Phase-transfer agents as catalysts for a nucleophilic substitution reaction in microemulsions

Article abstract of DOI:10.1002/chem.200306048

The reaction between 4-tert-butylbenzyl bromide and potassium iodide was carried out in microemulsions based on different nonionic surfactants, and the reaction rates were compared with those obtained in two-phase systems with added phase-transfer agent, either a quaternary ammonium salt or a crown ether. The reactions were relatively fast in the microemulsions and extremely sluggish in the two-phase systems without additional phase-transfer agent. Addition of a phase-transfer agent did not accelerate the reaction when a hydrocarbon was used as organic solvent, neither in the two-phase system nor in the microemulsion. When a chlorinated hydrocarbon was used as solvent, phase-transfer catalysis became effective and the rate obtained in the two-phase system with an equimolar amount of phase-transfer agent added was higher than that obtained in the microemulsion. When a catalytic amount of phase-transfer agent was used, the rate in the two-phase system was about the same as the rate obtained in the micro-emulsion without the phase-transfer agent. The combined approach, that is, use of a microemulsion as the reaction medium and addition of a phase-transfer agent, gave the highest reaction rate. The quaternary ammonium salt (tetrabutylammonium hydrogen sulfate) was a more efficient catalyst in the microemulsion system than the crown ether ([18]crown-6).

Full text of DOI:10.1002/chem.200306048

FULL PAPER
Phase-Transfer Agents as Catalysts for a Nucleophilic Substitution Reaction
in Microemulsions
[
a,c]
[b]
Maria Häger*
and Krister Holmberg
Abstract: The reaction between 4-tert-
butylbenzyl bromide and potassium
iodide was carried out in microemul-
sions based on different nonionic sur-
factants, and the reaction rates were
compared with those obtained in two-
phase systems with added phase-trans-
fer agent, either a quaternary ammoni-
um salt or a crown ether. The reactions
were relatively fast in the microemul-
sions and extremely sluggish in the
two-phase systems without additional
phase-transfer agent. Addition of a
phase-transfer agent did not accelerate
the reaction when a hydrocarbon was
used as organic solvent, neither in the
two-phase system nor in the micro-
emulsion. When a chlorinated hydro-
carbon was used as solvent, phase-
transfer catalysis became effective and
the rate obtained in the two-phase
system with an equimolar amount of
phase-transfer agent added was higher
than that obtained in the microemul-
phase-transfer agent was used, the rate
in the two-phase system was about the
same as the rate obtained in the micro-
emulsion without the phase-transfer
agent. The combined approach, that is,
use of a microemulsion as the reaction
medium and addition of a phase-trans-
fer agent, gave the highest reaction
rate. The quaternary ammonium salt
(tetrabutylammonium hydrogen sul-
fate) was a more efficient catalyst in
the microemulsion system than the
crown ether ([18]crown-6).
sion. When
a
catalytic amount of
Keywords: kinetics
·
microemul-
sion · nucleophilic substitution ·
phase-transfer catalysis
Introduction
solvents, such as DMSO and DMF, offer a simple and effi-
cient solution to the incompatibility. Not only are these liq-
uids excellent solvents, the reactivity when it comes to reac-
tions involving a nucleophilic attack is often high, because
the ionic reactant is only slightly solvated in these solvents
and consequently obtains a high nucleophilicity. However,
for toxicity and price reasons these solvents are unsuitable
for large-scale operations. An alternative approach is to use
a two-phase system consisting of water and a water-immisci-
ble organic solvent. The incompatible reactants dissolve in
the different phases, that is, the ionic reactant in the aque-
ous phase and the organic compound in the organic phase,
and the reaction can take place at the interface between the
two phases. The disadvantage with this approach is that the
interface is small, which usually causes low reaction rate.
Phase-transfer catalysis (PTC) can dramatically improve the
situation. In PTC a phase-transfer agent (PTA), usually a
quaternary ammonium or phosphonium salt (Q salt) or a
crown ether, is added to the two-phase system. These PTAs
have the ability to carry the ionic reactant into the organic
phase. Once in the organic phase, the reactant becomes
highly reactive because the degree of solvation is low; thus,
the anion behaves like a ꢀnakedꢀ ion. The choice of organic
solvent is often crucial for good extraction efficiency and
high nucleophilicity. The best solvents for PTC are those
that are aprotic and slightly polar. Chlorinated hydrocar-
For a chemical reaction to proceed it is necessary that the
reacting species can enter into intimate contact. Most reac-
tions are therefore carried out in homogeneous solutions.
