Steric, hydrogen-bonding and structural heterogeneity effects on the nucleophilic substitution of N-(p-fluorophenyldiphenylmethyl)-4-picolinium chloride in ionic liquids

Article abstract of DOI:10.1039/c3ob40105g

The nucleophilic substitution of N-(p-fluorophenyldiphenylmethyl)-4- picolinium chloride was investigated using water and a range of alcoholic nucleophiles in ionic liquid solvents. The reactivity patterns across the nucleophiles examined could be attributed to steric factors, which mediated the relative nucleophilicities. Reducing the hydrogen-bond acidity of the ionic liquid cation was found to generally increase the rate of reaction, however, the magnitude of this rate effect could be influenced by the steric bulk of the nucleophile and the structural heterogeneity of the ionic liquid. Preferential solvation phenomena in binary mixtures of ionic liquids were examined and suggest that the mechanism behind the hydrogen-bond solvation phenomenon arises from direct cation-mediated, rather than indirect anion-mediated, effects. The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c3ob40105g

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
Steric, hydrogen-bonding and structural heterogeneity
effects on the nucleophilic substitution of
N-(p-fluorophenyldiphenylmethyl)-4-picolinium
chloride in ionic liquids†
Cite this: Org. Biomol. Chem., 2013, 11,
2534
Cameron C. Weber, Anthony F. Masters and Thomas Maschmeyer*
The nucleophilic substitution of N-(p-fluorophenyldiphenylmethyl)-4-picolinium chloride was investigated
using water and a range of alcoholic nucleophiles in ionic liquid solvents. The reactivity patterns across the
nucleophiles examined could be attributed to steric factors, which mediated the relative nucleophilicities.
Reducing the hydrogen-bond acidity of the ionic liquid cation was found to generally increase the rate of
reaction, however, the magnitude of this rate effect could be influenced by the steric bulk of the nucleo-
phile and the structural heterogeneity of the ionic liquid. Preferential solvation phenomena in binary mix-
tures of ionic liquids were examined and suggest that the mechanism behind the hydrogen-bond
solvation phenomenon arises from direct cation-mediated, rather than indirect anion-mediated, effects.
Received 18th January 2013,
Accepted 21st February 2013
DOI: 10.1039/c3ob40105g
www.rsc.org/obc
been shown to possess an array of interesting structural fea-
tures that can influence reactivity. One of these is the for-
Introduction
Ionic liquids (ILs), frequently defined as salts that melt below mation of heterogeneous structures on elongation of the alkyl
100 °C,1,2 have often been proposed as attractive ‘green’ re- chain on the IL, with the creation of polar and non-polar
placements for volatile organic compounds (VOCs).3,4 Even domains.13–16 For imidazolium systems it has been found that
though questions have been raised as to how green ILs truly ILs, bearing alkyl chains longer than propyl, are capable of
are,4–6 the versatility of such solvents arising from the sheer forming such heterogeneous structures.17 ILs are also capable
number of potential cation and anion combinations has of forming liquid clathrates around aromatic solute molecules,
meant extensive work has been undertaken to assess their leading to the relatively greater solubility of aromatic com-
influence on reactions.7–11 The flexibility of IL choice means pounds relative to aliphatics in these solvents.18–21 Although
that, if the use of IL mixtures is included, the formation of a the use of such phenomena to account for unusual reaction
huge range of potential solvent systems is possible (estimated outcomes in ILs is still in its infancy, recently some examples
to be up to 1018 on the assumption that mixtures of ILs are have emerged implicating such effects in unusual reaction
considered to be ‘new’ solvent systems).12 To enable the outcomes.22–24 The consideration of these effects is, therefore,
rational selection of such solvents from such a vast pool, a important for the interpretation of any results obtained within
thorough understanding of the effects of the variation of each IL systems.
component and correspondingly of mixtures of those com-
ponents is required.
Nucleophilic substitution reactions have been investigated
within ILs utilising an array of substrates, ranging from SN1
In addition to directly influencing solvent–solute inter- substitutions of alkyl chlorides25–27 and aryl diazonium
actions through the choice of cation and anion pair, ILs have salts28,29 to SN2 reactions between amines and neutral or
charged electrophiles30–35 or halides and an array of
electrophiles.33,36–41 In these investigations, when multiple ILs
Laboratory of Advanced Catalysis for Sustainability, School of Chemistry,
were employed, the resultant reactivity could frequently be cor-
The University of Sydney, Sydney 2006, Australia.
