J. Geng and Z. B. Zhang, Chem. Commun., 2008, 505–
20 T. M. McDonald, J. A. Mason, X. Q. Kong, E. D. Bloch,
507.
D. Gygi, A. Dani, V. Crocella, F. Giordanino, S. O. Odoh, 38 H. B. Jansen and P. Ros, Chem. Phys. Lett., 1969, 3, 140–
W. S. Drisdell, B. Vlaisavljevich, A. L. Dzubak, R. Poloni, 143.
S. K. Schnell, N. Planas, K. Lee, T. Pascal, L. W. F. Wan, 39 S. F. Boys and F. Bernardi, Mol. Phys., 1970, 19, 553–
D. Prendergast, J. B. Neaton, B. Smit, J. B. Kortright, 566.
L. Gagliardi, S. Bordiga, J. A. Reimer and J. R. Long, Nature, 40 M. M. Szczesniak and S. Scheiner, J. Chem. Phys., 1986, 84,
2015, 519, 303–304. 6328–6335.
Dalton Trans.
This journal is © The Royal Society of Chemistry 2019