Unusual temperature dependence of salt effects for "on water" Wittig reaction: Hydrophobicity at the interface

Article abstract of DOI:10.1039/b809127g

An unusual variation with temperature of the salt effects in aqueous Wittig reaction is observed, suggesting that hydrophobic acceleration of reactions comprising "on water" reactants is fundamentally different from that for reactions with small non-polar solutes. The Royal Society of Chemistry.

Full text of DOI:10.1039/b809127g

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Unusual temperature dependence of salt effects for ‘‘on water’’ Wittig
reaction: hydrophobicity at the interfacew
Shraeddha Tiwari and Anil Kumar*
Received (in Cambridge, UK) 29th May 2008, Accepted 16th June 2008
First published as an Advance Article on the web 28th July 2008
DOI: 10.1039/b809127g
An unusual variation with temperature of the salt effects in
aqueous Wittig reaction is observed, suggesting that hydropho-
bic acceleration of reactions comprising ‘‘on water’’ reactants
is fundamentally different from that for reactions with small
non-polar solutes.
Water is known as ‘‘Nature’s solvent’’. The special effect of
water in enhancing the selectivities and yields of organic
Scheme 1 Wittig reaction of aldehydes 1a–d with stabilized phos-
phorus ylide 2.
reactions, and of Diels–Alder reactions in particular, has been
appreciated only after the pioneering contribution of Rideout
and Breslow.1 Different forces such as hydrophobic packing,2
difficult to carry out the kinetic studies with good accuracy in
refluxing water, an optimum temperature of 338 K was chosen
hydrogen bonding,3 enforced hydrophobic hydration4 etc. are
believed to be responsible for the origin of the special effect of
in addition to room temperature (298 K).z
At 338 K, the addition of 1 M LiCl and 1 M NaCl to the
water during organic and biochemical transformations. The
reaction medium led to an observable increase in the apparent
‘‘on water’’ method employed by Sharpless and co-workers
rate of reaction of 1a with 2 (Table 1). The presence of an
antihydrophobic agent such as GnCl lowered the rate to
recently has indicated the importance of interfacial processes
in determining the progress of aqueous reactions.5 A thorough
nearly half of the magnitude in water. In contrast, at 298 K,
understanding of the phenomenon of hydrophobicity at the
the use of prohydrophobic additives—LiCl and NaCl de-
molecular level, however is still elusive.
creased the rate of the reaction. This observation was contra-
Recently, ‘‘on water’’ Wittig reactions of stabilized and
dictory to all the previous reports about the rate-enhancing
effect of salts such as LiCl and NaCl.3c
semi-stabilized ylides in aqueous medium, leading to enhanced
rates and yields, have been reported.6 The use of 1.2 M
aqueous LiCl further enhances the yields. The increase in rates
on addition of LiCl is believed to be an evidence of the
dominance of hydrophobic effects against other solvent
effects.3c In the past, salt effects have also been employed as
mechanistic tools for homogeneous aqueous reactions. Herein
we highlight an unusual temperature-dependent effect of
prohydrophobic additives (‘‘salting-out’’ agents) and antihy-
drophobic additives (‘‘salting-in’’ agents)3d on the rates of
‘‘on water’’ Wittig reactions. The observations are explained
in terms of an interfacial process and are helpful in gaining an
insight into the complex phenomenon of hydrophobicity in
‘‘on water’’ reactions.
In order to confirm that the retarding effect of additives
was general, the kinetic studies were repeated for different
substrates—4-hydroxybenzaldehyde (1b), furfural (1c) and
butyraldehyde (1d) under identical conditions. Decrease in
rates is observed for all the three aldehyde substrates on
addition of prohydrophobic salts at 298 K. The results in-
dicate that the effect is not substrate-specific. The magnitude
of the relative rates depends on the concentration of the salt
Table 1 Relative rates krel (with respect to water) for the aqueous
Wittig reaction of 1a–d with 2 in different reaction media at 298 Kab
c
krel
The Wittig reactions of benzaldehyde (1a) and (ethoxy-
methylene)triphenylphosphorane (2) were carried out in water
and different aqueous salt solutions (Scheme 1). Two prohy-
drophobic salts LiCl and NaCl were chosen in addition to an
antihydrophobic salt guanidinium chloride, GnCl. The pre-
vious reports had mentioned refluxing conditions for carrying
out the Wittig reactions in 1.2 M aqueous LiCl.6 Since it was
Entry
Aldehyde T/K
1 M LiCl 1 M NaCl
1 M GnCl
1
2
3
4
5
a
1ad
1a
1b
1c
338
298
298
298
298
1.12
0.86
0.64
0.63
0.70
1.27
0.74
0.52
0.47
0.61
0.55
0.95
0.98
1.19
0.99
1d
Reactions carried out with 1 mmol aldehyde and 5 mmol of 2 in 10 mL
b
of reaction medium. The isolated yields (and E : Z ratios) at 298 K
after a reaction time of 3 h are 92% (93 : 7) in water, 89% (91 : 9) in aq.
