Tailoring ionic liquid catalysts: Structure, acidity and catalytic activity of protonic ionic liquids based on anionic clusters, [(HSO4)(H 2SO4)x]- (x = 0, 1, or 2)

Article abstract of DOI:10.1039/c4gc00415a

Aiming at inexpensive Bronsted-acidic ionic liquids, suitable for industrial-scale catalysis, a family of protonic ionic liquids based on nitrogen bases and sulfuric acid has been developed. Variation of the molar ratio of sulfuric acid, ?H2SO4, was used to tune acidity. The liquid structure was studied using 1H NMR and IR spectroscopies, revealing the existence of hydrogen-bonded clusters, [(HSO4)(H2SO4)]?, for ?H2SO4 0.50. Acidity, quantified by Gutmann Acceptor Number (AN), was found to be closely related to the liquid structure. The ionic liquids were employed as acid catalysts in a model reaction; Fischer esterification of acetic acid with 1-butanol. The reaction rate depended on two factors; for ?H2SO4 0.50, the key parameter was acidity (expressed as AN value), while for ?H2SO4 0.50 it was the mass transport (solubility of starting materials in the ionic liquid phase). Building on this insight, the ionic liquid catalyst and reaction conditions have been chosen. Conversion values of over 95% were achieved under exceptionally mild conditions, and using an inexpensive ionic liquid, which could be recycled up to eight times without diminution in conversion or selectivity. It has been demonstrated how structural studies can underpin rational design and development of an ionic liquid catalyst, and in turn lead to a both greener and economically viable process.

Full text of DOI:10.1039/c4gc00415a

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Tailoring ionic liquid catalysts: structure, acidity
and catalytic activity of protonic ionic liquids
Cite this: Green Chem., 2014, 16,
463
−
3
based on anionic clusters, [(HSO )(H SO ) ]
4
2
4 x
(x = 0, 1, or 2)†
a
a
b
b
Karolina Matuszek, Anna Chrobok,* Fergal Coleman, Kenneth R. Seddon and
b
Małgorzata Swadźba-Kwaśny*
Aiming at inexpensive Brønsted-acidic ionic liquids, suitable for industrial-scale catalysis, a family of
protonic ionic liquids based on nitrogen bases and sulfuric acid has been developed. Variation of the
1
molar ratio of sulfuric acid, χH2SO4, was used to tune acidity. The liquid structure was studied using H NMR
4 2 4
2 4
)(H SO >
)]−, for χH SO
and IR spectroscopies, revealing the existence of hydrogen-bonded clusters, [(HSO
.50. Acidity, quantified by Gutmann Acceptor Number (AN), was found to be closely related to the liquid
0
structure. The ionic liquids were employed as acid catalysts in a model reaction; Fischer esterification
of acetic acid with 1-butanol. The reaction rate depended on two factors; for χH2SO4 > 0.50, the key
parameter was acidity (expressed as AN value), while for χH2SO4 > 0.50 it was the mass transport (solubility
of starting materials in the ionic liquid phase). Building on this insight, the ionic liquid catalyst and reaction
conditions have been chosen. Conversion values of over 95% were achieved under exceptionally mild
conditions, and using an inexpensive ionic liquid, which could be recycled up to eight times without
diminution in conversion or selectivity. It has been demonstrated how structural studies can underpin
rational design and development of an ionic liquid catalyst, and in turn lead to a both greener and
economically viable process.
Received 9th March 2014,
Accepted 30th April 2014
DOI: 10.1039/c4gc00415a
www.rsc.org/greenchem
Introduction
Acid-catalysed processes are among the most important indus-
trial transformations, including large-scale alkylation, isomeri-
1
sation and esterification processes. The use of conventional
Brønsted acidic catalysts (HF, sulfuric or phosphoric acids,
zeolites) bears various drawbacks, such as high volatility, toxi-
city, waste generation and limited solubility/phase contact
2
with the starting materials. Brønsted acidic ionic liquids may
be promising greener alternatives to conventional Brønsted
acids for catalysis, helping to overcome some of these drawbacks.
Fig. 1 Locations of acidic protons (in red) on: (a) the 1-methylimidazo-
lium cation, (b) the 1-methyl-3-(3-sulfonic acid)propylimidazolium
cation, (c) the hydrogensulfate anion, and (d) the hydrogendi(acetate)
anion.
Structure and speciation of Brønsted acidic ionic liquids
Ionic liquids can be designed to carry a Brønsted acidic func-
tionality in the cation, anion, or both; the possible locations of
acidic protons within the structure of various ionic liquid ions
are shown in Fig. 1.
Here, we focus on ionic liquids prepared by proton trans-
fer between sulfuric acid and a base, where the acid was used
a
Department of Chemical Organic Technology and Petrochemistry,
Silesian University of Technology, Krzywoustego 4, 44-100 Gliwice, Poland.
E-mail: Anna.Chrobok@polsl.pl
3,4
b
QUILL, The Queen’s University of Belfast, Belfast, BT9 5AG, United Kingdom.
in various molar ratios, χH SO = 0.50, 0.67 or 0.75, see eqn (1).
E-mail: m.swadzba.kwasny@qub.ac.uk
2
4
†
Electronic supplementary information (ESI) available: Characterisation of ionic
liquids. See DOI: 10.1039/c4gc00415a
H2SO4 þ B ! ½HBꢀ½HSO4ꢀ χH SO ¼ 0:50
2
4
ð1aÞ
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 3463–3471 | 3463

View Article Online
Paper
Green Chemistry
2
H2SO4 þ B ! ½HBꢀ½ðHSO4ÞðH2SO4Þꢀ χH SO ¼ 0:67
ð1bÞ was typically used in stoichiometric or higher than stoichio-
2
4
1
2,18,25
metric quantities relative to the starting materials.
