Full Paper
we again found a size+polarizability anion effect. In summary,
+
+
the change from [BMIM] to [EMIM] reveals that the best IL
media are those bearing small and highly polarizable anions,
independent of the cation present in the IL: compare, for in-
ꢀ
1
ꢀ1
ꢀ1 ꢀ1
stance, k =0.47 M
s
and 0.44 M
s
for [BMIM]DCN and
N
[
EMIM]SCN, respectively.
On the other hand, previous studies have shown that
BMIM]DCN is a highly basic IL that in some reactions can pres-
[
ent catalytic properties, thereby qualifying it as a task-specific
[29]
solvent for S Ar reactions. Therefore, the possibility of estab-
N
lishing an advanced bond between the hydrogen atom of the
amine moiety of morpholine and the nitrogen atom of dicya-
namide is highly expected. As a result, the nucleophilicity of
the nitrogen atom in morpholine is enhanced. Note further
that the rate coefficient values in [BMIM]DCN with respect to
[
EMIM]DCN or [BMPyr]DCN is increased by approximately two
+
times. On the other hand, [EMIM] cations present significant
differences in electron delocalization patterns and number of
+
nitrogen atoms compared with [BMpyr] cations, yet the rate
coefficients are similar. In other words, the reactivity observed
in this system suggests that the “anion effect” outweighs the
“
cation solvent effect”. These results confirm the hypothesis
Scheme 1. Chemical structures for the series of cations and anions of the
ionic liquids used in this study.
proposed by Harper et al. in the sense that the structure of the
IL may become a determinant factor affecting the S Ar reactivi-
N
[12]
ty.
To interpret the solvation effects on the recorded kinetic re-
Table 1. Rate coefficient values for the reaction of morpholine with 1-
chloro-2,4-dinitrobenzene in ionic liquids at 258ꢁ0.18C.
sponses of the title reaction, we performed two different analy-
sis based on solvatochromic and NMR chemical shift measure-
ments. The first one, is one of the preferred models to analyze
ꢀ
1
ꢀ1
ꢀ1 ꢀ1
Ionic liquid
k
N
[m
s
]
Ionic liquid
N
k [m s ]
[16–25]
[
[
[
[
[
[
BMIM]DCN
BMIM]SCN
0.47ꢁ0.01
0.32ꢁ0.01
0.30ꢁ0.01
[EMIM]DCN
[EMIM]SCN
0.29ꢁ0.01
0.44ꢁ0.02
solvent effects in ILs.
It is based on a multivariate empirical
equation that includes the following parameters: hydrogen
bond acidity (a), measuring the ability of the IL to donate a HB
to the substrate; hydrogen bond basicity (b), measuring the
ability of the IL to accept a HB from the substrate; and a pa-
rameter p*, measuring the dipolar polarizability of the IL. Prior
to applying this technique to analyze the effect of the solvent
ꢀ
3
BMIM]CF
BMIM]PF
BMIM]NTf
3
SO
3
[EMIM]CF
[EMIM]NTf
[EMIM]FAP
3
SO
3
0.15ꢁ10
0.07ꢁ10
ꢀ
3
ꢀ3
3
6
0.09ꢁ10
2
0.02ꢁ10ꢀ
2
0.07ꢁ0.01
0.02ꢁ10
ꢀ
4
BMIM]FAP
[BMPyr]DCN
0.25ꢁ0.01
0.14ꢁ0.01
3
BM
2
IMNTf
2
0.08ꢁ10ꢀ
[BMPyr]CF
3
SO
3
0.06ꢁ10ꢀ
3
[BMPyr]NTf
2
on the kinetic responses observed for the model S Ar reaction,
N
we shall comment on the limitations that this model presents.
[30]
The main drawback was nicely explained by Welton et al.,
kN, is about 23.5 times faster for [BMIM]DCN compared with
BMIM]FAP, the IL containing the biggest anion of the series.
Note also that the comparison between [BMIM]DCN with
who emphasized that, in general, Kamlet–Taft (KT) parameters
miss the main ion–ion interactions expected for any substrate
dissolved in an IL at ambient conditions, because the probes
used to temperate the KT parameters are at most zwitterionic,
not ionic. This result is relevant as it affects the completeness
of the model severely because the best ILs are those that are
associated to a very low extent. In this sense, independent of
the technical quality of the KT parameters, all of them bear
a caveat emptor warning for all the potential users of these
scales. The second weakness of the KT technique is related to
the interpretation of the a and b parameters. Whereas a is as-
sociated with the HB acidity of the IL, and therefore associated
with the cation forming the IL, the b parameter is always asso-
ciated with the anion component of the IL. This simplistic as-
signation has been recently put into jeopardy, because in
[
[
BMIM]SCN gives only an improvement factor of 1.5 in favor of
ꢀ
DCN . This result suggests that, apart from anion size, there is
an additional effect that can be attributed to the electronic po-
ꢀ
larizability: DCN is expected to be significantly more polariza-
ꢀ
ble than SCN owing to the presence of a p-electron-rich
zone. Note finally that the comparison between [BMIM]NTf2
and [BM IM]NTf emphasizes the observed anion effect within
2
2
this series: the effect of blocking the acidic hydrogen atom at
C2 results in almost identically kinetic responses. This result
suggests that the probable “cation solvent effect” is marginal.
+
Within the [EMIM] series, some similar trends are main-
+
tained with respect to the [BMIM] series. Here, the main
+
change is the shortening of the alkyl chain at the cation, which
results in an inversion in the values of the rate coefficient in
[BMIM] based ILs, both the HB acidity and HB basicity may be
located at the cation, an effect that is, however, anion-depen-
ꢀ
+
[31]
favor of SCN this time. Finally, within the series of [BMPyr ],
dent through polarization effects.
&
&
Chem. Eur. J. 2016, 22, 1 – 6
2
ꢂ 2016 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!