Ionic Liquids as “Masking” Solvents of the Relative Strength of Bases in Proton Transfer Reactions

Article abstract of DOI:10.1002/cplu.201700514

Equilibrium constants for the proton transfer reaction between pyridines and trifluoroacetic acid were measured in room-temperature ionic liquids (ILs) of different cation–anion compositions. The experimental equilibrium constants for ion-pair formation were corrected according to the Fuoss equation. The calculated equilibrium constants for the formation of free ions were taken as a quantitative measure of the base strength in IL solutions and compared with the relative constants in water. The effect of IL composition is discussed for a series of fixed IL anions and fixed IL cations. Finally, the sensitivity of the proton transfer reaction to the electronic effects of the substituent groups on the pyridine ring was quantified by applying the Hammett equation. A more marked levelling effect on the base strength was observed in ILs than in water. The Hammett reaction constants ρ were then correlated with solvent parameters according to a multi-parametric analysis, which showed that both specific hydrogen-bond donor/acceptor and non-specific interactions play an important role, with α and permittivity being the main parameters affecting the ability of the IL to differentiate the strength of the base.

Full text of DOI:10.1002/cplu.201700514

Accepted Article
Title: Ionic liquids as "masking" solvents on the relative strength of
bases in proton transfer reactions.
Authors: Romina Zappacosta, Antonello Di Crescenzo, Valeria Ettorre,
Antonella Fontana, Marco Pierini, and Gabriella Siani
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemPlusChem 10.1002/cplu.201700514
Link to VoR: http://dx.doi.org/10.1002/cplu.201700514
A Journal of
www.chempluschem.org

10.1002/cplu.201700514
ChemPlusChem
Ionic liquids as “masking” solvents on the relative
strength of bases in proton transfer reactions.
Romina Zappacosta[a], Antonello Di Crescenzo[a], Valeria Ettorre[a], Antonella Fontana[a], Marco
Pierini[b], Gabriella Siani[b]*
a Dipartimento di Farmacia, Università “G. d’Annunzio” Via dei Vestini 31, 66013 Chieti.
b Dipartimento di Chimica e Tecnologie del Farmaco, Università degli Studi di Roma “La Sapienza”,
Piazzale Aldo Moro 5, 00185 Roma
siani@unich.it
Keywords: Ionic Liquids; Acid-base equilibria; Hammett equation
ABSTRACT Equilibrium constants for the proton transfer reaction between pyridines and
trifluoroacetic acid have been measured in some room temperature ionic liquids (ILs) of different cation-
anion compositions. The experimental equilibrium constant for ion pair formation have been corrected
according to the Fuoss equation. The calculated equilibrium constant for the formation of free ions have
been taken as a quantitative measure of the base strength in IL solutions and compared with the relative
constants in water. The effect of IL composition has been discussed for a series of fixed IL anion and
fixed IL cation. Finally, the sensitivity of the proton transfer reaction to the electronic effect of the
substituent groups on the pyridine ring has been quantified by applying the Hammett equation. A more
marked levelling effect on the base strength was observed in ILs than in water. The Hammett reaction
constants, , were then correlated with some solvent parameters, according to a multiparametric analysis
which shows that both specific hydrogen bond donor/acceptor and non-specific interactions play an
1
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
important role, with and permittivity being the main parameters affecting the IL ability to differentiate
the strength of the base.
INTRODUCTION
The ability to act as acids or as bases has always been regarded as a fundamental property of chemical
species. Indeed, acids and bases play a key role in a large part of organic chemistry, in biological
processes, in medicine and in the field of material science.[1-4] Being acids and bases practically
ubiquitous, it is of great practical importance to know their exact strength.
It is well known that acidity and basicity are influenced not only by the structural features of the
molecules, but also by the nature of the solvent in which the dissociation takes place.[5] Consequently,
rates and positions achieved by equilibria of all those processes in which acids and/or bases are involved
as reactants or as catalysts, change on passing from one solvent to another.
Indeed, the different solvation ability of solvents can drastically alter the relative and absolute
strengths of acids and/or bases, due to the changes in the solvation energies of the species involved in
the equilibrium.[6] A large number of experimental data has been accumulated over decades on acid-
base equilibria in a variety of non-aqueous solvents, from aprotic polar (DMSO, THF, acetonitrile) to
“very non-polar” media (heptane), in which self-consistent acidity and basicity scales have been
created.[7]
Recently great attention has been given to ionic liquids (ILs), which have emerged as alternative
media to molecular organic solvents due to their high solvation ability.[8]
ILs are low temperature molten salts composed entirely of ions, which possess many favourable
physicochemical properties such as negligible vapour pressure, high thermal stability and low
flammability,[9] so that they are often described as “green solvents”.[10] Their physicochemical properties
depend mainly upon the nature of the constituent ions so, due to a virtually infinite number of possible
combinations of cations and anions, task specific ILs could be designed simply by tuning these
properties through an appropriate choice of the cation-anion pairs.[11]
2
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
ILs have been applied in various field, such as electrochemistry,[12], gas capture,[13], soft materials,[14]
synthesis and catalytic reactions.[15] Indeed a number of organic reactions have been performed in ILs,
including reactions which require the presence of a base as reagent or catalyst, and both qualitative and
quantitative effects of the ILs have been investigated.[16-18] In many cases results were greatly
surprising, thus suggesting that, in general, an important task is to carry out “structure-properties
relationship” analyses useful to predict the peculiar behaviour of ILs when they are used as solvents.