More-
3
H2SO4 þ B ! ½HBꢀ½ðHSO4ÞðH2SO4Þ ꢀ χH SO ¼ 0:75
ð1cÞ over, it was attempted to correlate Brønsted acidity, expressed
2
2
4
as Hammett acidity function (H ), with the activities of ionic
0
For the χH SO4 = 0.50 composition, the acidity may arise
from the labile proton at the cation and also the hydrogensul-
fate anion, Fig. 1(a) and (c). With an excess of sulfuric acid
2
liquids as esterification catalysts. However, the correlation was
rather poor, suggesting that the acidity was not the sole, or
1
8,22
even dominant, factor affecting the process.
(χH SO4 = 0.67 and 0.75), it is expected that the acidic protons
2
incorporated in complex hydrogen-bonded anionic clusters,
Aim of this study
will be the main contributors to the acidity of the system. The
−
Here, it is attempted to correlate the liquid structure of
protonic ionic liquids with their physico-chemical properties
(in particular acidity and miscibility with reactants) and,
building on that, design an optimal ionic liquid to be used
as a solvent and catalyst in a model esterification reaction.
existence of hydrogen-bonded anionic clusters, [A(HA)
x
] , has
been demonstrated for halide-based ionic liquids, contain-
−
5
− 6
−
ing anions such as [Cl(HCl)] , [Cl(HCl)
(
2
] , [Br(HBr)] , [Br-
] , and analogous species
are responsible for the Brønsted superacidity of Lewis acidic
−
− 7
− 8
2 x
HBr) ] , [Cl(HBr)] , and [F(HF)
+
This is proposed as a generic approach to tailoring
2 3 2
[C mim]Cl–AlCl –HCl systems ([C mim] = 1-ethyl-3-methylini-
dazolium). Complex anions have also been detected in liquids
9
ionic liquids as a replacement of hazardous acids in indust-
rially-relevant reactions, which is why the cynosure is to gen-
erate both environmentally and economically sustainable
processes.
formed by carboxylic acid–base proton transfer, with the acid
−
10
in excess, e.g. [(CF
3
COO)(CF
and other hydrogen-bonded carboxylate
Finally, mesylate and triflate ionic liquids,
3
COOH)] ,
3
[(CH COO)-
−
11
(
CH
anions.
(CH SO
catalyse the dehydration of methanol to dimethyl ether. The
3 x
COOH) ] ,
1
2,13
−
−
[
3
3 3 3 3 3 3 3
)(CH SO H)] and [(CF SO )(CF SO H)] were used to
1
4
Experimental
−
structure of hydrogen-bonded anions, [(HSO )(H SO ) ] , has Materials
4
2
4 x
not been confirmed in ionic liquids, but their existence is well
documented, for example in the Earth’s stratosphere.
1
,8-Diazabicycloundec-7-ene (DBU), 1-methylpyrrolidine, 1,4-
1
5–17
diazabicyclo[2.2.2]octane (DABCO), 1-methylimidazole, tri-
ethylamine, octane, triethylphosphine oxide, sulfuric acid
Brønsted acidic ionic liquids by proton transfer in catalysis
(
95%), and glacial acetic acid were purchased from Sigma-
Ionic liquids prepared by proton transfer are synthesised in a Aldrich and used as received. 1-Butanol was purchased from
simple, inexpensive, solvent-free, one-step process. When such Avantor Performance Materials Poland S.A.
ionic liquids are based on sulfuric acid, they have negligible
vapour pressure and are intrinsically halide-free. Conse-
Synthesis
quently, they are truly sustainable solvents and catalysts,
Protonic ionic liquids. In a typical procedure, a base (1 mol
which have the potential for industrial applications. Ionic eq.) was placed in a round-bottomed flask equipped with a
−
liquids based on monomeric hydrogensulfate anions, [HSO ] , stirring bar. The flask was placed in an ice bath and stirred
4
1
8
have been extensively used for catalysis. Moreover, in some vigorously; then, the required amount of sulfuric acid (1, 2, or
cases, various amounts of sulfuric acid have been dissolved in 3 mol eq.) was added dropwise, and the mixture was stirred for
1
9
these ionic liquids to enhance acidity, but the structural at least 1 h. Subsequently, the ionic liquid produced was dried
under high vacuum (80 °C, 10− bar, overnight).
2
aspect of these complex liquids has never been reported.
In this study, protonic ionic liquids based on sulfuric acid
Esterification procedure. In a typical experiment, a mixture
are used in a model Fischer esterification, exploiting their of 1-butanol (1 mol) and ionic liquid (1, 5, 10, 12 or 15 mol%)
2
0,21
3
ability to adopt the dual roles of solvent and catalyst.
was placed in a round-bottomed flask (10 cm ) equipped
Fisher esterification is an equilibrium reaction, which achieves with a magnetic stirring bar. The appropriate amount of
good rates in a homogenous process, but it reaches equili- acetic acid (1–2 mol) was added to the flask, which was
brium at ca. 67% conversion. When Brønsted acidic ionic immersed in an oil bath and fixed with a reflux condenser,
liquids are used instead, poor solubility of the ester products and then the mixture was stirred vigorously (30, 50 or 90 °C,
in the ionic liquid layer causes phase separation, driving the 1–8 h, 1000 rpm). The reaction was monitored by GC analysis,
2
2
equilibrium towards the product.
with octane as internal standard. The ester product phase-
Previously, Fischer esterification reactions have been separated from the ionic liquid layer and was decanted after
carried out in ionic liquids containing protonic cations, the reaction.
