A study of the influence of ionic liquids properties on the kemp elimination reaction

Article abstract of DOI:10.1002/chem.200900148

The morpholino-induced elimination of 5-nitrobenzisoxazole into the relevant 2-cyano-4-nitrophenolate has been used as a sample reaction in order to investigate molecular properties of some room temperature ionic liquids. The kinetic study was carried out

Full text of DOI:10.1002/chem.200900148

DOI: 10.1002/chem.200900148
A Study of the Influence of Ionic Liquids Properties on the Kemp
Elimination Reaction
Francesca D’Anna,* Sandra La Marca, Paolo Lo Meo, and Renato Noto*[a]
Dedicated to the Centenary of the Italian Chemical Society
Abstract: The morpholino-induced
elimination of 5-nitrobenzisoxazole
into the relevant 2-cyano-4-nitropheno-
late has been used as a sample reaction
in order to investigate molecular prop-
erties of some room temperature ionic
liquids. The kinetic study was carried
out at 298 K by means of spectrophoto-
metric measurements. Ionic liquids,
which differ in both their cation and
anion properties, were used as solvent
systems. In particular, aliphatic (pyrro-
lidinium, piperidinium, and ammoni-
um) and aromatic (imidazolium and
pyridinium) ionic liquids were used.
For aromatic cations, imidazolium ions
having different hydrogen-bond donor
ability or a different alkyl-chain length
were taken into account. The anions
used were determined by using spec-
troscopic probes, such as 4-nitroaniline,
N,N-diethyl-4-nitroaniline, Nile Red,
and Reichardtꢀs dye. Finally, in order
to obtain information on the structural
organization of the solvent systems,
resonance light-scattering measure-
ments were carried out. The collected
data provide evidence that ionic liquids
are solvent media which exhibit pecu-
liar features, whose effects can be ra-
tionalized only considering all parame-
ters affecting their three-dimensional
structure.
ꢀ
ꢀ
ꢀ
chosen ([BF4 ], [PF6 ], [SbF6 ], and
ꢀ
[NTf2 ]; where NTf2 =bis(trifluorome-
thansulfonyl)imide) showed different
shape, size, and coordination ability.
Solvent parameters of all ionic liquids
Keywords: base catalysis · kinetics ·
ionic liquids · Kemp elimination
Introduction
ticular, the latter factor may be especially significant, be-
cause a wide range of cations and anions are available for
use and their different combinations give rise to dramatic
changes in the solvent properties. This is the reason why ILs
have been frequently labeled as “designer solvents”.[3]
A better knowledge of IL solvent effects also has a pre-
dictive role, as it can give information about the advantages
compared with commonly used solvents and it allows the
synthesis of new ILs with tailored properties for particular
chemical reactions. Furthermore, such a determination can
also help to clarify the different points of view of research-
ers working in this area, who alternatively rationalize data
collected in ILs either using classical solvent parameters,[4]
or pointing out differences in solvation effects due to their
organized structure.[5] In the light of this, we decided to use
a classical organic reaction, such as the elimination reaction,
to further deepen the study of ILs properties.
The understanding of solvent properties is one of the main
topics of organic chemistry and in particular of physical or-
ganic chemistry. Among the methods used to pursue this
task, the kinetic study of a probe reaction can provide
useful information.[1] This has been largely used for conven-
tional organic solvent and can also be applied whenever the
medium shows peculiar features such as those characteristic
ionic liquids (ILs). In the framework of our interest in
studying ILs properties such an approach has allowed us to
verify that the effects these solvent media may exert are
heavily dependent on both the nature of the reaction and
the cation/anion pair constituting the ionic liquid.[2] In par-
[a] Dr. F. D’Anna, S. La Marca, Prof. P. Lo Meo, Prof. R. Noto
Dipartimento di Chimica Organica “E. Paternꢁ”
Universitꢂ degli Studi di Palermo
The reaction used in this work is the morpholino-induced
elimination of 5-nitrobenzisoxazole (1a) into the corre-
sponding 2-cyano-4-nitrophenolate (2a) for which a concert-
ed E2 mechanism has been accepted (Scheme 1). This reac-
tion, studied previously in different solvent systems,[6] pro-
ceeds through a charge-separated and highly ordered transi-
Viale delle Scienze-Parco d’Orleans II, 90128 Palermo (Italy)
Fax : (+39)091596825
E-mail: fdanna@unipa.it
rnoto@unipa.it
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200900148.
7896
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7896 – 7902

FULL PAPER
ꢀ
ꢀ
tion state, in which both C H and N O bonds are exten-
sively cleaved.[6a] For all these reasons, according to our pre-
vious findings,[2b–d] this reaction should be a suitable probe
for our investigations (Scheme 1). We have recently studied
the amino-induced Kemp elimination of the benzisoxazole
(1b) in IL solution, evidencing a particular sensitivity of the
target reaction to the structure of the amine and hydrogen-
bond donor ability of the IL cation.[7] Furthermore, a posi-
tive effect exerted by ILs having an aliphatic cation has
been also detected.
cation and anion properties (Scheme 1). In particular, aro-
matic as well as aliphatic cations were used. Aromatic cat-
ions differ in their hydrogen-bond donor ability [bmim+ and
bm2im+; where bmim=1-(1-butyl)-3-methylimidazolium,
bm2im=1-(1-butyl)-2,3-dimethylimidazolium], in the size of
aromatic ring [bmpy+ =1-(1-butyl)-4-methylpyridinium],
but also in the alkyl chain length borne on the nitrogen
atom of the imidazolium cation [emim+, bmim+, hmim+,
moim+, and dmim+; where emim=1-ethyl-3-methylimida-
zolium, hmim=1-(1-hexyl)-3-methylimidazolium, moim=1-
methyl-3-(1-octyl)-imidazolium and dmim=1-(1-decyl)-3-
methylimidazolium]. Aliphatic cations with an acyclic
(bEt3N+) or cyclic structure [bmpyrr+ and bmpip+;
bmpyrr=N-(1-butyl)-N-methylpyrrolidinium and bmpip=
N-(1-butyl)-N-methylpiperidinium] and, in this latter case,
different in size and geometry of the ring, were used.