Cheng et al. have measured the pKa values of a series of carbon acids,[19] benzoic acids[20] and
sulphonamides[21] in three TF2N-based ILs and in BMIM OTf, by using the indicator overlapping
method and avoiding the previously suggested ion-pairing effect in ILs.[22,23] Indeed the carbanions of
the C-H acids used as substrate[19] or indicators[20,21] are sterically hindered and largely charge-
delocalized, properties which reduce the occurrence of ion-pair formation. In these studies, they
correlated the pKa variation of the same acid to the acid-strengthening ability of the different ILs which,
in turn, depends on the ionic composition of the IL. More recently, the equilibrium basicities of some
amines have been determined in two ILs in terms of acidic dissociations of the corresponding amine
triflide salts,[24] following the same general procedure based on the use of deprotonated carbon acids as
indicators for the measurements. They found that amines are stronger base in ILs than in DMSO and
water but weaker base than in acetonitrile. It is noteworthy that this type of dissociation involved
positively charged substrate instead of neutral ones, and, consequently, it involves a different solvated
initial state.
We have investigated the effect of some ILs on the thermodynamics[25] and the kinetics[26] of the base-
catalysed tautomerization reaction of 2-nitrocyclohexanone. This reaction, which has been largely
studied also in water[27] and in conventional solvents,[28] has proved to be a suitable probe for
investigating the effects of ILs at molecular level.[29]
Data collected in ILs seem to indicate that the thermodynamics of the above reaction is influenced
mainly by the nature of the IL ions, with the anion playing a major role, as well as by the alkyl chain
segregation in the heterogeneous organization of ILs.[25] Moreover, the reaction is faster in ILs than in
3
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
conventional organic solvents, with rate constants higher than those expected on the basis of the pKa
values.[26] This result could suggest a higher basicity of the catalyst in ILs than in conventional organic
solvents but we were aware that the pKa values collected in aqueous solution do not adequately describe
the acid-basic strength of species in ILs,[22, 30] due to the peculiar IL-solute interactions.
We have measured the equilibrium constants for the ion pair formation between some substituted
pyridines and trifluoroacetic acid in butylmethylimidazolium esafluorophosphate (BMIM PF6), by
means of a direct titration of the neutral base with the acid.[22] Actually, we followed the suggestion that
in ILs an acid-base reaction leads to the formation of ion pairs rather than free ions,[23] especially if ions
are not hindered and not largely charge-delocalized. Indeed, the relevant mathematical treatment of
experimental data, which considers the formation of ion pair, has provided a very good fit.[22] The
results obtained have shown that the basicity order of pyridines in BMIM PF6 and water is the same
while the effects of structural changes of solutes are less significant in the IL than in water.[22] Moreover,
a linear Brønsted type correlation has been obtained when the second order rate constant values for the
base-catalysed enolization of 2-nitrocyclohexanone in BMIM PF6 have been plotted as a function of the
calculated Kip values.
In order to gain new insights into the above investigation, other ILs have been considered in the present
work. The investigated ILs differ in anion and cation properties, such as coordination and hydrogen
bond accepting/donating ability, to shed new light on how the different IL composing ions can influence
the basicity strength of pyridines and the response of the reaction to the introduction of substituent
groups on the heterocyclic ring. Indeed, the comprehension of how the ionic constituents of the IL can
interact with acids and/or bases affecting the outcome of the reaction, is of fundamental importance to
tailor the best performing IL for a specific application.
4
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
RESULTS AND DISCUSSION
For a proton-transfer process between molecules with sufficiently strong proton-donating and proton-
accepting power, the following scheme is generally accepted:
AH + B
AH•••B
A-•••HB+
A- + BH+
In the first stage, hydrogen bond formation takes place, then the transfer of a proton from the donor to
the acceptor leads to the formation of an ion pair, which may dissociate to form ions, depending on the
relative permittivity of the solvent. When a proton transfer from an acid HA to a base B occurs in water
and, in general, in high polar solvents, the reaction’s equilibrium constant, Keq, refers to free ion
formation, as this type of solvents can strongly solvate both anions and cations.
In media of low polarity, the dissociation in free ions is less extensive or it is even negligible while ion-
ion associations, such as contact or solvent separated ion pairs, occur.
