Unveiling the dual role of the cholinium hexanoate ionic liquid as solvent and catalyst in suberin depolymerisation

Article abstract of DOI:10.1039/c3ra45910a

Disruption of the three-dimensional network of suberin in cork by cholinium hexanoate leads to its efficient and selective isolation. The reaction mechanism, which likely involves selective cleavage of some inter-monomeric bonds in suberin, was still unanswered. To address this question, the role of the ionic liquid during suberin depolymerisation and during cleavage of standard compounds carrying key chemical functionalities was herein investigated. A clear demonstration that the ionic liquid catalyses the hydrolysis of acylglycerol ester bonds was attained herein, both experimentally and computationally (DFT calculations). This behaviour is related to cholinium hexanoate capacity to activate the nucleophilic attack of water. The data showed also that the most favourable reaction is the hydrolysis of acylglycerol ester bonds, with the C2 position reporting the faster kinetics, whilst most of the linear aliphatic esters remained intact. The study emphasises that the ionic liquid plays the dual role of solvent and catalyst and leads to suberin efficient extraction through a mild depolymerisation. It is also one of the few reports of ionic liquids as efficient catalysts in the hydrolysis of esters.

Full text of DOI:10.1039/c3ra45910a

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Unveiling the dual role of the cholinium hexanoate
ionic liquid as solvent and catalyst in suberin
depolymerisation†
Cite this: RSC Adv., 2014, 4, 2993
Rui Ferreira,‡a Helga Garcia,‡a Andreia F. Sousa,bc Marina Guerreiro,a
Filipe J. S. Duarte,d Carmen S. R. Freire,b Maria Jose Calhorda,d
´
Armando J. D. Silvestre,b Werner Kunz,e Luıs Paulo N. Rebeloa and Cristina Silva
´
a
*
Pereira
Disruption of the three-dimensional network of suberin in cork by cholinium hexanoate leads to its efficient
and selective isolation. The reaction mechanism, which likely involves selective cleavage of some inter-
monomeric bonds in suberin, was still unanswered. To address this question, the role of the ionic liquid
during suberin depolymerisation and during cleavage of standard compounds carrying key chemical
functionalities was herein investigated. A clear demonstration that the ionic liquid catalyses the hydrolysis
of acylglycerol ester bonds was attained herein, both experimentally and computationally (DFT
calculations). This behaviour is related to cholinium hexanoate capacity to activate the nucleophilic
attack of water. The data showed also that the most favourable reaction is the hydrolysis of acylglycerol
ester bonds, with the C2 position reporting the faster kinetics, whilst most of the linear aliphatic esters
remained intact. The study emphasises that the ionic liquid plays the dual role of solvent and catalyst and
leads to suberin efficient extraction through a mild depolymerisation. It is also one of the few reports of
ionic liquids as efficient catalysts in the hydrolysis of esters.
Received 17th October 2013
Accepted 18th November 2013
DOI: 10.1039/c3ra45910a
www.rsc.org/advances
Suberin builds an apoplastic barrier that controls the ow of
water, gases and ions and protects against biological pathogens
and physical aggressions.9–11
Introduction
Renewable resources are increasingly regarded as substituents
for highly demanded petroleum-based chemicals.1 In this
context, the extraction, application and biosynthesis of plant
biopolyesters, namely cutin and suberin, have been extensively
studied.2–9 Suberin is ubiquitous in higher plants, and particu-
larly abundant in the outer barks of Quercus suber and Betula
pendula and in the peel of Solanum tuberosum. This hydro-
phobic polymeric material is deposited in the secondary cell
wall in the internal and the peripheral dermal tissues during
cell wall differentiation or as a response to stress and wounding.
Since the earliest descriptions, suberin is known to be
composed of aromatic and aliphatic monomeric units.12–14
Currently, the prevailing idea is that suberin comprises two
covalently linked domains, the major being the polyaliphatic
domain and the minor the polyphenolic one.10,15–18 Suberin is
generally accepted to be organised in situ as a lamellar type
structure,16–19 notwithstanding that some structural aspects are
still unanswered. The polyphenolic domain is ingrained on the
inner face of the primary cell wall.10,16–18,20,21 Especially due to its
high recalcitrancy, the compositional structure of this lignin-
like domain is not yet fully understood.22,23 Even so, it is known
to be composed mainly by hydroxycinnamic acids and their
derivatives (predominantly ferulic acid) and some vestigial
amounts of monolignols (viz. p-coumaryl, coniferyl and sinapyl
alcohols). Extensive cross-linking between these aromatic
monomers in the lignin-like structure, as well as to other cell
wall constituents, is done via stable carbon–carbon, amide and
ether bonds.9,23–25
a
´
´
Instituto de Tecnologia Quımica e Biologica, Universidade Nova de Lisboa, Apartado
127, 2781-901, Oeiras, Portugal. E-mail: spereira@itqb.unl.pt
bCICECO and Department of Chemistry, University of Aveiro, Campus de Santiago,
3810-193 Aveiro, Portugal
c
´
Department of Chemical Engineering, University of Coimbra, Polo II, Pinhal de
Marrocos, 3030-290 Coimbra, Portugal
d
´
´
ˆ
Departamento de Quımica e Bioquımica, CQB, Faculdade de Ciencias, Universidade
de Lisboa, 1749-016 Lisboa, Portugal
eInstitute of Physical and Theoretical Chemistry, University of Regensburg, D-93040
Regensburg, Germany
The polyaliphatic domain is composed mostly of long chains
(C16–C26) of alkanols, alkanoic acids, u-hydroxyalkanoic acids
and a,u-alkanedioic acids (and the corresponding mid-chain
unsaturated, epoxy or vic-diol derivatives) and glycerol. In
addition, the deposition of hydroxycinnamates leads to the
† Electronic supplementary information (ESI) available: Fig. S1 (thermal
microscopy analysis); Table S1 (GC-MS data) and S3 (cartesian coordinates and
computed total energies). See DOI: 10.1039/c3ra45910a
‡ Equally contributing authors.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 2993–3002 | 2993

View Article Online
RSC Advances
Paper
typical lamellar organisation of suberin of alternate aliphatic
and phenolic components.16 These monomers are in a parallel
alignment and linked via linear aliphatic ester or acylglycerol
ester bonds. Glycerol is a key cross-linker in the formation of a
three-dimensional network, connecting hydrophilic moieties
and both suberin domains.16,25
Knowledge on suberin is spread in diverse scientic disci-
plines, from biology, e.g. the understanding of the polymer
biosynthesis,4,7 to chemistry, e.g. the depolymerisation and
characterisation of suberin.17 Although only modestly exploi-
ted so far, suberin has been regarded as a source of monomers
and oligomers for the synthesis of novel macromolecular
materials.9,26–29 Of particular relevance is the abundance of
mid-chain hydroxy and epoxy fatty acids in suberin. These are
rare in other renewable resources and difficult to synthesise
chemically.