However, when the solubility characteristics of the reactants
are very different, it may be difficult to find a common sol-
vent for the reacting species. This is, for instance, often the
case when ionic species are to react with nonpolar organic
compounds. For such and many other cases, polar aprotic
[
a] Dr. M. Häger
Institute for Surface Chemistry
P.O. Box 5607, 114 86 Stockholm (Sweden)
Fax : (+46)820-8998
E-mail: maria.hager@surfchem.kth.se
maria.hager@chemeng.ntnu.no
[
b] Prof. K. Holmberg
Chalmers University of Technology
Department of Materials and Surface Chemistry
1296 Göteborg (Sweden)
4
Fax : (+46)31-160062
E-mail: kh@chem.chalmers.se
[
c] Dr. M. Häger
Norwegian University of Science and Technology
Ugelstad Laboratory, Department of Chemical Engineering
491 Trondheim (Norway)
7
Fax : (+47)7359-4080
5460
ꢁ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/chem.200306048
Chem. Eur. J. 2004, 10, 5460 – 5466

5
460 – 5466
bons, such as dichloromethane and chloroform, are com-
monly used for this purpose.
Use of a microemulsion as the reaction medium consti-
[1–7]
tutes an alternative to PTC.
A microemulsion is a ther-
modynamically stable mixture of water, oil (organic solvent)
and surfactant, with the surfactant forming a monolayer that
separates the oil and water domains. The curvature of the
surfactant film depends on various parameters, such as com-
ponent (water, oil and surfactant) ratio, temperature and
salt concentration. A microemulsion may consist of closed
oil domains in water or closed water domains in oil, or it
may be bicontinuous, that is, it may contain at the same
time continuous oil and water domains. All interfaces are
covered by surfactant. The domains are very small, in the
nanometer range, which leads to a very large total interfa-
cial area at which the reactants can collide and react.
The use of a two-phase system with added PTA and the
use of a microemulsion are two alternatives to overcome re-
agent incompatibility problems in organic synthesis. Both
routes have proven useful, but on entirely different ac-
counts. In PTC, the ionic reactant is carried into the organic
phase, in which it becomes highly reactive. In the micro-
emulsion approach there is no transfer of reactant from one
environment to another; the method relies on the creation
of the very large oil–water interface at which the reaction
occurs.
Scheme 1. Top: Nucleophilic substitution reaction between 4-TBBB and
KI. Bottom: Structures of TBAB (X=Br), TBAHS (X=HSO ) and
18]crown-6.
4
[
exists between 23 and 298C. Below 238C there is a two-
phase area with an L1 phase coexisting with oil (a so-called
Winsor I system). Above 298C a lamellar liquid crystalline
phase (L ) appears. Addition of the organic reactant, 4-
a
TBBB, moves the L1 region to lower temperatures, as is
shown in Figure 1. On the other hand, adding KI or a PTA
moves the L1 region to slightly higher temperatures, which
is shown in Figure 2.
In this paper we report results from an attempt to com-
bine the two approaches, that is, using a microemulsion as
reaction medium in presence of a PTA. There are only a
few reports in the literature on the use of PTAs in combina-
tion with microemulsions and none of these have dealt with
[
8–9]
the issue in a systematic way.
PTAs in micelles and macroemulsions have been studied in
However, the effect of
[10–11]
some detail by Battal and co-workers.
We here show
that PTC can indeed be used to accelerate microemulsion-
based reactions and we believe that the results presented
offer great promise for use of the combined approach in
preparative organic chemistry.
2 5
Figure 1. Phase diagram of the ternary D O/C12E /decane system in pres-
ence of 4-TBBB. The weight ratio between oil and surfactant is 48/52 and
the total weight fraction of oil+surfactant is 0.2. Raising the temperature
gives a transition from an oil-in-water microemulsion in equilibrium with
excess oil (L1+O phase) to an oil-in-water microemulsion (L1 phase)
and on to a lamellar liquid crystalline phase (L
cates the reaction composition (RC).
a
phase). The arrow indi-
Results
The reaction: The reaction studied is a typical nucleophilic
substitution involving a lipophilic organic compound and an
inorganic salt: the reaction between 4-tert-butylbenzyl bro-
mide (4-TBBB) and potassium iodide (KI) (Scheme 1). We
have previously investigated this reaction in some detail
using both microemulsions and surfactant liquid crystals as
reaction media, and we have shown that the reaction pro-
[12–13]
ceeds well in such microheterogeneous systems.
