Dimethylmalonyltrialkylphosphoranes: Probing the steric effect on phosphorus and its stereochemical consequence in esterification reactions of chiral secondary alcohols

Article abstract of DOI:10.1016/j.jorganchem.2004.10.046

High chemical yields and high levels of stereochemical inversion are demonstrated in the phosphorane-mediated esterification reaction of chiral alcohols with non-hindered carboxylic acids through the incorporation of sterically non-hindered alkyl groups o

Full text of DOI:10.1016/j.jorganchem.2004.10.046

Journal of Organometallic Chemistry 690 (2005) 2548–2552
www.elsevier.com/locate/jorganchem
Dimethylmalonyltrialkylphosphoranes: probing the steric effect
on phosphorus and its stereochemical consequence in
esterification reactions of chiral secondary alcohols
a,
a
a
b
b
J. Dyck *, S. Zavorine , A.J. Robertson , A. Capretta , V. Larichev ,
b
J. Britten , J. McNulty
,b
*
a
Cytec Canada Inc., P.O. Box 240, Niagara Falls, Ont., CANADA L2E 6T4
b
Department of Chemistry, McMaster University, Hamilton, Ont., CANADA L8S 4M1
Received 1 July 2004; revised 20 October 2004; accepted 22 October 2004
Abstract
High chemical yields and high levels of stereochemical inversion are demonstrated in the phosphorane-mediated esterification
reaction of chiral alcohols with non-hindered carboxylic acids through the incorporation of sterically non-hindered alkyl groups
of phosphorus.
Ó 2004 Elsevier B.V. All rights reserved.
Keywords: Esterification; Trialkylphosphoranes; Stereocontrol
The direct condensation and displacement reactions
of alcohols with various nucleophiles can be successfully
carried out using the versatile Mitsunobu reaction [1–4].
The Mitsunobu reaction is especially powerful when the
inversion of stereochemistry of a chiral secondary alco-
hol is required. However, the triphenylphosphine oxide
and hydrazine side products are notoriously difficult to
remove from the product adding considerable time
and resources and resulting in lower isolated yields of
the substitution product. In addition to its value as a
general fragment coupling technique for both inter-
and intra-molecular bond formations, the reliable ste-
reocourse of the reaction proceeding with inversion
was until recently accepted as the norm. Indeed, many
structural assignments have been made with the assump-
tion of inversion occurring during a Mitsunobu reaction
which does appear to be so in the vast majority of cases.
However, in recent years several accounts reporting
retention of stereochemistry have been disclosed [5–8]
involving the condensation of sterically hindered sec-
ondary alcohols under standard Mitsunobu conditions.
A somewhat modified view of the reaction mechanism
has evolved outlined in Scheme 1.
The first step in the reaction mechanism involves the
nucleophilic addition of triphenylphosphine to the azo-
dicarboxylate followed by proton transfer from the car-
boxylic acid to give the carboxylate 1a and activated
intermediate 3 as shown in Scheme 1. Since the product
ester is by and large the inversion product, the reaction
was thought [3,4] to proceed directly to the alkoxyphos-
phonium ion 4 (Path 1) through alcohol attack on 3.
Nucleophilic displacement by the carboxylate anion on
the activated alkoxyphosphonium ion 4 (inversion path)
then delivers the inverted ester [2a].
In order to obtain the retention product, a different
pathway must also be available. In this case, it has
been proposed that the reaction occurs through an
*
Corresponding author. Tel.: +9053745940; fax: +9053745819.
E-mail address: jeff.dyck@cytec.com (J. Dyck).
0022-328X/$ - see front matter Ó 2004 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2004.10.046