Suberin extraction from renewable resources is convention-
ally attained using alkaline hydrolysis30 or alkaline meth-
anolysis.17,31 These methods result in extensive ester bond
cleavage, i.e. depolymerisation. Although with very low extrac-
tion efficiency, comparable methods can be used to promote a
partial depolymerisation.17 An alternative extraction method,
using cholinium hexanoate, promotes efficient extraction of a
still partially cross-linked and highly polymerised suberin,
regardless of the renewable resource used, viz. cork,32–34 birch
outer bark34 and potato peel (unpublished work). The unique-
ness of such method is highlighted by the compositional
structure of suberin, hypothesised to be intimately related to
that of the in situ suberin.32 Emphasis should be also given to
the high biocompatibility and biodegradability of the ionic
liquid used in the extraction,35 which can be recycled and
reused throughout the process.32
Suberin depolymerisation induced by cholinium hexanoate
is likely to involve the selective cleavage of some bonds linking
suberin composing monomers. This is the underlying question
of the present study: to solve the chemical reaction mechanism
behind suberin depolymerisation in cholinium hexanoate
media. To address this question, we further investigated the
lability of key chemical functionalities of suberin and of stan-
dard compounds in the ionic liquid media. The chemical
mechanism proposed herein was also supported by DFT
calculations.
Chemicals
Cholinium hexanoate was synthesised by dropwise addition of
hexanoic acid to aqueous cholinium hydrogen carbonate
(Sigma ꢀ80% in water) in equimolar quantities, as previously
described.35 The ionic liquid purity was veried by 1H- and 13C-
NMR, CHNS elemental analysis and electrospray ionisation
mass spectrometry (ESI-MS). The ionic liquid was dried prior to
use by stir-heating in vacuum (40–50 ꢁC, ca. 0.01 mbar). The
water content, determined by Karl-Fischer titration, was ca.
0.2 wt%.
Poly(12-hydroxydodecanedioic acid) was prepared as
described before29 and 9-10-epoxy-18-hydroxyoctadecanoic acid
was extracted from Betula pendula.37
Octyl octanoate ($98%), glyceryl trioctanoate ($99%),
dimethyl sulfoxide (DMSO, 99.5%), sodium hydroxide (>97%)
and dichloromethane ($99.5%) were purchased from Sigma;
glycerol ($99.5%) from VWR, L-lactic acid (99.5%) from Fluka
and poly(lactic acid) with an L : D ratio of 96 : 4 and an Mw of
110 000 g molꢂ1 from Cargill-Dow Polymers.
Suberin extraction
The suberin extraction process followed a methodology which
has been previously described.32,34 Briey, the cholinium hex-
anoate (melting temperature, 60.57 ꢁC) was mixed with cork
(ionic liquid : cork z 9 : 1 wt/wt) and kept at 100 ꢁC during 1, 2,
4 or 8 hours, with stirring (each in triplicate). At the end of the
extraction process, DMSO was added to reduce the viscosity of
the mixture,32 facilitating its ltration through
a nylon
membrane with an exclusion pore size of 1.0 mm (Millipore, MA,
USA). The insoluble residue was then washed thoroughly with
an excess of water at 80 ꢁC. Precipitation of the extracted
suberin was obtained by keeping the ensuing ltrate (i.e. ionic
liquid, suberin, DMSO and the water added to wash the insol-
uble residue) at 4 ꢁC for 1 hour. Suberin was then recovered by
centrifugation (30 min at 4 ꢁC and 2450 ꢃ g), washed twice with
an excess of water and dried at 50 ꢁC, until constant weight was
attained. The aqueous phase (supernatant) was also concen-
trated in a rotary evaporator and its content in glycerol analysed
by high performance liquid chromatography (HPLC).
Attenuated total reectance-Fourier transform infrared
spectroscopy (ATR-FTIR)
ATR-FTIR spectra were collected on a Bruker IFS66/S FTIR
spectrometer (Bruker Daltonics, MA, USA) using a single
reection ATR cell (DuraDisk, equipped with a diamond
crystal). Data were recorded at room temperature, in the range
of 4000–600 cmꢂ1, by accumulating 128 scans with a resolution
of 4 cmꢂ1. Five replica spectra were collected for each sample in
order to evaluate reproducibility (OPUS v6.0).
Experimental
Cork
Granulated cork was obtained from the cork producers Amorim
˜
& Irmaos SA (Santa Maria de Lamas, Portugal). The samples
were ground to a ne powder (60 mesh) using a centrifuge mill
(Retsch) and the cork extractives removed by sequential Soxhlet
extraction with solvents of increasing polarity (dichloro-
methane, ethanol and water) as previously described by Gil
Nuclear magnetic resonance spectroscopy (NMR)
et al.36 The extractive-free cork powder, hereinaer dened 1H- and 13C-NMR analyses were recorded with a Bruker Avance
solely as cork, was further washed with an excess of deionised 400 Ultrashield Plus spectrometer. Spectra were run at 25 ꢁC
water for complete removal of low molecular weight using standard Bruker pulse programs. 13C CP/MAS NMR
compounds, and dried prior to use.
spectra were recorded at 9.4 T on a Bruker 400 spectrometer
2994 | RSC Adv., 2014, 4, 2993–3002
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
using 9 kHz spinning rate and MAS with proton 90ꢁ pulses of 100 ꢁC, with stirring (experimental triplicates and technical
4 ms. Chemical shis are given in ppm from glycine. The NMR duplicates).
spectra were processed and analysed with MestreNova v6.0
(MestreLab Research S.L.).
