Protic ionic liquids from di- or triamines: even cheaper Br?nsted acidic catalysts

Article abstract of DOI:10.1039/d1gc00515d

Dicationic and tricationic ionic liquids, synthesised by proton transfer from sulfuric acid and corresponding di- and triamines, were characterised and used as Br?nsted acidic catalysts in biodiesel production. It has been demonstrated that the new ionic liquids are less expensive than those based on monoamines (even on triethylamine) and are highly acidic. Their performance in a model catalytic esterification reaction surpassed that of sulfuric acid and other literature reports, due to the capability of the ionic liquids to absorb water while phase-separating the ester, thereby shifting the reaction equilibrium towards products. The overall catalytic performance was an interplay between acidity (quantified by Gutmann AN) and physical properties: density and viscosity, which enabled easy phase separation, thereby separating water from the ester. The substrate scope and recycling studies have demonstrated that the new ionic liquids are robust and recyclable catalysts.

Full text of DOI:10.1039/d1gc00515d

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Protic ionic liquids from di- or triamines: even
cheaper Brønsted acidic catalysts†
Cite this: Green Chem., 2021, 23,
4421
Alina Brzęczek-Szafran, *a Justyna Więcławik, a Natalia Barteczko,
a
Anna Szelwicka,a Emily Byrne, b Anna Kolanowska,a
a
Małgorzata Swadźba Kwaśny *b and Anna Chrobok
Dicationic and tricationic ionic liquids, synthesised by proton transfer from sulfuric acid and corres-
ponding di- and triamines, were characterised and used as Brønsted acidic catalysts in biodiesel pro-
duction. It has been demonstrated that the new ionic liquids are less expensive than those based on
monoamines (even on triethylamine) and are highly acidic. Their performance in a model catalytic esterifi-
cation reaction surpassed that of sulfuric acid and other literature reports, due to the capability of the
ionic liquids to absorb water while phase-separating the ester, thereby shifting the reaction equilibrium
towards products. The overall catalytic performance was an interplay between acidity (quantified by
Gutmann AN) and physical properties: density and viscosity, which enabled easy phase separation,
thereby separating water from the ester. The substrate scope and recycling studies have demonstrated
that the new ionic liquids are robust and recyclable catalysts.
Received 9th February 2021,
Accepted 6th May 2021
DOI: 10.1039/d1gc00515d
rsc.li/greenchem
from low vapour pressure.3 Coulombic forces in the ionic
liquid phase appear to facilitate separation from the organic
Introduction
Sustainable development of chemical processes necessitates phase, e.g. compared to sulfuric acid, which facilitates purifi-
solutions in which the principles of green chemistry meet cation and recycling. Approaches reliant on the immobilis-
economic requirements. The use of cheap, nontoxic and easily ation of Brønsted acidic ILs are also known, additionally
available reagents that can be regenerated and re-used, facile enabling the recycling.4–7
product separation and mild reaction conditions became the
key requirements that must be met when designing a new Brønsted acidic ILs, but the most economically favourable is
process.1,2 the acid–base reaction, generating protic ILs.8 In particular,
There is a number of synthetic approaches to prepare
Acid catalysis is one of the oldest branches of catalysis, of reactions of various amines with sulfuric acid give ILs with
great industrial importance, including numerous esterifica- fully protonated cations (therefore negligible vapour pressures)
tions, transesterifications, alkylations, isomerisations and and a very low cost.9 According to technoeconomic analysis
Beckmann rearrangements. Traditional mineral acids, such as from 2014, the cost of triethylammonium hydrogen sulfate or
H2SO4 or HF, are inexpensive and effective, but are plagued 1-methylimidazolium hydrogen sulfate could be as low as
with sustainability challenges, such as product purification, $1.24 per kg or $2.96–5.88 per kg, on par with the cost of
cost of recycling, corrosion issues and the potential for highly common organic solvents like acetone or ethyl acetate ($1.30–
harmful leakage.
$1.40 per kg).10 A very recent analysis of such ILs has shown
A prominent approach to the design of more sustainable that this cost can be even lower ($0.78 per kg or $1.46 per
acid catalysts was the use of Brønsted acidic ionic liquids (ILs). kg),11 when indirect costs of environmental impact, human
In contrast to HF or HCl, most Brønsted acidic ILs benefit health, etc. are considered in addition to direct costs. It has
been recently demonstrated that protic ILs based on sulfuric
acid may be truly sustainable solvents and catalysts, with
strong potential for industrial uptake.12–14
aFaculty of Chemistry, Silesian University of Technology, Krzywoustego 4, Gliwice,
44-100, Poland. E-mail: alina.brzeczek-szafran@polsl.pl
bThe QUILL Research Centre, School of Chemistry and Chemical Engineering,
Queen’s University Belfast, Belfast BT9 5AG, UK.