In ILs, the occurrence of ion pairing has been suggested (Scheme 1) so that the measured equilibrium
constant, Kip, has been referred to ion pair rather than free ion formation.[22, 23]
R
R
R = H
o-CH3
HO
O
Kip
C
O
O
N
H
+
N
p-CH3
o-OCH3
p-OCH3
C
CF3
CF3
Scheme 1. Ion pair formation in ILs
The relative strength of a series of pyridines has been ascertained in terms of the extent of their reaction
with trifluoroacetic acid (TFA) as the reference acid, in the ILs of Scheme 2. A series of ILs composed
of the same TF2N anion coupled with cations of different H-bonding donor abilities and a series of ILs
composed of the same BMIM cation coupled with anions of different coordinating abilities have been
5
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
selected to understand how changing the nature of the cation and anion of the IL can affect the outcome
of the proton transfer reaction.
The proton transfer reaction has been investigated by means of spectrophotometric titration, which turns
out to be a good approach to determine Kip values.
The increase of the pyridinium cation absorbance upon addition of increasing amounts of TFA was
followed until a constant value is reached.
R1 = C4H9
R1 = C6H13
R1 = C4H9
R1 = C4H9
R1 = C4H9
R2 = H
R2 = H
R2 = CH3
R2 = H
R2 = H
X- = TF2N -
X- = TF2N -
X- = TF2N -
1. BMIM TF2N
2. HMIM TF2N
3. BM2IM TF2N
4. BMIM BF4
5. BMIM PF6
R1
N
R2
-
X- = BF4
X-
N
+
-
X- = PF6
CH3
R1 = CH3
R2 = C4H9
X- = TF2N -
6. BMPyrr TF2N
X-
N
+
R2
R1
Scheme 2. Structures of the investigated ILs
The absorbance values registered at max of the pyridinium cation were plotted as a function of TFA
concentrations and then correlated with the concentrations of the involved species according to the
equation:
Abs = B[B] + IP[BH+•••TFA-]
(2)
where B and IP are the molar absorptivity of the free base and the ion pair, respectively. [B] and
[BH+·TFA-] are related to each other by Kip:
Kip = [BH+•••TFA-]/[B][TFA]
(3)
6
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
The assumption that, at an early stage of the proton transfer, ion-pairs are formed and then they
dissociate into free ions, depending on the different solvation properties of the solvents are supported by
the fact that the mathematical treatment, which considers the formation of ion pair, fits very well with
the measured experimental data as shown in Figure 1, in which the titration curve obtained for 3-
methoxypyridine in BM2IM TF2N is reported as an example.
0,7
0,65
0,6
0,55
0,5
0,45
0,4
0,35
0,3
0
5 10-5
0,0001 0,00015 0,0002 0,00025 0,0003
[TFA]
Figure 1. Absorbances at = 286 nm of a solution of 3-methoxypyridine after the addition of increasing
amounts of TFA in BM2IM TF2N, fitted into equation 2 (solid line).
On the contrary, no good correlation is obtained if the experimental data are mathematically treated
according to the formation of free ions.
The calculated Kip values are collected in Table 1.
7
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
Table 1 Experimental Kip values for the proton transfer reaction between pyridines and TFA in ILs at
25.0 ±0.1°C. The uncertainties arise from the fitting of the experimental data. Each titration has been
performed at least twice.
Kip/105
a
BMIM BF4
HMIM TF2N
0.692 ±0.031
3.49 ±0.49
3.65 ±0.41
-
BMIM TF2N
1.16 ±0.14
0.542 ±0.049
4.09 ±0.59
-
BMPyrr TF2N
0.325 ±0.038
0.523 ±0.032
1.89 ±0.07
-
BM2IM TF2N
0.660 ±0.033
0.282 ±0.023
2.41 ±0.15
BMIM PF6
0.818 ±0.030
1.01 ±0.15
1.26 ±0.09
1.84 ±0.60
2.74 ±0.40
0.106 ±0.008
0.564 ±0.059
0.620 ±0.052
1.03 ±0.12
3-methoxypyridine
Pyridine
3-methylpyridine
4-methylpyridine
6.87 ±0.77
9.58 ±1.23
15.8 ±1.76
6.86 ±0.44
15.0 ±1.9
3.82 ±0.28
4-methoxypyridine
a data from ref. 22
It is noteworthy that the basicity order in ILs and in water appears to be almost the same.
However, a direct comparison between the basicities of pyridines in ILs and in water is not correct as
the relative equilibrium constants, Kip and Ka, refer to different equilibria, i.e. the formation of ion pairs
and the formation of free ions, respectively.
To allow a rigorous comparison, the experimental Kip values have been corrected for the ion pair
dissociation into free ions.