J. Dyck et al. / Journal of Organometallic Chemistry 690 (2005) 2548–2552
2549
CO2Et
RCO2H
PPh3
N
N
EtO2C
1
(DEAD)
O
OPPh3
Inversion
O
O
R
R
RCO2
Ph3P
Path 1
R'
R"
1a
N
NH
CO2Et
3
R'
R'
R"
4
EtO2C
(-DEADH2)
"Cross-over"
(-DEADH2)
OH
R"
(-Ph3PO)
R'
O
2
CO2Et
NH
RCO2
O
OH
Path 2
N
Retention
PPh3
1a
R"
EtO2C
R
O
R'
R"
5
6
2
Scheme 1. Mechanism of the Mitsunobu esterification.
acyloxyphosphonium ion intermediate 5, described in
Path 2. This intermediate is generated by the more
nucleophilic carboxylate anion 1a attacking 3 rather
than the alcohol 2. It was postulated [3b,6] that there
may exist a ‘‘cross-over’’ reaction pathway mediated
by basic species present or generated during the reaction
such as the hydrazide anion 6. Such basic species would
be expected to promote the ‘‘cross-over’’ step resulting
in the formation of 4 leading to the inversion product.
This hypothesis was supported by the unambiguous
independent generation of acyloxyphosphonium ions
[9] and trapping with chiral secondary alcohols under
neutral or basic conditions where high levels of retention
or inversion of stereochemistry were observed
respectively.
We have been interested in the generation of new
phosphonium intermediates having the general structure
R3P+X (i.e., like 3), in which R is an alkyl or aryl sub-
stituent and X represents a good leaving group (as
XÀ) of controlled basicity. Other phosphorane reagents,
most significantly TsunodaÕs cyanomethylene reagents,
generate basic anions during the reaction and thus pro-
vide results similar to the original Mitsunobu class [10–
12] In contrast, our recently [6] reported phosphorane,
synthesized from tributylphosphine and chlorodimethyl-
malonate, produces an intermediate that expels the dim-
ethylmalonyl anion and is therefore significantly less
basic than 6. This tributylphosphorane reagent was
shown to promote the general condensation reaction
of a wide range of carboxylic acid and alcohol partners,
through generation of intermediates analogous to 5 un-
der conditions of controlled reaction basicity, providing
esters along with tributylphosphine oxide and dimethyl-
malonate [6]. In addition to the levels of stereocontrol
achievable (vide infra), the side products produced are
easily removed by aqueous base partition, eliminating
the purification difficulties inherent in the use of stan-
dard Mitsunobu reagents.
tion of a variety of carboxylic acids with chiral
secondary alcohols was carried out under a standard
set of conditions [6], some key observations are summa-
rized in Table 1. It is evident that electron deficient sub-
stituents not only increase the yield but also lead to high
levels of retention of stereochemistry in the final prod-
uct. The position of the substituent on the phenyl group
also affects the stereochemical outcome of the reaction,
with ortho groups increasing the level of inversion in
comparison to identical para substituents. Use of the
mechanistic pathway outlined in Scheme 1 proceeding
through the acyloxytri butyl phosphonium ion (analo-
gous to 5) and the base-induced ‘‘cross-over’’ allow a
rationalization of the overall substituent affects on the
stereochemical outcome (Table 1). Electron-rich carbox-
ylic acids are converted to their corresponding carboxyl-
ate ions which are expected to be stronger bases when
compared with electron-deficient acid. Therefore, elec-
tron-rich substituents would be expected to favour the
inversion product. In terms of the steric effect, sterically
hindered carboxylic acids render attack of the alcohol
on the acyl center of the acyloxyphosphonium ion more
difficult, allowing increased opportunity for ‘‘cross-
over’’ and the formation of the alkoxyphosphonium
ion leading to a relatively higher degree of inversion.
Lastly, with sterically hindered alcohols direct attack
at the acyl carbon atom is more likely lessening cross-
over to the alkoxyphosphonium ions, thereby favouring
the product of retention.
According to the revised mechanism, reaction inter-
mediates of type 5 may be considered ‘‘bifunctional’’ re-
agents that can allow alcohol or alkoxide attack at the
phosphorus atom, promoting crossover leading to the
normal inversion product, or may function as direct
acylating agents through alcohol attack at the carboxyl
carbon leading to the product of retention. As one of the
most important attributes of the Mitsunobu and related
redox condensation processes is the reliable high degree
stereocontrol achievable, particularly inversion of ste-
reochemistry, the preparation of analogs of these new
An extensive survey on the use of this new dimeth-
ylmalonyltributylphosphorane reagent on the condensa-