When an excess of sulfuric acid is used, the main source of
acidity becomes the anion, of
a
summary formula
[HSO4·nH2SO4]−. The liquid structure of such an ionic liquid
has been found to comprise of a 3D anionic network of hydro-
gen bonded [HSO4·nH2SO4]− units, with discrete pyridinium
cations, [Hpy]+.15 The acidity of systems with an excess of
E-mail: m.swadzba-kwasny@qub.ac.uk
†Electronic supplementary information (ESI) available: Thermogravimetric ana-
lysis curves, 1H NMR spectra, viscosity and density values. See DOI: 10.1039/
d1gc00515d
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 4421–4429 | 4421

View Article Online
Paper
Green Chemistry
Brønsted acid can be fine-tuned using an acid in various Acceptor number determination
molar ratios, and can match that of the parent acid.16–19
Moreover, when sulfuric acid is used in excess, the cost of the
entire IL (mostly dependent on the amine) becomes even
lower.
AN values were determined using a previously described pro-
cedure.16 Three samples were prepared for each IL (ca. 1 g)
spiked with triethylphosphine oxide (TEPO) at 1% wt, 2% wt
and 3% wt (exact masses of TEPO and IL were noted and used
for calculations). After the dissolution of TEPO in the IL, the
31P NMR spectra were registered at room temperature (in
several cases at 45 °C due to the high viscosity of samples), at
161.89 MHz and recorded using an Agilent 400 MHz spectro-
meter. As an external standard 85% phosphoric(V) acid solu-
tion in D2O was used. The measured 31P NMR chemical shift
for each ionic liquid-TEPO concentration arrangement was
In this work, a new generation of highly acidic protic ILs
based on sulfuric acid is presented; they are more polar to
increase the ease of separation and have a higher percentage
of sulfuric acid by weight, to further lower cost. This was
achieved using di- or triamines to generate di- or tricationic
ILs, respectively. To date, protic ILs based on diamines have
been studied as solvents for gas absorption (CO2, SO2), with
diamines such as (3-dimethylamino)-1-propylamine or N,N-di-
methylethylenediamine being employed as cations.20–22
Herein, we report a comprehensive physical and spectroscopic
study, followed by the first catalytic application of protic ILs
extrapolated to determine the infinite dilution of TEPO, δinf
For each IL the acceptor number was calculated based on the
formula, AN = 2.348·δinf
.
.
based on simple di- and triamines, and conclude with a rudi- Thermogravimetric analysis
mentary cost analysis, which we hope will inspire further
development of these interesting ILs.
Thermogravimetric analysis (TGA) was performed using a
LINSEIS STA PT1600 thermobalance (Selb). The samples
(20–40 mg) were heated in alumina crucibles, at 10 °C min−1
,
up to 800 °C, under the flow of argon (0.05 L min−1).
Experimental
Materials and methods
Esterification procedure
In a typical esterification experiment, a mixture of oleic acid
(1 g, 3.5 mmol), methanol (12 : 1–1 : 1, molar ratio of methanol
to oleic acid) and di- or tricationic IL (3, 5, 10, 20 or 30 mol%)
was placed in a round-bottomed flask (10 cm3) equipped with
a magnetic stirring bar and a reflux condenser. The mixture
was stirred at 1000 rpm at a reaction temperature of 30–70 °C
and allowed to react for 0.25–3 h. After this time the reaction
mixture was washed with hexane (3 cm3) to remove traces of
the product from the IL phase. The organic phase was ana-
lysed by GC analysis, with decane as an internal standard.
GC analysis was performed using a Shimadzu GC-2010 Plus
chromatograph equipped with a ZB-50 column (30 m ×
0.25 mm × 0.25 μm) and an FID detector (detecting tempera-
ture: 300 °C). The initial GC oven temperature was kept at
80 °C for 1 min, heated at 45 °C per minute up to 300 °C,
where it was kept for 4 min.
N,N-Dimethylaminopropylamine, 1-(3-aminopropyl)imidazole,
diethylenetriamine, sulfuric acid (98%), oleic acid (90%), bis
(hexamethylene)triamine, triethylphosphine oxide, and metha-
nol were purchased from Sigma-Aldrich and used as received.
Synthesis of polycationic protic ionic liquids
In a typical procedure, sulfuric acid (1 mol eq.) was poured
into a round-bottomed flask which was equipped with a stir-
ring bar and placed in an ice bath. An appropriate amount of a
base (1–9 mol eq.) was added dropwise, with vigorous stirring.
Subsequently, the mixture was stirred for at least 1 h at 70 °C,
followed by drying under a high vacuum for 12 h (70 °C, 10−2
bar). All products but DETA_3_1 (Tm = 53–61 °C) and
HMTA_3_1 (Tm = 103–112 °C) were room-temperature ILs.
Density and viscosity measurements
Densities were measured using a DDM2911 density meter at
three temperatures (25, 50 and 80 °C), with the standard devi-
ation across the three measurements for each experiment
being <0.00187%.
Kinematic viscosities were measured using a capillary
AVS370 viscosity measuring system at three temperatures (25,
50 and 80 °C), with the standard deviation across the three
measurements for each experiment being <0.65%.
Recycling
For the recycling experiment the scale was increased by a
factor of 20. A mixture of oleic acid (20 g, 70.8 mmol), metha-
nol (8 : 1 molar ratio of methanol to oleic acid) and HMTA_9_1
(5 mol%) were stirred at 60 °C for 1 h. The ester forming a sep-
arate phase was easily decanted, while the IL phase was rinsed
with hexane to remove traces of the product. Methanol and
water were distilled off and an HMTA_9_1 system was further
used in the next reaction cycle.
