Nucleophilic displacement reactions in ionic liquids: Substrate and solvent effect in the reaction of NaN3 and KCN with alkyl halides and tosylates

Article abstract of DOI:10.1021/jo026838h

Room-temperature ionic liquids have been used as environmentally benign solvents for the preparation of primary and secondary alkyl azides and nitriles under solid-RTIL phase-transfer conditions. The reaction of primary, secondary, and tertiary halides or tosylates with KCN and NaN₃ has been investigated in three ionic liquids ([bmim][PF₆], [bmim][N(Tf) ₂], and [hpyr] [N(Tf)₂]). The observed nucleofugacity scales for the reaction of NaN₃ are similar to those reported for the same process in cyclohexane, indicating that in these solvents it is possible to evidence the intrinsic ability to depart of leaving groups. Changes in the nature of the IL cation or anion determine significant modifications in reactivity of the investigated substrates. Reactivity has been interpreted considering a gradual shift of the mechanism from concerted S_N2 (primary substrates) to stepwise S_N1 (tertiary substrate, 3), through the nucleophilically assisted formation of an ion pair intermediate, in the case of 2d.

Full text of DOI:10.1021/jo026838h

Nu cleop h ilic Disp la cem en t Rea ction s in Ion ic Liqu id s: Su bstr a te
a n d Solven t Effect in th e Rea ction of Na N3 a n d KCN w ith Alk yl
Ha lid es a n d Tosyla tes†
Cinzia Chiappe,* Daniela Pieraccini, and Paola Saullo
Dipartimento di Chimica Bioorganica e Biofarmacia, via Bonanno 33, 56126 Pisa, Italy
cinziac@farm.unipi.it
Received December 11, 2002
Room-temperature ionic liquids have been used as environmentally benign solvents for the
preparation of primary and secondary alkyl azides and nitriles under solid-RTIL phase-transfer
conditions. The reaction of primary, secondary, and tertiary halides or tosylates with KCN and
NaN3 has been investigated in three ionic liquids ([bmim][PF6], [bmim][N(Tf)2], and [hpyr] [N(Tf)2]).
The observed nucleofugacity scales for the reaction of NaN3 are similar to those reported for the
same process in cyclohexane, indicating that in these solvents it is possible to evidence the intrinsic
ability to depart of leaving groups. Changes in the nature of the IL cation or anion determine
significant modifications in reactivity of the investigated substrates. Reactivity has been interpreted
considering a gradual shift of the mechanism from concerted SN2 (primary substrates) to stepwise
SN1 (tertiary substrate, 3), through the nucleophilically assisted formation of an ion pair
intermediate, in the case of 2d .
In tr od u ction
tion times, as well as product distribution. Since ionic
liquids have properties completely different from the
molecular solvents, it is evident that they may have
dramatic effects on reactions occurring within them.
Considering the number of available ionic liquids, and
the possibility to tailor new ionic liquids, surely the
chemist needs in addition to his intuition some general
rules to facilitate the choice. We have therefore decided
to study some typical organic reactions, such as electro-
philic additions6 and nucleophilic substitutions reactions,
to understand how the properties of these solvents may
affect chemical reactivity. In particular, in this work we
have investigated the possibility to obtain azides and
nitriles from alkyl halides and tosylates by a nucleophilic
substitution reaction using NaN3 or KCN in three ionic
liquids, [bmim][PF6], [bmim][N(Tf)2], and [hpyr][N(Tf)2]
(where bmim ) 1-butyl-3-methylimidazolium, hpyr )
1-hexylpyridinium, PF6 ) hexafluorophosphate, and
N(Tf)2 ) bis(trifluoromethylsulfonyl)imide).
The term ionic liquids (ILs) has been introduced in
recent years to describe a class of organic salts that are
liquid in their pure state at or near room temperature.
These liquids possess several properties, such as low
melting point, negligible vapor pressure, low coordinating
ability, and excellent thermal and chemical stability,
which make them attractive alternatives to traditional
solvents in a wide range of chemical processes,1 in liquid-
liquid extractions,2 and for ultralow volatility liquid
matrixes.3 Several excellent reviews describe many or-
ganic reactions that have successfully been performed
with high yields using these liquids as solvents.1 How-
ever, physical and chemical properties of some of these
solvents have only recently been investigated,4 and few
data have been reported to explain how the bulky
properties and the microscopic dynamic of these new
reaction media may affect reactivity.5 It is well-known
that the selection of the solvent is one of the most
important features for the success of a planned reaction;
solvent can affect equilibria and rate-determining reac-
† In memory of Professor Serena Catalano (1945-2002).
* Corresponding author.