LiCl, 89% (91 : 9) in aq. NaCl, 71% (82 : 18) in aq. GnCl and 84%
Physical Chemistry Division, National Chemical Laboratory, Pune,
India 411008. E-mail: a.kumar@ncl.res.in; Fax: +9120 2590 2636;
Tel: +9120 2590 2278
w Electronic supplementary information (ESI) available: Detailed
procedure, kinetic analysis, representative plots for rate determina-
tion. See DOI: 10.1039/b809127g
c
(83 : 17) in aq. LiClO4. krel = kobs/(kobs water
)
. The rate constants
d
agreed within an experimental error of ꢀ3%. The rates agreed
within an error of ꢀ6%.
ꢁc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 4445–4447 | 4445

View Article Online
Fig. 1 Relative rates for the Wittig reaction of 1a with 2 in (.) 1 M
NaCl at 338 K, (K) 1 M LiCl at 338 K, (J) 1 M LiCl at 298 K and
(,) 1 M NaCl solution at 298 K.
Fig. 3 Relative absorbance (an indicator for the extent of conversion)
against time plotted for the reaction of 1a with 2 at 298 K under (J)
homogeneous conditions and (’) heterogeneous conditions.
added. The effect of addition of LiCl and NaCl on the rates is
exactly opposite at the two temperatures: 298 K and 338 K
(Fig. 1). This contrasting effect of salt additives on the rates in
water and aqueous salt solutions cannot be explained by an
extension of current notion of ‘‘salting-out’’ and ‘‘salting-in’’
behaviour.1a,7 Since the rate acceleration on addition of LiCl
or NaCl is attributed to the enhancement of hydrophobic
effects, the contrasting results at 298 K underline the need for
further investigation of the role of hydrophobicity as a driving
force for the reaction.
evidence for the predominantly interfacial mechanism of the
reaction. This is a characteristic feature of ‘‘on water’’ reac-
tions.5 In such cases, the heterogeneous conditions are actually
found to ‘‘catalyze’’ the reactions by a notable magnitude. In
a recent report, Jung and Marcus have proposed that the
H-bonding between the substrate and the free OH groups of
the water molecules at the interface might be the basis for the
superior rates observed for ‘‘on water’’ reactions.8 The drastic
variations in the rates for aqueous interfacial reactions can be
explained on the basis of the difference in the spatial arrange-
ment of water molecules around a hydrophobically hydrated
solute and that around an extended macroscopic interface.9
According to the Lum–Chandler–Weeks (LCW) theory, the
hydration of a mesoscopic surface (radius r 4 100 A) involves
a ‘‘dewetting’’ of the surface.10 The local density of water at
such an interface is less than the bulk density of water. In fact,
the length-scale dependence of hydrophobic hydration is a
widely accepted physical phenomenon, which has been pre-
dicted theoretically11 and proved experimentally.12
The temperature dependent studies of the rates were carried
out in water and 3 M aqueous LiCl for the reaction of 1a with 2.
The rate enhancing effect of LiCl gradually appears to become
smaller in magnitude and at a temperature of B318 K, the rates
are similar in both the media. LiCl then acts as a retarding
additive if the temperature is decreased further (Fig. 2).
Since the ylide particles are not completely soluble in water or
the aqueous mixtures, the reaction can take place either at the
interface i.e. it takes place at the surface of the ylide particle or
in the bulk solution. In order to understand this aspect, the rates
were compared for an exclusively homogeneous system and an
identical but heterogeneous reaction mixture. The concentra-
tion of 1a did not change much after 40 min for the homo-
geneous reaction (i.e. when the ylide 2 was dissolved by
overnight stirring) at 298 K (Fig. 3). Under identical but
heterogeneous conditions, more than 25% of the 1a had under-
gone conversion to the product within 35–40 min. The rates
were determined for ylide samples having different particle size.