Glyceryl trioctanoate and octyl octanoate were mixed with
cholinium hexanoate at a 1.5 ratio of ionic liquid moles per
mole of ester bonds, during 1, 2, 4 or 8 hours. At the end of the
test, aer partition of the ensuing mixture in dichloromethane/
water, the organic layer was recovered, concentrated in a rotary
evaporator, dried at 50 ꢁC, and then analysed by GC-MS as
described in GC-MS – Method 1. The aqueous phases were also
concentrated in a rotary evaporator and their content in glycerol
analysed by HPLC.
The remaining standard compounds, namely 9,10-epoxy-18-
hydroxyoctadecanoic acid, poly(12-hydroxydodecanoic acid)
and poly(lactic acid) were mixed with cholinium hexanoate for 4
hours, using a mass ratio (standard: cholinium hexanoate) of
1 : 4, 1 : 3, 1 : 10, 1 : 10 and 1 : 10, respectively. For the rst two
compounds, at the end of the test, an excess of water was added,
the precipitate recovered by centrifugation, dried and analysed
by ATR-FTIR and/or NMR. In the test with poly(lactic acid), aer
the reaction, an excess of water was added, the insoluble residue
recovered by ltration, dried and then analysed by ATR-FTIR.
The content of lactic acid in the aqueous phase was determined
by HPLC.
Gas chromatography-mass spectrometry (GC-MS)
A Trace GC 2000 Series gas chromatograph equipped with a
Thermo Scientic DSQ II mass spectrometer was used. The GC-
MS was rst calibrated with pure reference compounds (repre-
sentative of the major classes of compounds present in suberin)
relative to n-hexadecane (internal standard). Compounds
identication was based on the equipment spectral library
(Wiley-Nist) and on previously published data, focussing their
EI-MS fragmentation patterns and/or retention times.30,31,38,39
Replicates were done to guarantee low variability and each
analysis repeated twice. Each suberin sample was analysed by
two complementary methods:
- Method 1, samples were converted to the corresponding
trimethylsilyl (TMS) derivatives and analysed by GC-MS. In
brief, suberin samples (ca. 15 mg) were reacted with 250 mL of
N,O-bis(trimethylsilyl)triuoroacetamide and 50 mL of trime-
thylchlorosilane in 250 mL of pyridine, during 30 min at 70 ꢁC.40
- Method 2, samples were submitted to alkaline hydrolysis
prior to the silylation, to release hydrolysable monomeric
constituents. Briey, suberin samples were treated with a
Thermal microscopy
ꢁ
solution of 0.5 M NaOH in methanol/water (1 : 1, v/v) at 95 C,
Thermal microscopy analyses were done using a Leitz Ortho-
plan polarizing microscope (Wetzlar, Germany) equipped with a
JVC digital camera (TK-C130), a Linkham hot stage with a
TMS90 temperature controller (ꢄ0.5 ꢁC) and a CS196 cooling
system.
during 4 hours.41 The mixture was cooled to room temperature,
acidied to pH 3–3.5 with 1 M HCl, and extracted three times by
dichloromethane/water partition. The combined organic
extracts were dried in a rotary evaporator, then trimethylsily-
lated as mentioned above, and analysed by GC-MS. The aqueous
phases were also concentrated in a rotary evaporator and their
content in glycerol analysed by HPLC.
Computational methods
Density functional theory (DFT)42,43 calculations were performed
with the Gaussian 09 soware package44 with the hybrid func-
tional PBE1PBE,45–48 and the 6-31 + G** basis set. Full geometry
optimisations including solvent effects (DMSO) were carried
out, using the polarisable continuum model (PCM).49 The
starting structures are based in the most stable approach
geometry of the ionic liquid and reactants. Harmonic vibra-
tional frequencies were calculated for all located stationary
structures to verify whether they were minima or transition
states. Zero-point energies and thermal corrections were taken
from unscaled vibrational frequencies. Free energies of activa-
High performance liquid chromatography
Samples were analysed by HPLC, using a Waters chromatogra-
pher, consisting of a 510 Pump, a 715 Autosampler and a
Temperature Control Module (Waters Chromatography, Mil-
ford, MA, USA), connected to a LKB 2142 Differential Refrac-
tometer detector (Bromma, Sweden). Data acquisition was
accomplished with the Millenium32 v3.05.01, 1998 system
(Waters). Chromatographic separation was undertaken at 60 ꢁC
using an Aminex HPX-87H column (300 ꢃ 7.8 mm) with 9 mm
particle size (Bio-Rad, Hercules, California). Elution was carried
out isocratically, at a ow rate of 0.5 mL minꢂ1, with 0.005 N of
H2SO4 and the sample volume injected was 20 mL. Lactic
acid (retention time ¼ 15.98 min) and glycerol (retention
time ¼ 15.20 min) quantications were done using an
external calibration curve with high purity standards and
within the quantication limits of 0.25–20.41 mg mLꢂ1 and
0.25–9.99 mg mLꢂ1, respectively.
ꢁ
tion are given at 25 C. All energies were calculated relative to
˚
the reagents. All bond lengths are in angstrom (A) and energies
in kJ molꢂ1
.
Results and discussion
The cholinium-hexanoate-based method for suberin extraction
combines partial depolymerisation of suberin with high selec-
tivity and extraction efficiency.32–34 To identify the chemical
bonds prone to be cleaved in the cholinium hexanoate media,
Reaction of standard compounds in cholinium hexanoate
media
Tests of cholinium hexanoate ability to cleave standard we carried out a comprehensive study of the lability of suberin
compounds were done in the same conditions used for cork, i.e. and of standard compounds carrying the key structural bonds
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 2993–3002 | 2995

View Article Online
RSC Advances
Paper
in suberin, namely linear aliphatic esters and acylglycerol suggested to report slightly lower mobility in comparison to those
esters, as depicted schematically in Fig. 1.
at d 30 ppm.31 This has been related to the presence of methylene
groups near oxygen vicinities (suggestive of linkage to lignin and
polysaccharides),31 or to the presence of crystalline/recalcitrant
domains.50,51 Other resonances typical of suberin can be noticed
in all spectra, namely signals at d 50–90 ppm, d 130 and 148 ppm,
assigned to carbons linked to oxygen, vinylic and quaternary
carbons, respectively. One cannot disregard the hypothesis that
cholinium hexanoate contamination contributed to the reso-
nance at d 54 ppm (more intense in the suberin samples extracted
for one or two hours, see * in Fig. 2A).