related with solvation parameters through linear solvation
E-mail: thomas.maschmeyer@sydney.edu.au; Fax: +61 2 9351 3329;
energy relationships (LSERs),42 implying that a combination of
Tel: +61 2 9351 2581
†Electronic supplementary information (ESI) available: Full details of the syn-
microscopic solvation effects could account for the observed
behaviour. The most frequently employed parameters for such
correlations are those described by Kamlet and Taft. These
thesis and characterisation of ILs and reaction products, mole fractions of ILs,
LFER details, Kamlet–Taft correlations for ethanol, 1-50 butanol and phenol,
preferential solvation data for benzyl alcohol, rate dependence on nucleophile
parameters account for solvation behaviour in terms of the
concentration and correlation of hydrogen-bond effect with Charton’s steric
parameter. See DOI: 10.1039/c3ob40105g
solvent’s hydrogen-bond acidity (α), hydrogen-bond basicity (β)
2534 | Org. Biomol. Chem., 2013, 11, 2534–2542
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
and dipolarity/polarisability (π*).43–45 A notable exception is
the nucleophilic substitution of a sulfonium salt with halide
nucleophiles, where the reactivity could be attributed to the
random distribution of ion-pairs within the IL, i.e. an ‘IL-
effect’.39–41
Another example of such an ‘IL-effect’ involves the nucleo-
philic substitution of N-(p-fluorophenyldiphenylmethyl)-4-
picolinium chloride ([Ar3CPic][Cl]), which was recently shown
to be influenced by the structural heterogeneity of the IL.22
Within that investigation, reaction rate enhancements were
observed with increasing IL alkyl chain length. For that reac-
tion system, rate increases were also detected upon reducing
the hydrogen-bond acidity of the IL solvent, although only
water was examined as a nucleophile.46 Importantly, the mag-
nitude of the rate increase resulting from the structural hetero-
geneity of the IL was found to be substantially greater than
that which could be attributed to hydrogen-bonding
effects.22,46 As the investigation into the interplay between the
hydrogen-bonding and structural effects of the solvent was
limited in these reports to water being used as a nucleophile,
it is necessary to extend this study to a wider range of hydro-
xylic nucleophiles to develop a more thorough understanding
of such solvent effects.
Furthermore, the previous investigations on this reaction
system have uncovered several interesting phenomena that
have not been fully explored thus far. Firstly, a linear relation-
ship between the rate constant and solvent composition was
observed in binary mixtures of 1-butyl-3-methylimidazolium
bis(trifluoromethanesulfonyl)imide ([BMIM][NTf2]) and 1-
butyl-2,3-dimethylimidazolium bis(trifluoromethanesulfonyl)-
imide ([BMMIM][NTf2]) for the hydrolysis of [Ar3CPic][Cl].46
This result is in stark contrast to the interaction of Reichardt’s
dye with these solvent mixtures, where significant preferential
solvation was observed. The underlying nature of this discre-
pancy was not further investigated in the original study.
Another unexplained experimental result was that the rate con-
stant for the hydrolysis of [Ar3CPic][Cl] in [BMIM][NTf2] was
significantly lower than anticipated from the fitted kinetic
model for the structural heterogeneity of the IL.22 Interestingly,
all the other IL and nucleophile combinations gave an excel-
lent fit to the model suggesting the poor fit discovered for
water and [BMIM][NTf2] was due to a fundamental difference
in solvent behaviour for this specific case. Finally, it was found
that using ethanol as a nucleophile compared to water and
methanol led to a smaller rate constant,22 despite an earlier
literature report47 on a closely related reaction finding that the
reaction involving ethanol proceeded substantially faster than
with water.
Fig. 1 Structure of cations and anion used in this investigation, with the corres-
ponding abbreviations.
Table 1 Kamlet–Taft parameters for all ILs investigated, data obtained from
ref. 48
Ionic liquid
α
β
π*
[EMIM][NTf2]
[BMIM][NTf2]
[BMPyrr][NTf2]
[BMMIM][NTf2]
0.63
0.61
0.43
0.38
0.23
0.23
0.24
0.26
1.00
0.99
0.95
1.02
imide [BMPyrr][NTf2] to enable further examination of the
hydrogen-bonding effects as well as the determination of
whether the aromaticity of the cation has any specific con-
sequences for aromatic nucleophiles. The structures of the
cations and the anion used in all the ILs discussed herein are
depicted in Fig. 1 with their corresponding Kamlet–Taft para-
meters detailed in Table 1.
Results and discussion
Scheme 1 outlines the general reaction investigated, including
the range of nucleophiles and solvents examined. It should be
noted that the mole fraction of IL employed for all reaction
systems was always greater than 0.73, so the IL constitutes the
bulk of the reaction medium. The precise IL mole fractions for
every nucleophile and IL combination are detailed in the ESI.†
[Ar3CPic][Cl] was generated in situ from p-fluorophenyldiphe-
nylmethyl chloride (FDMC) and the nucleophilic substitution
reactions monitored using 19F{1H} NMR, as has previously
been reported.22,46 4-Picoline was selected as the leaving group
for experimental convenience to enable data to be collected
To thoroughly explore the issues outlined above, many
different nucleophiles have been examined in the present
work. Furthermore, the ILs examined as solvents include the
combination of 1-ethyl-3-methylimidazolium bis(trifluoro-
methanesulfonyl)imide ([EMIM][NTf2]) and [BMIM][NTf2] to
investigate the consequences of structural heterogeneity,
[BMMIM][NTf2] to isolate hydrogen-bonding effects and
Scheme 1 Nucleophilic substitution of [Ar3CPic][Cl] in a range of ILs. R = H,
Me, Et, 1-Pr, 2-Pr, 1-Bu, Bn and Ph. IL = [EMIM][NTf2], [BMIM][NTf2], [BMPyrr]-
1-butyl-1-methylpyrrolidinium bis(trifluoromethanesulfonyl)- [NTf2] and [BMMIM][NTf2].
