Correlating ionic liquid solvent effects with solvent parameters for a reaction that proceeds through a xanthylium intermediate

Article abstract of DOI:10.1039/c9ob01807g

A unimolecular nucleophilic substitution reaction that proceeds through a xanthylium carbocation was studied in seven ionic liquid solvents. It was found that the general trend in the rate constant with changing proportion of ionic liquid in the reaction mixture was different to that seen for other unimolecular processes, with the rate constant increasing as more ionic liquid was added to the reaction mixture. A significant correlation was found between the natural logarithm of the rate constant and a combination of the Kamlet-Taft solvent parameters. This relationship indicated that the principal interaction involved hydrogen bonding between the ionic liquid and some species along the reaction coordinate. Further, this correlation enables prediction of the effects that other ionic liquids will have on this, and other, reactions that proceed through a similar intermediate.

Full text of DOI:10.1039/c9ob01807g

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Correlating ionic liquid solvent effects with solvent
parameters for a reaction that proceeds through a
xanthylium intermediate†
Cite this: Org. Biomol. Chem., 2019,
7, 9336
1
a
b
a
a
Alyssa Gilbert,
Götz Bucher,
Ronald S. Haines
and Jason B. Harper
*
A unimolecular nucleophilic substitution reaction that proceeds through a xanthylium carbocation was
studied in seven ionic liquid solvents. It was found that the general trend in the rate constant with chan-
ging proportion of ionic liquid in the reaction mixture was different to that seen for other unimolecular
processes, with the rate constant increasing as more ionic liquid was added to the reaction mixture. A sig-
nificant correlation was found between the natural logarithm of the rate constant and a combination of
the Kamlet–Taft solvent parameters. This relationship indicated that the principal interaction involved
hydrogen bonding between the ionic liquid and some species along the reaction coordinate. Further, this
correlation enables prediction of the effects that other ionic liquids will have on this, and other, reactions
that proceed through a similar intermediate.
Received 16th August 2019,
Accepted 7th October 2019
DOI: 10.1039/c9ob01807g
rsc.li/obc
1
2
the rate and selectivity, compared to traditional solvents.
One such example is the unimolecular nucleophilic substi-
Introduction
Ionic liquids have gained much attention in recent decades as tution (S 1) reaction between xanthydrol 1 and indole 2
N
1
13
alternatives to traditional solvents for organic synthesis. Ionic (Scheme 1). Previously, this reaction was performed in a
liquids, with a melting point below 100 °C, are generally com- range of molecular solvents as well as two ionic liquids:
2
posed of bulky, charge diffuse ions, which frustrate crystallisa- 1-butyl-3-methylimidazolium tetrafluoroborate ([bmim][BF ], 4)
4
3
tion and give them favourable properties compared to tra- and
1-butyl-3-methylimidazolium
hexafluorophosphate
], 5) (Fig. 1). It was found that in both cases where
flammability. There is also the potential to recycle ionic ionic liquids were used as solvents, the extent of conversion in
liquids after use, leading them to be considered as green sol- a given time was much greater than in any of the molecular
vents. Additionally, the ions can be varied in order to tune solvents, and the ionic liquids could be easily recycled without
4
ditional solvents such as negligible vapour pressure and non- ([bmim][PF
6
5
6
their physical properties, such as viscosity and miscibility, loss of yield after repeated use. No rationale was provided for
7
giving them the popular term ‘designer solvents’.
the observed increase in the rate of reaction.
Whilst ionic liquids are yet to be used on a wide scale for
In order for ionic liquids to achieve a more widespread use,
organic synthesis, there are many examples (see reviews by it is necessary to be able to understand the effects a given
8
9,10
Hallet and Welton, and Harper et al. ) where they have ionic liquid will have on a process. Some work has been
been used in place of traditional solvents due to their ability to carried out investigating how ionic liquids behave as solvents
1
1
14–19
be recycled, and as they can affect reaction outcome, such as for S 1 reactions,
however the outcome varies with the
N
nature of the substrate and the ionic liquid. Therefore, it is of
a
School of Chemistry, University of New South Wales, Sydney, NSW, 2052, Australia.
E-mail: j.harper@unsw.edu.au; Fax: +612 9385 6141; Tel: +61 2 9385 4692
b
WestCHEM, School of Chemistry, University of Glasgow, Joseph Black Building,
University Avenue, Glasgow G12 8QQ, UK
†
Electronic supplementary information (ESI) available: Experimental procedures
and NMR spectra for all synthesised compounds; details of the effects of ionic
liquids 4 and 5 on the reaction of species 1 and 2; attempted synthesis of acti-
vated xanthenes and a proposed mechanism to explain the observed products,
experimental details, rate constant data and stock solution compositions for
kinetic analyses; additional mole fraction dependence plot, Eyring plot, nucleo-
phile dependence plot and Kamlet–Taft correlation plots; complete multivariate
regression analyses. See DOI: 10.1039/c9ob01807g
Scheme 1 The reaction between xanthydrol 1 and indole 2 which has
previously been studied in a number of solvents including ionic liquids.