2550
J. Dyck et al. / Journal of Organometallic Chemistry 690 (2005) 2548–2552
Table 1
Chiral esterification reactions promoted by trinbutylphosphorane analogue
phosphoranes being more or less sterically congested
around phosphorus became of obvious interest. We
accordingly prepared the tri(iso)butyl and triethylphos-
phorane analogs. While both the butyl and (iso)butyl
analogs were isolated as viscous oils, the triethylphos-
phorane proved to be crystalline and as such a single
crystal X-ray structural determination was undertaken,
Fig. 1, confirming the structure.
To begin with, the reaction of 4-nitrobenzoic acid
with both (2S)-hexanol and (1R)-1-phenylpropanol
n
Fig. 1. X-ray structure (ORTEP) of dimethylmalonyltriethylphosphorane showing a distorted tetrahedral geometry about the phosphorus atom.

J. Dyck et al. / Journal of Organometallic Chemistry 690 (2005) 2548–2552
2551
Table 2
Chiral esterification reactions promoted by triiso and trinbutylphosphorane analogues
using both the n-butyl and (iso)butyl phosphoranes was
investigated. As can be seen from the results in Table 2,
more hindered substituents on phosphorus lead to a
higher degree of retention, with all other factors un-
changed. The lower yields reflect the overall slower rate
of the reaction with the more hindered phosphorane un-
der the standard reaction conditions. These stereochem-
ical results are fully in accord with our mechanistic
model in which attack at the phosphorus centre of inter-
mediate 5 becomes less favorable as steric congestion in-
creases on phosphorus leading to more direct acyl
transfer and a higher degree of retention of stereochem-
istry in the resulting ester. The condensation of a series
of carboxylic acid and chiral secondary alcohol partners
was then investigated with the less sterically congested
triethylphosphorane derivative.
Table 3
Chiral esterification reactions promoted by triethyl and trinbutylphosphorane analogues