(1) (a) Holbrey, J . D.; Seddon, K. R Clean Products Processes 1999,
1, 223. (b) Earle, M. J .; Seddon, K. R. Pure Appl. Chem. 2000, 72, 1391.
(c) Welton, T. Chem. Rev. 1999, 99, 2071. (d) Wasserscheid, P.; Keim,
M Angew. Chem., Int. Ed. 2000, 39, 3773. (e) Sheldon, R. Chem.
Commun. 2001, 2399. (e) Olivier-Bourbigou, H.; Magna, L. J . Mol.
Catal. A 2002, 182, 419. (f) Dupont, J .; de Souza, R. F.; Suarez, P. A.
Z. Chem. Rev. 2002, 102, 3667.
(2) Day, S.; J u, Y. H.; Barnes, C. E. J . Chem. Soc., Dalton Trans.
1999, 1201.
(3) Armstrong, D. W.; Zhang, L.-K.; He, L. Gross, M. L. Anal. Chem.
2001, 73, 3679.
(5) (a) Grodkowski, J .; Neta, P. J . Phys. Chem. A 2002, 106, 11130.
(b) Grodkowski, J .; Neta, P. J . Phys. Chem. A 2002, 106, 5468. (c)
McLean, A. J .; Muldoon, M. J .; Gordon, C. M.; Dunkin, I. R. Chem.
Commun. 2002, 1880. (d) Owes, G. S.; Abu-Omar, M. M. J . Mol. Catal.
A 2002, 3686, 1.
(4) Dzyuba, S. V.; Barsch, R. A. ChemPhysChem 2002, 3, 161.
Huddleston, J . G.; Visser, A. E.; Reichert, W. N.; Willauer, H. D.;
Broker, G. A.; Rogers, R. D. Green Chem. 2001, 3, 156.
(6) Chiappe, C.; Capraro, D.; Conte, V.; Pieraccini, D. Org. Lett. 2001,
3 (7), 1061. Chiappe, C.; Conte, V.; Pieraccini, V. Eur. J . Org. Chem.
2002, 2831.
10.1021/jo026838h CCC: $25.00 © 2003 American Chemical Society
Published on Web 07/25/2003
6710
J . Org. Chem. 2003, 68, 6710-6715

Nucleophilic Displacement Reactions in Ionic Liquids
It is worthy of note that, although many types of
reactions have been investigated in ILs, only few ex-
amples of nucleophilic substitution reactions have been
reported.7-9 The versatile reactivity and the synthetic
usefulness of the azido and cyano group is well docu-
mented, and there are several methods available for the
preparation of these compounds. The most common
methods utilize alkyl halides via the nucleophilic sub-
stitution with alkali metal azides or cyanides (mainly
NaN3 or KCN) in various solvents.10 However, these
methodologies often suffer from complex procedures, long
reaction times, and low yields, not to mention the usual
purification problems and the difficulties associated with
azide isolation from solvents (DMF or DMSO). To im-
prove the yields and to facilitate the product isolation,
nucleophilic displacement reactions are often carried out
under phase-transfer catalysis (PTC) conditions,11 which
favor the contact between the reagents and provide
activation of the nucleophilic anion. Unfortunately, in
conventional PTC the organic solvents used are environ-
mentally undesirable species. Recently, Eckert et al. have
demonstrated the possibility to use of [bmim][PF6] as a
catalyst and solvent for the cyanide displacement of
benzyl chloride. Furthermore, when this work was under
review, Afonso et al. have reported a study about the use
of the same IL as a catalyst and solvent for several
nucleophilic substitution reactions under aqueous-IL
phase-transfer conditions.12
that imidazolium cation can have on the nucleophilic
substitution reactions.
Since it is well-known14 that the presence of impurities
(unreacted base and halides) markedly affect the phys-
icochemical properties of RTILs, and that halides can
compete with added nucleophilic anions (CN- and N3-),
particular attention was paid to the purity of the ILs used
(see the Experimental Section). Furthermore, considering
that the presence of water can have dramatic effects on
reactivity, all ILs were subjected to a drying procedure
(2 h at 80 °C under vacuo) before use. The water content
of the dried ILs was determined by Karl-Fisher titra-
tion: [bmim][PF6] 348 ( 25 ppm, [bmim][N(Tf)2] 152 (
20 ppm, [hpyr][N(Tf)2] 120 ( 10 ppm.
To obtain information about the ability of ILs to affect
substitution reactions simple primary and secondary long
chain alkyl halides or tosylates (1a -d and 2a -c) and a
tertiary halide, 1-iodoadamantane (3), were chosen as
substrates, while KCN and NaN3 were used as nucleo-
phile sources. The reactions were carried out in screw
cup vials, in a recirculating heated bath, and were stirred
with magnetic stir bars. The substrate concentration was
0.5 M, and the amount of KCN or NaN3 was three times
the stoichiometric amount of alkyl halide or tosylate.