13
9336 | Org. Biomol. Chem., 2019, 17, 9336–9342
This journal is © The Royal Society of Chemistry 2019

View Article Online
Organic & Biomolecular Chemistry
Paper
Results and discussion
In the original study reported above, the ionic liquids
bmim][BF ] 4 and [bmim][PF ] 5 were used as solvents for the
reaction between xanthydrol 1 and indole 2; the largest extent of
[
4
6
13
conversion was reported when salt 4 was used. Initial studies
here (for full details, see ESI†) suggested that the rate constant
increase was marked; approximately 30-fold on moving from
Fig. 1 The ionic liquids used in this and the previous study for the reac-
4 6
tion shown in Scheme 1, ([bmim][BF ] 4 and [bmim][PF ], 5).
6
acetonitrile to [bmim][PF ] 5 and a further 30-fold on moving to
[
bmim][BF ] 4 (both salts χ = 0.99). Investigating the effect of the
4
interest to better understand what is occurring at the mole-
cular level in order to be able to predict the effects that
different ionic liquids will have on this reaction type.
proportion of the salt 4 in the reaction mixture gave an unusual
concentration dependence (Fig. S1, ESI†) with the rate constant
effectively constant on addition of any amount of ionic liquid 4
Previously, mole fraction dependence studies have been
used to identify how mixtures of ionic liquids in molecular sol-
(χ ≥ 0.02). However, these studies were complicated by the insolu-
1
4,19
bility of the substrate 1 (particularly compared to the concen-
vents affect the rate constant for a given process.
In the
13
trations reportedly used previously, ‡) and its notable acid
sensitivity.§ As such, a more reactive substrate was considered.
A variety of leaving groups (including chloride, bromide and
mesylate) on the xanthene were considered. However, difficulties
associated with the isolation of these reactive species (details of
attempted synthesis, along with a proposed mechanism, and
computational support for that mechanism are given in the ESI†)
led to consideration of the substrate 11, which was both isolable
and stable. The reaction proceeds as outlined in Scheme 2 with
nucleophile dependence studies showing that the reaction between
cases studied, it was found that for S 1 reactions the
maximum rate constant increase occurred when the reaction
mixture contained a small proportion of ionic liquid (ca.
N
2
–20 mol%); it was not necessary to use large amounts of ionic
liquids to get the greatest increase in the rate constant. In
2
0,21
addition, both activation parameters
parameters
and solvent specific
1
9,22,23
have been used to understand how the ionic
liquid is interacting with species along the reaction coordinate
to lead to these changes in the rate constant.
The work described herein will focus on identifying what
causes the increase in the rate of reaction shown in Scheme 1
when using an ionic liquid compared to a molecular solvent. In
addition, this study will consider how changing the constituents
of the ionic liquid affects reaction outcome, particularly focus-
ing on the ionic liquids 6–10 (Fig. 2). This, in turn, allows for
prediction of the effect a given ionic liquid will have on this,
and similar reactions, and will aid in extending the established
literature of the effects of ionic liquids as solvents for organic
reactions.
the acetate 11 and indole 2 occurs through an S 1 mechanism
N
(
see ESI†) allowing for comparison with previous S
N
1 cases.
Moreover, sensitivity to any trace of acid was not observed.
Mole fraction dependence studies were then performed to
determine the rate constant for the reaction between the
acetate 11 and indole 2 in various proportions of acetonitrile
and [bmim][BF ] 4 (Fig. 3).¶ The trend seen in this plot is
4
different to that seen previously for reactions of this type; pre-
viously, where the highest increase in the unimolecular rate
constant, k , occurred at low proportions of the ionic liquid in
1
the reaction mixture, with a decrease as the amount of ionic
1
4,17,19
liquid in the reaction mixture was increased.
In this case
Scheme 2 The reaction between xanthydrol acetate 11 and indole 2
considered herein.
Fig. 2 The ionic liquids considered in this work alongside those intro-
duced above in Fig. 1: 1-butyl-3-methylimidazolium trifluoromethane-
sulfonate ([bmim][OTf], 6), 1-butyl-3-methylimidazolium bis(trifluoro-
‡Numerous attempts to replicate the amounts used in the original study were
undertaken using higher temperatures and vigorous mixing but were unsuccess-
ful in each case.
methanesulfonyl)imide ([bmim][N(SO
imidazolium tetrafluoroborate ([bm
pyrrolidinium bis(trifluoromethanesulfonyl)imide ([bmpyr][N(SO
), and 1-butyl-2,3-dimethylimidazolium bis(trifluoromethanesulfonyl)
imide ([bm im][N(SO CF ], 10).