The anions were different in shape, size, and coordination
ability, which should cause a different packing and possibly
a different catalytic effect. In order to have a comparison of
the reactivity in conventional organic solvents (COS), the
reaction was also studied in 1,4-dioxane (the co-solvent used
in this work, see Experimental Section) and DMF solution.
Finally, in order to check for a possible correlation be-
tween the reaction rates and classical solvent parameters,
we used spectroscopic probes
The substrate 1a used in this work is an interesting probe,
because the organized structure of IL could affect the copla-
narity of the nitro group with the aromatic ring and, conse-
quently, its electronic effect. On the other hand, the higher
steric hindrance of the substrate, due to the presence of this
substituent, could induce a higher influence on the structural
organization of solvent media with respect to 1b.
The reaction rates were measured spectrophotometrically
(UV/Vis) at 298 K, at different morpholine concentrations
(0.008–0.02m), following the formation of 2a at 406 nm. In
order to obtain information about the activation parameters,
the investigation was also carried out at a fixed morpholine
concentration (0.008m) over a temperature range of 293–
313 K. The reaction was studied in different ILs with both
(Scheme 1) to determine polari-
ty (ET(30) and ENR),[8] polariza-
bility (p*),[9] hydrogen-bond
donor (a)[10] and acceptor abili-
ty (b)[11] of all used solvent
media.
Results and Discussion
First we analyzed the reactivity
data in COS. The data reported
in Table 1 show that the reac-
tion proceeds faster in DMF
than in 1,4-dioxane. Both sol-
vents are aprotic, but the higher
reactivity of 1a in DMF can be
confidently ascribed to its
higher polarity, allowing a more
effective stabilization of the
charge-separated
state.
transition
For all solvent systems, the
observed rate constants kobs
show a linear correlation with
the base concentration, accord-
ing to Equation (1):
kobs ¼ i þ kII½Amꢁ
ð1Þ
Second-order rate constant
Scheme 1. Schematic representation of the reaction studied, ILs, and spectroscopic probes used.
values (kII) as a function of the
Chem. Eur. J. 2009, 15, 7896 – 7902
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7897

R. Noto, F. D’Anna et al.
pyrrolidine and piperidine as the base catalyst.[7] We are
aware of the fact that, as the nature of both substrate and
base are different in the two cases, the present comparison
seems to be not homogeneous. However, it allows us to
draw some general conclusions about the sensitivity of two
reaction probes to solvent effects.
Table 1. Calculated second-order rate constants (kII) at 298 K for the
elimination reaction of 1a, in IL solution, in the presence of morpholine.
Entry
Solvent
kII/mꢀ1 sꢀ1
R
1
2
3
4
5
6
7
8
1,4-dioxane
DMF
0.0148ꢂ0.0006
0.729ꢂ0.050
0.403ꢂ0.017
0.250ꢂ0.014
0.210ꢂ0.005
0.236ꢂ0.007
0.191ꢂ0.007
0.299ꢂ0.009
0.197ꢂ0.006
0.202ꢂ0.008
0.213ꢂ0.003
0.736ꢂ0.007
0.227ꢂ0.011
0.370ꢂ0.011
0.194ꢂ0.009
0.247ꢂ0.010
0.338ꢂ0.022
0.995
0.990
0.995
0.994
0.998
0.998
0.995
0.998
0.998
0.996
0.999
0.997
0.994
0.998
0.995
0.995
0.989
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
In order to have a better understanding about the effect
exerted by ILs on the reactivity of 1a, we determined, for
all the ILs used in this work, polarity and Kamlet–Taft sol-
vent parameters (solvent parameters collected at 298 K are
reported in Table S4 in the Supporting Information). It is
noteworthy that, in order to have the same microenviron-
ment, all measurements were carried out using 1,4-dioxane
as a co-solvent (see Experimental Section). Analysis of the
data reported in the Table allows us to make some interest-
ing considerations. Firstly, the comparison with polarity and
Kamlet–Taft parameters, reported in the literature[13] for
some of ILs used in this paper, outlines that the co-solvent
presence has a negligible effect. According to our previous
report,[2e] this evidences that data collected in the binary
mixture can be referred to a partially organized solvent
medium, such as neat IL. Among the polarity parameters,
less significant variations were detected for ENR, indicating a
lower sensitivity of Nile Red to the structural variations of
the ILs considered in this paper. In Table S4 we do not
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
9
A
ACHTUNGTRENNUNG
10
11
12
13
14
15
16
17
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
A
ACHTUNGTRENNUNG
solvent used are reported in Table 1 (data collected at differ-
ent morpholine concentrations are reported in Table S3 in
the Supporting Information).
We checked for the possible occurrence of a spontaneous
reaction in the IL solution. We verified that the substrate re-
mains unchanged in the IL solution for at least 48 h. This
suggests that the uncatalyzed reaction did not proceed at an
appreciable rate (k ! 1.27ꢄ10ꢀ5 sꢀ1).