Nuclear magnetic resonance
Neat ILs were loaded into NMR tubes (5 mm, borosilicate
glass) containing sealed capillaries with d6-dimethylsulfoxide
(an external lock). The procedure was performed in a glovebox.
1H NMR spectra were recorded at 25 °C (except for API_2_1
Results and discussion
Design and synthesis of di- and triamino protic ionic liquids
and DMAPA 2_1 which were recorded at 50 °C due to high vis- Four aliphatic/aromatic di- and triamines (Fig. 1) were used as
cosity), using a Bruker AvanceIII 400 MHz spectrometer.
the precursors: N,N-dimethylaminopropylamine (DMAPA), 1-
4422 | Green Chem., 2021, 23, 4421–4429
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
Fig. 1 Cations of the di- and tricationic ILs.
(3-aminopropyl)imidazole (API), diethylenetriamine (DETA),
and bis(hexamethylene)triamine (HMTA).
The four amines were reacted with sulfuric acid at three
molar ratios (Table 1). Stoichiometries of each reaction were
designed considering the number of basic amine groups, to
achieve three compositions: equimolar acid : base ratio (χH SO
Fig. 2 Physical appearance of di- and tricationic ILs based on sulfuric
acid.
2
4
= 0.50), two-fold acid excess (χH SO = 0.67) and three-fold acid
2
4
systems were synthesised by the addition of an amine to sulfu-
ric acid. Despite the colouring, all 1H NMR spectra (S3–S6,
ESI†) showed clean compounds, and therefore all ILs were
used in this study.
excess (χH SO = 0.75). Consequently, the molar ratio of sulfuric
2
4
acid to di- or tri-amine molecules was dependent on the
number of amine groups in each basic molecule.
All ionic liquids were prepared via a very simple, one-step
solvent-free synthesis and dried using a Schlenk line before
further experiments were carried out. All products but
DETA_3_1 and HMTA_3_1 were room-temperature ionic
liquids, with colours ranging from colourless to very dark
brown. It has been found that the difference in the synthetic
procedure affected the colour of the ILs. Addition of an acid to
an amine resulted in coloured samples, as evidenced for
DMAPA_2_1 and DMAPA_3_1 systems (S1, ESI†). Applying a
reversed procedure (addition of an amine to an acid) gave col-
ourless samples of DMAPA systems, as depicted in Fig. 2. The
colouring of the samples was most probably caused by local
overheating of the mixture, which occurred during the syn-
thetic procedure. Based on this observation, all remaining
Economic considerations
A comprehensive economic assessment of the production cost
of an ionic liquid should include the total annualised cost
(TAC) of the scaled-up processes, comprising of the operating
expenses (OPEX) and the annualised capital expenses
(CAPEX).11 However, since modelling of the scaled-up syn-
thesis of the reported ILs was beyond the scope of this paper,
a simplified approach was adopted, which does not yield an
absolute cost of a real-life industrial process, but provides a
reasonable comparison with earlier reported ILs based on
monoamines. Using prices available directly from the Merck
catalogue,23 and assuming virtually the same synthetic pro-
cedure for all ILs, the cost of each IL was estimated based on
the price for 1 L of an amine and 2.5 L of sulfuric acid. From
this, the price of 1 kg of each acidic IL was calculated
(Table 2).
Table 1 The composition of the mixtures investigated in the study
together with their decomposition temperatures
Working under the assumption that bulk quantities of
amines would be orders of magnitude cheaper, but relative
b
Mixture
SA : A ratioa
χH SO
Td (°C)
2
4
DMAPA_2_1
DMAPA_4_1
DMAPA_6_1
2 : 1
4 : 1
6 : 1
0.50
0.67
0.75
311
282
262
Table 2 Comparison of the laboratory-scale synthesis cost of various
acidic systems
API_2_1
API_4_1
API_6_1
2 : 1
4 : 1
6 : 1
0.50
0.67
0.75
316
285
250
Reagent price
($ per L)
Laboratory costb
($ per kg)
Reagent
Product
DETA_3_1
DETA_6_1
DETA_9_1
3 : 1
6 : 1
9 : 1
0.50
0.67
0.75
303
286
252
DMAPA
DETA
25.1
66.9
160
246
81
DMAPA_6_1
DETA_9_1
HMTA_9_1
MIM_3_1
TEA_3_1
21.0
24.6
51.2
67.2
42.9
HMTA
HMTA_3_1
HMTA_6_1
HMTA_9_1
3 : 1
6 : 1
9 : 1
0.50
0.67
0.75
288
285
252
MIMa
TEAa
Sulfuric acid
35.5
a Molar ratio of sulfuric acid (SA) to amine (A). b Molar ratio of sulfuric a MIM – 1-methylimidazole; TEA – triethylamine. b Mass of the pro-
acid to each amine functionality.
ducts was calculated from the density measured at 25 °C.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 4421–4429 | 4423

View Article Online
Paper
Green Chemistry
differences in pricing would hold, the relative differences in of fine-tuning the density, viz. systems based on DETA, with
the lab-scale cost of the ILs provide a useful insight.
density values approaching those of neat sulfuric acid.