While the substrates were completely soluble in the ionic
liquids, solid KCN or NaN3 was present in the system at
all times. Therefore, before introducing the substrate, the
salt was stirred overnight in the ionic liquid in order to
reach an equilibrium concentration. Reactions were car-
ried out at 80 °C, under stirring, and at prefixed times
were stopped by product extraction with Et2O. The
product distribution was determined by NMR and GC-
MS analysis, and conversions were evaluated by GC after
addition of a proper amount of an internal standard. All
the reactions were carried out at least in triplicate. Table
1 illustrates the reaction with NaN3 while Table 2 reports
data related to reaction with KCN. It is worth noting that
the addition of water (5 equiv) did not affect the reaction
of primary and secondary substrates with NaN3 while it
gave higher yields of the substitution product in the
reaction of KCN. In contrast, substrate 3 gave always in
the presence of water relevant amounts of the corre-
sponding alcohol. These results show therefore that
anhydrous conditions are not necessary to obtain the
corresponding azides and nitriles, at least in the case of
primary and secondary substrates. Anyway, since one of
the aims of this investigation was to obtain information
about the relative reactivity of different substrates under
identical conditions, the reactions of substrates 1, 2, and
3 were carried out simultaneously using the same batch
of solvent (Scheme 1).
In this paper, we present new data about the reactions
of primary, secondary, and tertiary halides and tosylates
with KCN and NaN3 in three ionic liquids having
different properties in order to provide additional insight
on the mechanism of nucleophilic displacement reactions
under solid-RTILs PTC conditions.
Resu lts a n d Discu ssion
Meth od ology. To evaluate how the IL properties
affect the nucleophilic substitution reactions three dif-
ferent ionic liquids were used in this study; [bmim][PF6],
[bmim][N(Tf)2], and [hpyr] [N(Tf)2]. The [bmim]+ ILs have
been chosen since they are largely used in synthesis and
their physicochemical properties are significantly differ-
ent.4,13 For these ILs, in addition to the bulk physical
properties (viscosity, density and so on), some chemical
properties (dipolarity, hydrogen bond ability, and so on)
that may be important to rationalize the solvent-
substrate interactions have been recently reported.13 The
third IL has been chosen to evaluate the specific effects
(7) Wheeler, C.; West, K. N.; Liotta, C. L.; Eckert, C. A. Chem.
Commun. 2001, 887. Lancaster, N. L.; Welton, T.; Young, G. B. J .
Chem. Soc., Perkin Trans. 2 2001, 2267.
(8) Lancaster, N. L.; Saler, P. A.; Welton, T.; Young G. B. J . Org.
Chem. 2002, 67, 8855.
(9) J udeh, Z. M. A.; Shen, H.-Y. S.; Chi, B. C.; Feng, L.-C.; Selvasothi,
S. Tetrahedron Lett. 2002, 9381.
(10) (a) Scriven, E. F. V.; Turnbull, K. Chem. Rew. 1988, 88, 298.
(b) Stark, C. M. J . Am. Chem. Soc. 1971, 93, 195. (c) Landini, D.; Maia,
A.; Montanari, F.; Rolla, F. J . Org. Chem. 1983, 48, 3774.
(11) Varma, R. S.; Naicker, K. P. Tetrahedron Lett. 1998, 2915.
(12) Lourenc¸o, N. M. T.; Afonso, C. A. M. Tetrahedron 2003, 59, 789.
(13) Anderson, J . L.; Ding, J .; Welton, T.; Armstrong, D. W. J . Am.
Chem. Soc. 2002, 124, 14247. (b) Charmichael, A. J .; Seddon, K. R. J .
Phys. Org. Chem. 2000, 13, 591. (c) Aki, S. N. V. K.; Brennecke, J . F.;
Samanta, A. Chem. Commun. 2001, 413. (d) Muldoon, M. J .; Gordon,
C, M.; Dunkin, I. R. J . Chem. Soc., Perkin 2 2001, 433. (e) Fletcher, K.
A.; Storey, I. A.; Hendricks, A. E.; Pandey, S.; Pandey, S. Green Chem.
2001, 3, 210.