In IL solutions and in the presence of morpholine, we did
not have any evidence for the formation of intermediates
and we assumed that the mechanism of the base-catalyzed
elimination of 1a proceeds, as in water solution, through a
concerted E2 elimination.[6a]
report a and ET(30) values for [bEt3N]ACHTNUGTRNEUNG[NTf2]. Indeed, in
this case, the UV/Vis spectrum of Reichardtꢀs dye gave a
broad band, which did not allow us to calculate the corre-
sponding solvent parameter and polarity values. According
to previous literature reports,[13c] this experimental result
could be probably due to the low stability of Reichardtꢀs
dye 30 in the used solvent system. Alternatively, as suggest-
ed by a referee, this could be due to the fact that the betaine
dye is protonated because the solvents are not completely
acid-free.
The comparison among kII values calculated in IL solu-
tions and in COS shows that the reaction rate in DMF is
comparable with the one detected in [bmpyrr]ACTHUNRGTNEUNG[NTf2], but
higher than that obtained in all others ILs. The decrease in
reaction rate from DMF to ILs cannot be rationalized on
the grounds of a decrease in solvent polarity, because all ILs
used show ET(30) values higher than for DMF (ET(30)=
43.2 kcalmolꢀ1 for DMF;[12] ET(30)=48.1–56.9 kcalmolꢀ1 for
the ILs used, see Table S4 in the Supporting Information).
This result agrees with data previously obtained in the
presence of 1b,[7] and confirms that the hydrogen-bond
donor ability of the solvent exerts a deactivating effect on
the probe reaction. Indeed, all ILs used in this work have a
values higher than DMF (see Table S4, Supporting Informa-
tion). The above negative effect is operating also in IL solu-
tion. Indeed, kII values decrease on going from [bm2im]-
For [bmim+] ILs, the anion symmetry and coordination
ability seem to affect the ET(30), ENR, and p* values.
Indeed, the polarity of solvent media generally decreases
parallel with symmetry and anion coordination ability. For
ꢀ
aromatic [NTf2 ] ILs, the length of alkyl chain borne on ni-
trogen atom, seems to have an influence on the cation abili-
ty to give a hydrogen bond. In general, the highest a values
were detected for the shortest alkyl chains (n=2 and 4).
Probably, when the alkyl-chain length increases, a steric
effect due to its organization around the imidazolium cation
hampers hydrogen-bond formation by means of 2-H. Ac-
cording to the hydrophobic effect, polarity [ET(30) and p*
values] decreases on going from [emim]
[NTf2].
A different behavior was detected for [SbF6 ] ILs. In this
case, the lowest and highest a values were detected for
[hmim][SbF6] and [dmim][SbF6], respectively. This seems to
ACHTUNGRTEN[NUNG NTf2] to [dmim]-
A
R
ACHTUNGTRENNUNG
ꢀ
and 0.384 for [bmim]
ly).
ACHTUNGTRENNUG[NTf2] and [bm2im]CAHTUNGTERN[NUNG NTf2], respective-
A
ACHTUNGTRENNUNG
The comparison with data previously obtained provides
evidence that the decrease in the reaction rate, on going
from DMF to ILs, is affected by the nature of the substrate.
Indeed, the highest kII,DMF/kII,IL ratio is equal to 3.8 in the
case of 1a, but is equal to 1.4 in the case of 1b, both using
indicate a certain dependence of IL hydrogen-bond donor
ability on the three-dimensional structure of the solvent
medium that, in turn, is heavily affected by the anion
nature. Both ET(30) and p* values evidence a higher polari-
ty for [hmim]ACTHUNRGTNEUNG[SbF6]. On the grounds of this information, we
7898
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7896 – 7902

Ionic Liquids Properties
FULL PAPER
tried to correlate lnkII with each solvent parameter with un-
satisfactory results (see Figure S3 in the Supporting Infor-
mation).
of the aliphatic cation. However, in the series of aliphatic
cations also shape and geometry seem to be important fac-
tors. As a matter of fact kII values drastically decrease on
A satisfactory correlation according to lnkII =s [a or
ET(30)] + i (where s and i are the slope and the intercept,
respectively) was only found (see Figure S4 in the Support-
going from [bmpyrr]ACHTNGURTENU[NG NTf2] to [bmpip]ACHTUGNTNER[NUNG NTf2], evidencing
that the best catalytic effect is exerted by the planar aliphat-
ic cation. The importance of the effect due to the geometry
of aliphatic cations, such as pyrrolidinium and piperidinium,
has been recently claimed in order to explain the significant
differences in CMC and surface tension of surfactants ionic
liquids.[15]
ꢀ
ing Information) considering some of the [NTf2 ] based ILs.
In both cases a negative slope was calculated [s=ꢀ0.680ꢂ
0.090; i=ꢀ1.19ꢂ0.05; R=0.948, n=8 for lnkII vs a and s=
ꢀ0.04ꢂ0.01; i=0.39ꢂ0.34; R=0.920, n=8 for ln kII vs
ET(30)]. The significant slope value in the case of a once
again supports the hypothesis about the inhibitory effect ex-
erted by protic solvents on the target reaction.
ꢀ
Analysis of kinetic data concerning to [NTf2 ] aromatic
ILs allows us to point out some interesting considerations.
In these ILs, kII values range from 0.191mꢀ1 sꢀ1 for [emim]-
Any attempt to carry out multiparametric correlations
gave unsatisfactory results too, indicating that classical sol-
vent parameters are somehow inadequate to explain our re-
activity data. A similar result has been recently obtained by
Kumar et al.,[5k] and it has been ascribed to the effect of
temperature difference between kinetic and polarity meas-
urements. This is not applicable in our case, as all values
were collected at the same temperature. Probably, the lack
of correlation could account, particularly in the presence of
larger alkyl tails, for the occurrence of different structural
organizations (polar and nonpolar local environments).