In all cases, the amine is the cost-determining reagent in
Viscosities of all ILs were also measured at 25, 50 and
the mixture. However, in di- and triamines, the molar ratio of 80 °C, as a function of the sulfuric acid loading, χH2SO4 (S7,
the amine component would react with much more sulfuric ESI,† Fig. 3a). For each χH2SO4 value, ILs based on diamines
acid by mole (Table 1), and consequently by weight, while were less viscous than those based on triamines, due to an
retaining the ionic nature of the medium. In particular, increase in Coulombic interactions in triamine systems. The
DMAPA is available in large quantities, at a very low price, most dramatic change in viscosity occurred upon increasing
offering the potential for an ionic liquid much cheaper than the loading of sulfuric acid; for χH2SO4 = 0.50, all ILs were
the already inexpensive IL based on TEA (Table 2).
highly viscous (η = >600 mPa s at 80 °C, if measurable at all),
whereas mixtures with a high loading of sulfuric acid (χH2SO4
=
Physical properties
0.75) had viscosities nearly two magnitudes lower (η =
19–38 mPa s at 80 °C), which offers the perspective of a reason-
ably good mass transfer at these temperatures. Nevertheless,
even the least viscous IL, DMAPA_6_1, was much more viscous
than sulfuric acid (η = 104 vs. 23.8 mPa s at 25 °C).
The physical properties of ionic liquids are of key importance
owing to their use in catalytic processes. Density is a crucial
factor in separations, low viscosity is desired for good mass
transport and high thermal stability offers a wide window for
potential operating temperatures.
All these differences in viscosity can be explained in terms
of Coulombic interactions: with an increase in χH SO , negative
Densities of di- and tricationic ILs were measured at 25, 50
and 80 °C, as a function of the sulfuric acid loading, χH SO . All
2
4
charges are “diluted” by sulfuric acid included in the anionic
network surrounding the positively charged cations, which
reduces viscosity. Nevertheless, even for χH SO = 0.75, there is
2
4
data are tabularised in S7, ESI,† whereas a graphical compari-
son of densities at 80 °C is shown in Fig. 3b. In general,
density values ranged from ρ = 1.4 to 1.8 kg m−3, and
increased with an increase in the χH SO value, as expected
2
4
a highly ionic environment and strong Coulombic interactions,
manifesting themselves in viscosity more than four times
higher than that of sulfuric acid.
2
4
from the high density of sulfuric acid (ρ = 1.8 kg m−3 at 25 °C).
When comparing the density of diamines, DMAPA and API,
the latter systems were denser, most likely due to interactions
such as π–π stacking and hydrogen bonding between aromatic
imidazolium cations. Due to an interesting interplay between
the ratio of sulfuric acid (increasing density) and the contri-
bution of alkyl chains (decreasing density), for each χH SO
The thermal stability of all ionic systems was studied by
thermogravimetric analysis (TGA); TG curves are available in
S2 ESI.† Decomposition temperatures (Td) were determined as
the intersection point between the initial baseline and the
decomposition tangent line (Table 1). All ILs had good
thermal stability (Td ≥ 250 °C), which renders them suitable
for most liquid-phase catalytic applications. As expected,
thermal stability decreased somewhat upon increasing the
loading of sulfuric acid, suggesting the acid catalysed
decomposition of the cation.
2
4
value both the densest and the least-dense IL was based on a
triamine. For example, for χH SO = 0.75, densities were ρ =
2
4
1.58 kg m−3 for the HMTA mixture, and ρ = 1.79 kg m−3 for the
DETA mixture. This interplay opens up a very convenient way
Fig. 3 Viscosities (a) and densities (b) measured at 80 °C for di- and tricationic ILs.
4424 | Green Chem., 2021, 23, 4421–4429
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
For aliphatic systems, both an increase in the alkyl chain interesting question about the degree of protonation of each of
length (DETA, HMTA) and the presence of functional groups the amine sites, which were unfortunately outside of the scope
(DMAPA, DETA) decreased the thermal stability. A longer chain of this first paper, but could probably be answered by the com-
length increased the van der Waals interaction, consequently bination of more advanced NMR studies and neutron
decreasing the intramolecular interaction and thus lowering scattering.26,27
the thermal stability. Addition of an amine group also
decreased the thermal stability due to higher chemical reactiv-
Performance in catalysis
ity and easier decomposition of amino modified alkyl chains. An industrially relevant, model catalytic reaction was selected
Conversely, the presence of an imidazole ring increased the to validate the promising AN results: the esterification of fatty
thermal stability when compared to aliphatic cations, in line acids with methanol. The products of this reaction (a mixture
with earlier reports comparing the thermal stabilities of ionic of fatty acid methyl esters, FAMEs) are used in biodiesel, as a
+
liquids with [N1888
]
and [C8C1im]+ cations.24 Ultimately, sustainable, sulfur-free, biodegradable and non-toxic alterna-
however, these effects were only of significance for low acid tive to petroleum-based diesel.4 The process is conventionally
contents (χH SO = 0.50), and as a group the newly reported ILs operated under homogeneous catalysis conditions with strong
2
4
are characterised by reasonably high thermal stability, depend- alkali metal hydroxides or alkoxides, inducing potential sapo-
ing mainly on the ratio of sulfuric acid.
nification issues,28 which inspires the search for stable,
efficient, easily separable and inexpensive acid catalysts.