2
0,21
20,21,23,24
Fig. 1(a),
and protonic anions, Fig. 1(c),
the ionic
Recycling experiments. In a typical experiment where the
liquids adopting the dual roles of catalyst and solvent. The ionic liquid was recycled, reactions were scaled up by a factor
influence of a number of parameters (viz. reaction tempera- of ten. A mixture of 1-butanol (10.0 mol) and ionic liquid
3
ture, catalyst loading, reactant ratios and ionic liquid structure) (12 mol%) was placed in a round-bottomed flask (100 cm )
on the reaction has been investigated. The best results were equipped with a magnetic stirring bar. Acetic acid (15 mol)
−
achieved with ionic liquids based on the [HSO ] anion, which was added to the flask immersed in an oil bath and fixed with
4
3464 | Green Chem., 2014, 16, 3463–3471
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
a reflux condenser, and the mixture was stirred vigorously
(30 °C, 1 h, 1000 rpm). The ester product phase-separated
from the ionic liquid layer and was decanted after the reaction.
The ionic liquid was stirred under vacuum (70 °C, 10− bar,
2
6
h) to remove water, and subsequently used in the next cycle
of esterification.
Analyses
1
H NMR spectroscopy. In a glovebox, dry, neat ionic liquids
were loaded into NMR tubes (5 mm, borosilicate glass) con-
6
taining sealed capillaries with d -dimethylsulfoxide (an exter-
1
nal lock). H NMR spectra were recorded at 80 °C using a
Bruker AvanceIII 400 MHz spectrometer.
Acceptor number determination. For each ionic liquid,
three samples (ca. 1 g each) were weighed out in a glovebox,
and mixed with a P NMR probe molecule, triethylphosphine
oxide (TEPO), at different TEPO concentrations (3–10 mol%).
After dissolution of TEPO was ensured, the solutions were
Fig. 2 The structure of protonic cations used in this work.
3
1
between 0–100 °C (see ESI†), fulfilling the arbitrarily-assigned
2
8
definition of an ionic liquid. Systems based on the same
6
loaded into NMR tubes, containing sealed capillaries with d -
cations, but containing an excess of sulfuric acid (χH2SO4
>
dimethylsulfoxide (an external lock).
3
1
0.50), were all room-temperature ionic liquids, with glass tran-
P NMR spectra were recorded at 80 °C, at 161.98 MHz,
sitions T < 0 °C.
using a Bruker AvanceIII 400 MHz spectrometer. Phosphoric(V)
acid, 85% solution in water, was used as an external reference.
Three solutions of TEPO in hexane (ca. 5, 10 and 15 mol%)
were prepared, and then measured at 27 °C. For each TEPO-
g
2
+
In contrast, all systems based on [H DABCO]
or
2
2
+
2
[H DBU] were high-melting solids (doubly-charged cations
increase the lattice energy, promoting crystallisation), hence
these systems were not included in further studies.
3
1
ionic liquid system, the P NMR chemical shift for the infinite
Scanning (10 °C min− ) thermogravimetric analysis (TGA)
1
dilution of TEPO, δinf, was determined by extrapolation from
3
1
confirmed that the ionic liquids based on all four selected
the P NMR chemical shifts measured at different TEPO con-
centrations. The chemical shift of TEPO in hexane, extrapo-
lated to infinite dilution, δinf hex, was used as a reference
cations had good thermal stability (T > 200 °C), which makes
d
them suitable not only for this particular study, but for most
catalytic applications in organic chemistry. TGA curves are
illustrated in the ESI.†
(
δinf hex = 0 ppm).
The AN values for all samples were calculated from the
2
6,27
following formula: AN = 2.348·δinf
.
Structural analysis
FT-IR spectroscopy. Infrared spectra of neat samples were
recorded on a Perkin Elmer Spectrum 100 Series FT-IR spectro-
meter with a universal ATR accessory. Eight scans were
acquired for each sample.
Gas chromatography. GC analysis was performed using a
Perkin Elmer Clarus 500 chromatograph equipped with an
SPB™-5 column (30 m × 0.2 mm × 0.2 μm).
There were two foci of the structural studies: to confirm the
existence of hydrogen-bonded anionic clusters, and to study
the acidity of the labile protons. Since sulfuric acid is a strong
acid, and the produced liquids have good thermal stability
(
d
with T well above boiling points of the corresponding bases),
full protonation of the bases could be assumed for higher
temperatures as well as ambient. The structure of hydrogen-
bonded clusters of hydrogensulfate and sulfuric acid, [(HSO
(H SO ) ] (x = 0, 1, or 2), was anticipated to be very similar to
2 4 x
4
)-
−
Results and discussion
Synthesis and physical properties
the structure of hydrogensulfate-sulfuric acid clusters detected
commonly in the upper layers of the Earth’s atmosphere.