Deter m in a tion of Solven t P r op er ties a n d Nu cleo-
fu ga city. Table 1 shows that the reaction of primary and
secondary halides or tosylates with NaN3 provides cor-
responding azido derivatives as the sole products. Pri-
mary tosylate 1d reacts rapidly with NaN3 in all exam-
ined ILs although conversion depends on the solvent. The
solvent influence on the reaction rates is strictly related
to the nature of the leaving group. In the case of the tosyl
group reaction rates increase in the order [bmim][N(Tf)2]
< [hpyr][N(Tf)2] < [bmim][PF6]; thus, the highest rate is
(14) Seddon, K. R.; Stark, A.; Torres, M.-J . Pure Appl. Chem. 2000,
72, 2275.
J . Org. Chem, Vol. 68, No. 17, 2003 6711

Chiappe et al.
TABLE 1. Rea ction of Su bstr a tes 1-3 w ith Na N3 in ILs
a t 80 °C
TABLE 2. Rea ction of Su bstr a tes 1-3 w ith KCN in ILs
a t 80 °C
time conva
product
time conva product
substrate
solvent
(h)
(%)
selectivity
substrate
C8H17Cl
C8H17Br
C8H17Br
solvent
(h) (%) selectivity
4
8
C8H17Cl
C8H17Br
C8H17
1a [bmim][PF6]
1b [bmim][PF6]
1c [bmim][PF6]
1d [bmim][PF6]
7
7
7
3
7
30
20
99
100
100
100
100
5/6
100:0
>95:<5
7
1a [bmim][PF6]
1b [bmim][PF6]
1b [bmim][PF6] + water
1c [bmim][PF6]
7
7
7
7
7
5
100
18 100
95 100
95 100
95 100
9/6
I
C8H17OTs
C8H17
I
C8H17OTs
1d [bmim][PF6]
sec-C7H15Br
sec-C8H17OTs 2d [bmim][PF6]
2b [bmim][PF6]
7
3
56
100
sec-C7H15Br 2b [bmim][PF6]
sec-C8H17OTs 2d [bmim][PF6]
7
7
14 >95:ndb
100 0:100
10
AdI
3
[bmim][PF6]
7
42
90
4
AdI
3
[bmim][PF6]
7
ndb
C8H17Cl
C8H17Br
1a [bmim][N(Tf)2]
1b [bmim][N(Tf)2]
1c [bmim][N(Tf)2]
1d [bmim][N(Tf)2]
1d [bmim][N(Tf)2]
7
7
7
7
3
15
75
42
100
60
100
100
100
100
100
5/6
100:0
90:10
7
8
100
100
C8H17Cl
C8H17Br
1a [bmim][N(Tf)2]
1b [bmim][N(Tf)2]
1c [bmim][N(Tf)2]
1d [bmim][N(Tf)2]
7
7
7
7
4
7
C8H17
I
C8H17OTs
C8H17OTs
C8H17
I
37 100
75 100
9/6
C8H17OTs
sec-C7H15Br
sec-C8H17OTs 2d [bmim][N(Tf)2]
2b [bmim][N(Tf)2]
7
3
40
90
sec-C7H15Br 2b [bmim][N(Tf)2]
sec-C8H17OTs 2d [bmim][N(Tf)2]
7
7
8
>95:ndb
>90 0:100
10
AdI
3
[bmim][N(Tf)2]
7
55
90
AdI
3
[bmim][N(Tf)2]
7
ndb
4
8
0
10 100
50 100
31 100
9/6
C8H17Cl
C8H17Br
1a [hpyr][N(Tf)2]
1b [hpyr][N(Tf)2]
1c [hpyr][N(Tf)2]
1d [hpyr][N(Tf)2]
1d [hpyr][N(Tf)2]
7
7
7
7
3
6
60
25
90
88
100
100
100
100
100
5/6
100:0
90:10
7
C8H17Cl
C8H17Br
1a [hpyr][N(Tf)2]
1b [hpyr][N(Tf)2]
1c [hpyr][N(Tf)2]
1d [hpyr][N(Tf)2]
7
7
7
7
C8H17
I
C8H17
I
C8H17OTs
C8H17OTs
C8H17OTs
sec-C7H15Br 2a [hpyr][N(Tf)2]
sec-C8H17OTs 2b [hpyr][N(Tf)2]
7
7
10 >95:ndb
100 0:100
10
sec-C7H15Br
sec-C8H17OTs 2d [hpyr][N(Tf)2]
2b [hpyr][N(Tf)2]
7
3
50
100
AdI
3
[hpyr][N(Tf)2]
7
ndb
AdI
3
[hpyr][N(Tf)2]
7
31
90
a
Each value represents the mean of at least three determina-
a
Each value represents the mean of at least three determina-
tions. Average error: (5%. Isolated yields, given by the sum of
substrate and product(s), were always >90%. nd ) not detected.
b
tions. Average error: (5%. Isolated yields, given by the sum of
substrates and product(s), were always >90%.