The ATR-FTIR spectra of the suberin samples show two
major peaks at 2918, 2851 cmꢂ1 and a band between 3679 and
3034 cmꢂ1 (Fig. 2B). These are respectively associated to the
aliphatic chains and to the hydroxyl groups in carboxylic acids
and/or alcohols. The hydroxyl band increased along the
extraction time. This can be related to the release of hydroxyl
moieties, which reinforce the idea that suberin depolymerisa-
tion was continuous along the course of the extraction.
More information on the monomeric composition of suberin
along the extraction time can be retrieved using GC-MS analysis.
The underlying concept here is that suberin extracted with
cholinium hexanoate is still cross-linked and/or oligomeric.
Consequently, the high molecular weight of these structures
hampers their detection by GC-MS (see GC-MS – Method 1,
Experimental section). In agreement, very low identication
yields were attained (ca. 2–5 wt%), notwithstanding that the
main typical suberin monomers could be detected (data not
shown). Monoacylglycerol dimers were only detected in
the suberin samples recovered aer the rst two hours of
the extraction, reinforcing the idea that, until the end of the
extraction time, oligomers released to the ionic liquid media
were further depolymerised.
Suberin depolymerisation occurs through cleavage of ester
bonds: time-course analysis
The natural abundance of suberin in cork is typically ca. 50–60
wt%.10 The extraction of suberin from cork with cholinium
hexanoate was extremely fast, given that aer one hour the
recovery yield was 46.8 wt%. However, aer this phase the
suberin recovery yield showed a minor increase over time,
reaching 54.7 wt% at the eighth hour of extraction (Table 1).
Thermal microscopy analysis showed a rapid disruption of
cork during the rst hour of the extraction. No signicant
differences were observed for longer periods. Suberin extraction
was probably facilitated by this disruption, which exposes the
cell wall components to the ionic liquid phase (Fig. S1†).
Aiming to identify which chemical bonds in suberin were
cleaved in the cholinium hexanoate media, structural compo-
sition analyses of the samples recovered along the extraction
time were undertaken. Aer eight hours of extraction, the ester
features, namely the resonance at d 173 ppm (13C CP/MAS NMR,
Fig. 2A) and the carbonyl stretching at 1731 cmꢂ1 (ATR-FTIR,
Fig. 2B), were still detected in the corresponding spectra. The
intensity of the ester resonance at d 173 ppm decreased along
the extraction time, suggesting ester bond cleavage (ATR-FTIR,
Fig. 2B). In addition, the peak at 1715 cmꢂ1, which was barely
distinguishable in the rst hours of the extraction, for longer
periods became a dened peak. This peak is associated with
hydrogen bonded carbonyl groups in either esters or acids (see
magnication in Fig. 2B), corroborating the idea that ester bond
cleavage occurred. It becomes apparent that the ionic liquid led
to mild depolymerisation of suberin through continuous ester
bond cleavage.
Aiming to overcome the low GC-MS identication yields, any
hydrolysable chemical bonds in the ionic liquid extracted
suberin were cleaved by conventional alkaline hydrolysis before
the analysis (see GC-MS – Method 2, Experimental section).
Consequently, the identication yields of these samples were
signicantly higher, ca. 50 wt% (Table S1†). In general, the
Other important observations in the 13C CP/MAS NMR
spectra of the recovered suberin samples are the two major
resonances at d 33 and 30 ppm which are associated with
aliphatic methylenes (Fig. 2A). The ratio 33/30 ppm increased
over the extraction time. The carbons at d 33 ppm have been
Fig. 1 Schematic view of the experiments undertaken to determine the susceptibility of ester bonds and epoxy rings to undergo cleavage in the
cholinium hexanoate media. Cholinium hexanoate tests with cork (top) and with standard compounds (bottom). The standard compounds
reproduce main chemical functionalities present in suberin, namely acylglycerol ester bonds (A), aliphatic ester bonds (B) and epoxy rings (C).
2996 | RSC Adv., 2014, 4, 2993–3002
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
Table
1
Analysis of suberin extracted from cork by cholinium hexanoate. Values depicted correspond to suberin recovery yield
[100(mrecovered suberin/mcork)], solubility in dichloromethane [100(msoluble suberin/msuberin)] and alkaline hydrolysis recalcitrancy [100 ꢂ (100(mhydrolysable
suberin/msuberin)] and to the quantification of the glycerol released during suberin depolymerisation, namely in the cholinium hexanoate media
(glycerol[ChHex]) and by alkaline hydrolysis (glycerol[NaOH]). The second quantifies the hydrolysable glycerol that remained in suberin after the depo-
lymerisation of suberin in the ionic liquid media. Values in brackets stand for standard deviation between replicates
Time/h
1
2
4
8
Recovery yield/wt%
Solubility in CH2Cl2/wt%
Alkaline hydrolysis recalcitrancy/wt%
46.8 (4.5)
46.2 (0.3)
34.4 (3.9)
47.3 (0.9)
48.0 (1.1)
37.7 (12.4)
51.0 (5.8)
55.7 (2.9)
49.0 (7.8)
54.7 (14.8)
65.2 (1.7)
46.3 (0.1)
Glycerol mgglycerol/gsuberin
a
Glycerol[ChHex]
HPLC
GC-MS
Total
HPLC
GC-MS
Total
17.4 (3.0)
4.3 (0.4)
21.7
34.9 (4.5)
0.8 (0.3)
35.7
18.6 (3.1)
5.5 (2.3)
24.1
32.2 (3.2)
0.7 (0.2)
32. 9
51.1 (4.1)
2.8 (1.0)
53.9
7.6 (1.9)
0.4 (0.3)
8.0
63.0 (21.2)
2.7 (2.5)
65.7
0.0 (0.0)
0.3 (0.1)
0.3
b
Glycerol[NaOH]
Total glycerol
57.4
57.0
61.9
66.0
a
Glycerol[ChHex], glycerol released during suberin depolymerisation in the cholinium hexanoate media, calculated as the total of the glycerol
detected by HPLC (i.e. solubilised in water during the ltration and the precipitation steps) and by GC-MS (i.e. present in the sample).
b
Glycerol[NaOH], glycerol released during alkaline hydrolysis of suberin samples, calculated as the total of the glycerol detected by HPLC (i.e.
solubilised in the aqueous phase during dichloromethane/water partition aer the alkaline hydrolysis) and by GC-MS (in the sample aer
alkaline hydrolysis).