Tuneable ionic liquids have long been described as having
a great potential in biodiesel production, due to their ability to
act as both the catalyst and the solvent, which phase-separates
from the biodiesiel, but dissolves methanol and water, thereby
shifting the equilibrium of the reaction.29 Many acidic ionic
liquid systems have been explored in biodiesel production
from different feedstocks. They phase-separated directly in the
process or through precipitation after the process.30–32
Arguably, protic ionic liquids with acid in excess were the most
cost-effective medium,16,33 with the potential to catalyse esteri-
fication reactions at mild temperatures (25–60 °C). In this
work, esterification of oleic acid with methanol was chosen as
the model reaction.
Quantification of acidity
The acidity of the mixtures, expressed by the Gutmann accep-
tor number (AN), was determined by probing their strength to
protonate a weak base, triethylphosphine oxide (TEPO). The
strength of the interaction with TEPO was monitored by 31P
NMR spectroscopy, and compared to the chemical shift (δ
recorded for TEPO in hexane (AN = 0) and in 2 M solution of
SbCl5 (AN = 100).16,25
)
31P
The AN values (Table 3) were recorded for all investigated
systems except DETA_3_1 and HMTA_3_1, which were solid at
the temperature of the experiment (25 °C). Surprisingly,
systems where the sulfuric acid was equimolar to each amine
group (χH SO = 0.50) had a rather high acidity: AN = 113 and
2
4
110 for DMAPA and API, respectively, compared to ANCF COOH
3
Optimisation of the reaction conditions
= 105.5. This is in contrast with earlier findings for χH SO
=
2
4
The influence of catalyst loading, acid : methanol molar ratio,
reaction temperature and reaction time were studied using a
catalyst model system: HMTA_9_1 (Fig. 4).
0.50 mixtures of sulfuric acids with monoamines (2-picoline,
1-methylimidazole and N-methylpyrrolidine), which were
much milder acids (AN = 65.5, 73.3 and 96.5, respectively).16
Furthermore, for ILs based on diamines and triamines at
χH SO = 0.67, the AN values were in the range of 120–124, com-
As a general observation, at the beginning of each reaction,
all components of the reaction mixture (oleic acid, methanol
and the IL) formed a homogenous mixture. As the reaction
progressed, the reactants gradually formed a biphasic reaction
mixture: the bottom aqueous layer contained HMTA_9_1,
methanol and water from esterification, whereas the top
organic layer contained oleic acid and methyl oleate
(Scheme 1). The inclusion of water in the ionic liquid phase
provided an efficient method of shifting the reaction equili-
brium towards the product.
2
4
parable to the acidity of monoamine-based ILs at χH SO = 0.75.
2
4
Finally, the new ILs at χH SO = 0.75 had AN values of 124–126,
2
4
approaching that of methanesulfonic acid (ANCF SO H = 129.1).
3
3
These results prove that the acidities of ILs based on sulfu-
ric acid and diamines/triamines can be finely tuned by altering
the base and the sulfuric acid content. In addition, for
diamino and triamino protic IL mixtures, higher acidities can
be achieved at lower loading of sulfuric acid. This inspires an
Because the model reaction system is biphasic, the effect of
stirring was investigated first, under fixed reaction conditions
(10 : 1 methanol : oleic acid molar ratio, 60 °C, 5 mol% of the
catalyst), and three stirring speeds: 400, 1000 and 1400 rpm
(S8, ESI†). The reaction was limited by mass transfer at 400
rpm, but the results for 1000 and 1400 rpm were the same
across three sampling times, and therefore a stirring rate of
1000 rpm was selected for future experiments.
Table 3 AN values measured for acid : base mixtures, based on four
different bases and three mol ratios of H2SO4
AN
Amine precursor
χH SO = 0.5
χH SO = 0.67
χH SO = 0.75
2 4
2
4
2
4
DMAPA
API
DETA
HMTA
112.6
110.5
—
122.8
121.6
124.4
120.3
125.1
124.4
125.7
123.7
Subsequently, the influence of the catalyst loading was
studied, by altering the HMTA_9_1 quantity between 3 and
30 mol% with respect to oleic acid (Fig. 4a). It was found that
—
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 4421–4429 | 4425

View Article Online
Paper
Green Chemistry
Scheme 1 Experimental workflow of this study.
released water by the reduced quantify of the ionic liquid.
Nevertheless, at 5 mol% IL loading the conversion was very
close to that at 10 mol% loading, and therefore 5 mol% was
selected for further studies, taking into consideration the
economic aspect.