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2534–2542 | 2535

View Article Online
Paper
Organic & Biomolecular Chemistry
[BMIM][NTf2] and [BMMIM][NTf2] that suggests the underlying
cause for the apparent rate discrepancies. Since 2-propanol
results in rate constants more than an order of magnitude
smaller than those of 1-propanol, it is clear that effects other
than nucleophilicity are important. It would appear that steric
hindrance induced by α-substituents on the alcohol account
for such behaviour and explain the significant reduction in the
rate constant for 2-propanol and ethanol. The planarity of the
phenyl substituent in phenol could account for the similarity
in rate constants between phenol and benzyl alcohol, where
the anticipated increased nucleophilicity of benzyl alcohol is
offset by its larger effective steric bulk.50 The greater relative
acidity of phenol, however, means that deprotonation by the
excess 4-picoline to yield a more nucleophilic phenolate anion
cannot be eliminated as an alternative.
Table 2 Summary of pseudo-first-order rate constants obtained for the nucleo-
philic substitution of [Ar3CPic][Cl] at 294.3 K. Reported errors are standard devi-
ations from at least 3 replicate experiments
Rate constants (10−4 s−1
)
[EMIM]-
Nucleophile [NTf2]
[BMIM]-
[NTf2]
[BMPyrr]-
[NTf2]
[BMMIM]-
[NTf2]
H2O
MeOH
EtOH
1-PrOH
1-BuOH
BnOH
PhOH
2-PrOH
8.7 0.3
10.6 0.2
23.9 1.0
8.5 0.1
8.6 0.7
8.9 0.6
4.1 0.3
4.1 0.5
0.28 0.01
16.1 1.0
28.2 2.0
10.8 1.0
10.5 1.4
12.9 0.3
4.4 0.3
18.8 0.4
32.4 1.4
10.6 0.2
11.0 0.4
12.1 0.7
4.8 0.1
17.4 0.2
6.4 0.2
6.5 0.1
7.3 0.4
3.0 0.1
2.7 0.2
4.0 0.4
3.7 0.2
0.36 0.01
Attempts to fit these data to Charton steric parameters50,51
and the Mayr nucleophilicity parameters52 resulted in the
anticipated correlation. However, the lack of available nucleo-
philicity data meant that only 5 nucleophiles could be
included in each analysis. Consequently, these fits had large
standard errors and limited statistical significance (0.08 < p <
0.13) for the three parameters fitted. Therefore, these corre-
lations will not be discussed further in the main text, although
a more comprehensive analysis can be found in the ESI.† It
can be safely surmised, however, that the variation of rate con-
stants across the range of nucleophiles arises not only from
the order of nucleophilicity, but more significantly from the
steric hindrance of the α-substituents of the alcohol.
over a reasonable time scale near room temperature. The reac-
tion rate has been previously shown not to depend on the con-
centration of 4-picoline,46 nonetheless, the concentration of
4-picoline was maintained at 0.49 M for all reactions. The
nucleophile was used in excess, and all reactions monitored
gave excellent fits to the integrated first-order rate law. The
pseudo-first-order rate constants obtained are summarised in
Table 2. The initial nucleophile concentrations in all cases
were maintained at 0.515 M, meaning that the pseudo-first-
order rate constants are directly comparable across the range
of nucleophiles and ILs. To ascertain the specific dependence
of the reaction rate on nucleophile concentration, the concen-
tration of water, methanol and ethanol used for this reaction
were varied in [EMIM][NTf2] (see ESI†). From these results it is
apparent that a first order dependence on the nucleophile is
observed, demonstrating that the reaction studied is bimole-
cular, as has been reported previously.22,46
Cation effects
From Table 2, it is apparent that varying the solvent leads to
significant rate effects for most nucleophiles examined. For all
alcohols, with the exception of phenol, the reduction of the
hydrogen-bond acidity of the solvent through the use of
[BMMIM][NTf2] rather than [BMIM][NTf2] as a solvent, yielded
an increase in the rate constant. To compare the magnitude of
this increase, the rate constants obtained in [BMMIM][NTf2]
have been divided by those observed in [BMIM][NTf2] for the
corresponding nucleophile (Table 3). As the hydrogen-bond
acidity is the only significant difference between these ILs, it
can be concluded that this ratio is correlated with any ‘hydro-
gen-bond effect’ in the reaction.
Nucleophile effects
The results in Table 2 demonstrate significant variation across
the range of nucleophiles. Firstly, upon changing the nucleo-
phile from water to methanol the rate constant increases
within every solvent examined, as would be anticipated from
the greater nucleophilicity of methanol.49 When ethanol is
employed as a nucleophile, however, the observed rate con-
stant is significantly smaller than that observed for methanol
and water, an effect which cannot be attributed to the order of
nucleophilicity. Furthermore, as the length of the linear alkyl
chain on the nucleophile is extended, no additional kinetic
effects are observed with the rate constants for ethanol being
consistent with those for 1-propanol and 1-butanol within
most solvents.
Upon examination of Table 3, it is evident that hydrogen-
bonding has a larger effect on water than any of the other
Table 3 Ratios of rate constants observed in [BMMIM][NTf2] divided by those
in [BMIM][NTf2]. Reported errors are the propagated standard deviations
The electron-withdrawing effect of the aromatic substituent
of phenol and benzyl alcohol leads to the reduced nucleophili-
city of these species and correspondingly their reduced rate
constants relative to the aliphatic alcohols. However, the simi-
larity of such rate constants is surprising given that benzyl
alcohol would be expected to be substantially more nucleophi-
lic than phenol.