15–17
Protonic ionic liquids were prepared by the reaction between As shown in Fig. 3, the dinuclear and trinuclear clusters are
sulfuric acid and tertiary/aromatic amines, following eqn (1). believed bound with three and five strong hydrogen bonds,
The amines were selected such that the formed cations rep- respectively.
resented a variety of structures (aromatic, alicyclic, aliphatic)
To probe these assumptions, neat ionic liquids were
1
with a wide range of pK values (6.0–13.6), as shown in Fig. 2. studied using H NMR and FT-IR spectroscopies.
a
The anionic species were of the general formula [(HSO
(H SO ) ]
2 4 x
4
)-
(x = 0, 1, or 2), where x varied with the molar ratio studied using FT-IR spectroscopy. As a representative example,
of sulfuric acid, χ , eqn (1). spectra of the [Hmpyr][(HSO )(H SO ) ] system (χ = 0.50,
FT-IR spectroscopy. The structure of anionic species was
−
H SO
4
2
4 x
H SO
2 4
2
4
+
+
+
Ionic liquids based on [Hαmpy] , [Hmim] , [Hmpyr] or 0.67 and 0.75) are shown in Fig. 4. Singly bound S–O stretches
3
NH] and prepared by reaction of a base with an equimolar appear from 800 to 950 cm (black dashed line), double
amount of sulfuric acid (χ
+
−1
[Et
= 0.50) had melting points bonded SvO stretches fall in the 950–1300 cm− range (red
1
H SO4
2
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 3463–3471 | 3465

View Article Online
Paper
Green Chemistry
Fig. 3 The structure of anionic species expected to exist in protonic
ionic liquids based on sulfuric acid, depending on the mole fraction of
the acid, χH2SO4. For ease of comparison, atoms of the hydrogensulfate
anion are represented in blue, and those of sulfuric acid are in red.
1
Fig. 5 H NMR spectra (400.112 MHz, 80 °C, neat) of the [Hmpyr]-
(HSO )(H SO ] system, for χH2SO = 0.50, 0.67 and 0.75.
[
4
2
4 x
)
4
Concentrated sulfuric acid contains molecular H
2
SO
4
in
equilibrium with [HSO4]− and [H SO ] ions; characteristic
+
29
3
4
−
vibrations for both species of interest: H SO and [HSO ] were
2
4
4
3
0,31
detected (see Table 1), in agreement with the literature.
The χ
−
= 0.50 system is expected to contain only [HSO ] ;
H SO
4
2
4
indeed, the spectrum features one S–OH vibration at
44.2 cm− , symmetric SvO stretch at 1020.5 cm and a
1
−1
8
complex multiplet corresponding to various vibrations of the
−
1 30
{
SO } unit, with the most intense band at 1150.8 cm .
3
Complex anions present in χH SO = 0.67 and 0.75 have struc-
2
4
tures naturally similar to sulfuric acid, with both H SO and
2
4
−
−
[
HSO ] present. Vibrational frequencies for [HSO ] are inter-
4
4
mediate between those found in χH SO = 0.50 and in pure acid.
2
4
Bands originating from the S–OH bending vibrations are less
prominent and blue-shifted compared to the pure acid, since
this motion is restricted by very strong hydrogen bonding in
the clusters.
Fig. 4 FT-IR spectra of neat [Hmpyr][(HSO
0.50, 0.67 and 0.75, compared to the FT-IR spectrum of neat sulfuric
acid.
4 2 4 x
)(H SO )
] system, for χH2SO4
=
1
1
H NMR spectroscopy. H NMR spectra of the [Hmpyr]-
(HSO )(H SO ] system, for χH SO = 0.50, 0.67 and 0.75, are
shown in Fig. 5.
For χH2SO4 = 0.50, the most downfield-shifted signal (δ =
[
4
2
4 x
)
2
4
Table 1 Assignments of selected vibrational frequencies in the FT-IR
spectra of neat [Hmpyr][(HSO
and 0.75, and for sulfuric acid
4
)(H
2
SO
4
)
x
] system, for χH2SO = 0.50, 0.67
4
−
1
4
0.1 ppm) corresponds to the proton of the [HSO ] anion.
ν
S–O
ν
SvO
S–OH bend
Protonation of 1-methylpyrrolidine is evidenced by the weakly
shielded proton (δ = 8.0 ppm).
]−
]− [HSO
]−
χ
H SO4 [HSO
4
H SO
2 4
[HSO
4
4
H
2
SO
4
2 4
H SO
2
For χH SO = 0.67, the most downfield signal (δ = 11.4 ppm)
2
4
0
0
0
1
.50
.67
.75
.00
844.2
—
1020.5
1030.6
1036.8
1047.1
1150.8
—
—
—
—
integrates to three protons located on the dinuclear cluster,
871.6
874.7
885.6
923.1
928.1
942.7
1131.4 1388.0
1126.4 1369.4
1125.5 1347.5
−
−
[
(HSO )(H SO )] . They are deshielded relative to the [HSO ]
4
2
4
4
proton, because they are partaking in strong hydrogen bonds,
—
each shared between two oxygen atoms: S–O–H⋯O–S (see
1
6
Fig. 3); Hou et al. calculated (ab initio) the average hydrogen-
dashed line) and the SOH bend falls between 1350 and bond distance to be as low as 1.656 Å. Strong internal
400 cm−
1 18
.