2
CF
3
)
2
], 7), 1-butyl-2,3-dimethyl-
], 8), 1-butyl-1-methyl-
CF ],
§This sensitivity was briefly considered in the initial study as the cause of the
increase in observed yields due to acid being a hydrolysis product of both ionic
liquids 4 and 5. However it was neither definitively ruled out for the ionic
liquids nor considered in the other solvents.
2
im][BF
4
2
3 2
)
9
2
2
3
)
2
¶Although the uncertainties are large in this case, the trends observed are clear.
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem., 2019, 17, 9336–9342 | 9337

View Article Online
Paper
Organic & Biomolecular Chemistry
Table 1 The activation parameters determined for the reaction
between the acetate 11 and indole 2 in either acetonitrile or mixtures of
4
[bmim][BF ] 4 and acetonitrile
‡
−1 a
ΔS /J K mol−1 a
‡
−1
Solvent
CH CN
χ
4
ΔH /kJ mol
3
0
0.20
0.96
70.4 ± 2.3
70.3 ± 4.1
77.5 ± 5.2
−121 ± 7
−101 ± 12
−76 ± 16
[bmim][BF
bmim][BF
4
] 4
] 4
[
4
a
Uncertainties reported were obtained from the fit of the linear
regression for the respective Eyring plots.
due to the potential for cation–π interactions between the
cation of the ionic liquid and the acetate 11, along with poten-
] 4 ( ) in acetonitrile at 73.8 °C. Uncertainties reported were tial interactions with the acetate group.
4
1
Fig. 3 The dependence of the unimolecular rate constant (k ) of the
reaction between the acetate 11 and indole 2 on the mole fraction of
bmim][BF
obtained from the standard deviation of triplicate measurements.
[
While these temperature dependence studies give some
insight into what is likely occurring at a molecular level in the
reaction of substrate 11 in the ionic liquid 4, the activation
while the majority of the rate constant increase occurs at low parameters determined do not allow for a definitive expla-
mole fractions of ionic liquid (up to χ = 0.2), k continues to nation as to the particular microscopic interactions that lead
1
increase as the amount of ionic liquid in the reaction mixture to the change in k1 seen when moving from a molecular
is increased, albeit to a smaller degree at higher mole frac- solvent to the ionic liquid 4. Such interactions are of interest
tions; a trend that has been observed previously for bimolecu- as they might be used to predict the effect that a given ionic
2
4
lar nucleophilic substitution reactions and condensation liquid would have on this reaction; these interactions might be
2
5,26
processes.
1
The largest k is fifteen times that in aceto- elucidated by considering structural variations of the ionic
nitrile, although it can be seen that addition of any amount of liquid. [Bmim][PF ] 5 was selected due to its use in the pre-
6
ionic liquid in the reaction mixture leads to a significant vious study but also because it contains a different anion com-
increase in the rate constant.
pared to ionic liquid 4; other ionic liquids (Fig. 2) were also
In order to understand these changes in k when using the considered as they differed in the constituents of the ionic
1
ionic liquid 4, temperature dependence studies were per- liquid and therefore may give an indication as to how each
formed to determine the activation parameters for the process. component is contributing to the change in rate constant.
Studies were performed for the reaction in acetonitrile and
bmim][BF ] 4 at χ = 0.20 and 0.96 in order to look at the [bmim][BF ] 4 showed the majority of the rate constant
effects of both low and high mole fractions of ionic liquid and increase had occurred by χ ca. 0.2, with little change at higher
As the rate constant data in mixtures containing
[
4
4
4
1
7,19
to allow comparison to previous S 1 studies.
Table 1 mole fractions, the reaction of the acetate 11 in mixtures con-
shows the activation enthalpies and entropies at these solvent taining the ionic liquids 5–10 was studied only at low (χ ca.
compositions.