SCHEME 1
observed in the most viscous [bmim][PF6]. The situation
is completely different in the case of halides 1a -c for
which the highest rates are observed in [bmim][N(Tf)2];
the IL has the higher hydrogen bond acidity.13 The
reactivity scale as a function of the leaving group is,
however, the same in the three ILs: OTs > Br > I . Cl.
It is well-known that in molecular solvents the reactivity
sequence depends on the solvent, being mainly affected
by the ability of the medium to solvate both the entering
group and the activated complex. The nucleophilicity
-
scale for nucleophilic substitution by N3 in a series of
n-octyl derivatives is indeed I > Br > OTs > Cl in DMSO
and OTs > I > Br . Cl in cyclohexane.10c This behavior
has been explained10c considering that on going from the
polarizable DMSO to nonpolar, nonpolarizable cyclohex-
ane the solvation of the entering group and activated
complex becomes unimportant. Therefore, in cyclohexane
better than in any other solvent the observed reaction
rates reflect not only the intrinsic nucleophilicity of the
entering group but also the nucleofugacity of the leaving
group. The reactivity sequence found in the three ILs,
which exactly corresponds to that calculated15 for SN2
reactions in gas phase, seems to suggest the absence of
strong specific interactions between the examined ILs
and the activated complex.
that of the corresponding primary substrates, should be
noted. Secondary alkyl substrates generally react by an
SN2 mechanism, except that SN1 mechanism may become
important at high solvent polarity. Thus, they are gener-
ally enclosed in the borderline region between SN1 and
SN2 mechanisms. Alternative interpretations have been
given of this borderline region. For example, Sneen16
Related to this reaction, the high reactivity of the
secondary substrates, which is comparable or higher than
(15) Pearson, R. G. J . Org. Chem. 1987, 52, 2131.
6712 J . Org. Chem., Vol. 68, No. 17, 2003
(16) Sneen, R. A. Acc. Chem. Res. 1973, 6, 46.

Nucleophilic Displacement Reactions in Ionic Liquids
SCHEME 2
SCHEME 4
SCHEME 3
mechanism (upper pathway) and two mechanism involv-
ing preassociation: the concerted (lower pathway), and
the stepwise (central pathway). When the sandwich
intermediate complex (Nu•-R•+X-) break down to reac-
tants faster than Nu- diffuse away from it, i.e., k-i > k-d
,
the lowest energy pathway will be the preassociation
mechanisms.
It is evident from this brief summary that to establish
the reaction mechanism in the case of nucleophilic
substitution on secondary substrates is not easy. Their
reactions in molecular solvents have been object of
extensive discussions. However, the high reactivity of the
secondary substrates with respect the primary ones,
observed for the reaction of NaN3 in all examined ILs, is
in disagreement with a concerted SN2 mechanism. Among
the remaining possible reaction pathways, the distinction
between a free intermediates mechanism (SN1) and the
more concerted processes (SN2, SN2(intermediate), pre-
association) can be made on the basis of known19b criteria:
(i) The reaction of the free intermediate R+ with Nu-
is often diffusion controlled; therefore, the observation
of different rates with different nucleophiles under the
same conditions rules out rate determining diffusion
controlled trapping of the intermediate.19b In our experi-
considered four possible mechanisms to explain the
stereochemical and kinetic behavior of substitution reac-
tions: SN1, SN2, concurrent SN1 and SN2, and substitution
at the stage of intimate ion pairs. In this latter situation,
products are formed by a dipolar reaction on a reversible
formed ion pair. However, Sneen’s interpretations and
his conclusions have been criticized extensively. Consid-
ering the SN1 and SN2 reactions as “limiting” behavior,
subsequently, the SN2(intermediate) mechanism has been
proposed to account for relatively weakly nucleophilically
assisted processes showing ion-pair characteristics.17 The
term SN2(intermediate) should emphasize the possibility
that an SN2 reaction may proceed via a nucleophilically
solvated ion pair intermediate, which is energetically
more feasible than Sneen’s ion pairs mechanism. The
pentacoordinate intermediate proposed by this mechan-
ism is an ion pair similar to the activated complex in the
one stage SN2 reaction (Scheme 2).17
-
ments substrate 2b quickly reacts with N3 but, under
the same conditions, does not react with CN-. This
excluded a diffusion-controlled trapping of a free second-
ary carbocation.
Also, this mechanism has not been without criticism
and alternative reaction pathways have been proposed.