1
Key vibrational frequencies with assignments are hydrogen bonds in the dinuclear anion result in weaker
listed in Table 1.
external hydrogen-bonding interactions with the cation. As a
3466 | Green Chem., 2014, 16, 3463–3471
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
1
Table 2 Signals corresponding to N–H and S–O–H protons in H NMR
Table 3 AN values measured for protonic ionic liquids, based on four
spectra (400.112 MHz, 80 °C, neat) of protonic ionic liquids, based on
different bases and three mol ratios of H
Cation
S–O–H/ppm [Hαmpy]
2 4
SO
2 4 a
four different bases and three mol ratios of H SO , compared to pK of
the cations
χ
H SO
AN
2
4
+
Cation
pK
a
χ
H SO
δ
N–H/ppm
δ
0.50
0
0
65.5
116.7
120.4
73.3
119
120.7
96.5
117
2
4
.67
.75
+
32
33
34
34
[
[
[
[
Hαmpy]
6.0
0.50
13.35
12.64
12.04 (m)
11.16
10.77
10.57
7.99
6.76
10.45
11.47
10.96
+
[
[
[
Hmim]
0.50
0
0
.67
.75
0
0
.67
.75
+
a
Hmim]
7.0
0.50
+
Hmpyr]
0.50
0
0
.67
.75
11.20
10.97
10.11
11.39
10.89
0
0
.67
.75
+
121.5
—
118
Hmpyr]
10.5
10.7
0.50
+
Et
3
NH]
0.50
0
0
.67
.75
0
0
.67
.75
6.24 (m)
Solid at 80 °C
6.08
+
121
Et
3
NH]
0.50
0
0
.67
.75
11.44
11.03
5.58
a
Gutmann acceptor number measurements presented them-
selves as the perfect way to answer the above questions,
because they probe the strength of protonation of a weak base,
Very broad signal, corresponding to both protons in equilibrium
consequence, the signal of the N–H proton is shifted upfield TEPO, by the acid of interest. AN values were measured for all
by ca. 0.5 ppm; noteworthy, the shape of the peak is non- ionic liquids but [Et NH][HSO ], which was solid at the temp-
Lorentzian.
erature of the experiment (80 °C). In Table 3, AN values for ionic
For χH2SO4 = 0.75, the signal at 10.9 ppm integrates to five liquids with different cations and χH2SO values are compared.
3
4
4
protons corresponding to the trinuclear complex, [(HSO
4
)-
For systems with χH SO > 0.50, where hydrogen-bonded
H SO ) ] . These protons are partaking in hydrogen bonds, cluster anions are present, AN values appear to be indepen-
2 4
−
(
2
4 2
1
6
which are strong, but on average weaker (1.704 Å) than those dent of the cation, with AN = 118 ± 1 for χH SO = 0.67, and
2
4
of the dinuclear system (1.656 Å) – hence the upfield shift of AN = 120.5 ± 1 for χH2SO = 0.75. In contrast to the order of
4
1
this signal compared to χ
= 0.67. The signal corres- acidity that could be inferred from the H NMR chemical shifts,
H SO4
2
ponding to the N–H proton is shifted even more upfield and AN values for χH SO = 0.75 are consistently slightly higher than
2
4
split into a multiplet, indicating that it is strongly bound to those for χH SO = 0.67.
2
4
nitrogen (weakly interacting with the anion, and not
exchanging).
In systems where χH SO = 0.50, TEPO is exposed to two
2
4
−
35
weak acids: [HSO
]
(pK = 1.92) and a protonated amine
a
4
For all studied ionic liquids, there are two sets of protons in (see Table 2). Protonated strong bases form – by definition –
1
the 6–12 ppm region of the H NMR spectrum; one originating very weak acids (pK
> 10); these adducts cannot compete
a
−
from N–H protons in the cations, and the other from the S–O– effectively with [HSO
H protons in the anions. Their chemical shifts are listed in quently, the acceptor number for [Hmpyr][HSO
Table 2 and H NMR spectra are available in ESI.†
] for TEPO, as shown in Fig. 6(a); conse-
] is quite high,
4
4
1
+
+
AN = 96.5. In contrast, protons in [Hαmpy] and [Hmim] are
The chemical shifts of labile anionic protons, δS–O–H
,
more acidic; they may form strong hydrogen bonds with TEPO,
−
depend markedly on the χH SO4 value, but not significantly on competing with [HSO
4
] , Fig. 6(b).
2
the cation: for all studied systems, irrespective of the cation
structure, δS–O–H values are below 10.5 ppm for χH SO = 0.50, liquids based on sulfuric acid and different bases can be
The AN study demonstrates that the acidities of ionic
2
4
increase to ca. 11.5 ppm for χH SO4 = 0.67 and decrease to ca. tuned within a wide range, by changing the base and tuning
2
1
1.0 ppm for χ
= 0.75. The N–H protons are more shielded
χH SO4. For χH SO > 0.50, high acidities can be achieved, with
H SO4
2 2 4
2
for higher χH SO4 values; dinuclear and trinuclear anions are AN = ca. 120, which is comparable to values for other strong
2
−
26
less basic than [HSO
4
] , hence they have weaker interactions acids (ANCF COOH = 105.5, ANCF SO H = 129.1). Systems based
3
3
3
with the protonated bases. For aromatic cations, the δN–H
signals are strongly shifted downfield, even >13 ppm. However,
this arises from the interaction with the aromatic ring, and
despite strong deshielding, these protons are not very labile,
as demonstrated by the presence of non-Lorentzian peaks, and
the observation of multiplets in some spectra (ESI†) indicates
that they do not exchange rapidly on the NMR timescale.
The question arises, how do these measurements of chemi-
cal shifts relate to the bulk acidity of the studied ionic liquids?
By extension, how does the catalytic activity of ionic liquids,
i.e. their ability to protonate organic molecules, relate to the
Fig. 6 Proposed interaction of TEPO (a) with the [HSO
4
]− anion in the
presence of strong protonated base, which is effectively undissociated,
and does not effectively interact with TEPO; and (b) in the presence of a
weakly protonated base, which behaves like a very weak acid, causing a
‘leveling effect’ and weakening the overall acidity of the system.
1
observed H NMR chemical shifts?