Indeed, a proton transfer reaction between a generic acid HA and a generic base B can be represented
as in the Scheme 3
Kip
Kd
BH + A
B + HA
BH
A
Scheme 3
with a pre-equilibrium ion pair formation followed by the dissociation of the ion pair complex in free
…
ions. The 1:1 ion pair dissociation constant, Kd = [BH+][A-]/[BH+ A-], can be estimated theoretically
8
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
by using the Fuoss equation for the equilibrium between ion pairs and free ions for classical
electrolytes:[31]
Kd = 3000eb/(4Na3)
(4)
where b =-e2/(akT), N = 6.02×1023 mol-1, a is the inter-ion distance, e = 4.80×1010 esu,  is the
permittivity of the solvent, k = 1.38×10-6 erg/deg and T is the temperature in Kelvin.
In particular, the inter-ion distances (iiDs) of the 1:1 ion pairs formed between the considered
pyridines protonated at the nitrogen atom and the conjugate base of TFA were estimated from the
relevant molecular structures modelled through fast quantum mechanical calculations, taking into
suitable account the permittivity of the ILs in which they were generated (for details see Experimental
section). The calculated iiDs are reported in the Supporting Information.
The equilibrium constant for the free ion formation in ILs, KeqIL, has been derived from the
experimental Kip values and the calculated Kd values. Data treatment is reported in the Supporting
Information.
W
Table 2 Calculated KeqIL and Keq values for the proton transfer between pyridines and TFA in ILs and
in water
3-methylpyridine 4-methylpyridine 4-methoxypyridine
3-methoxypyridine pyridine
5.34(±0.24)×103 2.69(±0.38)×104
5.16(±0.62)×103 2.39(±0.22)×103
1.23(±0.14)×103 1.96(±0.12)×103
6.69(±0.33)×102 2.26(±0.18)×102
1.23(±0.09)×103 6.73(±0.71)×103
2.09(±0.23)×104
1.91(±0.28)×104
7.58(±0.28)×103
-
-
-
7.62(±0.98)×104
7.80(±0.87)×104
2.93(±0.19)×104
HMIM TF2N
BMIM TF2N
BMPyrr TF2N
BM2IM TF2N
BMIM BF4
2.27(±0.14)×103 7.29(±0.82)×103 1.71(±0.22)×104
7.18(±0.60)×103 1.18(±0.14)×104 4.34(±0.32)×104
1.66(±0.12)×101 3.41(±1.11)×101 6.49(±0.95)×101
a
9.43(±0.35)
2.3×103
8.52(±1.27)
4.5×103
BMIM PF6
1.45×104
3.165×104
1.26×105
H2O
a
data from ref. 25
9
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
Effect of changing the IL cation
To investigate how the cation can affect the proton transfer equilibrium, the values of KeqIL (Table 2)
obtained for the reaction in the ionic liquids HMIM TF2N, BMIM TF2N, BMPyrr TF2N and BM2IM
TF2N have been considered as a function of the cation type. Indeed, as the investigated ILs share the
same TF2N anion, any observed difference is primarily due to the change of the cation.
For a fixed anion, the change of the cation primarily affects the IL acidity. It has been demonstrated
that hydrogen bonding plays an important role in ILs.[32] In imidazolium-based ILs, H-bond formation
with the solvated solute is mainly due to the most acidic hydrogen in the 2 position of the imidazolium
ring,[33] especially in the presence of a weak H-bond acceptor anion such as TF2N anion. However other
protons on the imidazolium ring can be involved in hydrogen bonding as well as protons on a non-
aromatic ring such as pyrrolidinium cation even if, in this latter case, H-bonds are weaker.[34]
It is noteworthy that, for all the investigated pyridines, the calculated KeqIL increase on increasing the
HBD ability of the IL cation, quantitatively expressed by the solvent parameter following the order:
BM2IM TF2N (= 0.381) < BMPyrr TF2N (= 0.430) < BMIM TF2N (= 0.617) < HMIM TF2N
(= 0.650)
Indeed, a competition occurs for hydrogen bonding between the anions and cations which can
influence the proton transfer equilibrium position. The carboxylate anion is a better H-bond acceptor
than TF2N anion[35] and gives place to stronger interactions with the proton of the IL cation (Figure 2),
causing the equilibrium position to move forwards as much as the HBD ability of the cation increases.
CH2CH2CH2CH3
N
O
O
H
C
N
CF3
CH3
Figure 2. Hydrogen bonding interaction between TFA- and imidazolium cation
10
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
Effect of changing the IL anion.
Anions mainly affect the hydrogen bond accepting (HBA) ability of the ionic liquids, i.e. IL basicity,
quantitatively expressed by the value of the solvent parameter. It is reasonable that the higher the
ability of the anion to act as H-bond acceptor and to compete with TFA anion interfering in the H-bond
TF2N
network of Figure 2, the lower would be the value of KeqIL, decreasing along the series KeqPF6 > Keq
>
BF4
Keq
.