Therefore, bearing in mind the previous reports,[14] the con-
cept of liquid polarity should be used carefully, at least for
ILs. For example, many attempts carried out in these solvent
media have shown that their “polarity” is not an absolute
parameter but it is heavily dependent on the nature of the
used probe.[14a–b]
[NTf2] up to 0.299mꢀ1 sꢀ1 for [hmim]
ACHUTGTNRENNUG CAHTUNGTRENNUNG
values were obtained for [bmim]ACHTNUGTRENNUNG
AHCTUNGTRENNUNG
systems, a scarce contribution to p–p interactions from con-
tact surface area and of p-electron density.[16]
A more complex behavior may be observed if kinetic data
collected for aromatic ILs are analyzed as a function of the
ꢀ
alkyl-chain length. Indeed, for [NTf2 ] ILs, very similar kII
values (0.200ꢂ0.010mꢀ1 sꢀ1) were collected, only the kII
value (0.299mꢀ1 sꢀ1) for [hmim]
ACTHNUTRGEN[NUG NTf2] was different
ꢀ
(Figure 1). Differently, for [SbF6 ] ILs, kII values regularly
increase with alkyl-chain length on going from [bmim][SbF6]
up to [moim][SbF6] (entries 15–17, Table 1).
ACHTUNGTRENNUNG
AHCTUNGTRENNUNG
Therefore, the effect of the alkyl-chain length is also influ-
enced by the nature of the IL anion and seems to be quite
complex. Data collected herein cannot be explained on the
ꢀ
grounds of solvent viscosity. Indeed, for [SbF6 ] ILs, kII and
As a consequence of the high number of ILs used in this
work, for the sake of clarity, data collected will be discussed
as a function of the IL cation and anion structure.
h values show an inverse trend. Furthermore, [bmim]
is more viscous than [bmim][NTf2] (h =108 and 52 cP for
[bmim][SbF6] and [bmim]
[NTf2], respectively),[17] but kII
ACHTUNGTRENNUNG[SbF6]
AHCTUNGTRENNUNG
N
ACHTUNGTRENNUNG
values calculated in their solution are similar (entries 5 and
15, Table 1). The inadequacy of the viscosity parameter to
rationalize kinetic data obtained in IL solution has recently
been pointed out.[5k,18] On the one hand, it is well known
that ILs viscosity increases with alkyl chain length, as a con-
sequence of more significant van der Waals interactions.[19]
The cation effect: As we mentioned previously, the probe
reaction was studied for [NTf2 ] based ILs with different
cation structure. In particular, both aliphatic and aromatic
ꢀ
structures were taken into account. For these ILs the kII
values range from 0.191mꢀ1 sꢀ1 for [emim]
ACHTUNGTRENNUNG
0.736mꢀ1 sꢀ1 for [bmpyrr]
ACHTUNGTRENNUNG
ACHTUNGTRENNUNG
the target reaction is less aromatic than the initial state, the
result attests the importance of the initial-state stabilization
by aromatic cations through p–p interactions.
The comparison between data reported herein and the
data previously obtained in the case of 1b shows that the in-
crease in reaction rate on going from [bmim]
ACHTUNGRTEN[NUNG NTf2] to
+
[bmpyrr] =
G
3.5 for 1a; kbmpyrr+/kbmim+ =1.7 and 1.5 for 1b using pyrroli-
dine or piperidine, respectively, as the base catalyst).[7a] This
trend may be rationalized in terms of a different position of
the transition state along the reaction coordinate. Indeed, in
the case of 1a, the formation of a later and less aromatic
transition state can be hypothesized as a consequence of the
higher acidity of the hydrogen on C-3 of the benzisoxazole
ring. This should be more sensitive to the stabilizing effect
Figure 1. Plot of second-order rate constant values (kII) for the elimina-
&
&
tion reaction of 1a, in IL solution [NTf2]: , [SbF6]:
.
Chem. Eur. J. 2009, 15, 7896 – 7902
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7899

R. Noto, F. D’Anna et al.
This factor, hampering the contact between reagents, should
induce a decrease in the rate of the reaction. On the other
hand, as a consequence of the alkyl chain length, a different
packing degree of the solvent system could be assumed. In
some cases, as a function of this factor, a different solvation
of reagents through cation or alkyl chain has been claim-
ed.[5k,14] A longer and more flexible alkyl chain could induce
a decreasing order degree. In our case, a more disordered
aromatic solvent system should be less able to interact with
the initial state by means of p–p interactions and it should
induce its less effective stabilization, favoring the reaction.
On the whole, our results seem to indicate that, if a vis-
cosity effect is operating, the reaction rate is also affected
by ILs properties, which change as the anion changes. In
particular, the effect due to the structural organization of
solvent system seems to be significant.
In order to evaluate the importance of the latter factor,
we carried out some resonance light scattering (RLS) meas-
urements. This technique, introduced by Pasternack et al.[20]
allows to achieve information about the extent of aggrega-
tion of molecules containing chromophores. It has been
largely used to study porphyrins assemblies,[21] DNA assem-
blies,[22] and recently to investigate structural properties of
aqueous solution of ILs.[23]
Also in this case, in order to have the same microenviron-
ment, we collected RLS spectra of neat imidazolium ILs
having different alkyl-chain length, and of their mixtures ob-
tained in the presence of 1,4-dioxane (See Figure S5 in the
Supporting Information). In general, the addition of 75 mL
of the co-solvent induced an increase in the intensity of
RLS band, evidencing an increase in size of the aggregates,
but no change in the band shape was detected. The band in-
tensity was a function of the IL anion. In particular, for the
ꢀ
same cation, the intensity decreased on going from [NTf2 ]
ꢀ
to [SbF6 ]. As the intensity band can be correlated to the
Figure 2. RLS spectra of: a) 1,4-dioxane solutions of c: [bmin]
c: [hmin][NTf2], c: [moin][NTf2], c: [emin][NTf2], c: [dmin]-
[NTf2]; b) 1,4-dioxane solutions of c: [bmin][SbF6], c: [hmin][SbF6],
c: [moin][SbF6].