We
2 5
Figure 2. Phase diagram of the ternary D O/C12E /decane system in pres-
have also performed the reaction in a range of organic sol-
vents with different polarity (and without surfactant), and
by comparing the reaction rates obtained in the different
solvents we have been able to establish that the mechanism
is that of a second-order nucleophilic substitution reaction.
ence of KI (*), TBAHS (&) and [18]crown-6 (^). The weight ratio be-
tween oil and surfactant is 48/52 and the total weight fraction of oil+sur-
factant is 0.2. Raising the temperature gives a transition from an oil-in-
water microemulsion in equilibrium with excess oil (L1+O phase) to an
oil-in-water microemulsion (L1 phase) and on to a lamellar liquid crystal-
line phase (L
a
phase). The arrow indicates the reaction composition
(
RC). Addition of the second reactant, 4-TBBB induces a downward
Phase behaviour: The ternary system D O/decane/C E has
2
12
5
slope of both the upper and the lower phase boundary curves (see
Figure 1). Thus, the reaction composition lies within the one-phase
region in all cases.
previously been thoroughly investigated by Olsson and co-
[14–18]
workers.
An L1 phase (oil-in-water microemulsion)
Chem. Eur. J. 2004, 10, 5460 – 5466
www.chemeurj.org
ꢁ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5461

M. Häger and K. Holmberg
FULL PAPER
The phase behaviour of the system D O/C E /cyclohex-
2
12
5
ane was characterised and a partial phase diagram is shown
in Figure 3. An L2 phase (water-in-oil microemulsion) was
found in the oil-rich corner of the ternary phase diagram at
238C.
Figure 5. The reaction was monitored by following the decrease of the
ÀCH Br signal at 4.42 ppm and the increase of the ÀCH I signal at
2
2
4
.39 ppm.
and in the D O/decane two-phase system are shown in
Figure 6. As can be seen, the reaction runs well in the mi-
croemulsion, while in the two-phase system almost no prod-
2
2 5
Figure 3. Partial phase diagram of the ternary D O/C12E /cyclohexane
system at 238C. A water-in-oil microemulsion region (L2 phase) exists in
the oil-rich corner of the diagram. The arrow indicates the reaction com-
position (RC).
The ternary system D O/dichloromethane/C G at a con-
2
8
1
stant volume fraction of oil based on oil and water (0.5) has
[19]
been investigated by Egger and co-workers. The phase di-
agram is dominated by a large bicontinuous microemulsion
region, starting at a surfactant weight fraction of 0.2, as is il-
lustrated in Figure 4. The phase diagram was not markedly
affected by addition of the two reactants or of a PTA.
Figure 6. Top: Effect of Q salt (TBAHS or TBAB) on the reaction be-
tween 4-TBBB and KI at 238C in D
decane (&, ). Bottom: Effect of [18]crown-6 on the reaction between 4-
TBBB and KI at 238C in D O/decane
/decane (*,*) and D
). In both graphs open symbols denote reactions without PTA and
and black symbols reactions with PTA. The overall concentrations of 4-
TBBB, KI and [18]crown-6 are all 8.8mm.
2 5 2
O/C12E O/
/decane (*,*) and D
Figure 4. Phase diagram of the ternary D
surfactant weight fraction. The volume fraction of oil to water + oil is
.5. At low surfactant concentration a three-phase region (3) exists. On
2 8 1 2 2
O/C G /CH Cl system. a is the
&
0
2
O/C12
E
5
2
(
&,
&
increasing the surfactant concentration either an oil-in-water microemul-
sion in equilibrium with oil (2) or a water-in-oil microemulsion in equili-
¯
brium with water (2) forms depending on the temperature. At even
higher surfactant concentration a one-phase microemulsion (1) appears.
At higher concentration and a temperature lower than 158C a two-phase
region consisting of microemulsion and a lamellar liquid crystalline phase
uct was formed after 24 h reaction. Addition of an equimo-
lar amount of PTA (tetrabutylammonium hydrogen sulphate
(
1 + L
a
) is observed. (Redrawn from Ref. [19].) The arrow indicates the
reaction composition (RC).
(
TBAHS), tetrabutylammonium bromide (TBAB) or
[
18]crown-6) gave no effect, neither on the two-phase reac-
1
Reaction kinetics: The reaction was monitored by H N MR
tion nor on the reaction in microemulsion.
spectroscopy, following the rise of the ÀCH I signal and the
Decane was replaced by cyclohexane in order to enhance
the solubilizing power of the organic phase. In doing so, the
microemulsion had to be reformulated, and instead of an L1
phase (oil-in-water microemulsion) an L2 phase (water-in-
2
decay of the ÀCH Br signal, as described in the Experimen-
2
tal Section and illustrated by Figure 5. The reaction profiles
for the reactions in the D O/C E /decane microemulsion
2
12
5
5462
ꢁ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 5460 – 5466

Phase-Transfer Agents
5460 – 5466
oil microemulsion) was used as reaction medium. Reactions
were carried out in the D O/C E /cyclohexane microemul-
Phase-transfer catalysis is often performed with a catalyt-
ic, rather than an equivalent, amount of PTA. The reactions
in the D O/C G /CH Cl microemulsion and in the D O/
2
12
5
sion and in the D O/cyclohexane two-phase system in the
2
2
8
1
2
2
2
absence and presence of TBAHS. The reaction profiles were
very similar to those obtained with decane as oil component.