Fig. 2 13C CP/MAS NMR spectra (A) and ATR-FTIR spectra (B) of suberin extracted with cholinium hexanoate after 1, 2, 4 or 8 hours.
amounts of the monomers herein identied were coherent with recovered by dichloromethane/water partition, consequently
those typically reported for suberin extracted with conventional most of the hydrolysed glycerol remains in the aqueous
methods.17,30,31 Major compounds identied were u-hydroxy- phase. In fact, the total glycerol obtained herein is comparable
alkanoic acids, followed by a,u-alkanedioic acids and pheno- to that reported for conventionally extracted suberin
lics, and minor ones were alkanoic acids and alkan-1-ols (Fig. 3, (ca. 6 wt%, Table 1).52,53
Table S1†). Similar to previous reports,17 the most abundant
Suberin depolymerisation with cholinium hexanoate
monomers herein detected were 22-hydroxydocosanoic acid, released increasing quantities of glycerol, from 21.7 to 65.7
followed by ferulic acid, 9,10,18-trihydroxyoctadecanoic acid mgglycerol/gsuberin (glycerol[ChHex], Table 1). On the contrary, the
and 9,10-dihydroxyoctadecanedioic acid.
glycerol released from suberin during alkaline hydrolysis
Glycerol is a major component of suberin,10,17,52,53 however progressively decreases to zero along the extraction time
its high water solubility justies that the quantities detected (glycerol[NaOH], Table 1). This means that aer eight hours,
by GC-MS, either before or aer alkaline hydrolysis of suberin, virtually all hydrolysable acylglycerol ester bonds present in
were very low (Table 1). Most glycerol released from suberin, suberin were cleaved in the ionic liquid media. However, signif-
due to extraction with cholinium hexanoate, becomes solubi- icant amounts of suberin monomers were still released aer
lised in water during the ltration and the precipitation steps. alkaline hydrolysis. Both ester bonds present in suberin, viz.
Aer alkaline hydrolysis the hydrolysed suberin fraction is linear aliphatic and acylglycerol, are labile to the alkaline
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 2993–3002 | 2997

View Article Online
RSC Advances
Paper
from the cell wall to the ionic media. Along the eight hours of
extraction, suberin oligomeric structures in the ionic liquid
media undergo further depolymerisation and glycerol is
continuously released. In addition, few non-hydrolysable
suberin oligomeric structures were also removed from the cell
wall to the ionic liquid media. Aer eight hours of extraction,
linear aliphatic ester bonds were still preserved but most acyl-
glycerol ones were cleaved.
Preferential cleavage of acylglycerol esters in standard
compounds
Fig. 3 Quantification of suberin monomers in the samples recovered
after depolymerisation in cholinium hexanoate media and submitted
to alkaline hydrolysis prior to the GC-MS analysis (GC-MS – Method 2).
Detailed monomeric quantification in Table S1.†
Apparently cholinium hexanoate promoted efficient cleavage of
the acylglycerol ester bonds in suberin, but not of the linear
aliphatic ester bonds (see computational study below). To further
support these ndings, the ionic liquid ability to cleave standard
compounds carrying linear aliphatic or acylglycerol ester bonds
was analysed by monitoring along time the reaction of glyceryl
trioctanoate or octyl octanoate, respectively (Fig. 1). These stan-
dard compounds showed different reaction kinetics (Fig. 5). Octyl
octanoate was fairly resistant to cleavage. Aer eight hours of
reaction only ca. 30 wt% was hydrolysed to octanoic acid and
octanol. On the contrary, aer the same reaction time, more than
80 wt% of the glyceryl trioctanoate was cleaved, releasing glyc-
erol, octanoic acid and the corresponding diacyl- and monoacyl-
glycerols (Fig. 5). Based on the low amounts of 2-mono-
acylglycerol released, it is likely that the most favourable cleavage
occurred in the acyl group at position C2 of glycerol. This is
suggestive of possible steric hindrance and/or chemical selec-
tivity in the reaction promoted by the ionic liquid. This partially
explains why amongst the few monomers identied by GC-MS in
the suberin samples extracted with cholinium hexanoate none
was a 2-monoacylglycerol.32,34
hydrolysis. Thus, it seems that the ionic liquid is efficient towards
acylglycerol ester bonds, but fails to cleave most of the linear
aliphatic esters bonds.
Suberin solubility in dichloromethane and its recalcitrancy
to alkaline hydrolysis increased over the extraction time, from
46.2 to 65.2 wt% and from 34.4 to 46.3 wt%, respectively for
samples extracted for two and eight hours (Table 1). Upon rapid
removal from the cell wall to the ionic liquid media, the suberin
oligomeric structures undergo continuous depolymerisation –
hence solubility in dichloromethane increases. At the same
time, non-hydrolysable suberin oligomeric structures are
progressively removed from the cell wall – hence recalcitrance
increases. All aforementioned data was integrated to propose a
model for suberin extraction from cork in cholinium hexanoate
media (Fig. 4). In brief, aer one hour most suberin oligomeric
structures, except few non-hydrolysable ones, were removed
Fig. 4 Schematic view of how cholinium hexanoate leads to suberin depolymerisation along the extraction time. S, suberin deposited in the cell
wall; IL Media, Ionic liquid media.
2998 | RSC Adv., 2014, 4, 2993–3002
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
glyceryl trioctanoate and octyl octanoate (Fig. 5 and S4†), as well
as from suberin32,34 (GC-MS Method 1, data not shown). This
brings further light to this process: the ester bonds were
hydrolysed, implying that the ionic liquid plays the dual role of
solvent and catalyst (see computational study below). Water was
present in the reaction media, since it was ca. 0.2 wt% in
cholinium hexanoate and ca. 2.4 wt% in cork (determined
previously by thermogravimetric analyses).34 Thus the molar
fraction of water was 4.9% relative to the ionic liquid (0.42 wt%
in the reaction media and 8.4 wt% relative to suberin).