ACHTUNGTRENNUNG[NTf2],
size of aggregates, the presence of more extended aggre-
gates in [NTf2 ] ILs could be supposed. The difference in
G
N
ACHTUNGTRENNUNG
ꢀ
G
E
ACHTUNGTRENNUNG
band intensity was more significant for [moim+] ILs (see
Figure 2) and in these solvent media, the higher difference
in reaction rates was detected (see Figure 1). In particular,
the larger extent of aggregation seems to disfavor the reac-
tion.
AHCTUNGTRENNUNG
the balance between viscosity and structural organization
effect. Indeed, similar values of kII were detected, with the
exception of [hmim]ACHTNUTRGNE[NUG NTf2], despite of the corresponding ag-
The alkyl-chain length also seems to play an important
role. Indeed, for [NTf2 ] ILs, the band intensity, decreased
gregates size and viscosity values.
ꢀ
on going from [emim]
G
E
The anion effect: The probe reaction was also used in four
[bmim+] ILs differing in the size, symmetry, and coordina-
tion ability of the anion. Data reported in Table 1 (en-
tries 3–5, 15) show that kII values decrease along the series:
[BF4 ] > [PF6 ] > [NTf2 ] > [SbF6 ]. The experimental
trend seems to indicate that the reaction is favored by
anions having a higher symmetry and coordination ability,
on going from [hmim]
ACHTUNGTRENNUNG
ꢀ
For [SbF6 ] ILs, in the spectral range between 300 and
450 nm, where intensity variations seem to be more signifi-
ꢀ
ꢀ
ꢀ
ꢀ
cant, a decrease in band intensity going from [bmim]
to [moim][SbF6] was detected (Figure 2b).
[SbF6]
ACHTUNGTRENNUNG
Different results were obtained as a function of the IL
anion. Indeed, for [SbF6 ] ILs, the structural effect seems to
ꢀ
ꢀ
such as [BF4 ] and [PF6 ]. However, with the same anion
ꢀ
ꢀ
prevail on the viscosity effect. In this case, the increase in
alkyl-chain length, despite of the higher viscosity, induces a
decrease in the extent of aggregation giving rise to less or-
ganized solvent systems that favor the reaction. Differently,
symmetry ([PF6 ] and [SbF6 ]), the anion size plays also a
role. On the whole, these results agree with data previously
collected for 1b,[10] and evidence once again the importance
of the role played by the IL anion in reactions that proceed
through an ammonium charged transition state (i.e., its sta-
ꢀ
in the case of [NTf2 ] ILs, kII values seem to be a result of
7900
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7896 – 7902

Ionic Liquids Properties
FULL PAPER
bilization by means of the proton coordination on the am-
monium nitrogen).[2c,24] Furthermore, for 1a, a higher sensi-
tivity of the reaction to the nature of the IL anion can be
evidenced. Indeed, kACHTUGNRTEG[NNUN BF ]/kACHTUNGRTNEU[GN NTf ] = 1.6 and 1.9 for 1b and 1a,
respectively. Probably, for 1a, a later transition state, having
a more marked ammonium character, feels more the stabi-
lizing effect due to the IL anion.
A more articulate situation can be observed considering
the values of the activation parameters as a function of dif-
ferent cation structure. In this case, activation parameters
ꢀ
span in a narrow range. For [NTf2 ] ILs mean values of
4
2
¼
¼
DH =(41ꢂ3) kJmolꢀ1 and DS =(ꢀ158ꢂ7) JKꢀ1 molꢀ1
ꢀ
can be calculated, whereas for [SbF6 ] ILs mean values of
¼
¼
DH =(44ꢂ3) kJmolꢀ1 and DS =(ꢀ147ꢂ10) JKꢀ1 molꢀ1
can be calculated. These findings are not surprising, because
they confirm that in such solvents only weak interactions
are operating.
Activation parameters: It has been reported that, for kinet-
ics carried out in RTILs, a sharp curvature in Arrhenius or
Eyring plots could be observed as a consequence of some
significant structural changes in the RTIL.[25] Therefore, for
a careful analysis of the temperature effect the reaction was
carried out at five temperatures going from 293 up to 313 K.
In all cases considered, an excellent linear correlation of
Conclusions
The overall collected data confirm that ILs are solvents
which possess peculiar features. In particular, the rationali-
zation of their solvent effects heavily depends on the used
probe reaction. Indeed, although in some cases analysis by
using classical solvent parameters has been proven suitable
in order to understand ILs effects, in our case the whole of
parameters affecting the three-dimensional structure of sol-
vents must be taken into account in order to explain experi-
mental trends.
logACHTUNGTRENNUNG(kA,R/T) versus 1/T was obtained, indicating that the
above upsetting effect is not operating in the analyzed tem-
perature range, and that the calculated activation parame-
ters depend only on the nature of the elimination process.
In Table 2 activation parameter values collected for the mor-
pholino-elimination of 1a in different solvent media are re-
ported (data collected at different temperatures are avail-
able in Table S5 in the Supporting Information).
Firstly, we analyzed activation parameters values in COS.