TBAHS did not accelerate the reactions.
CH Cl two-phase system were therefore repeated with a
2 2
0.01 mole equivalent of PTA (based on the reactants). The
reaction profiles are shown in Figure 8. Since TBAHS and
The organic phase was changed to dichloromethane in
order to further improve the solvency and extraction effi-
ciency. A microemulsion with dichloromethane as organic
component cannot be formulated with normal nonionic sur-
factants of alcohol ethoxylate type, because such surfactants
have too high solubility in chlorinated hydrocarbons. In-
stead, a formulation based on a sugar surfactant, octyl glu-
coside (C G ), was employed. Figure 7 shows the profiles for
8
1
Figure 8. Top: Effect of a catalytical amount of TBAHS on the reaction
between 4-TBBB and KI at 238C in D
CH Cl (&, ). Bottom: Effect of a catalytical amount of [18]crown-6 on
the reaction between 4-TBBB and KI at 238C in D O/C /CH Cl
(*,*)
and D O-CH Cl (&, ). In both graphs open symbols denote reaction
2 8 1 2 2 2
O/C G /CH Cl O-
(*,*) and D
&
2
2
2
8
G
1
2
2
&
2
2
2
without PTA and black symbols reactions with PTA. The overall concen-
trations of both 4-TBBB and KI are 20mm, and of PTA 0.2mm.
Figure 7. Top: Effect of Q salt on the reaction between 4-TBBB and KI
at 238C in D O/C /CH Cl
: (*) without Q-salt, (*) with TBAHS, (~)
with TBAB; and in D O-CH Cl : (&) without Q-salt, ( ) with TBAHS,
^) with TBAB. The overall concentrations of 4-TBBB, KI and Q salt
are all 20mm. Bottom: Effect of [18]crown-6 on the reaction between 4-
TBBB and KI at 238C in D O/C /CH Cl O/CH Cl
(*,*) and D
). Open symbols denote reactions without [18]crown-6 and black
2
8
G
1
2
2
TBAB gave similar results when used in equimolar amounts,
only TBAHS was tested in catalytic amounts. As can be
seen, the reactions in both the two-phase system and the mi-
croemulsion were considerably slower when the PTA addi-
tion was reduced to a catalytic amount. Reaction in the mi-
croemulsion without a PTA was approximately as fast as re-
action in the two-phase system with a PTA. The combina-
tion of a microemulsion as the reaction medium and addi-
tion of a PTA gave the highest reaction rate. The Q salt was
more effective than the crown ether.
&
2
2
2
(
2
8
G
1
2
2
2
2
2
(
&,
&
symbols reactions with [18]crown-6. The overall concentrations of 4-
TBBB, KI and [18]crown-6 are all 20mm.
the reactions performed in the D O/C G /CH Cl micro-
2
8
1
2
2
emulsion and in the D O/CH Cl two-phase system in the
2
2
2
absence and presence of an equimolar amount of a PTA
TBAHS, TBAB or [18]crown-6). The reaction profiles in
the absence of a PTA were similar to those obtained in the
All rate constants, calculated as the slope from the
straight line obtained when plotting the reverse overall con-
centration of the substrate against reaction time are com-
piled in Table 1.
(
decane-based reaction media. The reaction rate in the D O/
2
C G /CH Cl microemulsion was approximately the same as
Benzyl bromide is known to hydrolyse into benzyl alco-
hol. This side reaction would distort the monitoring of the
reaction, since there would then not be a full equivalence in
8
1
2
2
in the D O/C E /decane microemulsion, and the two-phase
2
12
5
reaction was equally as sluggish as the corresponding reac-
tion in the D O/decane system. However, addition of a PTA
terms of NMR signals between disappearence of ÀCH Br
2
2
increased the reaction rate in the two-phase system greatly.