While the capacity of ionic liquids to play the dual role of
solvent and catalyst has been well described,54 the complex set
of chemical and physical interactions, including Coulombic
and dipole forces, hydrogen-bonding and acid–base interac-
tions, behind the augmented reactivity in an ionic liquid media
are not yet fully understood. Up to now, few hydrolysis reactions
in ionic liquid media have been reported, associated to the
nucleophilic activation of water.54,55 It has been demonstrated
that, when present at low concentrations, water is dispersed in
the ionic liquid as single molecules.54,55 They form specic
electrostatic interactions and hydrogen-bonding with the ions.
As the concentration of water increases, self-association of the
water molecules and ion clustering probably occurs, weakening
the catalytic effect of the ionic liquid. The chemistry of water in
these domains is still a puzzling scientic question.
It was demonstrated that the hydrolysis kinetics is inu-
enced by the ionic liquid hydrogen bonding donor and
acceptor ability.56 Cholinium carboxylates are good hydrogen
bond acceptors (Kamlet Ta parameters b > 0.9).57 Accord-
ingly, the mechanism of suberin hydrolysis in cholinium
hexanoate media might be related to the nucleophilic activa-
tion of water. In fact, increasing the molar fraction of water in
the reaction media from 4.9% to 40%, led to a dramatic
reduction in the extraction efficiency of cholinium hexanoate
(from 51.0 ꢄ 5.8 to 25.3 ꢄ 2.4 wt%). Water in excess hampered
the ionic liquid catalytic effect and/or acted as an anti-solvent.
Hence, during suberin depolymerisation from cork, a ne
balance of the water concentration in the reaction media
needs to be attained.
Fig. 5 Compounds detected after treatment of glyceryl trioctanoate
and octyl octanoate with cholinium hexanoate during 1, 2, 4 and 8
hours. All compounds were identified and quantified by GC-MS.
Glycerol, octanol, octanoic acid and error bars were not represented
for ease of visualisation (average standard error ¼ 27.8 mg). At time
zero, glyceryl trioctanoate and octyloctanoate were assumed to
represent the only compounds present in the mixture.
The lack of efficiency of cholinium hexanoate to promote
cleavage of linear aliphatic esters was further conrmed using
linear aliphatic polyesters. Aer four hours, no alterations were
detected in the ATR-FTIR spectra of poly(lactic acid) (Fig. S2†),
in particular in the peaks assigned to C]O stretching. More-
over, aer this reaction time the mass of the poly(lactic acid)
recovered was ca. 91.61 ꢄ 0.26 wt%, agreeing with the amount
of lactic acid detected (6.03 ꢄ 0.63 wt%). The ATR-FTIR spectra
of poly(12-hydroxydodecanoic acid) (Fig. S3†) was also unal-
tered aer four hours in the ionic liquid media.
Since the epoxy ring is a signicant non-structural function-
ality in suberin its lability in the ionic liquid media was also
preliminary analysed using 9,10-epoxy-18-hydroxyoctadecanoic
acid. NMR analysis showed that while the epoxy ring was partially
preserved aer 4 hours, some secondary alcohol signals were also
detected (1H: 3.5 ppm; 13C: 72 ppm, data not shown). This
observation, which was also preliminarily conrmed by GC-MS
and ESI-MS (data not shown), can be associated with the opening
of the epoxy ring into a vic-diol and/or methoxyhydrin, as sug-
gested above by the 13C CP/MAS NMR data of suberin samples.
Suberin is a structural component of the cork cell wall, thus
cannot be removed without impairing its integrity (Fig. S1†).
The other major polymers in the cork cell wall are likely to be
also exposed to the ionic liquid media, namely polysaccharides
and lignin, which are linked via ether bonds and ether/C–C
bonds, respectively. Preliminary data obtained with a-cellulose
and alkali lignin, suggested that cholinium hexanoate was
unable to promote hydrolysis of their inter-monomeric bonds
(data not shown). This further supports cholinium hexanoate
selectivity towards suberin.32–34
Cholinium hexanoate catalyses the hydrolysis of esters:
computational study
According to the experimental results, cholinium hexanoate
catalysed the hydrolysis of esters originating acids and alcohols.
A density functional theory (DFT) study42,43 was performed in
order to determine the reaction mechanism. All the interme-
diates and transitions states (TSs) along several possible reac-
tion pathways were calculated (S3†), and the lowest energy
pathway obtained, shown in Scheme 1, consists of two steps. In
the rst step, the hydroxyl group of the cholinium cation (IL+)
establishes one hydrogen bond with the oxygen of the carbonyl
group of the ester, favouring the nucleophilic attack of water at
the carbon (TS-1). The intermediate 2a has the newly formed
The role of cholinium hexanoate as catalyst in the hydrolysis
of ester bonds
The cleavage of ester bonds promoted by cholinium hexanoate OH bond. In the second step, one proton is transferred from 2b
led to the release of free alcohols and carboxylic acids from to the O(R2) atom, weakening the C–O bond (TS-2). The
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 2993–3002 | 2999

View Article Online
RSC Advances
Paper
Fig. 6 Calculated transition states (TSs) structures for the two steps of
the hydrolysis of octyl octanoate catalysed by cholinium hexanoate. All
bond lengths are in angstrom (A˚).
of cholinium acetate,33 this is most probably due to the choli-
nium hexanoate superior solvent ability towards aliphatic
chains. This further emphasises the dual role of the cholinium
hexanoate, both as solvent and as catalyst.
Scheme 1 Proposed mechanism for the hydrolysis of esters catalysed
by cholinium alkanoates. TS and IL stand for transition state and ionic
liquid, respectively.
The experimental results also showed that the hydrolysis
catalysed by cholinium hexanoate was faster in triacylglycerol
esters than in monoacylglycerol esters (Fig. 5), and that tri-
acylglycerol esters were preferential hydrolysed at the C2 posi-
tion (Table 3).
products, the acid 3 and the alcohol 4, are formed and the ionic
liquid regenerated, closing the catalytic cycle.