Nucleophile
kBMMIM/kBMIM
H2O
1.77 0.05
1.35 0.08
1.25 0.02
1.29 0.12
1.36 0.12
1.18 0.09
0.89 0.11
1.27 0.06
MeOH
EtOH
PrOH
BuOH
BnOH
PhOH
2-PrOH
It is the significant decrease in the rate constant observed
for 2-propanol compared with the other nucleophiles in both
2536 | Org. Biomol. Chem., 2013, 11, 2534–2542
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
nucleophiles. This is somewhat counter-intuitive as water
Notably in all of these correlations except for water, the
should be the weakest hydrogen-bond acceptor, with the excep- value for [EMIM][NTf2] (α = 0.63) is below the trend, suggesting
tion of phenol. The result implies that in the sterically hin- that the discrepancy in the rate constants between [EMIM]-
dered transition state, interactions between the cation and the [NTf2] and the other ILs examined is likely due to a systematic
nucleophile are somewhat restricted for the alcohols relative to factor other than hydrogen-bond acidity. This is consistent
the substantially less hindered water molecule. Alternatively, with the excellent fits to the domain models obtained in our
the observed hydrogen-bond effect may arise from the weaker earlier communication for all alcohols except for water,22
cation–anion interactions, in turn leading to stronger nucleo- implying that such a breakdown in the dependence on hydro-
phile–anion and electrophile–anion interactions. The evidence gen-bond acidity can be attributed to the structural heterogen-
for or against each of these potential mechanisms will be dis- eity of the IL. Interestingly, the good correlation observed for
cussed in more detail within the next section.
water implies that it is sufficiently dissociative in 0.515 M con-
As discussed above, it is evident that hydrogen-bonding can centrations in [BMIM][NTf2], such that the inherent heterogen-
account for at least some of the apparent solvent-induced rate be- eity of this IL is affected. The other noteworthy result is that
haviour and it is clear (from Table 1) that the α parameter is the no significant outlier behaviour is observed for [BMPyrr][NTf2]
only Kamlet–Taft parameter that varies significantly upon chang- with any of the aromatic alcohols, confirming that the sol-
ing the IL cation. Therefore, an attempt to correlate the observed vation of aromatics is not strongly influenced by the aromati-
rate constant with the α parameter of each IL was made. Fig. 2 city of the cation itself. This supports the view that the strong
depicts some representative examples of such correlations and it interactions between aromatic molecules and ILs reported in
is apparent that while this fit is appropriate for water, it breaks the literature result primarily from ion–quadrupole rather than
down significantly for most of the other alcohols.
π–π interactions.20,21,53
Fig. 2 Linear correlation between the observed pseudo-first-order rate constants and the Kamlet–Taft α parameter of the IL solvent for (clockwise from top left):
water, methanol, 1-propanol and benzyl alcohol. Similar correlations for the other nucleophiles are given in the ESI.†
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2534–2542 | 2537

View Article Online
Paper
Organic & Biomolecular Chemistry
Preferential solvation
representing other preferential solvation phenomena has also
been demonstrated.46,54,55
In addition to the results in the neat ILs; water, methanol,
ethanol and 1-butanol were examined in a range of binary
[BMIM][NTf2] : [BMMIM][NTf2] mixtures. These experiments
were conducted in an attempt to isolate the origin of the pre-
ferential solvation differences observed previously between
water as a nucleophile in this reaction system and Reichardt’s
dye as a solvatochromic probe.46 It was hypothesised that the
different behaviour observed may have arisen from either the
difference in size and, hence, rate of diffusion of the two mole-
cules, or the variation in their hydrogen-bond accepting
ability. The magnitude of any preferential solvation pheno-
menon would also yield additional information as to the
nature of the hydrogen-bond effect discussed previously.
RS1 þ S2 Ð RS2 þ S1
ð1Þ
From eqn (1), by assigning [BMIM][NTf2] as S2 and
[BMMIM][NTf2] as S1, it can be shown that the observed rate
constant in binary mixtures of these ILs (kobs) can be fitted to
the equilibrium constant (K), the individual rate constants in
[BMIM][NTf2] and [BMMIM][NTf2] (kBMIM and kBMMIM respecti-
vely) and the mole fraction of [BMIM][NTf2] (xBMIM) through
eqn (2).
KxBMIMðkBMIM ꢀ kBMMIM
1 ꢀ xBMIM þ KxBMIM
Þ
kobs
¼
þ kBMMIM
ð2Þ
The rate constants observed in these binary mixtures were
fitted to a simple preferential solvation model that models the
solvation of the solute molecule (R) as an equilibrium between
solvent 1 (S1) and solvent 2 (S2), as outlined in eqn (1). Such a
model has been traditionally utilised for the preferential sol-
vation of solvatochromic dyes, however, its utility for
The fitted data and equilibrium constants obtained for the
nucleophiles water, methanol, ethanol and 1-butanol are
depicted in Fig. 3. It is worth noting that the values of K
obtained from these fits are the inverse of those reported in
our earlier publication46 as S2 is now assigned to [BMIM]-
[NTf2]. This modification was employed as it results in values
Fig. 3 Rate constants observed in [BMIM][NTf2] : [BMMIM][NTf2] binary solvent systems fit to the model described in eqn (2). The solid line represents the fitted
model, the dashed line indicating ideal behaviour, i.e. no preferential solvation, is provided as a guide to the eye. Fitted K parameters and their standard error are
reported.