0.20) and high (χ ca. 0.96) mole proportions of salt in the reac-
N
Comparison of the activation parameters between each tion mixture. For clarity, Table 2 shows the rate constants
solvent composition found no change in the enthalpies of acti- determined at each of these solvent compositions in these
vation between that of acetonitrile and [bmim][BF
χ = 0.20 and 0.96∥ and an increase in the entropy of activation ing all data can be found in Fig. S4 in the ESI†). Immediately
on going from acetonitrile to [bmim][BF ] 4 at χ = 0.96. This apparent is that the rate constants observed are in all cases
increase in the entropy of activation indicates that the ionic larger than in acetonitrile (k = (2.35 ± 0.30) × 10
4
] 4 at both ionic liquids 4–10 (the mole fraction dependence plot contain-
4
−
4
−1
1
s
), with
liquid is ordering about the starting material 11 and that this little (if any) difference between the two different proportions
ordering is disrupted as the reaction proceeds, leading to the of the ionic liquid in the reaction mixture. The trend seen in
entropy change observed.** This interaction is not unexpected the rate constants as constituents of the ionic liquid are varied
is generally the same at both low and high mole fractions of
salt in the reaction mixture; that is, k
1
generally decreases in
3 2
CF ) ]
the order: [bmim][OTf] 6, [bmim][BF ] 4, [bmim][N(SO
4
2
∥
Due to the large uncertainties obtained for the activation parameters, there can
be no discussion on the difference between the acetonitrile and ionic liquid
cases due to enthalpy changes. These uncertainties are the result of variation in [bm
the initial k data, as discussed above.
*The alternative to this explanation is that the ionic liquid is interacting less
7
, [bmim][PF ] 5, [bm im][BF ] 8, [bmpyr][N(SO CF ) ] 9,
6
2
4
2
3 2
2
im][N(SO
2
CF
3
)
2
] 10. Analysis of this trend can be split into
1
two parts: the effect of the cation and the effect of the anion.
*
Focussing first on the effect of the cation, the trend in k
1
+
with the transition state than acetonitrile. This argument is considered unlikely
given the charged nature of the ionic liquid and the extent of charge delocalisa-
tion that would be expected in the transition state for this type of process. See,
−
when the [N(SO
2
CF
3
)
2
] anion is present is as follows: [bmim]
+
+
>
[bmpyr] > [bm
2
im] . This order suggests that to have a large
15,17
for example, previous work on similar processes.
increase in k relative to the molecular solvent, both a deloca-
1
9338 | Org. Biomol. Chem., 2019, 17, 9336–9342
This journal is © The Royal Society of Chemistry 2019

View Article Online
Organic & Biomolecular Chemistry
Paper
Table 2 The rate constants determined for the ionic liquids 4–10 at the trend in this parameter again follows that seen in the rate
both low and high mole fractions for the reaction between the acetate
constant above. While there was a reasonable correlation
between the natural logarithm of k and the α parameter at
1
11 and indole 2 in acetonitrile at 73.8 °C
/10−4 s−1 a
k
1
both low and high proportions of ionic liquid in the reaction
mixture (R = 0.78 and 0.80, respectively; Fig. S6 and S7†), cor-
Ionic liquid
χ
salt
2
[
[
[
[
[
[
[
bmim][OTf] 6
bmim][BF ] 4
bmim][N(SO
bmim][PF ] 5
bm im][BF ] 8
bmpyr][N(SO CF
bm im][N(SO CF
0.20
.96
0.20
.97
0.20
.95
0.19
.97
0.20
.95
0.19
.96
0.19
.92
15.3 ± 0.2
17.4 ± 0.8
10.4 ± 1.4
14.7 ± 3.2
7.3 ± 0.6
11.3 ± 2.1
6.7 ± 0.1
9.7 ± 1.3
4.3 ± 0.3
6.5 ± 1.6
3.6 ± 0.5
3.8 ± 0.1
2.5 ± 0.2
2.6 ± 0.5
1
relations between the natural logarithm of k and either the β
0
2
or π*‡‡ parameters were very poor in all cases (R < 0.40;
Fig. S8–S11†).§§ However the trend observed above with the
anions suggests some contribution from the β parameter.
Due to these observations, multivariate regression analysis
was considered for mixtures containing either low or high
mole fractions of salt to determine whether a better correlation
could be obtained. This methodology has been used previously
4
0
2
3 2
CF ) ] 7
0
6
0
2
4
0
2
3 2
) ] 9
0
with Kamlet–Taft parameters to determine ionic liquid
19,32
2
2
3 2
) ] 10
effects.
Here, the following equation was used:
0
a
lnðk1Þ ¼ intercept þ aα þ bβ þ cπ*
ð1Þ
Uncertainties reported were obtained from the standard deviation of
triplicate measurements.
where a, b and c are coefficients of each parameter. In this
study, multivariate regression analyses were used (as described
3
3
by Welton et al. ) to determine any correlation between the
change observed in the rate constants and the Kamlet–Taft
parameters for each ionic liquid 4–10, as well as the relative
contributions of each parameter. A correlation was determined
to be significant when p ≤ 0.05 for each coefficient, including
the intercept (for full analysis, see ESI†). Focussing first on
solvent mixtures with a small amount of ionic liquid present
+
+
lised charge ([bmim] vs. [bmpyr] ) and the presence of an
+
+
acidic proton ([bmim] vs. [bm im] ) are beneficial. Such a
2
trend is consistent with the potential interactions described
above (which would involve the cation interacting with the π
system or the acetate). In the cases of ionic liquids with
+
different anions, all containing the [bmim] cation, a more
1
coordinating anion results in an increase in k as the trend
(
χ ca. 0.20), an excellent correlation was found:
follows [OTf]− > [BF4] > [N(SO CF ) ] ≈ [PF ] . Interestingly,
−
−
2
3 2
6
this suggests an anionic interaction not proposed earlier.