To explain the reaction mechanism in the borderline
situation it has been proposed that ion pairs could be
formed by heterolysis of the bond between carbon and
leaving group before the rate-determining nucleophilic
attack, and “internal return” to starting materials occurs
more rapidly than the attack of the nucleophile.18 Alter-
natively, a preassociation mechanism with yet another
timing association of the nucleophile with the substrate
occurring before the rate-limiting step has been consid-
ered.19 On the basis of this model, as suggested19 by
J ancks, a clear distinction between the different mech-
anisms can be made considering the lifetime of the
intermediate rather than on the character of the transi-
tion state. When the intermediate does not exist, or it is
too unstable to diffuse away through the solvent, the
reaction must occur through a preassociation mechanism
in which the reactants are assembled in an encounter
complex before the first bonding-making and bonding-
breaking step occurs. The preassociation mechanism can
be either concerted, with no intermediate, or stepwise,
with intermediate. Scheme 3 reports for a substitution
reaction in which Nu- is the nucleophile the classical SN1
(ii) A free intermediate must show constant partition-
ing, regard less of its source, in its reaction with different
nucleophiles.19b In our experiments, substrate 2d gives,
in the same solvent and at the same temperature, the
substitution product with N3- and an elimination product
with CN-.
(iii) A free carbocation intermediate gives generally
racemization having the same reactivity on both sides
of the trigonal carbon.19b The reaction of the optically pure
tosyl derivative of (S)-2-hexanol (11) with NaN3 in the
three examined ionic liquids gives the corresponding (R)-
2-azidohexane20 (12) through a process characterized by
a practically complete inversion (ee > 93%, determined
by GC on a chiral column) (Scheme 4).
Analogously, the triflate ester of diacetone-D-glucose
(13) gives by reaction with NaN3, in [bmim][PF6] at 80
°C, exclusively the azido derivative21 (14) having D-allo
configuration which arises from complete inversion
(Scheme 5).
It is worth noting that when the same reaction was
carried out in DMF a ca. 1:1 mixture of azido derivative
14 and elimination product 15 was obtained. Thus, IL
reduces elimination in favor of substitution.
Finally, 2-bromoadamantane practically does not react
with NaN3 in all the three ILs under reaction conditions
which give a conversion >50% from 2b. This halide is a
(17) Bentley, T. W.; Bowen, C. T.; Morten, D. H.; Schleyer, P. v. R.
J . Am. Chem. Soc. 1981, 103, 5466 and references therein.
(18) Allen, A. D.; Kanagasabapathy, V. M.; Tidwell, T. J . Am. Chem.
Soc. 1985, 107, 4513.
(19) (a) J encks, W. P. Acc. Chem. Res. 1980, 13, 161. (b) J encks, W.
P. Chem. Soc. Res. 1981, 10, 345.
(20) Levene; Rothen; Kuna. J . Biol. Chem. 1937, 120, 789.
(21) Nayak, U. G.; Whistler, R. L. J . Org. Chem. 1969, 34, 3819.
Hanessian, S.; Vate`le, J .-M. Tetrahedron Lett. 1981, 3579.
J . Org. Chem, Vol. 68, No. 17, 2003 6713

Chiappe et al.
complex in the CN- reaction, which probably occur
through a transition state characterized by a more
advanced C-X breaking. Finally, secondary bromide 2b
reacts slower with CN- than with N3-, while tosylate 2d
gives under the same conditions exclusively the corre-
sponding elimination product(s). The reactivity order of
2b (N3- . CN-) is opposite to that observed in molecular
solvents for displacement reactions on methyl halides,
but the same as for nucleophilic reactions with carboca-
tions.24 This behavior confirms the large amount of
carbocation character of the transition state. Further-
more, the different elimination/substitution ratio found
in the reaction of 2b and 2d with KCN strongly supports
the involvement of ion pair intermediates in the reaction
of secondary substrates. For the first-order SN1 reactions
the leaving group has nothing to do with the competition
between elimination and substitution. Analogously in SN2
reactions the elimination/substitution ratio is not greatly
dependent on leaving group and, in particular, when OTs
is the leaving group generally there is much more
substitution.25 Only in the case of ion-pair intermediates
leaving group can affect this ratio.
SCHEME 5
SCHEME 6
typical secondary alkyl substrate for which nucleophilic
assistance is hindered by the cage structure and it is
generally used as model for “limiting” behavior. Substitu-
tion can occur exclusively through SN1 mechanism
(Scheme 6).
In conclusion, all of the results exclude the involvement
of a free carbocation intermediate in the reaction of
secondary substrates with NaN3 in the examined ILs
suggesting a more concerted mechanism. We therefore
retain that the reaction proceeds through the rear-side
nucleophilic attack of azide with a SN2 mechanism in the
case of primary substrates. This mechanism gradually
shifts toward a pure SN1, which surely occurs in the case
of the bridgehead cage compound 3, passing through the
nucleophilically assisted formation of an ion pair inter-
mediate in the case of 2d . It is noteworthy that the
reactions of primary and secondary substrates might
occur even with preassociation.22
Finally, the data of secondary substrates show that in
the examined ILs CN- anion is a hard anion prone to
abstract a proton giving the elimination adduct, while
-
the N3 anion is a softer anion which interacts prefer-
entially with an sp3 carbon to give a substitution product.