This is not unexpected, since chlorinated hydrocarbons are
known to function well as solvents for PTC. Addition of a
PTA also increased the rate of the reaction in the micro-
emulsion to an appreciable extent.
and appearance of ÀCH I. Only in the systems with di-
2
chloromethane and with TBAHS present did a peak appear
in the NMR spectrum that can be attributed to the alcohol,
that is, to ÀCH OH methylene protons (at 4.60 ppm). The
2
ÀCH Br and ÀCH I peaks appear at 4.52 and 4.49 ppm, re-
2
2
Chem. Eur. J. 2004, 10, 5460 – 5466
www.chemeurj.org
ꢁ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5463

M. Häger and K. Holmberg
FULL PAPER
Table 1. Calculated rate constants for the reaction between equimolar overall amounts of 4-TBBB and KI in tions performed in the two-
absence and in presence of a PTA in both microemulsions and two-phase systems. The volume fraction of oil
phase systems in absence of
based on oil and water is denoted f
denoted a . The reactions in microemulsions were carried out at the compositions indicated by arrows in Fig-
ures 2–4. All reactions were performed at 238C at a stirring rate of around 900 rpm.
v
and the weight fraction of surfactant based on water, surfactant and oil is
PTA are much slower than
those occurring in the micro-
emulsions. As can be seen from
Table 1, the overall rate con-
stants for reactions in the mi-
croemulsions based on the dif-
ferent organic solvents are of
the same magnitude. One may
note that there does not seem
to be a direct relationship be-
tween the oil–water interfacial
area and the reaction rate. The
interfacial area of a microemul-
sion is largely determined by
the amount of surfactant in the
system. The microemulsions
based on decane, cyclohexane
and dichloromethane have sur-
factant weight fractions 0.10,
0.11 and 0.29, respectively. Thus
the microemulsion based on the
chlorinated solvent has a much
larger internal interface than
the other two microemulsions;
however, the rate constant in
this medium is in-between
those in decane and in cyclo-
hexane.
m
Reaction system
C
4-TBBB
À3
PTA
C
PTA
À3
f
V
a
m
k
overall
À3
À3
À3
À3
À1 À1
[10 moldm
]
[10 moldm
]
[10 dm mol
s
]
microemulsion
D
D
D
D
D
D
D
D
D
D
D
D
2
2
2
2
2
2
2
2
2
2
2
2
O/C12
O/C12
O/C12
O/C12
O/C12
O/C12
E
E
E
E
E
E
5
5
5
5
5
5
/decane
/decane
/decane
/decane
/cyclohexane 21.01
/cyclohexane 21.02
8.75
8.74
8.75
8.75
none
0.12 0.10
0.12 0.10
0.12 0.10
0.12 0.10
0.92 0.11
0.92 0.11
0.50 0.29
0.50 0.29
0.50 0.29
4.5
4.5
4.5
4.9
2.1
2.1
2.8
7.3
3.5
[18]crown-6
TBAB
TBAHS
none
TBAHS
none
[18]crown-6 19.20
[18]crown-6
TBAB
8.74
8.74
8.75
2.1
O/C
O/C
O/C
O/C
O/C
O/C
8
8
8
8
8
8
G
1
/CH
/CH
/CH
/CH
/CH
/CH
2
2
2
2
2
2
Cl
Cl
Cl
Cl
Cl
Cl
2
2
2
2
2
2
19.20
19.20
19.20
19.20
19.20
19.20
G
G
G
G
G
1
1
1
1
1
0.19
19.20
19.20
0.19
0.50 0.29 40.6
0.50 0.29 61.5
0.50 0.29
TBAHS
TBAHS
4.3
two-phase
D
D
D
D
D
D
D
D
D
D
D
2
2
2
2
2
2
2
2
2
2
2
O/decane
9.8
9.8
9.8
none
0.12
0.12
0.12
0.12
0.92
0.92
0.50
0.50
0.50
0.50
0.50
O/decane
O/decane
O/decane
O/cyclohexane
O/cyclohexane
[18]crown-6
TBAB
TBAHS
none
TBAHS
none
[18]crown-6 27.6
[18]crown-6
TBAHS
9.8
9.8
9.8
9.8
22.7
22.7
27.6
27.6
27.6
27.6
27.6
2.3
O/CH
O/CH
O/CH
O/CH
O/CH
2
2
2
2
2
Cl
Cl
Cl
Cl
Cl
2
2
2
2
2
0.1
6.1
0.7
28.6
2.0
0.28
27.6
0.28
TBAHS
An attempt to use a PTA to
spectively, in the system with dichloromethane. The integral
speed up the reaction in the decane/water two-phase system
gave little effect. This is not surprising, since it is known
that a solvent such as decane, which has a low dielectric con-
stant (e=1.99), is usually not useful in phase-transfer cataly-
sis. The results are in line with those of a previous work, in
which a nucleophilic substitution reaction was performed in
a dodecane/water system with added PTA and compared to
of the ÀCH OH peak was approximately 5% of the com-
2
bined integrals of the ÀCH Br and ÀCH I peaks; hence, the
2
2
hydrolysis affects the results only marginally. For all the re-
actions, except the one in dichloromethane with TBAHS as
phase-transfer agent, there was very good agreement be-
tween the integrals of the rising ÀCH I peak and the declin-
2
[20]
ing ÀCH Br peak. The reason why the side reaction, that is,
the reaction in a microemulsion.