Overall, these trends†† indicate that there is the potential to
correlate these ionic liquid effects on the rate constant with
solvent specific parameters.
lnðk Þ ¼ ꢀ10:58 þ 4:14α þ 3:17β
ð2Þ
1
with p-values of 2.07 × 10−7, 6.79 × 10−5 and 4.8 × 10−4 for the
intercept, α and β, respectively. This correlation can be seen
pictorially in Fig. 4. This outcome shows that both the hydro-
gen bond donating and accepting ability of the solvent affect
, with the donating ability slightly more dominant in control-
ling the rate constant. Therefore, in order to observe a large
increase in k relative to the molecular solvent, an ionic liquid
able to hydrogen bond donate and accept to a high degree is
required.
In solvent mixtures where the ionic liquid is almost solely
diluted by reagents (χ ca. 0.96), multivariate regression ana-
lyses were again used in the same manner and an excellent
correlation was again observed:
Kamlet–Taft solvent parameters have been used in previous
studies as a tool to correlate changes in the reaction outcome
when using ionic liquids as solvents (with good correlations
found with both the hydrogen bond accepting ability of
k
1
1
9,27,28
1
the solvent and the polarizability).
These solvent para-
meters can be determined for any type of solvent but are quite
useful in ionic liquid studies as they also give an indication as
to how each component of the ionic liquid is behaving. The
solvent parameters include α, the hydrogen bond donating
29
ability of the solvent (often associated with the cation of the
ionic liquid); β, the hydrogen bond accepting ability of the
3
0
solvent (often associated with the anion of the ionic liquid);
lnðk1Þ ¼ ꢀ12:91 þ 4:92α þ 2:52β þ 2:33π*
ð3Þ
3
1
and π*, the polarisability of the solvent (associated with both
constituents of the ionic liquid).
with p-values of 6.59 × 10− , 0.00065, 0.020 and 0.05, respect-
ively. Interestingly in this case, the contribution from β
decreases relative to α and the π* contribution becomes more
significant.¶¶ This change indicates that while the hydrogen
5
Given the trends that are summarised in Table 2, it was
expected that the α parameter of the solvent would be impor-
+
+
tant as this value increases in the order [bm im] < [bmpyr] <
[
2
+
bmim] ; the same trend as that seen above in the rate con-
stant data. Additionally, given the trend seen for the anion, it
was also expected that β may correlate to the changes in k as the variation of the cation and anion of the ionic liquid.
‡
‡This lack of correlation is unsurprising given the interactions indicated from
1
§
§All Kamlet–Taft parameters used in this study were taken from Crowhurst
43
et al.
†
†The exception in each case is [bm
2
im][BF
4
] 8, which does not fit the trends for
¶¶This correlation also includes the acetonitrile data as Kamlet–Taft parameters
are known for this solvent and can be directly compared to the χ = 0.96 case
where ionic liquids 4–10 are diluted almost solely by reagents only.
either cation or anion variation, further indicating that both constituents of the
ionic liquid are responsible for the changes in reaction outcome.
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem., 2019, 17, 9336–9342 | 9339

View Article Online
Paper
Organic & Biomolecular Chemistry
1
9
methane. In that example, the correlation between the rate
constant and the Kamlet–Taft solvent parameters was exclu-
sively with the π* parameter. That correlation indicated that
the change in reaction outcome when using ionic liquids was
due primarily to non-specific interactions; that is, the general
charge–charge interactions that would occur between the ionic
liquid and the transition state, as well as any π–π and cation–π
interactions between the ionic liquid and either the starting
material or the transition state. That correlation is in contrast
to that seen for this study; the dependence of the rate constant
change on both the α and β parameters in the case introduced
here indicate that the dominant interactions responsible for
these changes are specific hydrogen bonding interactions
between the ionic liquid and the substrate.∥∥ The major differ-
ences between these two systems are the nature of the leaving
1
Fig. 4 The relationship between the natural logarithm of k at 73.8 °C
and a combination of the α and β parameters for the reaction between
the acetate 11 and indole 2 in mixtures of each of the ionic liquids 4–10
in acetonitrile at χ = 0.20. Uncertainties reported were obtained from group, the presence of oxygen in the xanthene 11 substrate
the standard deviation of triplicate measurements and transformed on and the planar nature of the cationic intermediate (cf. freely
calculating the natural logarithm.