The activation barriers obtained for the hydrolysis of four
types of linear aliphatic esters (methyl and ethyl acetate, butyl
butyrate, and octyl octanoate) catalysed by cholinium acetate or
hexanoate are depicted in Table 2 and Fig. 6, and show that the
second step of the reaction is the limiting one. The energy
difference between TS-1 and TS-2 (D (TS-2–TS-1)) increases with
the size of the ester carbon chain, as the linear ester model
becomes closer to reality. This arises essentially from the
destabilisation of TS-2, since the close proximity of the carbon
chain R1 and R2 increases the steric repulsion and thus the
activation energy. The calculated activation energies are higher
in the reaction catalysed by cholinium hexanoate than by
cholinium acetate, reecting the slightly higher pKa of hexanoic
acid (4.88 vs. 4.76 of acetic acid). This difference is, however, too
small and not signicant considering the nature of the model,
as the substrates are linear aliphatic esters and the trends may
easily be reversed. Since cholinium hexanoate efficiency
The reaction mechanism was studied considering the
hydrolysis of glyceryl triacetate, catalysed by cholinium
Table
hydrolysis of glyceryl triacetate catalysed by cholinium acetate
3
Gibbs activation energies (kJ molꢂ1) calculated for the
Transition state
Attack A (C1)
Attack B (C2)
TS-1
TS-2
147.8
151.8
4.0
144.3
146.5
2.2
towards the extraction of suberin was 1.5 fold higher than that D (TS-2–TS-1)
Table 2 Gibbs activation energies (kJ molꢂ1) calculated for the transition states (TSs) presented in Scheme 1
Ester
Ionic liquid
Transition state
Methyl acetate
Ethyl acetate
Butyl butyrate
Octyl octanoate
Cholinium acetate
TS-1
TS-2
D (TS-2–TS-1)
TS-1
TS-2
151.6
151.9
0.3
153.6
155.0
1.4
153.7
157.4
3.7
154.3
161.5
7.2
154.8
160.5
5.7
156.4
164.2
7.8
154.1
163.2
9.1
155.5
166.6
11.1
Cholinium hexanoate
D (TS-2–TS-1)
3000 | RSC Adv., 2014, 4, 2993–3002
This journal is © The Royal Society of Chemistry 2014

View Article Online
Paper
RSC Advances
acetate, at both positions C1 and C2 (Table 3). The second step
(TS-2) is again the limiting one for the reaction at both carbon
centres. In addition, the reaction at C2 is faster than at C1
8 M. A. Bernards, Can. J. Bot., 2002, 80, 227–240.
9 E. C. Lulai and D. L. Corsini, Physiol. Mol. Plant P., 1998,
53(4), 209–222.
(lower activation energies), in agreement with the experi- 10 H. Pereira, Cork: Biology, Production and Uses, Elsevier,
mental results (Fig. 5). Amesterdam, 1st edn, 2007.
Finally, the activation energies for the hydrolysis of glyceryl 11 R. Franke, I. Dombrink and L. Schreiber, Front. Plant Sci.,
triacetate catalysed by cholinium acetate (Table 3) are lower 2012, 3, 1–8.
than those of the linear aliphatic esters (Table 2). This can be 12 P. E. Kolattukudy and V. P. Agrawal, Lipids, 1974, 9, 682–691.
correlated to the higher electrophilicity of the ester oxygen atom 13 R. Riley and P. E. Kolattukudy, Plant Physiol., 1975, 56, 650–
in the triacylglycerol than in the linear aliphatic ester. These
654.
data suggest how cholinium hexanoate might promote a mild 14 P. E. Kolattukudy, Science, 1980, 208, 990–1000.
depolymerisation of suberin from cork.
15 M. A. Bernards, in Encyclopedia of Life Sciences, John Wiley &
Sons, Ltd, Chichester, 2002, http://www.els.net.
16 M. A. Bernards, Can. J. Bot., 2002, 80, 227–240.
17 A. Gandini, C. Pascoal and A. J. D. Silvestre, Prog. Polym. Sci.,
2006, 31, 878–892.
Conclusions
Two distinct phases describe the extraction of suberin from cork
with cholinium hexanoate. First, rapid removal of nearly whole 18 J. Graça and S. Santos, Macromol. Biosci., 2007, 7, 128–135.
suberin, still highly polymerised and cross-linked, occurs. This is 19 P. Sitte, Protoplasma, 1962, 54, 555–559.
followed by slow removal of recalcitrant structures, non-hydro- 20 R. E. Stark and J. R. Garbow, Macromolecules, 1992, 25, 149–
lysable and displaying a high cross-link density, concomitantly
with continuous depolymerisation of labile oligomers dispersed 21 M. Lopes, A. Sarychev, C. Neto and A. Gil, Solid State Nucl.
in the ionic liquid media (Fig. 4). The ionic liquid plays the dual Magn. Reson., 2000, 16, 109–121.
role of solvent and catalyst and leads to suberin efficient extrac- 22 M. A. Bernards and N. G. Lewis, Phytochemistry, 1998, 47,
tion through a mild depolymerisation. Disruption of the three- 915–933.
dimensional network of suberin in the ionic liquid media, 23 M. A. Bernards, M. L. Lopez, J. Zajicek and N. G. Lewis, J.
progresses by preferential cleavage of acylglycerol ester bonds
Biol. Chem., 1995, 270, 7382–7386.
(with the faster kinetics in C2), whilst the linear aliphatic ester 24 J. Negrel, B. Pollet and C. Lapierre, Phytochemistry, 1996, 43,
bonds remain largely preserved. This reaction mechanism leads 1195–1199.
