Unexpected Barbier-Grignard allylation of aldehydes with magnesium in water [17]

Full text of DOI:10.1021/ja981020s

9102
J. Am. Chem. Soc. 1998, 120, 9102-9103
Grignard allylation of aldehydes (1) with magnesium and allyl
halides (2) proceeds smoothly in water (eq 1).
Unexpected Barbier-Grignard Allylation of
Aldehydes with Magnesium in Water
Chao-Jun Li* and Wen-Chun Zhang
Department of Chemistry, Tulane UniVersity
New Orleans, Louisiana 70118
ReceiVed March 26, 1998
To start our investigation, we reacted allyl bromide with
benzaldehyde and magnesium turnings in 0.1 N aqueous HCl for
3 h at room temperature. TLC analysis of the ether extract clearly
showed a spot that corresponds to the desired allylation product.
Subsequently, 1H NMR measurement of the crude reaction
mixture showed about 28% of the allylation product (3), together
with 66% of the pinacol coupling product (4),11 and 6% benzyl
alcohol. This promising result prompted us to examine factors
that influence the reaction. We then examined in greater detail
the effect of the solvent system on the magnesium reaction by
using various combinations of water and THF as the reaction
solvent together with a small amount of iodine to initiate the
reaction. Workup of the reactions involved extraction with diethyl
ether, drying over magnesium sulfate, and careful removal of the
low boiling solvent in vacuo. In the case where no organic
cosolvent was involved, the reaction mixture was extracted with
CDCl3 and the extract was examined directly. 1H NMR spectra
of the reaction products under various solvent combinations
revealed a very interesting phenomenon. The results are listed
in Table 1. In freshly distilled THF dried over sodium/
benzophenone, the reaction between benzaldehye, allyl bromide,
and magnesium turnings together with a small amount of iodine
occurred almost quantitatively, generating the expected allylation
product. The addition of a small amount of water to the freshly
dried THF did not affect the progress of the reaction. The
expected reaction proceeded effectively until about 7% of water
in THF was used, which suddenly blocked the reaction progress.
After repeating several times, we identified a midpoint (entry 4)
at which ca. 56% conversion of the starting benzaldehyde was
oberved. However, when the composition of the solvent is
changed to water alone, a smooth reaction started again, generating
the allylation product albeit with a low conversion. The low
conversion could be attributed to the formation of magnesium
hydroxide on the metal surface which blocks further reactions.
Additional reactions using 0.1 N HCl or NH4Cl solutions as the
reaction solvent gave a quantitative conversion, generating a
mixture of the allylation and pinacol coupling products. Other
aromatic aldehydes gave similar results; whereas reactions of
aliphatic aldehydes gave more complicated mixtures. The use
of 0.1 N NH4Cl aqueous solution as the solvent was found to be
superior to the use of 0.1 N HCl. The use of a catalytic amount
of InCl3 did not affect the reaction.12 Changing the allyl bromide
to allyl iodide further increased the formation of the allylation
product (58%). Only a minute amount of the allylation product
was observed with allyl chloride. In the absence of HCl, NH4-
Cl, or iodine, virtually no reaction was observed with allyl
bromide. Experiments revealed that, with iodine, the reaction of
allyl bromide proceeded through an allyl iodide intermediate. The
formation of such an allyl iodide intermediate, however, is not
necessary in 0.1 N NH4Cl or 0.1 N HCl.
An important step in the history of modern chemistry was the
introduction of magnesium for carbon-carbon bond formations1
by Barbier and Grignard about a century ago,2 through the addition
of an organometallic reagent to a carbonyl group. The study of
magnesium-based reactions since then has sparked the develop-
ment of new reagents based on electronically more negative and
more positive metals as well as semi-metallic elements for various
synthetic purposes to tailor reactivities and selectivities (chemo,
regio, and stereo).3 For carbonyl additions based on organomag-
nesium reagents, it is generally accepted that strict anhydrous
reaction conditions are required for a smooth reaction.4 The
presence of moisture inhibits the reaction. Various methods, such
as using dibromoethane5 or iodine initiators, mechanical activa-
tion6 and ultrasonic irradiations,7 have been developed to help
initiate the reaction. More recently activated magnesium has also
been developed.8 Because of economical and environmental
concerns, the use of water as a solvent for metal-mediated
carbon-carbon bond formations has generated considerable
interests. Within the last several years, various metals have been
developed to mediate Barbier-Grignard-type reactions.9 For a
long time, we have been intrigued by the possibility of performing
classical Barbier-Grignard reactions by using magnesium in
water. The study would possibly extend the scope of aqueous
metal reactions as well as increase the understanding of the
mechanism of the classical Barbier-Grignard reaction. However,
in view of the high reactivity of organomagnesium reagents
toward water, it is doubtful that magnesium could be used for
such a purpose. Nevertheless, this question has constantly haunted
our minds and led us to test the magnesium-mediated allylation
reaction of benzaldehyde in water. Historically, the allylation of
carbonyl compounds with allylmagnesium reagents had not been
well-established for decades until Gilman and McGlumphy
developed a new procedure to prepare Grignard reagents from
allyl halides.10 Herein we report the observation that Barbier-
(1) Kharasch, M. S.; Reinmuth, O. Grignard Reactions of Nonmetallic
Substances, Prentice-Hall: New York, 1954.