rotating diphenylmethane cation). This indicates that the
oxygen in the aromatic system and the oxygen atoms present
on the leaving group are likely the site for the hydrogen
bond donating ability of the solvent is still the most dominant bonding interactions that result in the effect on the rate con-
1
factor in affecting k , and the hydrogen bond accepting ability stant for this system. In the bromodiphenylmethane system,
is still important, the general polarizability of the solvent also with its halide leaving group, equivalent interactions with the
3
5
needs to be considered when using large proportions of ionic halogen would be expected to be less.
liquid in the reaction mixture. These contributions are not
unexpected given the highly delocalised, aromatic nature of
the substrate 11; with a large amount of ionic liquid present in
the system, it is logical that non-specific interactions such as
Experimental
π–π interactions and charge–charge interactions become more Xanthydrol 1 was recrystallised from ethanol and stored at
significant. However, in both solvent compositions, the hydro- 253 K until use. Indole was purchased from Merck and used
gen bond donating ability of the solvent is most significant in without further purification. 9H-Xanthen-9-yl acetate 11 was
affecting the rate constant, and this is reflected in all multi- prepared through reaction of xanthydrol 1 with acetic anhy-
variate regression analyses performed in this study.
dride and stored at 253 K until use. Acetonitrile was purified
3
6
In combination with the activation parameter data, some according to literature methods and stored over 3 Å sieves
comments can be made about the nature of the key solute– prior to use. The ionic liquids 4–10 were synthesised using
3
7–41
solvent interactions in the mixtures that affect the rate con- modified literature methods
through alkylation of the
stant. The increase in both activation parameters on moving to corresponding heterocycle to give the precursor halide salt, fol-
high proportions of the salt 4 indicates that the solvent effects lowed by a salt metathesis with the appropriate anion. All
are dominated by interaction with the starting material 11. ionic liquids were dried under reduced pressure (<0.2 mbar)
Based on the importance of hydrogen bond donation indi- for at least 7 hours and found to have <200 ppm water using
cated above and that this feature is associated with the cation, the Karl Fischer titration method. Ion chromatography was
the key interaction is between the cation of the ionic liquid used to determine <0.1 mol% residual halide in each ionic
and the xanthene 11. The site on the substrate could be any liquid. For full synthetic procedures, see ESI.†
electron rich area (e.g. the electronegative oxygen atoms and
Kinetic analyses were carried out at 347 K with either the
−1
the delocalised π systems). Whilst modelling data (such as electrophile 1 or 11 (ca. 0.01 mol L ), the nucleophile 2
recent examples of organisation about an amine nucleo- (ca. 0.02 mol L− ), and the desired amount of acetonitrile and
1
34
phile ) might further highlight the location(s) of these inter- one of the ionic liquids 4–10 to make up the appropriate
actions, comparisons with other systems allow the sites to be solvent composition. In the case of the acetate 11 system, tri-
identified (see below). Less significant interactions implied by ethylamine (ca. 0.05 mol L− ) was also added. The reaction
the lesser contributions of both β and π* parameters in the fits between xanthydrol 1 and indole 2 was monitored by following
above can also be interpreted in terms of interactions involving the depletion of the xanthylic signal in the starting material 1
the anion (likely with the comparatively electron poor areas of
1
the molecule) and non-specific interactions.
With the above argument in mind, it is useful to contrast
the correlation presented here to those for the previously
∥
∥The less significant dependence on π* at high mole fractions in this study
indicates similar non-specific interactions also occur but these are not solely
responsible for driving the change in the rate constant, unlike that seen pre-
considered unimolecular substitution of bromodiphenyl- viously for the bromodiphenylmethane system.
9340 | Org. Biomol. Chem., 2019, 17, 9336–9342
This journal is © The Royal Society of Chemistry 2019

View Article Online
Organic & Biomolecular Chemistry
Paper
1
at δ 6.0 in the H NMR spectrum, either to completion or
using initial rates methodology. The reaction between the
Conflicts of interest
acetate 11 and indole 2 was monitored by following the There are no conflicts to declare.
increase of the xanthylic signal of the product. at δ 5.8 in the
1
H NMR spectrum. Rate constants were determined by inte-
grating these signals over time. Activation enthalpies and
Acknowledgements
entropies were determined by fitting the obtained rate con-
4
2
stants to the Eyring equation. Further details of experimental AG acknowledges the support of the Australian Government
and kinetic procedures, including exact amounts of reagents through the receipt of Research Training Program
used, can be found in the ESI.†
Scholarship. JBH acknowledges financial support from the
a
Microsoft Excel (version 16.25) was used for multivariate Australian Research Council Discovery Project Funding
regression analyses of Kamlet–Taft parameters and the natural Scheme (Project DP180103682). The authors also acknowledge
logarithm of the obtained rate constants.
the NMR facility and the Bioanalytical Mass Spectrometry facil-
ity, both within the Mark Wainwright Analytical Centre at the
University of New South Wales; the former for NMR support
and the latter for carrying out mass spectrometric analysis.