Surprisingly, the lowest Keq values were obtained in BMIM PF6 although PF6 anion is less able to act
-
as H-bond acceptor than TF2N- and BF4 . This result suggests that even if there is evidence of specific
cation H-bonding effect, it cannot be the only factor responsible for the observed changes so other
contributions have to be considered. An anomalous behavior of the PF6 anion has been observed also by
Harper[36] in his study on a SN2 reaction, performed in a series of ILs containing different anions, for
which he cannot directly correlate the activation parameters with the nature of the anion. He related the
“special case” of PF6 anion to its size and shape, responsible of the greatest cross-linking of BMIM PF6
in comparison with the other considered salts. He concluded that the effect of changing the IL anion is
less easily rationalizable than the effect of changing the cation.
Effect of IL on the sensitivity of the proton transfer reaction to the change in pyridine
substituent. Hammett equation[37] has been applied to the data of Table 2 with the aim to correlate the
IL
observed changes in the Keq with the changes in substituent on the pyridine ring. Hammett equation
relates the relative magnitude of the equilibrium constants to a constant  and an empirical electronic
substituent parameter , as follows:
log(KeqX/KeqH) = 


where KeqX and KeqH are the equilibrium constants for the proton transfer from TFA to a X-substituted
pyridine and to the unsubstituted pyridine, respectively.
11
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
The reaction constant  measures how sensitive is the reaction to the substitution. The greater the
magnitude of , the more sensitive the reaction is to the substituent effects.
The Hammett plots for the reactions in the series of ILs sharing the same anion and the same cation
are reported in Figures 3 and 4, respectively. For the sake of comparison, the Hammett plot obtained for
the reaction in water has been reported in Figure 4. The  values obtained as the slopes of the plots have
been collected in Table 3, together with  and  solvent parameters, the Kirkwood’s function f() = (-
1)/(2+1) and the Hildebrand’s  values.
Figure 3. Hammett plots for the proton transfer between TFA and substituted pyridines in HMIM
TF2N (♦), BMIM TF2N (▲), BMPyrr TF2N (■) and BM2IM TF2N (●). Error bars represent standard
deviations.
12
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
Figure 4. Hammett plots for the proton transfer between TFA and substituted pyridines in BMIM
TF2N (▲), BMIM PF6 (■), BMIM BF4 (♦) and water (●). Error bars represent standard deviations.
Table 3 Physical properties of ILs and  values for the proton transfer between pyridines and TFA in ILs
and in water.
a



f(
0.650b
0.617c
0.430b
0.381c
0.627c
0.634c
1.17d
0.259b
0.243c
0.250b
0.239c
0.376c
0.207c
0.18d
0.475
0.471
0.470
0.455
0.478
0.429
0.490
11.5
11.1
10.8
11.9
12.8
12.7
23.4
-2.94±0.66
-3.01±0.16
-3.50±0.55
-3.66±0.19
-3.81±0.37
-2.17±0.23
-4.35±0.26
HMIM TF2N
BMIM TF2N
BMPyrr TF2N
BM2IM TF2N
BMIM BF4
a
BMIM PF6
H2O
a
b
c
d
determined according to the procedure described in ref. 29
determined according to the procedure described in ref. 33a
data from ref. 33a
data from ref. 38
13
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
The linearity of the Hammett plots demonstrates that the use of the Fuoss equation is a valid approach
despite the limitations of the Fuoss-type treatment as it is derived for spherical ions and it takes into
account only the permittivity of the solvent neglecting all the specific solute-solvent interactions.
It clearly appears that, for ILs with the same anion, the sensitivity to the substituent effect on the
pyridine basicity linearly decreases as the IL cation ability to act as hydrogen bond donor increases (the
greater the value of , the less negative the value of , see Figure S2 in Supporting Information).
Indeed, the ability of the IL cation to interact via H-bonding with the TFA anion levels the strength of
the base involved in the pre-equilibrium of the ion-pair formation and reduces the electronic effect of the
substituent on the described equilibrium.
The attempt to include in the same correlation the  values obtained for the ILs with fixed cation and
different anions failed, suggesting that other contributions must be involved in the observed solvent
effect.
However, a quite linear trend was found for the above three ILs in a vs. plot (R2 = 0.893, slope= -
8.70, intercept= 0.60), which shows that the sensitivity to the substituent effect decreases as a response
to the decrease of the HBA ability of the solvent (the lower the value of , the less negative the value of
).