The poor effect of a
2
hydrolysis, occurs when TBAHS and not [18]crown-6 is used
as PTA is still unclear. TBAHS did not increase the pH of
the water phase.
PTA in such a system is due to unfavourable extraction
1
equilibrium of the ion pair into the organic phase. H N MR
spectroscopy showed that almost no PTA was present in the
organic phase after extracting a solution of KI and TBAHS
or of tetrabutylammonium iodide (TBAI) in D O with
2
Discussion
decane.
Changing the solvent to cyclohexane, which has a slightly
higher dielectric constant (e=2.00), still gave no appreciable
effect of the added PTA. Also in this system, H NMR spec-
troscopy revealed that the extraction equilibrium was very
unfavourable.
PTC is known to function well in chlorinated hydrocar-
bons. Use of dichloromethane, with a dielectric constant of
8.90, as organic solvent resulted in a favourable extraction
equilibrium with a high concentration of PTA in the organic
phase. Another reason why aprotic and relatively polar liq-
uids, such as dichloromethane, are suitable solvents for
PTC-based reactions of the type studied here is that the nu-
cleophile will be poorly solvated and, thus, much higher in
nucleophilicity than when present in water or other protic
The solubility of 4-TBBB in the aqueous phase is very low.
The solubility of the iodide in the organic phase is also ex-
tremely low. One may, consequently, assume that the nucle-
ophilic substitution reaction only takes place at the inter-
face, that is, no reaction occurs in the bulk phases. The drop-
let size in a microemulsion is in the range of 10–50 nm,
while the droplet size in a typical (macro)emulsion is in the
micrometer range. The droplets obtained during vigorous
stirring of the two-phase systems used in this work are prob-
ably even larger, since the oil–water interfacial tension is
high in such surfactant-free systems. Consequently, the dif-
ference in total interfacial area between the two reaction
media is very large. It is, thus, not surprising that the reac-
1
5464
ꢁ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 5460 – 5466

Phase-Transfer Agents
5460 – 5466
solvents. As is shown in Figure 7, the reaction rate is faster
in the two-phase systems with added PTA (used in equimo-
lar amount) than in the microemulsion without PTA. Addi-
tion of a PTA to the microemulsion speeds up this reaction
to the same level as for the two-phase systems with PTA.
The observation that the reaction rates are the same in the
two-phase system and in the microemulsion when a PTA is
present (in equimolar amounts) indicates that the extraction
of the iodide ion into the organic phase is fast and efficient
and that the reaction in both systems mostly occurs in the
bulk organic phase or domain, not at the interface. Thus, the
rate-limiting step is the reaction as such, not the transport of
the iodide ion across the interface. The situation may be dif-
ferent for a faster organic reaction. The extraction may then
be the rate-limiting step and the much larger interface of
the microemulsion may be more important. The situation
should also be different in systems with lower extraction ef-
ficiency. In such systems the transport of the nucleophile
across the interface may be the rate-determining event and
the larger oil–water interface of the microemulsion could
then be taken advantage of. This reasoning nicely explains
results from a previous work in which the ring-opening of a
lipophilic epoxide with sodium hydrogen sulfite was studied
in a microemulsion and in a two-phase system. The reaction
in the microemulsion was much faster than the reaction in
the two-phase system when an equimolar amount of PTA
The phase behaviour of the system did not change during
the course of the reaction, neither for the case of equimolar
amount of PTA and nucleophilic reagent nor for the case of
catalytic amount of PTA. The mixture remained a clear and
low viscous one-phase medium, indicative of a microemul-
sion.
Conclusions
The present work shows that the reaction between 4-tert-bu-
tylbenzyl bromide and potassium iodide can be performed
at a reasonable rate in a microemulsion based on an aliphat-
ic hydrocarbon. PTC was not effective in such systems.
However, when the solvent was changed to a chlorinated
hydrocarbon, a PTA accelerated the reaction both in a two-
phase system and in a microemulsion. The combined ap-
proach of using a microemulsion as reaction medium and
addition of a catalytic amount of a PTA resulted in consider-
ably higher reaction rate than either of the approaches
alone. This could be of practical interest in preparative or-
ganic chemistry in which reagent incompatibility problems
often occur and in which PTC is frequently used as a way to
bring the reactants into contact.