The influence of the reactant ratio was investigated by
varying the methanol : oleic acid molar ratio from 1 : 1 to 12 : 1
(Fig. 4b). The reaction with a 1 : 1 ratio proceeded at a very low
rate, reaching 61.4% yield after 5 h. Increasing the methanol
content (12 : 1) improved the shift of the reaction equilibrium,
resulting in full substrate conversion after 30 min. The most
significant improvement was recorded upon increasing the
methanol excess from 6 : 1 to 8 : 1 (Fig. 4b), and therefore the
8 : 1 ratio was used in all further experiments.
The influence of reaction temperature was studied within
the 30–60 °C range (Fig. 4c). The reaction could be carried out
even at very mild temperatures (30–40 °C), achieving the same
results as other acidic ILs at ca. 60–90 °C.32,34,35
Fig. 4 Yield of methyl oleate in the HMTA_9_1-catalysed esterification
reaction, as a function of the reaction time: (a) various catalyst loadings
(60 °C, methanol : oleic acid molar ratio: 10 : 1); (b) various methanol :
oleic acid molar ratios (60 °C, 5 mol% catalyst loading); and (c) various
reaction temperatures (5 mol% catalyst loading, methanol : oleic acid
molar ratio: 8 : 1).
Screening of the ionic liquid catalysts
Screening esterification experiments were performed for three
out of four ionic liquid systems, as it has been found that
DETA does not mix with methanol. In line with previous
decreasing the catalyst loading from 30 to 10 mol% increased reports,19,22 the activity of investigated systems depended
the yield from 70.4 to 99.1% after 30 min. This counter-intui- largely on the sulfuric acid content (χH SO ). All ILs with an
2
4
tive result can be justified in terms of a higher concentration excess of sulfuric acid (χH SO = 0.67 and 0.75 gave very high
2
4
of methanol at a lower dosage of HMTA_9_1, resulting in more yields (>87% after 1 h, 5 mol% loading and 8 : 1 reactant ratio),
favourable reaction kinetics. The reaction catalysed with con- whereas all the χH SO = 0.50 compositions gave ∼50% yields
2
4
centrated sulfuric acid (at 10 mol% loading) gave 88.1% yield, under the same reaction conditions. This general trend can be
compared to the 99.1% yield obtained using the HMTA_9_1 attributed to the lower acidities of χH SO = 0.50 ILs (Table 3).
2
4
catalyst. This can be attributed to the capability of the ionic
liquid to absorb water and thus shift the reaction equilibrium.
It has been reported that the relationship between AN and
conversion is described by a Boltzmann distribution: at a
Further lowering of the catalyst dosage to 5, and then to certain acidity level there was a dramatic increase in conver-
3 mol%, resulted in a decrease in the yield. It is thought that sion, and past this value, a further increase in AN brought no
this decrease is not due to an insufficient amount of acid cata- significant improvement.36 However, in this study, other
lyst, but rather due to the lower capacity for absorbing the factors were found to contribute significantly to the overall
4426 | Green Chem., 2021, 23, 4421–4429
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
Fig. 5 The influence of various di- and tricationic ILs on the yield of
methyl oleate. Reaction conditions: 60 °C, 5 mol% catalyst loading,
methanol : oleic acid molar ratio: 8 : 1, 1 h.
outcome: HMTA_9_1 gave by far the best performance (Fig. 5),
while having an AN value marginally lower than those of other
ILs at the same composition (Table 3).
Notably, among all χH SO
= 0.75 ionic liquids, the
2
4
HMTA_9_1 system had the lowest density and simultaneously
the highest viscosity, providing the best separation between
the ester phase and ionic mixture phase. In conclusion, the
overall outcome was reliant on both the acidity and favourable
physical properties of the catalytic IL.
The performance of the acid-rich HMTA system in the ester-
ification of the oleic acid was compared to the performance of
other reported Brønsted acidic ILs (Table 4). This extensive
comparison demonstrates that the HMTA system offers con-
siderable process advantages in terms of higher conversions
and shortened reaction times due to efficient water removal to
the ionic phase, thereby shifting the equilibrium of the reac-
tion toward the ester formation. It has been shown that
HMTA_9_1 enabled operation at exceptionally mild reaction
temperatures (30–40 °C).
Substrate scope
The versatile character of the ionic liquid systems was con-
firmed by esterification of various free fatty acids (FFAs),
including palmitic acid, stearic acid and lauric acid, with
methanol (Fig. 6 and S9, ESI†). Under the optimised reaction
conditions (8 : 1 methanol : acid molar ratio, 60 °C, 5 mol% of
HMTA_9_1 catalyst, 1 h) all esters formed a biphasic system,
ready for separation, and were isolated in high yields >93.1%.
Catalyst recycling
Recovery and recycling of the ionic liquid catalyst were studied
using HMTA_9_1 as the catalyst. Eight consecutive runs of
esterification of oleic acid in methanol (1 h per run) were
carried out. After each run a biphasic mixture was formed,
enabling easy product separation. The upper layer containing
the product was decanted, while the lower IL layer was washed
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 4421–4429 | 4427

View Article Online
Paper
Green Chemistry
acid and di- or triamines; this design led to a very high molar
excess of sulfuric acid when compared to the amine, up to
9 molecules of H2SO4 per molecule of amine. Even accounting
for the increased molecular mass of triamines, this translates
up to 80–90 mass% of an ionic liquid being sulfuric acid.