The comparison between data collected in 1,4-dioxane and
in DMF shows that, enthalpy being equal, in the latter sol-
vent system the reaction is entropy favored. Activation pa-
rameters calculated in DMF are similar to those calculated
Experimental Section
Materials: Commercial 1,4-dioxane, [bmim]ACHTNUTRGENNUG[BF4], [bmim]ACHUTGNTNER[NUNG PF6], 4-nitroa-
niline, N,N-diethyl-4-nitroaniline, Nile Red, and Reichardtꢀs dye 30 were
purchased and used without further purification. Substrates 1a and 2a
were prepared according to a previously reported procedure.[26] [bmim]-
in [bmpyrr]ACHTUNGTRENNUNG[NTf2]. In all other cases, the comparison IL/
DMF solution shows that the morpholino-induced elimina-
tion of 1a is enthalpy disfavored, but entropy favored. En-
thalpy values in ILs solutions range from 33.6 kJmolꢀ1 up to
G
G
A
U
ACHTUNGTRENNUNG[NTf2],
R
A
U
ACHTUNGTRENNUNG
52.1 kJmolꢀ1,
while
entropy
values
range
from
R
G
ACHTUNGTRENNUNG
ꢀ128 JKꢀ1 molꢀ1 up to ꢀ173 JKꢀ1 molꢀ1.
thesis of the commercially available corresponding chloride and NaSbF6
or LiNTf2, according to a procedure previously reported.[27] [dmim][Cl],
[bmpip][Cl], and [bEt3N][Cl] were synthesized by the reacting N-methyli-
midazole, N-methylpiperidine, or triethylammine with an excess of the
haloalkane (1.4 equivalents). Upon completion of the reaction, there was
no evidence for the presence of unreacted amine. The excess of haloal-
kane was removed in vacuo. The chloride salts were purified by washing
For [bmim][X] ILs, both the changes in enthalpy and en-
tropy values can be rationalized on the grounds of different
cation–anion interactions and consequently, on the grounds
of different IL anion ability to stabilize the ammonium
charged transition state.
with diethyl ether. This step was repeated several times. [dmim]
[bmpip][NTf2], and [bEt3N]
ACHTUNGTRENNUNG
G
ACHTUNGTRENNUNG
Table 2. Activation parameter values for the morpholino-induced elimi-
nation reaction of 1a, in different solvents.
corresponding chloride and LiNTf2 according to a reported procedure.[27]
All ionic liquids were dried on a vacuum line at 608C for at least 2 h,
before use, then stored in a dryer under argon and over calcium chloride.
Morpholine was freshly distilled before use.
¼
¼
Entry
Solvent
DH /kJmolꢀ1
DS /JKꢀ1molꢀ1
1
2
3
4
5
6
7
8
1,4-dioxane
DMF
33ꢂ3
ꢀ208ꢂ9
ꢀ179ꢂ2
ꢀ128ꢂ7
ꢀ140ꢂ10
ꢀ148ꢂ1
ꢀ153ꢂ4
ꢀ150ꢂ10
ꢀ159ꢂ5
ꢀ173ꢂ6
ꢀ164ꢂ10
ꢀ157ꢂ4
ꢀ171ꢂ7
ꢀ149ꢂ2
ꢀ160ꢂ4
ꢀ161ꢂ7
ꢀ134ꢂ6
ꢀ146ꢂ6
32.3ꢂ0.7
52ꢂ2
Kinetic measurements and calculations: UV/Vis spectra and kinetic
measurements were carried out by using a spectrophotometer equipped
with a Peltier temperature controller, able to keep the temperature con-
stant within 0.1 K. The sample for a typical kinetic run was prepared by
mixing into a quarz cuvette (optical path 0.2 cm), 500 mL of IL, 50 mL of
a solution of substrate in 1,4-dioxane, and then 25 mL of a concentrated
solution of amine in 1,4-dioxane, previously thermostated. The substrate
concentration was constant and equal to 2.0ꢄ10ꢀ4 m; the morpholine con-
centration ranged from 8.7–22ꢄ10ꢀ3 m. All measurement solutions were
perfectly clear. The reactions were studied over at least three half-lives.
In all cases, the correlation coefficients were higher than 0.9998. The ap-
parent first-order rate constants obtained were reproducible within
ꢂ3%. All kinetic data were analyzed by means of the KALEIDA-
GRAPH 4.0 software package.
A
ACHTUNGTRENNUNG
A
U
48ꢂ3
A
U
44.6ꢂ0.2
43ꢂ1
A
ACHTUNGTRENNUNG
A
U
45ꢂ4
A
U
41ꢂ2
9
A
U
37ꢂ2
10
11
12
13
14
15
16
17
A
U
40ꢂ3
A
U
42ꢂ1
A
U
34ꢂ2
A
U
44.1ꢂ0.6
39ꢂ1
A
ACHTUNGTRENNUNG
A
U
41ꢂ2
A
U
48ꢂ2
Determination of solvent parameters: The determination of solvent pa-
rameters was carried out by mixing into a quartz cuvette (optical path
A
U
44ꢂ2
Chem. Eur. J. 2009, 15, 7896 – 7902
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
7901

R. Noto, F. D’Anna et al.
0.2 cm), 500 mL of IL and 75 mL of a concentrated solution of spectro-
scopic probe in 1,4-dioxane. The concentration of probe was equal to
2.0ꢄ10ꢀ4 m. The obtained solution was thermostated at 298 K.
vert, Helv. Chim. Acta 2000, 83, 2183–2191; i) F. Hollfelder, A. J.
Kirby, D. S. Tawfik, J. Org. Chem. 2000, 65, 5866–5874; j) P.
Lemma-Gray, Biotechnol. Bioeng. 2001, 74, 524–527; k) J. E. Klijn,
J. B. F. N. Engberts, J. Am. Chem. Soc. 2003, 125, 1825–1833; l) J. E.
Klijn, J. B. F. N. Engberts, Org. Biomol. Chem. 2004, 2, 1789–1799;
m) Y. Hu, K. N. Houk, K. H. Kikuchi, K. Hotta, D. Hilvert, J. Am.
Chem. Soc. 2004, 126, 8197–8205; n) F. P. Seebeck, D. Hilvert, J.