[8]
was present in both systems. In that reaction system, the
extraction constant was considerably lower than in the pres-
ent system due to two contributing factors. Firstly, the sol-
vent was trichloroethylene, which is less polar than dichoro-
methane. Secondly the nucleophile was hydrogen sulfite,
which is a much more hydrophilic anion than iodide, which
Experimental Section
Materials: The nonionic surfactants used, penta(ethylene glycol)monodo-
decyl ether (C12
chased from Nikko Chemicals and Anatrace, respectively. The solvents,
decane, cyclohexane, dichloromethane (CH Cl ) and deuterochloroform
5 8 1
E ) and n-octyl-b-d-glucopyranoside (C G ), were pur-
2
2
+
À
means that Q HSO is less readily extracted into the or-
(CDCl ), were all from Aldrich. The reagents, potassium iodide (KI) and
4-tert-butylbenzylbromide (4-TBBB), were supplied by Merck and Al-
drich, respectively. Deuterium oxide (D O), tetrabutylammonium bro-
2
3
3
+
À
ganic phase than Q I .
In industrial applications of PTC equimolar amounts of
the PTA and the nucleophilic reagent is seldom used. Only
a catalytic amount of the PTA is usually employed, which
means that the Q salt or the crown ether is expected to go
back and forth across the interface to pick up and deliver
new nucleophiles into the organic phase. One may antici-
pate that the size of the interface will then be more crucial
than when the PTA is used in equimolar amounts. As can
be seen in Figure 8, the use of a catalytic, rather than equi-
molar, amount of either a Q salt or a crown ether slows
down the reaction rate both in the D O/C G /dichlorome-
mide (TBAB), tetrabutylammonium hydrogen sulfate (TBAHS) and tet-
rabutylammonium iodide (TBAI) were all supplied by Aldrich. The
crown ether, [18]crown-6, was from Lancaster. The water was Millipore
filtered. All chemicals used had a purity of ꢀ99%, except 4-TBBB
(
97%) and TBAHS (97%). All chemicals were used without any further
purification.
2 5 2 5
Phase diagram: Both the D O/C12E /decane and the D O/C12E /cyclohex-
ane systems were studied directly and through crossed polarisers in a
transparent water bath, in which the temperature was regulated by a
thermostat. The temperature was varied by one degree at a time and the
mixture was allowed to reach equilibrium before determining the self-as-
sembly structure formed. The microemulsion region was determined by
titration of D O into a stock solution containing surfactant and organic
2
solvent. The titrated amount was controlled by weight with an accuracy
of 0.0001 g.
2
8
1
thane microemulsion and in the D O/dichloromethane two-
2
phase system. The reduction is more pronounced in the two-
phase system. The combined approach, that is, use of a mi-
croemulsion as reaction medium and addition of a PTA,
gives the fastest reaction. It seems that when the PTA is
used in catalytic amount, transport of the nucleophile across
the interface becomes a crucial event. An interpretation of
the results is that in such a system part of the reaction takes
place at the interface, illustrated by the curve for the reac-
tion in the microemulsion in the absence of a PTA, and part
of the reaction occurs in the bulk organic phase, as evi-
denced by the increase in reaction rate when the PTA is
added to the microemulsion. In all the reactions, the crown
ether was slightly less effective than the Q salts.
Chemical reaction: The nucleophilic substitution reaction between 4-
TBBB and KI was performed at 238C in various microemulsions in an
NMR tube and in two-phase systems in bottles by using vigorous agita-
tion. The substrate and the nucleophile were always used in equimolar
concentration. Stock solutions of KI in D O and of 4-TBBB in oil were
2
prepared. The concentration of the PTA was the same as of the reactants,
if not otherwise mentioned.
Three different microemulsions were used. The first was based on D O,
2
5
decane and C12E . The concentration of 4-TBBB in decane was 64.3mm,
giving an overall concentration of 8.8mm. The microemulsion was made
by first mixing the oily stock solution with the surfactant and then adding
v
the aqueous stock solution. The volume fraction of the oil, f , based on
the oil and water volume, was 0.15. The surfactant weight fraction, a ,
m
based on total weight was 0.1.
Chem. Eur. J. 2004, 10, 5460 – 5466
www.chemeurj.org
ꢁ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
5465

M. Häger and K. Holmberg
FULL PAPER
2 5
The second microemulsion was D O/cyclohexane/C12E . The concentra-
[
[
[
1] F. M. Menger, J. U. Rhee, H. K. Rhee, J. Org. Chem. 1975, 40, 3803–
tion of 4-TBBB in cyclohexane was 24.7mm, giving an overall concentra-
tion of 21mm. The f value was 0.92 and the a value 0.11. The same for-
mulation procedure as described above was used.