Since elimination in the 1-adamantyl nucleus is highly
improbable, due to the large amount of strain which
would be present in the resulting bridgehead olefin,
iodide 3 gives neither the corresponding nitrile, nor the
elimination product, but a complex mixture of unidenti-
fied products or the corresponding alcohol in the presence
of water.
Theoretical calculations indicate the presence of an
encounter complex nucleophile-substrate (or -product),
corresponding to an energy minimum before and after
the energy maximum, on the reaction coordinate for the
SN2 substitution process in gas phase.23 In solution,
generally the complexes become less important on in-
creasing solvent polarity due to the solvation of the anion.
However, if the energy of solvation of N3- anion is much
smaller in the ionic liquid than in water or protic
solvents, as recently suggested5a for Br-, the formation
of the encounter complex, and therefore, the preassocia-
tion mechanism, might become much more favorable in
these solvents than in the molecular ones.
Related to the reaction of KCN, it is noteworthy that,
as expected, our primary substrates react slower with
KCN than benzyl chloride7 in all investigated ionic
liquids. The reaction is furthermore much more sensitive
than that of NaN3 to the nature of the solvent and to the
presence of water and, maybe in agreement with this
feature, the higher conversions have been obtained in
[bmim][PF6], the IL containing the higher amount of
water. It is also worth noting that the nucleofugacity
scale becomes with this nucleophile OTs > I > Br . Cl
in [bmim]+ ILs and I > OTs > Br . Cl in [hpyr][N(Tf)2].
This seems to indicate the presence of more specific
interactions solvent nucleophile and/or solvent activated
Con clu sion s
In conclusion, this work further confirms the prelimi-
nary report indicating that ionic liquids may be used as
solvents for nucleophilic substitution reactions between
organic compounds and inorganic salts.7,8,12 Furthermore,
it shows for the first time that it is possible to obtain in
these media primary, secondary, and tertiary azides,
without problems of purification and isolation. It is worth
noting that in the reaction of NaN3 elimination does not
compete with substitution even in the case of the steri-
cally hindered substrates, as the triflate ester of diac-
etone-D-glucose. This behavior is in contrast with that
observed in DMF. Furthermore, the data obtained under
strictly comparable conditions show that the course of
the substitution reactions depends on the nature of the
nucleophile, on the nature of the leaving group and on
the properties of the reaction medium. Changes in the
nature of IL cation or anion determine significant modi-
fications in the reactivity of the same substrates with the
same nucleophile. Although ILs should be considered
polar solvents (polarity measured13 using dye probes
ranges between that of short-chain alcohols and aceto-
nitrile-DMSO) the results reported in this work indicate
that concerted or rear-side nucleophilically assisted reac-
(22) Richard, J . P.; J encks, W. P. J . Am. Chem. Soc. 1984, 106, 1383.
(23) Chandrasekhar, J .; J orgensen, W. L. J . Am. Chem. Soc. 1985,
107, 2974. Safi, B.; Choho, K.; Geerlings, P. J . Phys. Chem. A 2001,
105, 591.
(24) Pearson, R. G.; Sobel, H.; Songstand, J . J . Am. Chem. Soc. 1968,
90, 319. Ritchie, C. D.; Gandler, J . J . Am. Chem. Soc. 1979, 101, 7318.
(25) Mearch, J . Advanced Organic Chemistry: reactions, mecha-
nisms and structure, IVth ed.; J . Wiley & Sons: New York, 1992.
6714 J . Org. Chem., Vol. 68, No. 17, 2003

Nucleophilic Displacement Reactions in Ionic Liquids
Syn th esis of Tosyl Der iva tives of P r im a r y a n d Sec-
on d a r y Alcoh ols. Gen er a l P r oced u r e. To a solution of
p-toluensulfonyl chloride (5 g, 0.027mol) in 25 mL of dichlo-
romethane, containing a catalytic amount of pyridine (0.5 mL),
was added the proper alcohol (1-octanol, 2-octanol, or (S)-(+)-
2-hexanol) (0.018 mol). The reaction was followed by TLC. At
the end of the reaction, the pyridinium chloride (visible as long
white needles) was removed by filtration, and the resulting
filtrate was washed with an aqueous solution of HCl (0.1 N),
then with NaHCO3 (aq), and finally with water until neutral-
ization. The organic layer was then dried (MgSO4) and filtered,
and the solvent was removed by distillation at reduced
pressure.
tions (with or without preassociation) are preferred with
respect the formation of free ionic intermediates. Ionic
liquids seem therefore to behave as ionizing but not
dissociating solvents. Work is in progress to assess the
ionic intermediates stability in these new reaction media.