Conclusions
This study sought to understand the effects of ionic liquids on
the reaction between xanthydrol 1 and indole 2. However, Notes and references
due to difficulties with data reproduction and acid
1
2
T. Welton, Chem. Rev., 1999, 99, 2071–2084.
P. Wasserscheid and W. Keim, Angew. Chem., Int. Ed., 2000,
9, 3772–3789.
R. Hayes, G. G. Warr and R. Atkin, Chem. Rev., 2015, 115,
357–6426.
G. Cevasco and C. Chiappe, Green Chem., 2014, 16, 2375–
385.
sensitivity, instead the unimolecular nucleophilic substitution
reaction*** between the acetate 11 and indole 2 was studied in
a range of ionic liquids that varied in their constituent ions. It
was found that the trend in the reaction outcome with solvent
composition was different to that seen previously for this
mechanism type; the unimolecular rate constant increased on
addition of any amount of ionic liquid to the reaction mixture
with the largest increase occurring at a high mole fraction.
Activation parameters determined for this system indicated
that interactions with the starting material were key in deter-
mining this reaction outcome.
3
3
4
6
2
5
6
K. Ghandi, Green Sustainable Chem., 2014, 4, 44–53.
M. J. Earle and K. R. Seddon, Pure Appl. Chem., 2000, 72,
1391.
7
8
M. Freemantle, Chem. Eng. News, 1998, 76, 32–37.
J. P. Hallett and T. Welton, Chem. Rev., 2011, 111, 3508–
Further insight into the changes observed in the rate con-
stant on moving to an ionic liquid as the solvent was found
with the use of Kamlet–Taft parameters; excellent correlations
3576.
9
S. T. Keaveney, R. S. Haines and J. B. Harper, in
Encyclopedia of Physical Organic Chemistry, Wiley, New
Jersey, 2017.
1
between the natural logarithm of k and a combination of the
parameters were found. Most dominant in affecting the rate
constant was the hydrogen bond donating ability of the
solvent, however both the hydrogen bond accepting ability and
the polarizability of the ionic liquid were also important.
Different outcomes between this reaction and a previous
example of the same mechanism were found to be due to the
different manner in which the ionic liquid was interacting
with the substrate. Not only does this provide an understand-
ing of how the ionic liquid is interacting with the species
along the reaction coordinate, it also allows for prediction of
the effects additional ionic liquids will have on reactions
which occur through similar substrates. Therefore, this adds
to the growing body of literature on how and why ionic liquids
affect reaction outcome for unimolecular nucleophilic substi-
tution reactions.
1
0 A. Gilbert, R. S. Haines and J. B. Harper, in Reference
Module in Chemistry, Molecular Sciences and Chemical
Engineering, Elsevier, 2018.
1
1 N. L. Mai, K. Ahn and Y.-M. Koo, Process Biochem., 2014,
49, 872–881.
1
2 R. R. Hawker and J. B. Harper, in Advances in Physical
Organic Chemistry, ed. I. H. Williams and N. H. Williams,
Academic Press, 2018, pp. 49–85.
1
1
1
3 L.-Y. Liu, B. Wang, H.-M. Yang, W.-X. Chang and J. Li,
Tetrahedron Lett., 2011, 52, 5636–5639.
4 B. Y. W. Man, J. M. Hook and J. B. Harper, Tetrahedron
Lett., 2005, 46, 7641–7645.
5 H. M. Yau, S. A. Barnes, J. M. Hook, T. G. A. Youngs,
A. K. Croft and J. B. Harper, Chem. Commun., 2008, 3576–
3578.
1
1
6 E. Kochly, N. Lemon and A. Deh-Lee, Molecules, 2016, 21, 60.
7 S. T. Keaveney, B. P. White, R. S. Haines and J. B. Harper,
Org. Biomol. Chem., 2016, 14, 2572–2580.
*
**It should be noted that throughout this manuscript, the reaction has been
discussed in terms of being an S 1 process involving the acetate 11. Formally, it
could also be considered an S Ar process on the indole 2. However, given the
N
E
18 X. Creary, E. D. Willis and M. Gagnon, J. Am. Chem. Soc.,
kinetic analyses described (particularly independence of rate on indole 2 con-
centration) we suggest that the nomenclature used is reasonable.