2538 | Org. Biomol. Chem., 2013, 11, 2534–2542
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
of K greater than 1 when preferential solvation by the more
hydrogen-bond acidic IL occurs. Such an analysis was also
attempted for the preferential solvation of benzyl alcohol,
however, the small magnitude of the hydrogen-bonding effect
and large relative experimental errors precluded a meaningful
fit of the model in this case. The data for benzyl alcohol are
included in the ESI.† Good fits were observed for all other
nucleophiles examined, and it appears that all alcohols investi-
gated exhibit some preferential solvation by the more hydro-
gen-bond acidic [BMIM]+ cation. Water, on the other hand, is
only marginally preferentially solvated and it appears the pre-
ference is for solvation by the [BMMIM]+ cation, although such
an effect lies only just outside of experimental error. It does
not appear that a significant trend can be observed within the
alcohols as such, but it is apparent that methanol is more pre-
ferentially solvated than either ethanol or 1-butanol.
The observation that methanol is more strongly preferen-
tially solvated than ethanol disproves the hypothesis regarding
molecular size being the main determining factor. That is, the
hypothesis that the more rapid diffusion of water within the IL
medium resulted in a more fully averaged solvation environ-
ment than is observed for more bulky compounds such as
Reichardt’s dye.46 In fact, methanol is even more strongly pre-
ferentially solvated within this binary solvent system than is
Reichardt’s dye, despite the dye being both significantly larger
and a better hydrogen-bond acceptor.46 To explain such a dis-
crepancy the origin of the preferential solvation phenomenon
needs to be discussed in relation to the specific transition
state structure of this reaction (vide infra).
Fig. 4 A simplified structure representing some potential steric interactions in
the proposed reaction transition state. The solvent has been omitted for clarity.
group dissociation in the rate determining step. Alternatively,
the presence of the bulky [NTf2]− anion in the IL system could
contribute to this discrepancy as the interaction of this anion
with the [Ar3CPic]+ cation would further crowd the transition
state compared to the interaction with the substantially
smaller chloride anion in nitromethane.
From the preceding discussion, it is apparent that the
hydrogen-bond effect has less of an impact on the alcohols
than water despite the former, with the exception of phenol,
being better hydrogen-bond acceptors. Consequently, two
hypotheses have been presented. Firstly, that the steric conges-
tion in the transition state means that alcohols, which are
more sterically hindered, are less able to be affected by the sol-
vating cation in the reaction transition state and therefore
exhibit a smaller rate enhancement upon reduction of the
hydrogen-bond strength of the medium. Alternatively, the
‘hydrogen-bond effect’ being observed could be a result of
reduced cation–anion interactions within the ILs, leading to
stronger substrate–anion interactions. As water and alcohols
have been established to form strong hydrogen-bonds with the
anion in IL systems,56,57 such an effect could result in
increased interactions between the nucleophile and [Ar3CPic]+
cation.
To distinguish between these mechanisms, it is important
to consider the preferential solvation of the aliphatic alcohols
by [BMIM][NTf2] with respect to [BMMIM][NTf2]. If the nature
of the hydrogen-bond effect were anion-dominated then it
would be anticipated that the nucleophiles most capable of
hydrogen-bonding to the anion should be most strongly
affected. However, water does not exhibit any significant pre-
ferential solvation whereas methanol, ethanol and 1-butanol
do, implying that a property of these alcohols renders their
interaction with [BMIM][NTf2] significantly more favourable
than with [BMMIM][NTf2]. This supports the direct interaction
of the cation with the transition state, as anion-directing effects
should influence water more strongly, while the stronger
hydrogen-bond accepting properties of the alcohols could
account for their preferential solvation by [BMIM][NTf2]. The
observation that the preferential solvation effect is more pro-
nounced for methanol than either ethanol or 1-butanol can
potentially be accounted for, at least in part, in terms of sterics
Consideration of the reaction mechanism
With the foregoing in mind, it is instructive to consider the
mechanism of the reaction in more detail. Clearly, a tra-
ditional SN2 process is unlikely due to the steric hindrance of
the p-fluorophenyldiphenylmethyl group. Therefore, dis-
sociation of the 4-picoline must precede the nucleophilic
attack. As discussed previously, no rate dependence with
4-picoline concentration is observed,46 indicating the absence
of leaving group return. Furthermore, we have found a first-
order dependence on the nucleophile concentration for a
range of nucleophiles, illustrating that it is truly a bimolecular
process. Collectively these results imply that although 4-pico-
line dissociation must occur first, nucleophilic attack must
proceed prior to the solvent separation of the carbocation and
4-picoline. Correspondingly, the transition state of such a
process would be exceptionally congested, resulting in the sig-
nificant steric dependence observed for the nucleophile. Fig. 4
gives a simplified representation of some potential steric inter-
actions in the transition state.