In order to find, if possible, a more general correlation among the several different contributions of
solute-solvent interactions to the observed trend of the  sensitivity, classical multiparameter regression
analyses were performed, based on equations of the following type:
|| =  Ci × Di

where Di represents the generic i selected descriptor of solvation interactions, and Ci corresponds to its
regression coefficient. The , , f( and  parameters reported in Table 3 were tested as proper Di
descriptors. The choice was motivated by the need to monitor the ability of the solvent to act as HBD
and/or HBA media, to perform electrostatic shielding of atomic charges and to take into account the
cavitational energy required to host the solute molecules inside a proper cavity generated within the
14
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
solvent. The quality of each performed regression was checked by carrying out the relevant F-test and t-
test statistical evaluations (the first, useful to assess in percentage the probability of casualness of the
obtained regression, the second one needed to establish the statistical significance of each employed Di
parameter). The regression resulting by the insertion of all the four descriptors in equation 8, although
characterized by absence of casualness (probability lesser than 0.12%) and a very good correlation
coefficient (R2 =0.981), provided the indication that the descriptor is not statistically significant. On
the contrary, the correlation obtained taking into account only the descriptors ,  and f() (equation 9)
resulted again of good quality (R2 =0.978), devoid of randomness (probability lesser than 0.0032%) and
all the descriptors were significant under a statistical point of view:
|| = - 3.62  +6.89  +7.32 f()


From equation 9, it clearly appears that the sensitivity to the substituent effect increases as: a) the IL
acidity decreases (negative value of the  regression coefficient); b) the IL basicity increases (positive
value of the regression coefficient); c) the solvent permittivity increases. The latter property is the
factor playing the major (though not dominant) role (46% of relative contribution to , calculated on a
percentage basis from the absolute values of the regression coefficient of equation 9). In particular, it
acts in favor of an increase of the sensitivity as a response to an increase of solvent polarity. The
residual percentage contributions coming from the HBD and HBA capacity of solvent can be assessed in
31% and 23%, respectively.
The higher sensitivity to the substituent effect of the acid-base equilibrium observed for water ( =
-4.35) with respect to the investigated ILs is probably a consequence of the high influence of solvent
permittivity (see Table 3).
15
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
CONCLUSIONS
The basic properties of pyridine and some substituted pyridines have been investigated in some ILs
with different cation-anion combinations by protonation with TFA. The formation of ion pairs has been
suggested and the relative values of Kip show a trend that almost follows the basicity order of pyridines
in water. The experimental Kip values have been corrected for the ion pair dissociation into free ions by
the use of the Fuoss equation. It is noteworthy that, for a fixed IL anion, the calculated KeqIL increase on
increasing of the HBD ability of the IL cation. A different and more complex behavior was detected in
ILs with fixed cation and different anions as no simple correlation with solvent parameters can be found.
The application of the Hammett equation has allowed to measure the susceptibility of the reaction to
the electronic characteristics of the substituents. It turns out that pyridines are better differentiated in
water than in ILs which, definitely, act masking differences in the intrinsic strength of the bases. In the
series of the investigated ILs, the sensitivity to substituent effects depends on the nature of IL composing
anions and cations.
A multiparametric analysis has demonstrated that the observed effects depend on the IL hydrogen
bond donor acidity and hydrogen bond acceptor basicity, acting in opposite ways. However, the IL
permittivity seems to have a major positive influence on the differentiating ability of the solvent.
The evidenced effect of ILs and, particularly, of the IL composition on the relative strength of bases
has to be taken into account when tailoring ionic liquids as solvents for reaction involving bases both as
reagents and catalyst.
16
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
EXPERIMENTAL SECTION
Materials: Dichloromethane and trifluoroacetic acid were commercial samples of AnalaR grade and
were used without further purification. Commercial pyridine, 3-methylpyridine, 4-methylpiridine, 3-
methoxypyridine, and 4-methoxypyridine have been distilled in the presence of NaOH before each
experiment. Commercial ILs of Scheme 1 were dried 2 h at 80.0±0.1°C under vacuum before each
experiment. The purity of imidazolium salts was checked by ESI-MS and UV spectrophotometry
(purified [BMIM]+ salts containing [Cl]- or [Br]- < 0.1 ppm have no absorption band in the 250-300 nm
region). The water amount in ILs was determined by the Karl Fisher technique by using an apparatus
composed of a standard titrator and a coulometer. A water content within 120-150 ± 20 ppm was found
for ILs 1, 2, 3 and 6, whereas water contents of 250 ± 23 ppm and 351 ± 26 ppm were found for IL 4 and
IL 5, respectively.
UV-Vis titrations: The UV-Vis spectra were recorded with a UV-Vis spectrophotometer with a spectral
resolution set to 0.333 nm. The spectrophotometer was provided with a thermostated cell holder.
A small aliquot of a stock solution of the base in CH2Cl2 was transferred in a quartz cuvette with a light
path of 1 mm, containing 200 l of the appropriate IL. The concentrations of pyridine and substituted
pyridines were kept constant at 2×10-4 and 1×10-4 mol dm-3, respectively, in all experiments. Aliquots of
a 0.1 mol dm-3 stock solution of TFA in CH2Cl2 were then added to the cuvette to obtain final TFA
concentrations lying in the range 0-3.5·10-4 mol dm-3. A spectrum was acquired after equilibrium had
been established at each addition of TFA. The end-point was determined by plotting the absorbance
values versus the titrant concentration. All spectra were recorded at 25.0 ± 0.1 °C.