(Table 2) A smaller particle size leads to a greater surface area
exposed to the reaction medium, and hence, is reflected in the
higher apparent rate constants. The observations are a strong
The medium effects for ‘‘on water’’ reactions could be
explained on the basis of such size-dependent hydration
thermodynamics. The surface of the solid ylide particles,
which is the ‘‘site’’ of the interfacial Wittig reactions, can be
considered as a macroscopic surface. Thus the orientation of
water molecules around the ylide particles will be very differ-
ent from the structured ‘‘iceberg’’-like model used for the
study of hydrophobicity of small non-polar solute molecules.13
Addition of salts such as NaCl that are known to enhance the
structure of water (structure-making or kosmotropes)14 should
lead to an increase in the cost of hydration of the ylide particle,
Table 2 Apparent rate constants (kobs) for Wittig reaction of 1a with
2a samples having different particle size at 298 K
Particle diametera/mm
104kobsb/sꢂ1
162.7
131.7
70.5
64.8
39.4
a
3.47
5.28
5.37
5.51
5.95
Samples of different particle size prepared by using a rolling ball mill
b
and their diameter determined using a particle size analyzer. Reac-
tions carried out with 1 mmol of 1a and 5 mmol of 2 in 10 mL of
reaction medium.
Fig. 2 Eyring plots for the Wittig reaction of 1a with 2a in water (K)
and 3 M LiCl (’).
ꢁc
This journal is The Royal Society of Chemistry 2008
4446 | Chem. Commun., 2008, 4445–4447

View Article Online
as reflected by the greater free energy of cavitation for the
‘‘more structured’’ salt solutions.15
prohydrophobic and antihydrophobic salts is capable of in-
fluencing hydrophobicity at the interface.
This may cause a further depletion of the water density
around the surface, or in other words, a greater extent of
‘‘dewetting’’. The addition of NaCl is reported to increase the
free energy of hydration per unit surface area, DGhydration/A (A
is the surface area).12b On the basis of SFM experiments
measuring hydrophobic adhesion forces, Kurutz and Xu
stated that addition of NaCl had no effect on the hydrophobic
force experienced by the ‘‘supramolecular’’ surface, in contrast
to its solution-phase effect on solubility of small solute
molecules.16
Notes and references
z Brief experimental procedure for kinetic analysis: For a standard
kinetic run, the 1 mM aldehyde solution (1 mmol in 10 mL) was
allowed to equilibrate at the desired temperature. The temperature was
controlled using a constant temperature bath with an accuracy of
ꢀ 0.01 K. The reaction was initiated by addition of 2 (5 mmol in
10 mL) into the above aldehyde solution. The reaction progress was
monitored by following the decrease of the aldehyde concentration
using UV spectrophotometry. (See ESIw for details of analytical
method used). The reaction mixture was heterogeneous in nature
and vigorous stirring was required to minimise aggregation. The
pseudo first-order rate constant thus obtained is, in fact, an apparent
rate constant due to the heterogeneity of the medium. The rate
constants were reproducible to within ꢀ 3% at 298 K and ꢀ 6% at
338 K.
Applying the same logic to the ‘‘on water’’ reaction kinetics,
the addition of salts such as NaCl and LiCl might lead to a
lesser number of water molecules at the interface available for
H-bonding. The resultant change in the extent of interaction
(wider ‘‘dewetted’’ region or fewer water molecules at the
interface) may lead to a weaker ‘‘on water’’ effect, provided
the reaction is indeed accelerated by the highly specific inter-
facial arrangement of water molecules. The fewer number of
water molecules at the interface will translate into a lower
extent of ‘‘catalysis’’ by dangling –OH groups and hence, a net
slowing down of the reaction at 298 K.
1 D. C. Rideout and R. Breslow, J. Am. Chem. Soc., 1980, 102, 7816.
2 R. Breslow and U. Maitra, Tetrahedron Lett., 1983, 24, 1901; R.
Breslow, Acc. Chem. Res., 1991, 24, 159 and references cited
therein; for ‘‘prohydrophobic’’ and ‘‘antihydrophobic’’ terminol-
ogy, see: R. Breslow and R. V. Connors, J. Am. Chem. Soc., 1995,
117, 6601; R. Breslow and T. Guo, J. Am. Chem. Soc., 1995, 117,
9923.
At higher temperature, water is intrinsically ‘‘less struc-
tured’’ i.e. at 338 K as compared to 298 K. The interfacial
arrangement of water molecules is greatly disturbed at such
temperatures. Consequently, although salts such as NaCl or
LiCl will still continue to perturb the interfacial structure but
to a weaker extent. In this case, the effect of additives on the
rates can possibly be dominated by other physicochemical
processes. For example, at higher temperatures, it is probable
that the salts exert their effect through controlling the solubi-
lity equilibrium of the ylide. The addition of LiCl or NaCl
should increase the proportion of the undissolved ylide—
favouring the faster ‘‘on water’’ reaction as compared to the
sluggish homogeneous reaction of the dissolved ylide. Pre-
liminary solubility studies in our laboratory have indicated
that presence of LiCl and NaCl lead to a decrease in the
solubility of the ylide as compared to that in water alone, in
accordance with the conventional picture. This means that the
addition of NaCl, LiCl should ‘‘salt-out’’ the ylide. However a
thorough kinetic and solubility analysis is required for any
conclusive explanation of the phenomenon. The effect of
antihydrophobic additives also needs a closer examination.