Such significant steric effects were not observed for the
nucleophilic substitution of a comparable pyridinium salt in
nitromethane using water and ethanol.47 The most likely expla-
nation lies with the reduced basicity of the pyridine leaving
group used in the earlier study relative to 4-picoline used in
the present study. This reduced basicity would decrease the
strength of the C–N bond and increase the extent of leaving
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2534–2542 | 2539

View Article Online
Paper
Organic & Biomolecular Chemistry
Fig. 5 Schematic depiction of the potential cation–nucleophile interactions in the transition state for (a) water and (b) an alcohol. The IL anions, explicit O–H bonds
and hydrogen-bonding interactions between the nucleophile and 4-picoline have been omitted for clarity.
and the availability of the nucleophilic oxygen for hydrogen- nucleophile reduces the magnitude of the effect more than the
bonding interactions with the IL cation. The direct role of the hydrogen-bond basicity of the nucleophile. Furthermore, the
IL cation in the transition state is depicted schematically in ability to exhibit nearly ideal reaction behaviour within binary
Fig. 5.
mixtures of [BMIM][NTf2] and [BMMIM][NTf2] has been shown
It is important distinguish between the steric effects being to be unique to water, with alcohols exhibiting significant pre-
introduced to account separately for the observed trends in ferential solvation behaviour (where such investigations could
nucleophilicity and the hydrogen-bonding effect, as the substi- be conducted), implicating the influence of the IL cation in
tuents implicated in these effects are subtly different. In the the transition state of the reaction. Finally, the dependence on
former case, it is the additional steric bulk of the α-substituents the hydrogen-bond acidity of the solvent indicated that the aro-
on the alcohol which reduces the number of allowed confor- maticity of the cation had no specific influence on reaction
mations of the nucleophile as it approaches the sterically rates, while structural heterogeneity led to significant dispar-
crowded carbocationic centre, thereby reducing the observed ities between the rates predicted and observed for the alcohols
reaction rate. In the latter case, the ability of the cation to examined. Water was the exception with a good correlation
solvate the nucleophile is dependent upon the sterics of any between hydrogen-bond acidity and reaction rate for all the ILs
substituent attached directly to the alcoholic group. In fact, investigated, likely as a result of the partial dissociation of the
a reasonable correlation (R2 = 0.868) can be observed between ions and breakdown of the IL structure. These results will
Charton’s ν parameter and the observed hydrogen-bond effect hopefully contribute to a greater understanding of the reaction
(see ESI†). This correlation is somewhat biased by the extreme mechanism of such sterically crowded species and illuminate
phenol and water ν values, as the other alcohols examined the effect that IL selection can have on nucleophilic substi-
possess similar steric parameter values and the magnitude of tution reactions, as well as revealing the influence even rela-
the hydrogen-bond effect observed for these alcohols was tively small concentrations of water may have on IL structure
largely within experimental error. Nonetheless, this correlation and reactivity.
is consistent with the proposal that the magnitude of the
hydrogen-bond effect is predominantly determined by the
ability of the solvent to directly interact with the nucleophilic
oxygen in the sterically crowded transition state. Collectively,
these observations illustrate that the IL cation can have a sig-
Experimental
General procedures
nificant and direct effect on reaction rates, even for similarly
charged transition states where such interactions may not be
expected.
[EMIM][NTf2], [BMIM][NTf2], [BMMIM][NTf2], [BMPyrr][NTf2]
and FDMC were all prepared according to literature procedures
or modifications thereof58,59 and characterised by NMR,
ESI-MS and melting point analysis where appropriate. 4-Pico-
line (Fluka) was passed through activated neutral alumina and
stored over 4 Å molecular sieves; methanol, ethanol, 1-propa-
nol, 1-butanol, 2-propanol and benzyl alcohol were dried over
Conclusion
The rate of nucleophilic substitution of [Ar3CPic][Cl] in a range calcium hydride, distilled and stored over 3 Å molecular sieves
of ILs has been shown to depend not only on the nucleophili- under nitrogen. Phenol was dried and stored in vacuo over
city, but also the steric bulk of the α-substituents of the P2O5. All ILs were dried at 80 °C in vacuo for at least 24 h prior
alcohol. In addition, the effect of solvent hydrogen-bond to kinetics experiments. Karl-Fischer titration found that this
acidity on the rate has been examined and surprisingly procedure yielded water contents consistently below 200 ppm.
suggests that the steric bulk of the substituents on the In addition, hydrolysis side-products, where they could be
2540 | Org. Biomol. Chem., 2013, 11, 2534–2542
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
detected, never accounted for more than 3% of product
formation observed and were typically within the error of
integration.
References
1 P. Wasserscheid and W. Keim, Angew. Chem., Int. Ed., 2000,
39, 3772–3789.
2 H. Weingartner, Angew. Chem., Int. Ed., 2008, 47, 654–670.
3 M. J. Earle, J. M. S. S. Esperanca, M. A. Gilea, J. N. C. Lopes,
L. P. N. Rebelo, J. W. Magee, K. R. Seddon and
J. A. Widegren, Nature, 2006, 439, 831–834.
4 M. Deetlefs and K. R. Seddon, Green Chem., 2010, 12,
17–30.
5 R. P. Swatloski, J. D. Holbrey and R. D. Rogers, Green
Chem., 2003, 5, 361–363.
6 P. J. Scammells, J. L. Scott and R. D. Singer, Aust. J. Chem.,
2005, 58, 155–169.
7 T. Welton, Chem. Rev., 1999, 99, 2071–2083.
8 J. P. Hallett and T. Welton, Chem. Rev., 2011, 111,
3508–3576.
9 A. Stark, Top. Curr. Chem., 2009, 290, 41–81.
10 C. Chiappe and D. Pieraccini, J. Phys. Org. Chem., 2005, 18,
275–297.