Estimation of iiD distances: All quantum mechanical optimizations were performed with the software
package SPARTAN 10, v. 1.1.0. The inter-ion distances (iiD) used in eq. (4) have been estimated from
the structures of the 1:1 ion pairs formed between the considered pyridines protonated at the nitrogen
atom and the conjugate base of TFA, which were modelled performing SCF optimizations at the HF/3-
17
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
21G level of theory. More in particular, as a function of the particular position and type of substitution
on the pyridinium cycle, the number of optimized ion pair structures was: a) one for the pristine pyridine
and 4-methylpyridine bases; b) two for 3-methylpyridine and 4-methoxypyridine bases; c) four for the
3-methoxypyridine base. According to the SM8 solvation model implemented in Spartan, the SCF
optimization of the just quoted ten ion-pair structures was carried out by simulating their dissolution in
four solvents of wide different permittivity, whose values range from 8.9 to 37.5, that is to say, in
CH2Cl2 (=8.9); acetone (=20.7); nitromethane (=35.9); acetonitrile (=37.5). Thus, for each selected
solvent and pyridinium compound, the final iiD was computed as the value averaged from those
estimated by the centers of mass of the couple of ions constituting the analyzed structure of the
optimized supramolecular isomers, according to their relevant Boltzmann populations. In Figure S3 of
Supporting Information it is reported a graph in which, for each of the five pyridinium species, the found
iiD values have been plotted as a function of the permittivity of the solvents in which they were
modelled. Eventually, the final searched iiDs data, attributable to each pyridinium compound within
each of the considered 1-6 ILs, were assessed by suitable interpolation on the plots collected in Figure
S3, starting from the known  values of the investigated ILs.
Multiparameter Linear Analyses: All multiparameter linear regression analyses, as well as the related
F-test and t-test statistical evaluations, were performed by the dedicated functions implemented in the
Microsoft Office Excel 2013 program. Probability related to the t-Student distribution in the t-test was
set to 0.05.
Supporting information for this article is given via a link at the end of the document.
18
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
Notes and references
[1] R. P. Bell, The proton in chemistry, Chapman and Hall, London, 2nd ed., 1973 and references
therein.
[2] T. Akiyama, Chem. Rev. 2007, 5744-5758.
[3] R.J. Bergeron, J.S. McManis, W.R. Weimar, K.M. Schreier, F. Gao, Q. Wu, J. Ortiz-Ocasio, G.
R. Luchetta, C. Porter and J. R. T. Vinson, J. Med. Chem. 1995
[4] V. Amendola, M. Bonizzoni, D. Esteban-Gomez, L. Fabbrizzi, M. Licchelli, F. Sancenon and
A. Taglietti, Coord. Chem. Rev. 2006 250, 1451-1470.
, 38, 2278-2285.
,
[5] B. G. Cox, Acids and Bases: Solvent Effects on Acid-Base Strength, Oxford university press,
2013 and references therein.
[6] W. E. Waghorne, Chem. Soc. Rev. 1993
[7] a) F. B. Bordwell, Acc. Chem. Res. 1988
Greaves, D. G. Gusev, J. G. Hinman, S. E. Landau, A. J. Lough and R. H. Morris, J. Am. Chem.
Soc. 2000 122, 9155-9171; c) T. Rodima, I. Kaljurand, A. Pihl, V. Mäemets, I. Leito and I. A.
Koppel, J. Org. Chem. 2002 67, 1873-1881; d) I. Kaljurand, T. Rodima, I. Leito, I. A. Koppel and
R. J. Schwesinger, Org. Chem. 2000 65, 6202-6208; e) I. Leito, T. Rodima, I. A. Koppel, R.
Schwesinger and V. M. Vlasov, J. Org. Chem. 1997 62, 8479-8483.
[8] R. D. Rogers and K. R. Seddon, Science 2003 302, 792-793.
[9] a) K. R. Seddon, Chem. Eng. 2002 730, 33-35; b) C. Chiappe and D. Pieraccini, J. Phys. Org.
Chem. 2005 18, 275-297; c) H. Zhang, B. R. Bakshi and E. S. Demessie, Environ. Sci. Technol.
2008 42, 1724-1730.
[10] M. Earle and K. R. Seddon, Pure Appl. Chem. 2000
[11] R. Giernoth, Angew. Chem. Int. Ed. 2010 49, 2834-2839.
,
22, 285-292
,
21, 456-463; b) K. Abdur-Rashid, T. P. Fong, B.
,
,
,
,
,
,
,
,
,
72, 1391-1398.
,
[12] K. Kirchner, T. Kirchner, V. Ivaništševa and M. V. Fedorova, Electrochim. Acta 2013, 110,
762-771.
19
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
[13] C. Wang, X. Luo, H. Luo, D.-E. Jiang, H. Li and S. Dai, Angew. Chem. Int. Ed. 2011, 50,
4918-4922.
[14] S. Li, J. L. Baňuelos, P. Zhang, G. Feng, S. Dai, G. Rotherc and P. T. Cummingsa, Soft Matter
2014, 10, 9193-9200.