In conclusion, the results show that the presence (or ab-
sence) of rate acceleration on addition of prohydrophobic salts
at any temperature need not be a conclusive evidence for the
predominance of (or lack of) ‘‘hydrophobic effect’’, at least for
heterogeneous aqueous reactions. The salting effects on the
kinetics of ‘‘on water’’ reactions differ significantly from those
observed for homogeneous aqueous reactions. The length-
scale dependence of hydrophobicity must be considered before
interpreting any salting phenomenon. This observation is
relevant to numerous chemical processes which are known
to take place at a ‘hydrophobic’ interface. The results give an
indication of the complex manner in which the presence of
3 J. F. Blake and W. L. Jorgenson, J. Am. Chem. Soc., 1991, 113,
7430.
4 W. Blokzijl, J. B. F. N. Engberts and M. J. Blandamer, J. Am.
Chem. Soc., 1990, 112, 1197; W. Blokzijl and J. B. F. N. Engberts,
J. Am. Chem. Soc., 1991, 114, 5440; W. Blokzijl and J. B. F. N.
Engberts, Angew. Chem., Int. Ed. Engl., 1993, 32, 1545; J. W.
Wijnen and J. B. F. N. Engberts, J. Org. Chem., 1997, 62, 2039; A.
Maijer, S. Otto and J. B. F. N. Engberts, J. Org. Chem., 1998, 63,
8989; S. Otto, J. B. F. N. Engberts and J. C. T. Kwak, J. Am.
Chem. Soc., 1998, 120, 9517; S. Otto and J. B. F. N. Engberts, Pure
Appl. Chem., 2000, 72, 1365; J. B. F. N. Engberts and M. J.
Blandamer, Chem. Commun., 2001, 1701; T. Rispens and
J. B. F. N. Engberts, J. Org. Chem., 2002, 67, 7369.
5 S. Narayan, J. Muldoon, M. G. Finn, V. V. Fokin, H. C. Kolb and
K. B. Sharpless, Angew. Chem., Int. Ed., 2005, 44, 3275.
6 J. Dambacher, W. Zhao, A. El-Batta, R. Anness, C. Jiang and M.
Bergdahl, Tetrahedron Lett., 2005, 46, 4473; J. Wu, D. Zhang and
S. Wei, Synth. Commun., 2005, 35, 1213; J. Wu and D. Zhang,
Synth. Commun., 2005, 35, 2543; A. El-Batta, C. Jiang, W. Zhao,
R. Anness, A. L. Cooksy and M. Bergdahl, J. Org. Chem., 2007,
72, 5244.
7 W. F. McDevit and F. A. Long, J. Am. Chem. Soc., 1952, 74, 1773;
M. R. J. Dack, Chem. Soc. Rev., 1975, 4, 211; R. Breslow and
T. Guo, Proc. Natl. Acad. Sci. USA, 1990, 87, 167.
8 Y. Jung and R. A. Marcus, J. Am. Chem. Soc., 2007, 129, 5492.
9 F. H. Stillinger, J. Solution Chem., 1973, 2, 141.
10 K. Lum, D. Chandler and J. D. Weeks, J. Phys. Chem. B, 1999,
103, 4570; D. M. Huang and D. Chandler, Proc. Natl. Acad. Sci.
USA, 2000, 97, 8324.
11 D. Chandler, Nature, 2005, 437, 640; S. Rajamani, T. M. Truskett
and S. Garde, Proc. Natl. Acad. Sci. USA, 2005, 102, 9475; G.
Graziano, J. Phys. Chem. B, 2006, 110, 11421.
12 A. Poynor, Phys. Rev. Lett., 2006, 97, 26601; M. Mezger, H.
Reichert, S. Schroder, J. Okasinski, H. Schroder, H. Dosch, D.
Palms, J. Ralston and V. Honkimaki, Proc. Natl. Acad. Sci. USA,
2006, 103, 18401.
13 H. S. Franks and M. W. Evans, J. Chem. Phys., 1945, 13, 507.
14 R. Leberman and A. K. Soper, Nature, 1995, 378, 364.
15 S. S. Pawar, U. Phalgune and A. Kumar, J. Org. Chem., 1999, 64,
7055.
16 J. W. Kurutz and S. Xu, Langmuir, 2001, 17, 7323.
ꢁc
This journal is The Royal Society of Chemistry 2008
Chem. Commun., 2008, 4445–4447 | 4447