11 J. B. Harper and M. N. Kobrak, Mini-Rev. Org. Chem., 2006,
3, 253–269.
12 H. Niedermeyer, J. P. Hallett, I. J. Villar-Garcia, P. A. Hunt
and T. Welton, Chem. Soc. Rev., 2012, 41, 7780–7802.
13 W. Jiang, Y. Wang and G. A. Voth, J. Phys. Chem. B, 2007,
111, 4812–4818.
14 Y. Wang and G. A. Voth, J. Am. Chem. Soc., 2005, 127,
12192–12193.
15 J. N. A. Canongia Lopes and A. A. H. Padua, J. Phys. Chem.
B, 2006, 110, 3330–3335.
16 K. Shimizu, M. F. Costa Gomes, A. A. H. Padua,
L. P. N. Rebelo and J. N. Canongia Lopes, J. Mol. Struct.
(THEOCHEM), 2010, 946, 70–76.
17 O. Russina, A. Triolo, L. Gontrani, R. Caminiti, D. Xiao,
L. G. Hines, R. A. Bartsch, E. L. Quitevis, N. Plechkova and
K. R. Seddon, J. Phys.: Condens. Matter, 2009, 21, 424121.
18 J. D. Holbrey, W. M. Reichert, M. Nieuwenhuyzen,
O. Sheppard, C. Hardacre and R. D. Rogers, Chem.
Commun., 2003, 1636–1637.
19 J. K. D. Surette, L. Green and R. D. Singer, Chem. Commun.,
1996, 2753–2754.
20 C. G. Hanke, A. Johansson, J. B. Harper and R. M. Lynden-
Bell, Chem. Phys. Lett., 2003, 374, 85–90.
21 J. B. Harper and R. M. Lynden-Bell, Mol. Phys., 2004, 102,
85–94.
22 C. C. Weber, A. F. Masters and T. Maschmeyer, Angew.
Chem., Int. Ed., 2012, 51, 11483–11486.
Kinetics experiments
Solutions of ROH were prepared by mixing known volumes of
ROH and IL under a dry nitrogen atmosphere using gas-tight
syringes to yield the desired ROH concentration, with the
exception of phenol for which a gravimetric method was
employed. Volume additivity of these solutions was assumed,
as published data for these and closely related systems indi-
cate excess volumes of these binary mixtures would be smaller
than the anticipated error from the volumetric methods
employed.60–62 For example, the volume contraction of 1 mL of
a 0.55 M methanol in [EMIM][NTf2] solution would be less
than 0.15 μL (i.e. the error would be less than 0.015%). All
solutions were stirred for at least an hour prior to use to
ensure thorough mixing.
For the general kinetics protocol, an NMR tube equipped
with a Young’s valve was charged with FDMC (15–25 mg,
0.05–0.08 mmol) under a flow of nitrogen. 4-Picoline (50 μL)
was added to dissolve the substrate. Approximately 5 minutes
before the reaction was to be monitored, the solution was
diluted with ionic liquid containing ROH (0.98 mL of 0.55 M
ROH) and cooled on ice. The sample was then mixed using a
Vortex mixer and the reference capillary (1-bromo-4-fluoro-
benzene in acetone-d6 (0.50 M)) inserted co-axially. The tube
was immediately inserted into the NMR probe which had been
pre-cooled to 294.3 K, the temperature of the NMR tube
allowed to equilibrate, and 19F{1H} NMR spectra subsequently
obtained at that temperature every 65 seconds for 10–40
spectra. For reactions with 2-propanol, the above procedure
was followed except a spectrum was obtained every 21 minutes
for a total of 37 spectra.
All reactions afforded the anticipated alcohol/ether, as con-
firmed by either comparison with an independently syn-
thesised compound or isolation of the reaction product. The
details and characterisation data for all compounds are given
in the ESI.†
Preferential solvation experiments
IL solutions were prepared by directly weighing dry samples of
[BMIM][NTf2] and [BMMIM][NTf2] in the desired molar ratios.
The resultant mixture was further dried and the kinetics
experiments conducted using the protocols outlined above.
23 S. G. Jones, H. M. Yau, E. Davies, J. M. Hook,
T. G. A. Youngs, J. B. Harper and A. K. Croft, Phys. Chem.
Chem. Phys., 2010, 12, 1873–1878.
Acknowledgements
The authors would like to acknowledge Dr Ian Luck for his 24 S. Puttick, A. L. Davis, K. Butler, L. Lambert, J. El harfi,
assistance with NMR experiments and the University of Sydney
for funding. CCW would like to thank the University of Sydney
D. J. Irvine, A. K. Whittaker, K. J. Thurecht and P. Licence,
Chem. Sci., 2011, 2, 1810–1816.
for receipt of the Vice Chancellors Research Scholarship. TM 25 H. M. Yau, S. A. Barnes, J. M. Hook, T. G. A. Youngs,
thanks the Australian Research Council for a Future Fellowship
(FT0990485).
A. K. Croft and J. B. Harper, Chem. Commun., 2008,
3576–3578.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem., 2013, 11, 2534–2542 | 2541

View Article Online
Paper
Organic & Biomolecular Chemistry
26 B. Y. W. Man, J. M. Hook and J. B. Harper, Tetrahedron 45 M. J. Kamlet, J. L. Abboud and R. W. Taft, J. Am. Chem.
Lett., 2005, 46, 7641–7645. Soc., 1977, 99, 6027–6038.