[15] J. P. Hallett and T. Welton, Chem. Rev. 2011, 111, 3508-3576.
[16] J. S. Yadav, B. V. S. Reddy, A. K. Basak and A. V. Navsaiah, Tetrahedron Lett. 2003, 44,
2217-2220.
[17] a) F. D’Anna, V. Frenna and R. Noto, J. Org. Chem. 2006
Marca and R. Noto, J. Org. Chem. 2008 73, 3397-3403.
[18] G. Ranieri, J. P. Hallett and T. Welton, Ind. Eng. Chem. Res. 2008
, 71, 5144-5150; b) F. D’Anna, S. La
,
,
47, 638-644.
77, 7291-7298.
78, 12487-12493.
[19] H. Deng, X. Li, Y. Chu, J. He and J.-P. Cheng, J. Org. Chem. 2012
,
[20] Z. Wang, H. Deng, X. Li, P. Ji and J.-P. Cheng, J. Org. Chem. 2013
,
[21] Z. Wang, X. Li, P. Ji and J.-P. Cheng, J. Org. Chem., 2016, 81, 11195-11200.
[22] G. Angelini, P. De Maria, C. Chiappe, A. Fontana, M. Pierini and G. Siani, J. Org. Chem.
2010
[23] a) F. D’Anna and R. Noto, Tetrahedron 2007
R. Noto, J. Org. Chem. 2009 74, 6224-6230.
[24] C. Mao, Z. Wang, P. Ji and J.-P. Cheng, J. Org. Chem. 2015
, 75, 3912-3915.
,
63, 11681-11685; b) F. D’Anna, P. Vitale and
,
,
80, 8384-8389.
[25] R. Zappacosta, A. Di Crescenzo, P. Di Profio, A. Fontana and G. Siani, J. Org. Chem. 2015
,
80, 2333-2338.
[26] G. Angelini, P. De Maria, C. Chiappe, A. Fontana, C. Gasbarri and G. Siani, J. Org. Chem.
2009 74, 6572-6576.
,
[27] G. Angelini, P. De Maria, A. Fontana, M. Pierini and G. Siani, J. Org. Chem. 2007, 72, 4039-
4047.
[28] G. Siani, G. Angelini, P. De Maria, A. Fontana and M. Pierini, Org. Biomol. Chem. 2008, 6,
4236-4241.
20
This article is protected by copyright. All rights reserved.

10.1002/cplu.201700514
ChemPlusChem
[29] G. Angelini, C. Chiappe, P. De Maria, A. Fontana, F. Gasparrini, D. Pieraccini, M. Pierini and
G. Siani, J. Org. Chem. 2005 70, 8193-8196.
,
[30] K. E. Johnson, R. M. Pagni and J. Bartmess, Monatsh. Chem Chem. Monthly 2007, 138, 1077-
1101.
[31] a) R.M. Fuoss, J. Am. Chem. Soc. 1958
Greaves, D. G. Gusev, J. G. Hinman, S. E. Landau, A. J. Lough and R. H. Morris, J. Am. Chem.
Soc. 2000 122, 9155-9171.
[32] K. Dong, S. Zhang and J. Wang, Chem. Commun. 2016
[33] a) L. Crowhurst, P. R. Mawdsley, J. M. Perez-Arlandis, P. A. Saler and T. Welton, Phys.
Chem. Chem. Phys. 2003 , 2790-2794; b) J.P. Canal, T. Ramnial, D.A. Dickie and J.A.C.
Clyburne, Chem. Commun. 2006 61, 1809-1818; c) E. Ruiz, V. R. Ferro, J. Palomar, J. Ortega and
J. J. Rodriguez, J. Phys. Chem. B 2013 117, 7388-7398.
[34] a) S. Gründemann, A. Kovacevic, M. Albrecht, J. W. F. Robert and H. Crabtree, Chem.
Commun. 2001, 2274-2275; b) S.T. Handy and M. Okello, J. Org. Chem. 2005 70, 1915-1918; c)
K. A. Kurnia, F. Lima, A. F. M. Claudio, J. A. P. Coutinho and M. G. Freire, Phys Chem. Chem.
Phys. 2015 17, 18980-18990.
, 80, 5059-5061; b) K. Abdur-Raschid, T. P. Fong, B.
,
,
52, 6744-6764 and references therein.
,
5
,
,
,
,
[35] R. Klein, O. Zech, E. Maurer, M. Kellermeier and W. Kunz, J. Phys. Chem. B 2011, 115,
8961-8969.
[36] S. T. Keaveney, D. V. Francis, W. Cao, R. H. Haines, J. B. Harper, Aust. J. Chem. 2014, 68,
31-35.
[37] L. P. Hammett, Chem. Rev. 1935, 17, 125-136.
[38] M. J. Kamlet, J. L. M. Abboud, M. H. Abraham, R. W. Taft, J. Org. Chem. 1983
,
48, 2877-
2887.
21
This article is protected by copyright. All rights reserved.