(2) Barbier, P. Comptes. Rendus 1898, 128, 110. Barbier, P. J. Chem. Soc.
1899, 76, Pt. 1, 323; Grignard, V. Comptes. Rendus 1900, 130, 1322.
(3) For representative monographs and reviews, see: Wakefield, B. J.
Organomagnesium Methods in Organic Chemistry, Academic Press: New
York, 1995. Blomberg, C. The Barbier Reaction and Related One-Step
Processes; Springer-Verlag: New York, 1993. Lai, Y. H. Synthesis 1981,
585; Courtois, G.; Miginiac, L. J. Organomet. Chem. 1974, 69, 1. Normant,
H. AdV. Org. Chem., Methods Results 1960, 2, 1. Ioffe, S. T.; Nesmeyanov,
A. N. The Organic Compounds of Magnesium, Beryllium, Calcium, Strontium
and Barium; North-Holland: Amsterdam, 1976.
(4) Wakefield, B. J. Organomagnesium Methods in Organic Synthesis, Best
Synthetic Methods series, Katritzky, A. R., Meth-Cohn, O., Rees, C. W., Eds.;
Academic Press: San Diego, 1995.
(5) Pearson, D. E.; Cowan, D.; Beckler, J. D. J. Org. Chem. 1959, 24,
504.
(6) Shaw, M. C. J. Appl. Mechanics 1948, 15, 37.
(7) Sprich, J. D.; Lewandos, G. S. Inorg. Chim. Acta 1983, 76, L241.
(8) For reviews, see: Rieke, R. D. Science 1989, 246, 1260. Bogdanovic,
B. Acc. Chem. Res. 1988, 21, 261. For a review on graphite-metal compounds,
see: Csuk, R.; Glanzer, I.; Furstner, A. AdV. Organomet. Chem. 1988, 28,
85.
To explain the unusual phenomenon of the solvent change, we
postulate that in freshly dried THF, the normal reaction occurs
between the organo halide and magnesium generating the
(9) For reviews, see: Li, C. J.; Chan, T. H. Organic Reactions in Aqueous
Media; John Wiley & Sons: New York, 1997. Li, C. J., Tetrahedron 1996,
52, 5643. Chan, T. H.; Isaac, M. B. Pure Appl. Chem. 1996, 68, 919. Lubineau,
A.; Auge, J.; Queneau, Y. Synthesis 1994, 741. Li, C. J. Chem. ReV. 1993,
93, 2023.
(10) Gilman, H.; McGlumphy, J. H. Bull. Soc. Chim. Fr. 1928, 43, 1322.
(11) For a recent review on pinacol coupling, see: Wirth, T. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 61-63.
(12) For representing examples of using InCl3 in Barbier-type reactions,
see: Li, X. R.; Loh, T. P. Tetrahedron: Asymmetry 1996, 7, 1535.
S0002-7863(98)01020-8 CCC: $15.00 © 1998 American Chemical Society
Published on Web 08/20/1998

Communications to the Editor
J. Am. Chem. Soc., Vol. 120, No. 35, 1998 9103
Table 1. Magnesium-Mediated Barbier-Grignard Allylation of Benzaldehyde in Various Solventsa
entry
allyl halide
solvent
mediator
3/4/benzyl alcohol
conversion (%) of aldehyde
1
2
3
4
5
6
7
8
9
allyl bromide
allyl bromide
allyl bromide
allyl bromide
allyl bromide
allyl bromide
allyl bromide
allyl bromide
allyl iodide
THF/H2O (10:0.04)
THF/H2O (10:0.2)
THF/H2O (10:0.55)
THF/H2O (10:0.67)
THF/H2O (10:0.7)
H2O
0.1 N HCl
0.1 N NH4Cl
0.1 N NH4Cl
Mg/I2
Mg/I2
Mg/I2
Mg/I2
Mg/I2
Mg/I2
Mg/I2
Mg
100%/0/0
100%/0/0
98%/0/0
9%/43%/4%
0/0/0
13%/2%/1%
28%/66%/6%
41%/52%/7%
58%/34%/8%
quantitative
quantitative
98
56
0
16
quantitative
quantitative
quantitative
Mg
a Entries 1-6: benzaldehyde:allyl halide:magnesium ) 1:3:20. Entries 7-9: benzaldehyde:allyl halide:magnesium ) 1:3:10 (the use of a
stoichiometric amount of magnesium resulted in a low conversion).