27 R. Bini, C. Chiappe, D. Pieraccini, P. Piccioli and 46 C. C. Weber, A. F. Masters and T. Maschmeyer, J. Phys.
C. S. Pomelli, Tetrahedron Lett., 2005, 46, 6675–6678. Chem. B, 2012, 116, 1858–1864.
28 R. Bini, C. Chiappe, E. Marmugi and D. Pieraccini, Chem. 47 E. Gelles, E. D. Hughes and C. K. Ingold, J. Chem. Soc.,
Commun., 2006, 897–899. 1954, 2918–2929.
29 K. K. Laali, T. Okazaki and S. D. Bunge, J. Org. Chem., 2007, 48 M. A. Ab Rani, A. Brant, L. Crowhurst, A. Dolan, M. Y. Lui,
72, 6758–6762.
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.
30 H. M. Yau, A. G. Howe, J. M. Hook, A. K. Croft and
J. B. Harper, Org. Biomol. Chem., 2009, 7, 3572–3575.
31 H. M. Yau, A. K. Croft and J. B. Harper, Faraday Discuss., 49 S. Minegishi, S. Kobayashi and H. Mayr, J. Am. Chem. Soc.,
2012, 154, 365–371. 2004, 126, 5174–5181.
32 L. Crowhurst, N. L. Lancaster, J. M. Perez Arlandis and 50 M. Charton, J. Am. Chem. Soc., 1975, 97, 1552–1556.
T. Welton, J. Am. Chem. Soc., 2004, 126, 11549–11555.
33 N. L. Lancaster, J. Chem. Res., 2005, 413–417.
51 M. Charton, J. Am. Chem. Soc., 1969, 91, 615–618.
52 H. Mayr and M. Patz, Angew. Chem., Int. Ed., 1994, 33, 938–957.
34 A. Skrzypczak and P. Neta, Int. J. Chem. Kinet., 2004, 36, 53 K. Shimizu, M. F. Costa Gomes, A. A. H. Padua,
253–258.
L. P. N. Rebelo and J. N. Canongia Lopes, J. Phys. Chem. B,
2009, 113, 9894–9900.
54 E. Bosch and M. Roses, J. Chem. Soc., Faraday Trans., 1992,
88, 3541–3546.
55 R. D. Skwierczynski and K. A. Connors, J. Chem. Soc., Perkin
Trans. 2, 1994, 467–472.
56 J. N. Canongia Lopes, M. F. Costa Gomes and
A. A. H. Padua, J. Phys. Chem. B, 2006, 110, 16816–16818.
57 M. Lopez-Pastor, M. J. Ayora-Canada, M. Valcarcel and
B. Lendl, J. Phys. Chem. B, 2006, 110, 10896–10902.
58 A. K. Burrell, R. E. Del Sesto, S. N. Baker, T. M. McCleskey
and G. A. Baker, Green Chem., 2007, 9, 449–454.
59 S. T. Bowden and T. L. Thomas, J. Chem. Soc., 1940,
1242–1249.
35 R. Bini, C. Chiappe, C. S. Pomelli and B. Parisi, J. Org.
Chem., 2009, 74, 8522–8530.
36 N. L. Lancaster, P. A. Salter, T. Welton and G. B. Young,
J. Org. Chem., 2002, 67, 8855–8861.
37 N. L. Lancaster and T. Welton, J. Org. Chem., 2004, 69,
5986–5992.
38 N. L. Lancaster, T. Welton and G. B. Young, J. Chem. Soc.,
Perkin Trans. 2, 2001, 2267–2270.
39 J. P. Hallett, C. L. Liotta, G. Ranieri and T. Welton, J. Org.
Chem., 2009, 74, 1864–1868.
40 J. P. Hallett, C. L. Liotta, G. Ranieri and T. Welton, ECS
Trans., 2009, 16, 81–87.
41 M. Y. Lui, L. Crowhurst, J. P. Hallett, P. A. Hunt,
H. Neidermeyer and T. Welton, Chem. Sci., 2011, 2, 60 A. Wandschneider, J. K. Lehmann and A. Heintz, J. Chem.
1491–1496. Eng. Data, 2008, 53, 596–599.
42 L. Crowhurst, R. Falcone, N. L. Lancaster, V. Llopis-Mestre 61 J. Lachwa, P. Morgado, J. M. S. S. Esperanca,
and T. Welton, J. Org. Chem., 2006, 71, 8847–8853.
43 R. W. Taft and M. J. Kamlet, J. Am. Chem. Soc., 1976, 98,
2886–2894.
44 M. J. Kamlet and R. W. Taft, J. Am. Chem. Soc., 1976, 98,
377–383.
H. J. R. Guedes, J. N. Canongia Lopes and L. P. N. Rebelo,
J. Chem. Eng. Data, 2006, 51, 2215–2221.
62 Y. Deng, P. Husson, J. Jacquemin, T. G. A. Youngs,
V. L. Kett, C. Hardacre and M. F. Costa Gomes, J. Chem.
Thermodyn., 2011, 43, 1708–1718.
2542 | Org. Biomol. Chem., 2013, 11, 2534–2542
This journal is © The Royal Society of Chemistry 2013