The ionic nature of the ILs promoted a very favourable
phase behaviour: absorbing water while phase-separating the
ester. As a consequence, the reaction equilibrium was shifted
towards products, the ester could be simply decanted and the
ionic liquid was easy to recover, overcoming the main chal-
lenges attributed to the use of neat sulfuric acid (which com-
monly dissolves in the reaction medium, inducing separation
and purification issues). The new IL mixtures have demon-
strated excellent performance (yields >99% in 1 h), taking
advantage of the continuous water removal to the IL phase,
while operating under very mild conditions (30–60 °C, atmos-
pheric pressure).
Fig. 6 Esterification of different FFAs, in the presence of HMTA_9_1.
Reaction conditions: 60 °C, 5 mol% catalyst loading, methanol : acid
molar ratio: 8 : 1, 1 h.
Conflicts of interest
There are no conflicts to declare.
Acknowledgements
This work was financed by the National Centre for Research
and Development, Poland (under the Leader program, grant
agreement LIDER/24/0100//L-9/17/NCBR/2018). A. B-S., J. W.,
A. S. and N. B. also acknowledge the National Agency for
Academic Exchange of Poland (under the Academic
International Partnerships program, grant agreement PPI/
APM/2018/1/00004) for financial support of the internship at
QUILL Research Centre, School of Chemistry and Chemical
Engineering, Queen’s University Belfast, which enabled the
execution of a part of the experimental work.
Fig. 7 The yield of methyl oleate in consecutive cycles of oleic acid
esterification with methanol, in the presence of an HMTA_9_1 system,
compared to the percentage recovery of the catalyst in the consecutive
cycles. Reaction conditions: 60 °C, 5 mol% catalyst loading, methanol :
oleic acid molar ratio: 8 : 1, 1 h.
Notes and references
1 J. A. Gladysz, in Pure and Applied Chemistry, Walter de
Gruyter GmbH, 2001, vol. 73, pp. 1319–1324.
2 D. Kralisch, D. Ott and D. Gericke, Green Chem., 2015, 17,
123–145.
3 M. Vafaeezadeh and H. Alinezhad, J. Mol. Liq., 2016, 218,
95–105.
4 F. Su and Y. Guo, Green Chem., 2014, 16, 2934–2957.
5 Z. Wu, Z. Li, G. Wu, L. Wang, S. Lu, L. Wang, H. Wan and
G. Guan, Ind. Eng. Chem. Res., 2014, 53, 3040–3046.
6 C. Fehér, E. Kriván, J. Hancsók and R. Skoda-Földes, Green
Chem., 2012, 14, 403–409.
with hexane to remove traces of the ester. Organic phases were
combined, and the solvent was removed. From the IL layer,
unreacted methanol and esterification water were removed by
vacuum distillation; then the ionic system was reused. After
five consecutive runs, the yield of the ester started to decline
slightly, but it could be attributed to the gradual mass loss of
the recycled IL, which occurred while washing with hexane
(Fig. 7). A further improvement in recyclability is probably a
chemical engineering challenge, rather than a chemical one.
7 Q. Zhang, J. Luo and Y. Wei, Green Chem., 2010, 12, 2246–
2254.
8 A. S. Amarasekara, Chem. Rev., 2016, 116, 6133–6183.
9 A. George, A. Brandt, K. Tran, S. M. S. N. S. Zahari,
D. Klein-Marcuschamer, N. Sun, N. Sathitsuksanoh, J. Shi,
V. Stavila, R. Parthasarathi, S. Singh, B. M. Holmes,
Conclusions
A group of new, very low cost, highly Brønsted acidic ionic
liquids has been developed. They are synthesised from sulfuric
4428 | Green Chem., 2021, 23, 4421–4429
This journal is © The Royal Society of Chemistry 2021

View Article Online
Green Chemistry
Paper
T. Welton, B. A. Simmons and J. P. Hallett, Green Chem., 25 V. Gutmann, The Donor-Acceptor Approach to Molecular
2015, 17, 1728–1734. Interactions, Plenum Press, New York, 1978.
10 L. Chen, M. Sharifzadeh, N. Mac Dowell, T. Welton, 26 M. Hasani, S. A. Amin, J. L. Yarger, S. K. Davidowski and
N. Shah and J. P. Hallett, Green Chem., 2014, 16, 3098–3106.
C. A. Angell, J. Phys. Chem. B, 2019, 123, 1815–
11 H. Baaqel, I. Díaz, V. Tulus, B. Chachuat, G. Guillén-
1821.
Gosálbez and J. P. Hallett, Green Chem., 2020, 22, 3132– 27 J. Stoimenovski, E. I. Izgorodina and D. R. MacFarlane,
3140. Phys. Chem. Chem. Phys., 2010, 12, 10341–10347.
12 U. Dorosz, N. Barteczko, P. Latos, K. Erfurt, E. Pankalla and 28 S. Nasreen, M. Nafees, L. A. Qureshi, M. S. Asad, A. Sadiq
A. Chrobok, Catalysts, 2019, 10, 37.
and S. D. Ali, in Biofuels - State of Development, InTech,
13 M. Przypis, K. Matuszek, A. Chrobok, M. Swadźba-Kwaśny
and D. Gillner, J. Mol. Liq., 2020, 308, 113166.