Am. Chem. Soc. 2005, 127, 1307–1312.
RLS measurements: Resonance light scattering measurements were car-
ried out on a spectrofluorophotometer employing a synchronous scan-
ning mode in which the emission and excitation monochromators were
preset to identical wavelengths. The resonance light scattering spectrum
was recorded from 300 to 600 nm setting both excitation and emission slit
widths at 1.5 nm.
[7] F. D’Anna, S. La Marca, R. Noto, J. Org. Chem. 2008, 73, 3397–
3403.
[8] a) C. Reichardt, Solvents and Solvent Effects in Organic Chemistry,
3rd ed., Wiley-VCH, Weinheim, 2003; b) P. K. Kipkemboi, A. J.
Easteal, Aust. J. Chem. 1994, 47, 1771–1781.
[9] M. J. Kamlet, J. L. M. Abboud, R. W. Taft, J. Am. Chem. Soc. 1977,
99, 6027–6038.
[10] Y. Marcus, Chem. Soc. Rev. 1993, 22, 409–416.
[11] M. J. Kamlet, R. W. Taft, J. Am. Chem. Soc. 1976, 98, 377–383.
[12] C. Reichardt, Chem. Rev. 1994, 94, 2319–2358.
Acknowledgements
We thank the University of Palermo and MUR (PRIN 20066034372) for
financial support.
[13] a) L. Crowhurst, P. R. Mawdsley, J. M. Perez-Arlandis, P. A. Salter,
T. Welton, Phys. Chem. Chem. Phys. 2003, 5, 2790–2794; b) C.
Reichardt, Green Chem. 2005, 7, 339–351; c) B. R. Mellein,
S. N. V. K. Aki, R. L. Ladewski, J. F. Brennecke, J. Phys. Chem. B
2007, 111, 131–138.
[14] a) D. W. Armstrong, L. He, Y. S. Liu, Anal. Chem. 1999, 71, 3873–
3876; b) J. Ding, V. Desikan, X. Han, T. L. Xiao, R. Ding, W. S.
Jenks, D. W. Armstrong, Org. Lett. 2005, 7, 335–337; c) M. Deetlefs;
K. R. Seddon, Chem. Today 2006, 24, 16–23; d) Z. Hu, C. J. Margu-
lis, Acc. Chem. Res. 2007, 40, 1097–1105; e) A. A. H. Pꢆdua, M. F.
Costa Gomes, J. N. A. Canongia Lopes, Acc. Chem. Res. 2007, 40,
1087–1096.
[15] M. Blesic, A. Lopes, E. Melo, Z. Petrovski, N. V. Plechkova, J. N.
Canongia Lopes, K. R. Seddon, L. P. N. Rebelo, J. Phys. Chem. B
2008, 112, 8645–8650.
[16] W. J. Steed, J. L. Atwood in Supramolecular Chemistry, Wiley-VCH,
Weinheim, 2000, pp. 26–28.
[17] J. G. Huddleston, A. E. Visser, W. M. Reichert, H. D. Willauer,
G. A. Broker, R. D. Rogers, Green Chem. 2001, 3, 156–164.
[18] K. Baba, H. Ono, E. Itoh, S. Itoh, K. Noda, T. Usui, K. Ishihara, M.
Inamo, H. D. Takagi, T. Asano, Chem. Eur. J. 2006, 12, 5328–5333.
[19] a) R. A. Mantz, P. C. Trulove in Ionic Liquids in Synthesis (Eds.: P.
Wasserscheid, T. Welton), Wiley-VCH, Weinheim, 2008, pp. 72–88;
b) A. Ahosseini, A. M. Scurto, Int. J. Thermophys. 2008, 29, 1222–
1243.
[20] R. F. Pasternack, P. J. Collings, Science 1995, 269, 935–939.
[21] See for example: a) R. F. Pasternack, K. F. Schaefer, Inorg. Chem.
1994, 33, 2062–2065; b) Y. Liu, M. Fan, S. Zhang, J. Yao, J. Phys.
Org. Chem. 2007, 20, 874–877; c) Y. T. Wang, W. J. Jin, Spectrochim.
Acta Part A 2008, 70, 871–877; d) J. J. Wu, N. Li, K. A. Li, F. Liu, J.
Phys. Chem. B 2008, 112, 8134–8138.
[22] See for example: a) D. H. Cheng, X. W. Chen, J. H. Wang, Z. L.
Fang, Chem. Eur. J. 2007, 13, 4833–4839; b) J. Li, M. Li, X. Li, J.
Tang, J. Kang, H. Zhang, Y. Zhang, Colloids Surf. B 2008, 67, 79–
84; c) J. Zheng, X. Wu, M. Wang, D. Ran, W. Xu, J. Yang, Talanta
2008, 74, 526–532.
[23] H. Zhang, K. Li, H. Liang, J. Wang, Colloids Surf. A 2008, 329, 75–
81.
[24] L. Crowhurst, N. L. Lancaster, J. M. Pꢇrez-Arlandis, T. Welton, J.
Am. Chem. Soc. 2004, 126, 11549–11555.
[25] a) A. J. Arduengo III, R. L. Harlow, M. Kline, J. Am. Chem. Soc.
1991, 113, 361–363; b) S. T. Handy, M. Okello, J. Org. Chem. 2005,
70, 1915–1918.
[26] a) H. Lindemann, H. Thiele, Justus Liebigs Ann. Chem. 1926, 449,
63; b) F. Hollefelder, A. J. Kirby, D. S. Tawfik, K. Kikuchi, D. Hil-
vert, J. Am. Chem. Soc. 2000, 122, 1022–1029.
[1] C. Reichardt, Solvents and Solvent Effects in Organic Chemistry, 3rd
ed., Wiley-VCH, Weinheim, 2003, pp. 136–295.