2005, 127, 18114–18120.
This journal is © The Royal Society of Chemistry 2019
Org. Biomol. Chem., 2019, 17, 9336–9342 | 9341

View Article Online
Paper
Organic & Biomolecular Chemistry
1
2
9 A. Gilbert, R. S. Haines and J. B. Harper, Org. Biomol. 32 R. Bini, C. Chiappe, V. L. Mestre, C. S. Pomelli and
Chem., 2019, 17, 675–682. T. Welton, Org. Biomol. Chem., 2008, 6, 2522–2529.
0 H. M. Yau, S. T. Keaveney, B. J. Butler, E. E. L. Tanner, 33 M. A. A. Rani, A. Brant, L. Crowhurst, A. Dolan, M. Lui,
M. S. Guerry, S. R. D. George, M. H. Dunn, A. K. Croft and
J. B. Harper, Pure Appl. Chem., 2013, 85, 1979.
1 P. Pavez, D. Millan, J. Morales, M. Rojas, D. Cespedes and
N. H. Hassan, J. P. Hallett, P. A. Hunt, H. Niedermeyer,
J. M. Perez-Arlandis, M. Schrems, T. Welton and
R. Wilding, Phys. Chem. Chem. Phys., 2011, 13, 16831–
16840.
2
2
2
2
2
2
2
2
2
3
3
J. G. Santos, Org. Biomol. Chem., 2016, 14, 1421–1427.
2 L. Crowhurst, R. Falcone, N. L. Lancaster, V. Llopis-Mestre 34 S. T. Keaveney, J. B. Harper and A. K. Croft, ChemPhysChem,
and T. Welton, J. Org. Chem., 2006, 71, 8847–8853. 2018, 19, 3279–3287.
3 J. Alarcón-Espósito, R. Contreras, R. A. Tapia and 35 A. Kovács and Z. Varga, Coord. Chem. Rev., 2006, 250, 710–
P. R. Campodónico, Chem. – Eur. J., 2016, 22, 13347–13351. 727.
4 R. R. Hawker, R. S. Haines and J. B. Harper, Chem. 36 W. L. F. Armarego and C. Chai, Purification of Laboratory
Commun., 2018, 54, 2296–2299. Chemicals, Elsevier, 2009.
5 S. T. Keaveney, K. S. Schaffarczyk McHale, R. S. Haines and 37 L. Cammarata, S. G. Kazarian, P. A. Salter and T. Welton,
J. B. Harper, Org. Biomol. Chem., 2014, 12, 7092–7099. Phys. Chem. Chem. Phys., 2001, 3, 5192–5200.
6 S. T. Keaveney, R. S. Haines and J. B. Harper, Org. Biomol. 38 E. E. L. Tanner, H. M. Yau, R. R. Hawker, A. K. Croft and
Chem., 2015, 13, 3771–3780. J. B. Harper, Org. Biomol. Chem., 2013, 11, 6170–6175.
7 M. Gazitua, R. A. Tapia, R. Contreras and 39 F. F. C. Bazito, Y. Kawano and R. M. Torresi, Electrochim.
P. R. Campodonico, New J. Chem., 2014, 38, 2611–2618. Acta, 2007, 52, 6427–6437.
8 P. Pavez, D. Millán, J. I. Morales, E. A. Castro, C. A. López 40 S. N. Baker, E. B. Brauns, T. M. McCleskey, A. K. Burrell
and J. G. Santos, J. Org. Chem., 2013, 78, 9670–9676. and G. A. Baker, Chem. Commun., 2006, 2851–2853.
9 R. W. Taft and M. J. Kamlet, J. Am. Chem. Soc., 1976, 98, 41 G. Qu, W. He, Y. Cai, X. Huang and P. Ning, Carbohydr.
886–2894. Polym., 2016, 148, 390–396.
0 M. J. Kamlet and R. W. Taft, J. Am. Chem. Soc., 1976, 98, 42 H. Eyring, J. Chem. Phys., 1935, 3, 107–115.
2
3
77–383.
43 L. Crowhurst, P. R. Mawdsley, J. M. Perez-Arlandis,
P. A. Salter and T. Welton, Phys. Chem. Chem. Phys., 2003, 5,
2790–2794.
1 M. J. Kamlet, J. L. Abboud and R. W. Taft, J. Am. Chem.
Soc., 1977, 99, 6027–6038.
9342 | Org. Biomol. Chem., 2019, 17, 9336–9342
This journal is © The Royal Society of Chemistry 2019