14 C. L. Chambon, V. Fitriyanti, P. Verdía, S. M. Yang,
S. Hérou, M. M. Titirici, A. Brandt-Talbot, P. S. Fennell and
2018.
29 M. J. Earle, N. V. Plechkova and K. R. Seddon, in Pure and
Applied Chemistry, De Gruyter, 2009, vol. 81, pp. 2045–
2057.
J. P. Hallett, ACS Sustainable Chem. Eng., 2020, 8, 3751– 30 J. Gao, Y. Zhu, W. Liu, S. Jiang, J. Zhang and W. Ma, ACS
3761. Omega, 2020, 5, 12110–12118.
15 M. Swadźba-Kwaźny, J. D. Holbrey, T. Youngs, R. Guiney, 31 D. Fang, J. Yang and C. Jiao, ACS Catal., 2011, 1, 42–47.
E. L. Byrne and A. McGrogan, Liquid structure of sulfuric 32 J. Chen, M. Li, M. Li, X. Lin and T. Qiu, ACS Sustainable
acid and derived protic ionic liquids, STFC ISIS Facility, 2020,
DOI: 10.5286/ISIS.E.RB2010681.
16 K. Matuszek, A. Chrobok, F. Coleman, K. R. Seddon and
M. Swadźba-Kwaśny, Green Chem., 2014, 16, 3463–3471.
Chem. Eng., 2020, 8, 6956–6963.
33 P. Y. S. Nakasu, P. Y. S. Nakasu, C. J. Clarke, S. C. Rabelo,
A. C. Costa, A. Brandt-Talbot and J. P. Hallett, ACS
Sustainable Chem. Eng., 2020, 8, 7952–7961.
17 A. S. Khan, Z. Man, M. A. Bustam, A. Nasrullah, Z. Ullah, 34 F. Yang, J. Zhang, Y. Guo, L. Feng, S. Wei and S. Chen,
A. Sarwono, F. U. Shah and N. Muhammad, Carbohydr.
Polym., 2018, 181, 208–214.
18 M. P. Atkins, M. J. Earle, K. R. Seddon, M. Swadźba-Kwaśny
and L. Vanoye, Chem. Commun., 2010, 46, 1745–1747.
19 A. S. Khan, Z. Man, M. A. Bustam, C. F. Kait, A. Nasrullah,
Chem. Eng. Technol., 2019, 42, 174–181.
35 F. F. Roman, A. E. Ribeiro, A. Queiroz, G. G. Lenzi,
E. S. Chaves and P. Brito, Fuel, 2019, 239, 1231–1239.
36 K. Matuszek, S. Coffie, A. Chrobok and M. Swadźba-
Kwaśny, Catal. Sci. Technol., 2014, 7, 1045.
Z. Ullah, A. Sarwono, P. Ahamd and N. Muhammad, 37 Y. Li, S. Hu, J. Cheng and W. Lou, Chin. J. Catal., 2014, 35,
J. Cleaner Prod., 2018, 170, 591–600. 396–406.
20 T. Oncsik, R. Vijayaraghavan and D. R. Macfarlane, Chem. 38 Y. Zhao, J. Long, F. Deng, X. Liu, Z. Li, C. Xia and J. Peng,
Commun., 2018, 54, 2106–2109. Catal. Commun., 2009, 10, 732–736.
21 R. Vijayaraghavan, T. Oncsik, B. Mitschke and 39 T. Chang, L. He, X. Zhang, M. Yuan, S. Qin and J. Zhao,
D. R. MacFarlane, Sep. Purif. Technol., 2018, 196, 27–31. Chin. J. Catal., 2015, 36, 982–986.
22 H. Zhang, B. Jiang, N. Yang, N. Zhang, L. Zhang, Z. Huang, 40 B. Aghabarari, N. Dorostkar, M. Ghiaci, S. G. Amini,
X. Xiao and X. Tantai, Energy Fuels, 2019, 33, 8937–8945.
23 Sigma Aldrich, https://www.sigmaaldrich.com/united-
E. Rahimi and M. V. Martinez-Huerta, J. Taiwan Inst. Chem.
Eng., 2014, 45, 431–435.
states.html, (accessed 19 January 2021).
24 P. S. Kulkarni, L. C. Branco, J. G. Crespo, M. C. Nunes,
41 X. Lin, X. Ling, J. Chen, M. Li, T. Xu and T. Qiu, Green
Chem., 2019, 21, 3182–3189.
A. Raymundo and C. A. M. Afonso, Chem. – Eur. J., 2007, 42 A. H. M. Fauzi and N. A. S. Amin, Energy Convers. Manage.,
13, 8478–8488.
2013, 76, 818–827.
This journal is © The Royal Society of Chemistry 2021
Green Chem., 2021, 23, 4421–4429 | 4429