Scheme 1. Postulated rationale for magnesium mediated
carbon-carbon bond formation in water
magnesium reagent. Subsequent reaction with carbonyl com-
pounds generates the allylation product. However, in a homo-
geneous mixture of THF and water, the water molecule would
be evenly distributed throughout the solution; hence the contact
between water and the metal surface would prohibit the formation
of the organomagnesium reagent. The solvent system with ca.
7% water may correspond to the point where the surface is
completely covered by water molecules. With water alone, allyl
bromide might be squeezed on to the metal surface due to
hydrophobic effect;13 locally and microscopically, magnesium
would not be in direct contact with water. The formation of
allylmagnesium can thus occur as in dry organic solvents, which
leads to the generation of allylation products.
The mechanisms of magnesium mediated Barbier and Grignard
reactions have been studied intensively by several groups.14,15 It
is generally believed that radicals on the metal surface are
involved in the organomagnesium reagent formation reaction.14
There is still no conclusion as to the freedom of these free
radicals.15 To gain mechanistic information of the present
reaction, we quenched the allylation process in 0.1 N aqueous
NH4Cl. The reaction mixture was extracted with CDCl3 and the
1H NMR spectrum was taken directly. The measurement shows
the formation of the three products (allylation, pinacol, and benzyl
alcohol) reflecting our previous experiments. In addition, 1,5-
hexadiene was clearly formed. However, no propene formation
was detected with NMR.16 To explain the experimental observa-
tions, we tentatively proposed Scheme 1. The transfer of electrons
from magnesium to the allyl halide (path a), the aldehyde (path
b), or both generates the corresponding radical anions 5 and 6.
Then, reaction of either 5 with the aldehyde or 6 with the allyl
halide generates the corresponding allylation product 3. Reaction
of 5 with the allyl halide gives the Wurtz-coupling product 8;
reaction of 6 with the aldehyde gives pinacol product 4. The
reaction of these radical anions with water led to the corresponding
reduced products 7 and 9. Subsequent experiments show that in
the absence of the allyl halide, the corresponding pinacol coupling
product 4 was obtained.
by Alexander.17 These two forms will also lead to either the
protonation of the carbanion (overall reduction of the halide) or
Wurtz-type coupling;18 whereas the intermediate reacts with
aldehydes through the usual six-membered-ring mechanism as
proposed by Young and Roberts19 (represented by the symbol
SE2′).20 The radical intermediate leads to the formation of 1,6-
hexadiene, pinacol product, and benzyl alcohol.21 A detailed
exploration of the origin and synthetic utilities of this magnesium-
mediated reaction are under investigation.
Acknowledgment. We are grateful to the support provided by the
NSF-EPA joint program for a Sustainable Environment. Acknowledgment
is also made to the donors of the Petroleum Research Fund (administered
by the American Chemical Society), LEQSF, and the NSF CAREER
Award program for partial support of this research. We also thank Prof.
H. Ensley for discussions and the anonymous reviewers for their
suggestions.
Supporting Information Available: Experimental procedures for the
magnesium mediated allylation of benzaldehyde in water (1 page, print/
PDF). See any current masthead page for ordering information and Web
access instructions.
Alternatively, the reaction of allyl bromide on the metal surface
generates an organometallic intermediate that is in an equilibrium
with the charge-separated form and the radical form as proposed
JA981020S
(13) Tanford, C. The Hydrophobic Effect: Formation of Micelles and
Biological Membranes, 2nd ed.; John Wiley & Sons: New York, 1980.
Breslow, R. Acc. Chem. Res. 1991, 24, 159.
(14) Gomberg, M.; Bachmann, W. E. J. Am. Chem. Soc. 1927, 49, 236.
(15) Lawrence, L. M.; Whitesides, G. M. J. Am. Chem. Soc. 1980, 102,
2493. Walborsky, H. M.; Zimmermann, C. J. Am. Chem. Soc. 1992, 114,
4996. Walling, C.; Acc. Chem. Res. 1991, 24, 255. Garst, J. F. Acc. Chem.
Res. 1991, 24, 95. Walborsky, H. M. Acc. Chem. Res. 1990, 23, 286.
(16) Subsequent GC-MS analysis of the gaseous component revealed the
presence of propene.
(17) Alexander, E. R. Principles of Ionic Organic Reactions, John Wiley
& Sons: New York, 1950; p 188.
(18) For a recent review, see: Billington, D. C. In ComprehensiVe Organic
Synthesis; Trost, B. M., Fleming, I., Eds.; Pergamon Press: Oxford, 1991;
Vol. 1.
(19) Young, W. G.; Roberts, J. D. J. Am. Chem. Soc. 1946, 68, 1472.
(20) Felkin, H.; Roussi, G. Tetrahedron Lett. 1965, 4153. Benkeser, R.
A.; Broxterman, W. E. J. Am. Chem. Soc. 1969, 91, 5162. Sakurai, H.; Kudo,
Y.; Miyoshi, H. Bull. Chem. Soc. Jpn. 1976, 49, 1433.
(21) Ashby, E. C.; Goel, A. B. J. Am. Chem. Soc. 1977, 99, 310.