Exp er im en ta l Section
Gen er a l Rem a r k s. 1H and 13C NMR spectra were recorded
in CDCl3. GC-MS spectra were carried out on a 30 m DB5
capillary column using an instrument equipped with an ion
trap detector. GC analysis were carried out using an ECONO-
CAP EC-5 column (30 m) or a 25 m Chiraldex GTA column.
1-Chlorooctane (99%), 1-bromooctane (99%), 1-iodooctane (98%),
1-octanol (99%), 2-octanol (97%), 2-bromoheptane, 2-bromoad-
amantane (98%), 1-iodoadamantane (98%), (S)-(+)-2-hexanol
(99%), and diacetone-D-glucose (98%) were used without
purification. [bmim][PF6], [bmim][N(Tf)2], and [hpyr][N(Tf)2]
were prepared following the reported procedures:26 attention
was paid to the elimination of bases and Cl- ions which may
be present in the solvents as impurities. To eliminate the
eventually present traces of unreacted bases [bmim][Cl] and
[hpyr][Cl] were carefully recrystallized. The presence of 1-me-
thylimidazole in [bmim][Cl] was checked using the recently
reported27 colorimetric method. To remove chloride anion, after
metathesis reactions, all ILs were washed several times with
deionized water until no chloride was found in the aqueous
phase (silver nitrate test in the aqueous phase). The purity of
imidazolium salts was finally checked by UV measuring the
absorption spectra between 240 and 400 nm. Purified [bmim]+
salts (containing Cl- < 0.1 ppm) have practically no absorption
band in the 250-300 nm region.28 After drying (2 h at 80 °C
under vacuo) the water amount in ILs was determined by Karl
Fisher technique using an apparatus composed of a stand
titrator and a coulometer. 3-O-(Trifluoromethanesulfonyl)-1,2:
5,6-di-O-isopropylidene-R-D-glucofuranose was prepared as
previously reported.29
Su bstitu tion Rea ction s. Gen er a l P r oced u r es. Reactions
were carried out in a screw cup with Teflon-faced rubber
septum vials under magnetic stirring. To a suspension of KCN
or NaN3 (3 mmol) in 2 mL of ionic liquid, stirred overnight at
80 °C, was added a proper amount of substrate (1 mmol). The
mixture was stirred at 80 °C for the times reported in Tables
1 and 2, and then the products were extracted with Et2O (1
mL × 5). The combined organic layers were dried (MgSO4) and
diluted to an exactly known volume. A portion of this solution,
exactly measured, was analyzed by GC-MS after addition of
an appropriate amount of an internal standard (benzonitrile
or 1-chlorohexane). The remaining solution was analyzed by
NMR after removal of the solvent by distillation in vacuo.
The reaction of 3-O-(trifluoromethanesulfonyl)-1,2:5,6-di-O-
isopropylidene-R-D-glucofuranose with NaN3 in DMF at 80 °C
was carried out using the same procedure described above.
Products were extracted as previously reported.21
Ack n ow led gm en t. We acknowledge the financial
contribution of MIUR and Universita` di Pisa.
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
spectra of 1d , 2d , (S)-2-hexyltosylate, 1-azodooctane, 1-cya-
nooctane, (R)-2-azidohexane, 2-azidoheptane, and 2-azidooc-
tane. This material is available free of charge via the Internet
at http://pubs.acs.org.
(26) Fuller, R. T.; Carlin, H. C.; De Long, D.; Haworth, J . Chem.
Soc., Chem. Commun. 1994, 299. Cammarata, L.; Kazarian, S. G.;
Salter, P. A.; Welton, T. Phys. Chem. Chem. Phys. 2001, 3, 5192.
Chiappe, C.; Pieraccini, D. Green Chem. 2003, 5, 193.
(27) Holbrey, J . D.; Seddon, K. R.; Wareing, R. Green Chem. 2001,
3, 33.
J O026838H
(28) Billard, I.; Moutiers, G.; El Azzi, A.; Gaillard, C.; Mariet, C.;
Lu¨tzenkinchen, K. Inorg. Chem. 2003, 42, 1726.
(29) Binkley, R. W.; Ambrose, M. G.; Hehemann, D. G. J . Org. Chem.
1980, 45, 299.
J . Org. Chem, Vol. 68, No. 17, 2003 6715