154.
to the efficient recovery of a high cross-linked and thus esteried 25 J. Graça, Phytochem. Rev., 2010, 9, 85–91.
suberin, thus different from that extracted with conventional 26 N. Cordeiro, M. N. Belgacem, A. Gandini and C. Pascoal
methods which promote thorough ester cleavage. The ionic
liquid catalyses the hydrolysis of esters, hence the water 27 N. Cordeiro, A. Blayo, M. N. Belgacem, A. Gandini, C. Pascoal
concentration in the reaction media is a decisive parameter. At Neto and J. F. LeNest, Ind. Crops Prod., 2000, 11, 63–71.
high concentrations, water weakens the catalytic effect of choli- 28 A. F. Sousa, G. Alessandro, A. J. D. Silvestre, C. P. Neto,
Neto, Ind. Crops Prod., 1999, 10, 1–10.
nium hexanoate and/or acts as anti-solvent. A DFT study of the
reaction mechanism showed that cholinium hexanoate catalyses
J. J. C. C. Pinto, C. Eckerman and B. Holmbom, J. Polym.
Sci., Part A: Polym. Chem., 2011, 49, 2281–2291.
the nucleophilic attack of water in two steps. This work should 29 A. Sousa, A. Gandini, A. Silvestre and C. Pascoal Neto,
inspire the isolation of biopolyesters from numerous renewable ChemSusChem, 2008, 1, 1020–1025.
resources, aiming at the development of strategies for their val- 30 R. Ekman and C. Eckerman, Pap. Ja Puu-Pap. Timber, 1985,
orisation. A new chemical reaction for the hydrolysis of lipids 67, 255–273.
catalysed by ionic liquids is now available. Ionic liquid chemistry 31 M. H. Lopes, A. M. Gil, A. J. D. Silvestre and C. P. Neto, J.
opens the possibility to achieve designer solvents of high
Agric. Food Chem., 2000, 48, 383–391.
performance displaying high selectivity towards a specic carbon 32 R. Ferreira, H. Garcia, A. F. Sousa, M. Petkovic, P. Lamosa,
centre in an ester bond.
C. S. R. Freire, A. J. D. Silvestre, L. P. N. Rebelo and
C. Silva Pereira, New J. Chem., 2012, 36, 2014–2024.
33 H. Garcia, R. Ferreira, M. Petkovic, J. L. Ferguson,
Notes and references
˜
M. C. Leitao, H. Q. N. Gunaratne, K. R. Seddon,
1 A. Gandini, Green Chem., 2011, 13, 1061–1083.
2 P. E. Kolattukudy, Annu. Rev. Plant Physiol., 1981, 32, 539–567.
3 A. Heredia, Biochim. Biophys. Acta, 2003, 1620, 1–7.
4 K. Ranathunge, L. Schreiber and R. Franke, Plant Science,
2011, 180, 399–413.
L. P. N. Rebelo and C. Silva Pereira, Green Chem., 2010, 12,
367–369.
34 R. Ferreira, H. Garcia, A. F. Sousa, C. S. R. Freire,
A. J. D. Silvestre, L. P. N. Rebelo and C. Silva Pereira, Ind.
Crops Prod., 2013, 44, 520–527.
5 L. Schreiber, Trends Plant Sci., 2010, 15, 546–553.
6 M. Pollard, F. Beisson, Y. Li and J. Ohlrogge, Trends Plant
Sci., 2008, 13, 236–246.
35 M. Petkovic, J. L. Ferguson, H. Q. N. Gunaratne, R. Ferreira,
˜
M. C. Leitao, K. R. Seddon, L. P. N. Rebelo and C. Silva
Pereira, Green Chem., 2010, 12, 643–649.
7 F. Beisson, Y. Li-Beisson and M. Pollard, Curr. Opin. Plant 36 A. M. Gil, M. Lopes, J. Rocha and C. P. Neto, Int. J. Biol.
Biol., 2012, 15, 329–337.
Macromol., 1997, 20, 293–305.
This journal is © The Royal Society of Chemistry 2014
RSC Adv., 2014, 4, 2993–3002 | 3001

View Article Online
RSC Advances
Paper
¨
37 A. Olsson, M. Lindstrom and T. Iversen, Biomacromolecules,
R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth,
P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels,
O. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and
D. J. Fox, Gaussian 09, Revision B. 1, Gaussian, Wallingford,
CT, 2009.
2007, 8, 757–760.
38 N. Cordeiro, M. N. Belgacem, A. J. D. Silvestre, C. P. Neto and
A. Gandini, Int. J. Biol. Macromol., 1998, 22, 71–80.
39 P. C. R. O. Pinto, A. R. Sousa, A. J. D. Silvestre, C. P. Neto,
A. Gandini, C. Eckerman and B. Holmbom, Ind. Crops 45 J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett.,
Prod., 2009, 29, 126–132.
1996, 77, 3865–3868.
40 R. Ekman, Holzforschung, 1983, 37, 205–211.
41 A. F. Sousa, P. C. R. O. Pinto, A. J. D. Silvestre and C. P. Neto,
J. Agric. Food Chem., 2006, 54, 6888–6893.
42 W. Koch and M. C. Holthausen, A chemist's guide to density
functional theory, Wiley-VCH, New York, 2nd edn, 2001.
43 R. G. Parr and W. Yang, Density-functional theory of atoms and
molecules, Oxford University Press, Oxford, 1989.
44 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
46 J. P. Perdew, K. Burke and M. Ernzerhof, Phys. Rev. Lett.,
1997, 78, 1396.
47 C. Adamo and V. Barone, J. Chem. Phys., 1999, 110, 6158–
6170.
48 S. E. Wheeler, A. Moran, S. N. Pieniazek and K. N. Houk, J.
Phys. Chem. A, 2009, 113, 10376–10384.
49 J. Tomasi, B. Mennucci and R. Cammi, Chem. Rev., 2005,
105, 2999–3093.
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, 50 J. W. Turner, B. E. Hartman and P. G. Hatcher, Org.
B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, Geochem., 2013, 57, 41–53.
X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, 51 A. P. Deshmukh, A. J. Simpson, C. M. Hadad and
J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, P. G. Hatcher, Org. Geochem., 2005, 36, 1072–1085.
R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, 52 J. Graça and H. Pereira, Holzforschung, 1997, 51, 225–234.
O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery, Jr, 53 H. Pereira, BioResources, 2013, 8, 2246–2256.
J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, 54 J. Hallett and T. Welton, Chem. Rev., 2011, 111, 3508–3576.
E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, 55 S. Chowdhury, R. S. Mohan and J. L. Scott, Tetrahedron, 2007,
J. Normand, K. Raghavachari, A. Rendell, J. C. Burant,
S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, 56 C. Weber, A. Masters and T. Maschmeyer, J. Phys. Chem. B,
M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, 2012, 116, 1858–1864.
J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, 57 J. Restolho, J. L. Mata and B. Saramago, Fluid Phase Equilib.,
63, 2663–2389.
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski,
2012, 322–323, 142–147.
3002 | RSC Adv., 2014, 4, 2993–3002
This journal is © The Royal Society of Chemistry 2014