2552
J. Dyck et al. / Journal of Organometallic Chemistry 690 (2005) 2548–2552
The reactions of this triethylphosphorane were inves-
chemical yields up to 80%. To achieve high levels of
retention, the use of the tributylphosphorane or
tri(iso)butylphosphorane in conjunction with an elec-
tron deficient carboxylic acid appears to be the optimal
combination. All of these results are in accord with the
modified view [2c] of the mechanism outlined in Scheme
1. Applications of these reagents towards the synthesis
of natural products are in progress and will be reported
in due course.
tigated side by side with the tributyl analog and the re-
sults summarized in Table 3. Entries 1 and 2 were
conducted under conditions expected to provide a high
degree of retention (electron deficient, non-hindered
acid). As can be seen, the triethylphosphorane provides
a higher degree of inversion even here. Most signifi-
cantly, entries 4–7 show that high degrees of inversion
of configuration can be obtained using the triethylphos-
phorane for the coupling of chiral secondary alcohols
and electron rich, sterically unencumbered acids (no
ortho-substituents), particularly with the use of 4-meth-
oxybenzoic acid. These results contrast to the use of the
tributylphosphorane analog, with which only the use of
electron-rich hindered ortho-substituted acids provided
such high inversions.
Appendix A. Supplementary data
Supplementary data associated with this article can
be found, in the online version at doi:10.1016/
j.jorganchem.2004.10.046.
Overall, these results clearly demonstrate that the ste-
ric bulk of the alkyl groups present on the phosphorus
atom of the phosphorane play a significant but predict-
able role on the stereochemical outcome of the reaction.
Less bulky groups favor inversion of stereochemistry
while bulky butyl groups favor retention of configura-
tion. These results reported for the uncongested triethyl-
phosphorane provide further structural evidence in full
agreement with the current mechanistic picture of the
reaction.
References
[1] O. Mitsunobu, M. Yamada, Bull. Chem. Soc. Jpn. 40 (1967)
2380.
[2] (a) For comprehensive reviews see: O. Mitsunobu, Synthesis
(1981) 1;
(b) D.L. Hughes, Org. React. 42 (1992) 335;
(c) R. Dembinski, Eur. J. Org. Chem. (2004) 2763.
[3] (a) D.L. Hughes, R.A. Reamer, J.J. Bergan, E.J.J. Grabowski, J.
Am. Chem. Soc. 110 (1988) 6487;
In conclusion, the new dimethylmalonyltrialkylphos-
phorane class of esterification reagent has been extended
and the steric effect of the substituent on phosphorus
probed. We feel that these reagents have much potential;
they are easy to prepare and can be handled in air for
short periods of time. They provide halogen, acid and
base free reaction media for esterification reactions
and thus tolerant of a variety of functional groups. They
have been successfully employed in the coupling of a
wide range of primary and secondary alcohols with var-
ious carboxylic acids [6]. The ester product can be puri-
fied with relative ease using as aqueous base-organic
partition sequence to remove the water and base soluble
side products. Lastly, a major advantage of this new
class of reagent is the controlled levels of inversion or
retention that can be achieved through careful choice
of reagents. To achieve high levels of inversion on a chi-
ral secondary alcohol, two optimum combinations of re-
agents are recommended. The use of 2,6-disubstituted
benzoic acids in conjunction with the tributylphospho-
rane results in a formation of the corresponding benzo-
ates with chemical yields up to 90% and high inversion.
Alternatively, the combination of p-methoxybenzoic
acid with the triethylphosphorane reagent provides the
corresponding inverted p-methoxybenzoates with high
(b) C. Ahn, R. Correia, P. DeShong, J. Org. Chem. 67 (2002)
1751.
[4] (a) For further mechanistic discussions see: M. Varasi, K.A.M.
Walker, M.L. Maddox, J. Org. Chem. 52 (1987) 4235;
(b) D.D. Camp, I.D. Jenkins, Aust. J. Chem. 41 (1988) 1835;
(c) D. Crich, H. Dyker, R.J. Harris, J. Org. Chem. 54 (1989) 257;
(d) D. Camp, I.D. Jenkins, J. Org. Chem. 54 (1989) 3045;
(e) J.A. Dodge, J.I. Trujillo, M. Presnell, J. Org. Chem. 59 (1994)
234;
(f) D.L. Hughes, Org. Prep. Proced. Int. 28 (1996) 127;
(g) W. Adam, N. Narita, Y. Nishizawa, J. Am. Chem. Soc. 106
(1984) 1843.
[5] C. Ahn, P.J. DeShong, Org. Chem. 67 (2002) 1754.
[6] J. McNulty, A. Capretta, V. Laritchev, J. Dyck, A. Robertson, J.
Org. Chem. 68 (2003) 1597.
[7] A.B. Smith III, I.G. Safonov, R.M. Corbett, J. Am. Chem. Soc.
124 (2002) 11102.
[8] X. Liu, Y. Wu, J.K. De Brabander, Angew. Chem., Int. Ed. 42
(2003) 1648.
[9] J. McNulty, A. Capretta, V. Laritchev, J. Dyck, A. Robertson,
Angew. Chem., Int. Ed. 42 (2003) 4051.
[10] (a) T. Tsunoda, F. Ozaki, S. Ito, Tetrahedron Lett. 35 (1994)
5081;
(b) K. Uemoto, A. Kawahito, N. Matsushita, I. Sakamoto, H.
Kaku, T. Tsunoda, Tetrahedron Lett. 42 (2001) 905.
[11] (a) F. Zaragoza, H. Stephensen, J. Org. Chem. 66 (2001) 2518;
(b) F. Zaragoza, J. Org. Chem. 67 (2002) 4963.
[12] S. Ito, T. Tsunoda, Pure Appl. Chem. 71 (1999) 1053.