The third microemulsion was D O/dichloromethane/C
tion of 4-TBBB in dichloromethane was 55.2mm, giving an overall con-
centration of 20mm. The f value was 0.5 and the a value 0.29. The
same formulation procedure as described above was used.
The reaction was also performed in two-phase systems based on D
either decane, cyclohexane or CH Cl with similar ratio between the oil
3
805.
2] F. M. Menger, A. R. Elrington, J. Am. Chem. Soc. 1991, 113, 9621–
624.
3] J. H. Ramsden, R. S. Drago, R. Riley, J. Am. Chem. Soc. 1989, 111,
958–3961.
V
m
9
2
8 1
G . The concentra-
3
V
m
[
[
[
4] R. Schomäcker, Nachr. Chem. Tech. Lab. 1992, 40, 1344–1351.
5] R. Schomäcker, J. Chem. Res. 1991, 92–93.
6] S. Gutfelt, J. Kizling, K. Holmberg, Colloids Surf. A. 1997, 128, 265–
2
O and
2
2
2
71.
phase and the aqueous phase as in the microemulsion based on the same
organic solvent. Furthermore, the concentrations of reactants and PTAs
[
7] M. Häger, K. Holmberg, in Adsorption and Aggregation of Surfac-
tants in Solution (Eds.: K. L. Mittal, D. O. Shah), Marcel Dekker,
New York, 2002, p. 327.
were the same as well. The reaction was quenched by addition of CDCl
in a volume ratio of 1:1 CDCl /reaction mixture after different time inter-
vals. The aqueous phase was removed after centrifugation at 3000 rpm
for 5 min and the organic phase was dried by Mg SO . After filtration the
3
3
[
[
8] M. Häger, K. Holmberg, Tetrahedron Lett. 2000, 41, 1245–1248.
9] G. Bode, M. Lade, R. Schomäcker, Chem. Eng. Technol. 2000, 23,
2
4
4
05–409.
10] C. Siswanto, T Battal, O. E. Schuss, J. F. Rathman, Langmuir 1997,
3, 6047–6052.
11] T. Battal, C. Siswanto, J. F. Rathman, Langmuir 1997, 13, 6053–
057.
12] M. Häger, K. Holmberg, U. Olsson, Langmuir, 2004, 20, 6107–
115.
13] M. Häger, F. Currie, K. Holmberg, Colloids Surf.,A in press.
14] U. Olsson, P. Schurtenberger, Langmuir 1993, 9, 3389–3394.
15] M. S. Leaver, U. Olsson, Langmuir 1994, 10, 3449–3454.
16] K. Fukuda, U. Olsson, U. Wꢂrz, Langmuir 1994, 10, 3222–3229.
17] U. Olsson, H. Wennerström, Adv. Colloid Interface Sci. 1994, 49,
1
organic phase was analysed by H NMR spectroscopy.
[
[
[
1
All reactions were analysed by H NMR spectroscopy, by using a
1
5
00 MHz Varian Inova spectrometer. The reaction rate was obtained by
following the consumption of 4-TBBB and the formation of the product.
The specific proton signals that were monitored were the protons of the
6
methylene groups at 4.42 (ÀCH
Br) and 4.39 ppm (ÀCH
2
I), recorded in
2
6
2 5
D O/C12E
/decane. In chlorinated hydrocarbons these values were shifted
[
[
[
[
[
slightly to higher ppm values. The reactions performed in the microemul-
sions were analysed in situ every five minutes, while the two-phase reac-
tions were analysed after the workup procedure. All reactions were
monitored during three hours, at which time the yield was calculated.
1
13–146.
18] U. Olsson, P. Schurtenberger, Prog. ColloidPolym. Sci. 1997, 104,
57–159.
19] H. Egger, T. Sottmann, R. Strey, A. Berkessel, Tenside, Surfactants,
Deterg. 2002, 39, 17–22.
[
[
1
Acknowledgement
We would like to thank the Swedish Foundation for Strategic Research
for financing M.H. through its Colloid and Interface Technology Pro-
gramme. M.H. also thanks the Institute for Surface Chemistry AB for fi-
nancial support. Furthermore, we thank the Swedish NMR Center for
granting spectrometer time and Charlotta Damberg at the NMR Center
for help with the instruments.
[20] K. Andersson, J. Kizling, K. Holmberg, S. Byström, Colloids Surf. A
1998, 144, 259–266.
Received: December 19, 2003
Revised: June 21, 2004
Published online: September 20, 2004
5466
ꢁ 2004 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 5460 – 5466