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 3463–3471 | 3467

View Article Online
Paper
Green Chemistry
Table 4 Miscibility of organic reactants and products (visual obser-
vation) with the ionic liquid layer, under the reaction conditions (30 °C)
Miscibility with the ionic liquid phase at
a
30 °C
Scheme 1 Model esterification reaction of acetic acid with 1-butanol.
Cation
χ
H SO
BuOH
MeCOOH
MeCOOBu
2
4
on the [HSO
4
]− anion (χH SO4 = 0.50) offer milder and tunable
+
2
[Hαmpy]
0.50
0.67
Y
Y
Y
N
N
N
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
Y
N
N
N
N
N
N
N
N
N
N
N
N
2
6
acidities, lower than that of formic acid (AN = 83.6).
0.75
+
Esterification of 1-butanol catalysed by protonic ionic liquids
[Hmim]
0.50
0
0
.67
.75
Esterification of acetic acid with 1-butanol (Scheme 1) is an equi-
librium reaction: without catalyst, the maximum conversion is [Hmpyr]
limited to 66.7%.
+
0.50
0.67
18,36
In a reaction catalysed by sulfuric acid
0.75
(homogenous liquid-phase system), it was been possible to shift
+
[
Et
3
NH]
0.50
0.67
the equilibrium further to the right, achieving 77.5% conversion.
Brønsted acidic ionic liquids can adopt the dual roles of solvent
and catalyst; the formed ester phase-separates from the reaction
0.75
a
Y – miscible with ionic liquid; N – immiscible with ionic liquid.
18,36
mixture, which allows for a further shift in equilibrium.
All the ionic liquids studied here are inexpensive and easily
prepared, presenting themselves as industrially viable cata-
lysts. Furthermore, the AN studies indicate that their acidity
can be easily tuned. Having gained a good understanding of
their structure and quantified their acidity by AN, now we set
out to understand how the ionic liquid properties relate to
their catalytic activity, and demonstrate their applicability in
practice, in an catalytic process.
i.e. low yields were obtained for χH SO = 0.50 and much higher
2
4
for χH SO = 0.67 or 0.75; indeed, higher yields of ester were
2
4
obtained for χ
= 0.75 than for χ
= 0.67, in agreement
H SO
4
H SO
2
4
2
+
with increasing AN. However, ionic liquids based on [Hmim]
do not appear to follow this trend.
To understand this discrepancy between AN values and
+
reactivity in [Hmim] -catalysed reactions, the phase behaviour
of the reaction mixture components was investigated by simple
visual observation (Table 4). It was observed that, at 30 °C,
acetic acid was miscible with all studied ionic liquids, and the
product (ester) always phase-separated from the mixture. The
crucial difference was observed for 1-butanol, which was
Key parameters affecting the reaction
Screening esterification experiments were performed under
very mild conditions (30 °C, 12 mol% of ionic liquid, 2 h reac-
tion time). Yield of the ester produced using different ionic
liquids (circles) and AN values measured for those ionic
liquids (bar chart) are compared in Fig. 7. In general, there is a
good correlation between AN values and the yield of the ester;
weakly miscible/immiscible with ionic liquids based on the
+
[
Hmim] cation, but miscible with other systems (see Table 4).
Reduced miscibility of 1-butanol had a detrimental impact on
mass transport, and in consequence on the yield of butyl
acetate, as clearly shown in Fig. 7.
To confirm that difference in conversion derives from struc-
tural variations in the anion, and not merely from changing
concentrations of sulfuric acid, an experiment was performed
whereby the reactants : sulfuric acid ratio remained constant,
but the amount of base (2-picoline) in the system was varied,
to form χ
= 0.50, 0.67 and 0.75 compositions. After 2 h
H SO
2
4
reaction at 30 °C, with molar ratio of alcohol to acid 1 : 1, the
following yields of ester were obtained for the respective χH SO
2
4
values: 20, 68 and 74%. This clearly demonstrates that differ-
ences in the ionic liquid structure affect the conversion,
despite identical nominal acid concentration.
The influence of process parameters on the model reaction
Having understood that the reaction rate depends on the
χ
H SO value of the ionic liquid and potential mass transport
2 4
constraints (i.e. miscibility of an ionic liquid with 1-butanol),
and taking into consideration that the reaction equilibrium
can be shifted towards the product by using excess of one of
the reactants and by phase-separation of the ester, reaction
conditions could be studied.
Fig. 7 Yield of butyl acetate in reactions catalysed with ionic liquids
based on sulfuric acid and four different bases, compared with AN
values measured for these ionic liquids.
3468 | Green Chem., 2014, 16, 3463–3471
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
Fig. 9 Yield of the ester product from the esterification of 1-butanol
3 4 2 4 2
with acetic acid, at 30 °C, catalysed with [Et NH][(HSO )(H SO ) ], as a
Fig. 8 Yield of the ester product for the esterification of 1-butanol with
function of time; curves correspond to different reactants ratios and
different ionic liquid loading (expressed as mol% per 1-butanol), as
shown on the graph.
3 4 2 4 2
acetic acid, at 30 °C, catalysed by [Et NH][(HSO )(H SO ) ], as a function
of time; curves correspond to different ionic liquid loadings (expressed
as mol% per 1-butanol), as shown on the graph.
The reaction was carried out at 30 °C, which was demon-
strated to be a viable temperature, much lower than literature
reports on similar reactions in ionic liquids, where tempera-
1
8,22,25
tures around 80–90 °C are used.