[2] a) F. D’Anna, V. Frenna, R. Noto, V. Pace, D. Spinelli, J. Org.
Chem. 2005, 70, 2828–2831; b) F. D’Anna, V. Frenna, V. Pace, R.
Noto, Tetrahedron 2006, 62, 1690–1698; c) F. D’Anna, V. Frenna, R.
Noto, V. Pace, D. Spinelli, J. Org. Chem. 2006, 71, 9637–9642; d) F.
D’Anna, V. Frenna, R. Noto, V. Pace, D. Spinelli, J. Org. Chem.
2006, 71, 5144–5150; e) F. D’Anna, V. Frenna, S. La Marca, R.
Noto, V. Pace, D. Spinelli, Tetrahedron 2008, 64, 672–680; f) F.
D’Anna, S. Marullo, R. Noto, J. Org. Chem. 2008, 73, 6224–6228.
[3] a) P. N. Naik, S. J. Nara, J. R. Harjani, M. M. Salunke, Can. J. Chem.
2003, 81, 1057–1060; b) J. Mc Nulty, J. J. Nair, S. Cheekoori, V. Lari-
chev, A. Capretta, A. J. Robertson, Chem. Eur. J. 2006, 12, 9314–
9322; c) P. Nockemann, B. Thujs, S. Pittois, J. Thoen, C. Glorieux, K.
Van Hecke, L. Van Meervelt, B. Kirchner, K. Binnemans, J. Phys.
Chem. B 2006, 110, 20978–20992; d) T. P. Wells, J. P. Hallett, C. K.
Williams, T. Welton, J. Org. Chem. 2008, 73, 5585–5588.
[4] a) R. Bini, C. Chiappe, V. Llopsis Mestre, C. S. Pomelli, T. Welton,
Org. Biomol. Chem. 2008, 6, 6437–6440; b) G. Ranieri, J. P. Hallett,
T. Welton, Ind. Eng. Chem. Res. 2008, 47, 638–644; c) P. M. Manci-
ni, G. G. Fortunato, C. G. Adam, L. R. Vottero, J. Phys. Org. Chem.
2008, 21, 87–95.
[5] a) C. G. Hanke, A. Johansson, J. B. Harper, R. M. Lynden-Bell,
Chem. Phys. Lett. 2003, 374, 85–90; b) J. D. Holbrey, W. M. Reich-
ert, M. Nieuwenhuyzen, O. Sheppard, C. Hardacre, R. D. Rogers,
Chem. Commun. 2003, 476–477; c) J. Dupont, J. Braz. Chem. Soc.
2004, 15, 341–350; d) J. B. Harper, R. M. Lynden-Bell, Mol. Phys.
2004, 102, 85–94; e) C. S. Consorti, P. A. Suarez, . R. F. de Souza,
R. A. Burrow, D. H. Farrar, A. J. Lough, W. Loh, L. H. M. da Silva,
J. Dupont, J. Phys. Chem. B 2005, 109, 4341–4349; f) M. Deetlefs,
C. Hardacre, M. Nieuwenhuyzen, O. Sheppard, A. K. Soper, J. Phys.
Chem. B 2005, 109, 1573–1598; g) J. Dupont, P. A. Z. Suarez, Phys.
Chem. Chem. Phys. 2006, 8, 2441–2452; h) J. Lachwa, I. Bento,
M. T. Duarte, J. N. C. Lopes, L. P. N. Rebelo, Chem. Commun. 2006,
2445–2447; i) R. M. Lynden-Bell, M. G. Del Pꢅpolo, T. G. A.
Youngs, J. Kohanoff, C. G. Hanke, J. B. Harper, C. C. Pinilla, Acc.
Chem. Res. 2007, 40, 1138–1145; j) C. Hardacre, J. D. Holbrey, M.
Nieuwenhuyzen, T. G. A. Youngs, Acc. Chem. Res. 2007, 40, 1146–
1155; k) S. Tiwari, N. Khypse, A. Kumar, J. Org. Chem. 2008, 73,
ˇ
ˇ
´
9075–9083; l) M. Meciarovꢆ, M. Cigꢆn, S. Toma, A. Gꢆplovsky, Eur.
J. Org. Chem. 2008, 4408–4411.
[6] a) M. L. Casey, D. S. Kemp, K. G. Paul, D. D. Cox, J. Org. Chem.
1973, 38, 2294–2301; b) D. S. Kemp, D. D. Cox, K. G. Paul, J. Am.
Chem. Soc. 1975, 97, 7312–7318; c) K. K. S. N. Thorn, D. Hilvert, J.
Am. Chem. Soc. 1996, 118, 8184–8185; d) F. Hollfelder, A. J. Kirby,
D. S. Tawfik, J. Am. Chem. Soc. 1997, 119, 9578–9579; e) P. G.
McCraken, C. G. Ferguson, D. Vizitiu, C. S. Walkinshaw, Y. Wang,
G. R. J. Thatcher, J. Chem. Soc. Perkin Trans. 2 1999, 911–912; f) J.
Perꢇz-Juste, F. Hollfelder, Org. Lett. 2000, 2, 127–130; g) F. Holl-
felder, A. J. Kirby, D. S. Tawfik, K. Kikuchi, D. Hilvert, J. Am.
Chem. Soc. 2000, 122, 1022–1029; h) K. Hotta, K. Kikuchi, D. Hil-
[27] L. Cammarata, S. G. Kazarian, P. A. Salter, T. Welton, Phys. Chem.
Chem. Phys. 2001, 3, 5192–5200.
Received: January 19, 2009
Published online: June 27, 2009
7902
www.chemeurj.org
ꢃ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2009, 15, 7896 – 7902