The ionic liquid of
choice was [Et NH][(HSO )(H SO ) ]; it dissolves both starting
3
4
2
4 2
materials, has high acid content (χH2SO4 = 0.75), thus providing
high reaction rate, and is based on a very inexpensive amine,
which makes it industrially viable.
Reaction rate has been studied as a function of the ionic
liquid content (1–15%, expressed as mol% per 1-butanol
content) to find the optimum catalyst loading (Fig. 8).
As shown in Fig. 8, for ≥12% of ionic liquid, the reaction rate
remains constant; moreover, even 5% of ionic liquid provides a
very good reaction rate, with 80% yield of butyl acetate after 2 h.
However, it is important to note that the final yield of the ester
is highest for 10–12% loadings of ionic liquid. It appears that for
lower ionic liquid content, the ester may not phase-separate
Fig. 10 Yield of butyl acetate from the esterification of 1-butanol with
completely, pushing the equilibrium to the left, and for higher acetic acid (30 °C, 12% of [Et NH][(HSO )(H SO ) ], 50 mol% excess of
3
4
2
4 2
loadings (>12%), acidic hydrolysis of the ester starts counter- acetic acid, 2 h) in consecutive recycling cycles.
balancing the reaction upon its formation. In conclusion, 12%
loading of ionic liquid has been selected as the optimum.
excess of acetic acid, 2 h). After each catalytic run, the product
was decanted from above the ionic liquid layer, and unreacted
starting materials were removed from the ionic liquid solution
by vacuum distillation. In eight consecutive runs, no mass loss
of the ionic liquid was observed. Yield of butyl acetate, although
appearing to decline slightly, remained within the error bars of
the measurement for all eight runs (mean yield = 98%, Fig. 10).
The influence of reactants ratio on reaction rate and on the
final yield of butyl acetate has been studied at 30 °C, using two
−
different loadings of [Et
3
NH][(HSO
4
)(H
2
SO
4
)
2
] : 12% and, as a
comparison, 5% (Fig. 9). As expected, the use of 50 mol%
excess of acetic acid resulted both in higher reaction rates and
in increased conversions. Under chosen conditions (12% of
ionic liquid, 50% excess of acetic acid), butyl acetate was syn-
thesised in greater than 95% yield.
Conclusions
Recycling study
Recycling of the ionic liquid was studied under the reaction The results discussed above expand and develop a family of
−
conditions (30 °C, 12% of [Et NH][(HSO )(H SO ) ] , 50 mol% inexpensive, easily prepared protonic ionic liquids based on
3
4
2
4 2
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 3463–3471 | 3469

View Article Online
Paper
Green Chemistry
sulfuric acid and various bases. The key development was to
4 A. Stark and K. R. Seddon, “Ionic Liquids”, in Kirk-Othmer
Encyclopaedia of Chemical Technology, ed. A. Seidel, John
Wiley & Sons, Inc., Hoboken, New Jersey, 5th edn., 2007,
vol. 26, p. 836.
5 T. A. Zawodzinski and R. A. Osteryoung, Inorg. Chem., 1988,
27, 4383.
use variable amounts of sulfuric acid (χ
= 0.67 and 0.75),
H SO4
2
which led to the formation of hydrogen-bonded anionic clus-
−
4 2 4 x
ters, [(HSO )(H SO ) ] (x = 1 or 2). The structure of the liquids
1
was studied and confirmed by H NMR and IR spectroscopies.
The acidity of these ionic liquids was quantified using
Gutmann acceptor numbers. It was found that, for χH SO
=
6 J. Campbell and K. Johnson, Inorg. Chem., 1993, 32,
3809.
2
4
0
.50, AN values may vary in a wide range depending on the
base used; they were higher in systems based on stronger
7 G. Driver and K. E. Johnson, Green Chem., 2003, 5, 163.
8 G. A. Olah, T. Mathew, A. Goeppert, B. Török, I. Bucsi,
X.-Y. Li, Q. Wang, E. R. Marinez, P. Batamack,
R. Aniszfeld and G. K. S. Prakash, J. Am. Chem. Soc.,
2005, 127, 5964.
bases, where the N–H proton was strongly bound to the base
−
and did not compete for the probe with acidic anion, [HSO ] .
4
For χH SO4 = 0.67 and 0.75, acidity was independent of the
2
cation, but depended only on the χH SO4 value. These ionic
2
liquids were highly acidic (AN ≈ 120).
9 G. P. Smith, A. S. Dworkin, R. M. Pagni and S. P. Zingg,
J. Am. Chem. Soc., 1989, 111, 525; G. P. Smith, A. S. Dworkin,
R. M. Pagni and S. P. Zingg, J. Am. Chem. Soc., 1989, 111,
5075.
The two key parameters which determine the reaction rate
of a model esterification reaction are: acidity of the ionic
liquid (well correlated with AN values) and solubility of start-
ing materials in the ionic liquid phase. For χH SO = 0.50, 10 M. Yoshizawa, W. Xu and C. A. Angell, J. Am. Chem. Soc.,
2
4
where acidity (AN) is lower, it plays the crucial role. For χH SO
=
2003, 125, 15411.
2
4
0
.60 and 0.75, where acidity is consistently very high, it ceases 11 K. M. Johansson, E. I. Izgorodina, M. Forsyth,
to be the main distinguishing parameter, and it is the miscibil-
ity of ionic liquid with alcohol that becomes the limiting
D. R. MacFarlane and K. R. Seddon, Phys. Chem. Chem.
Phys., 2008, 10, 2972.
factor. This miscibility is governed by the cation of the ionic 12 F. Kohler, H. Atrops, H. Kalali, E. Liebermann,
liquid, as shown in Table 4. Importantly, AN values gave more
E. Wilhelm, F. Ratkovics and T. Salamon, J. Phys. Chem.,
1981, 85, 2520.
insight into, and correlated better with, the observed catalytic
1
activity than did the H NMR chemical shifts of the labile 13 (a) K. Bica and R. D. Rogers, Chem. Commun., 2010, 46,
protons.
Building on these fundamental structural results, esterifica-
tion of acetic acid with 1-butanol was studied. The ionic liquid
of choice, [Et NH][(HSO )(H SO ], dissolved both starting
1215; (b) J. Stoimenovski, D. R. MacFarlane, K. Bica and
R. D. Rogers, Pharm. Res., 2010, 27, 521; (c) J. Stoimenovski,
P. M. Dean, E. I. Izgorodina and D. R. MacFarlane, Faraday
Discuss., 2011, 154, 335.
3
4
2
4 2
)
materials well, and was highly acidic, thereby ensuring good 14 M. P. Atkins, M. J. Earle, K. R. Seddon, M. Swadźba-Kwaśny
reaction kinetics. Prepared from triethylamine and sulfuric and L. Vanoye, Chem. Commun., 2010, 46, 1745.
acid, it was of very low cost; moreover, it was easily recycled 15 (a) J. Curtius, K. D. Froyd and E. R. Lovejoy, J. Phys. Chem.
multiple times without significant loss of activity. Using 50%
excess of acetic acid, it was possible to consistently achieve
conversions above 95% under exceptionally mild conditions.
A, 2001, 105, 10867; (b) E. R. Lovejoy and J. Curtius, J. Phys.
Chem. A, 2001, 105, 10874.
16 G.-L. Hou, W. Lin, S. H. M. Deng, J. Zhang, W.-J. Zheng,
F. Paesani and X.-B. Wang, J. Phys. Chem. Lett., 2013, 4,
779.
1
7 T. I. Yacovitch, N. Heine, C. Brieger, T. Wende, C. Hock,
D. M. Neumark and K. R. Asmis, J. Phys. Chem. A, 2013,
Acknowledgements
1
17, 7081.
8 C. Chiappe, S. Rajamania and F. D’Andreab, Green Chem.,
013, 15, 137; and references therein.
This work was financed by the Ministry of Science and Higher
Education, National Science Centre (grant no. UMO-2012/06/
M/ST8/00030).
1
1
2
9 (a) P. Wasserscheid, M. Sesing and W. Korth, Green
Chem., 2002, 4, 134; (b) W. Keim, W. Korth and
P. Wasserscheid, Ionic liquids, World Pat. WO 0016902,
2
000.
Notes and references
2
0 C. Chiappe and S. Rajamani, Eur. J. Org. Chem., 2011, 5517.
1
H. J. Arpe, Industrial Organic Chemistry, Wiley-VCH, Wein- 21 K. E. Johnson, R. M. Pagni and J. Bartmess, Monatsh.
heim, 5th edn, 2010. Chem., 2007, 138, 1077.
Catalysis from A to Z: A Concise Encyclopedia 4th completely 22 D. J. Tao, Y. T. Wu, Z. Zhou, J. Geng, X. B. Hu and
revised and enlarged edn., ed. B. Cornils, W. A. Herrmann, Z. B. Zhang, Ind. Eng. Chem. Res., 2011, 50, 1989.
C.-H. Wong and H.-W. Zanthoff, Wiley-VCH, Weinheim, 23 J. P. Belieres and C. A. Angell, J. Phys. Chem. B, 2007, 111,
Germany, 2013. 4926.
M. Freemantle, An Introduction to Ionic Liquids, RSC Publi- 24 M. Yoshizawa, W. Xu and C. A. Angell, J. Am. Chem. Soc.,
cations, Cambridge, UK, 2010. 2003, 125, 15411.
2
3
3470 | Green Chem., 2014, 16, 3463–3471
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
2
2
2
2
2
3
5 C. Zhang, X.-Y. Pan, M.-J. Yu, L. Jin and G. Wu, Chem. Eng. 31 T. I. Yacovitch, T. Wende, L. Jiang, N. Heine, G. Meijer,
J., 2012, 209, 464.
6 V. Gutmann, The Donor-Acceptor Approach to Molecular
Interactions, Plenum Press, New York, 1978.
D. M. Neumark and K. R. Asmis, J. Phys. Chem. Lett., 2011,
2, 2135.
32 R. H. Linnell, J. Org. Chem., 1960, 25, 290.
7 J. Estager, A. A. Oliferenko, K. R. Seddon and M. Swadźba- 33 T. C. Bruice and G. L. Schmir, J. Am. Chem. Soc., 1958, 80,
Kwaśny, Dalton Trans., 2010, 39, 11375–11382. 148.
8 D. R. MacFarlane and K. R. Seddon, Aust. J. Chem., 2007, 34 H. K. Hall Jr., J. Am. Chem. Soc., 1957, 79, 5441.
6
0, 3.
9 T. C. Waddington, Non-Aqueous Solvents, Nelson, London,
969.
0 G. E. Walrafen and D. M. Dodd, Trans. Faraday Soc., 1961,
7, 1286.
35 C. E. Housecroft and A. G. Sharpe, Inorganic Chemistry,
Pearson Education Limited, Harlow, 2005.
36 T. Iwasaki, Y. Maegawa, T. Ohshima and K. Mashima,
Esterification, in Kirk-Othmer Encyclopedia of Chemical
Technology, John Wiley & Sons, Inc., 4th edn, 2000.
1
5
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 3463–3471 | 3471