Reduction of β-hydroxyketones by Sml2/H2O/ Et3N

Article abstract of DOI:10.1021/ol047835p

(Chemical Equation Presented) Reduction of a series of β- hydroxyketones by Sml₂/H₂O/Et₃N provided 1,3-diols in quantitative yields. The reactions were exceedingly clean with no byproduct formation, negating the need for f

Full text of DOI:10.1021/ol047835p

ORGANIC
LETTERS
2005
Vol. 7, No. 1
119-122
Reduction of
SmI2/H2O/Et3N
â-Hydroxyketones by
Todd A. Davis,† Pramod R. Chopade,† Go
1
ran Hilmersson,‡ and
Robert A. Flowers, II*,§
Department of Chemistry & Biochemistry, Texas Tech UniVersity,
Lubbock, Texas 79409, Department of Chemistry, Go¨teborg UniVersity,
SE-412 96 Go¨teborg, Sweden, and Department of Chemistry, Lehigh UniVersity,
Bethlehem, PennsylVania 18015
rof2@lehigh.edu
Received October 20, 2004
ABSTRACT
Reduction of a series of
â-hydroxyketones by SmI2/H2O/Et3N provided 1,3-diols in quantitative yields. The reactions were exceedingly clean
with no byproduct formation, negating the need for further purification. Most reactions provided moderate to excellent diastereoselectivity
with syn-diols as the major isomer in most instances.
The pioneering discovery by Inanaga1 describing the use of
HMPA in accelerating SmI2-mediated reactions has resulted
in widespread applications of the SmI2-HMPA mixture. This
combination is utilized in several reduction and reductive
coupling reactions. Due to its ability to enhance reaction
outcomes, HMPA is the preferred cosolvent in SmI2-
mediated reactions. There is, however, one serious drawback
to the use of HMPA. It is a suspected human carcinogen
and, therefore, extreme caution needs to be exercised during
its use. Several nontoxic alternatives to HMPA have been
reported recently. Cabri2 and co-workers have explored the
use of 1,1,3,3-tetramethylguanidine (TMG), 1,1,3,3-tetra-
methylurea (TMU) and Et3N as cosolvents in halide-olefin
coupling reactions. Curran3 has reported the effects of water
and DMPU in the reduction of ketones. These additives have
proven to be useful in several reactions but unfortunately
do not have the broad applicability of HMPA, and as a
consequence, the search for an alternative is ongoing.
In this regard, the work reported by Hilmersson4 and co-
workers is of considerable significance. They have reported
the use of an SmI2/H2O/Et3N mixture in the reduction of
ketones. These reactions are instantaneous and provide
excellent yields of reduced products. A comparison of the
H2O/Et3N method and the HMPA/alcohol method in reduc-
tion of ketones indicates that H2O/Et3N is approximately 100
times faster. This method has also been applied in the
reduction of R,â-unsaturated esters, imines, and conjugated
olefins with excellent results.5,6 These examples clearly show
the utility of SmI2/H2O/amine mixtures in reduction of
several functional groups. Moreover, they provide better
yields and require much less time than the HMPA/alcohol
systems. The workup and subsequent purification of products
are straightforward since the byproducts precipitate during
the course of the reaction. Therefore, the combination of
amine and H2O provides an excellent alternative to HMPA
in SmI2-based reactions.
Initial mechanistic studies show that water and Et3N do
not accelerate the reactions separately. The acceleration is a
result of the Et3N-H2O mixture. Other amines such as
N,N,N′,N′,-tetramethylethylenediamine (TMEDA) and N,N,N′,
N′′,N′′-pentamethyldiethylenetriamine (PMDTA) have the
same effect as Et3N, while replacement of water by alcohols
has a deleterious impact on the rates of reduction. It has been
† Texas Tech University.
‡ Go¨teborg University.
§ Lehigh University.
(1) Inanaga, J.; Ishikawa, M.; Yamaguchi, M. Chem. Lett. 1987, 1485.
(2) Cabri, W.; Candiani, I.; Colombo, M.; Franzoi, L.; Bedeschi, A.
Tetrahedron Lett. 1995, 36, 949.
(4) Dahlen, A.; Hilmersson, G. Tetrahedron Lett. 2002, 43, 7197.
(5) Dahlen, A.; Hilmersson, G. Tetrahedron Lett. 2003, 44, 2661.
(6) Dahlen, A.; Hilmersson, G. Chem. Eur. J. 2003, 9, 1123.
(3) Hasegawa, E.; Curran, D. P. J. Org. Chem. 1993, 58, 5008.
10.1021/ol047835p CCC: $30.25
© 2005 American Chemical Society
Published on Web 12/15/2004

Sm(II) reductant. The diastereoselectivity was determined
by gas chromatography, and the stereochemistry of the 1,3-
diols was determined utilizing the protocol described by
Rychnovsky.10 The results are shown in Table 1.
Table 1. Reduction of â-Hydroxyketones by SmI2/H2O/Et3N
All reactions were quantitative, and the precipitation of
byproducts Sm(OH)3 and HNEt3+ I- made purification quite
simple. Filtration of the precipitate and rotary evaporation
of solvent provided clean product, and no further purification
was necessary. Inspection of the results in Table 1 shows a
number of interesting trends. Most reactions provided the
syn diastereomer predominantly, although two cases (entries
9 and 12) provided the anti diastereomer as the major
product. When R2 was Ph, the diastereoselectivity decreased
dramatically, whereas Me, Et, and i-Pr substituents resulted
in good to excellent diastereoselectivity.
entry
R1
R2
ds (syn:anti)a
1
2
3
4
5
6
7
8
9
t-Bu
t-Bu
t-Bu
t-Bu
t-Bu
Ph
Ph
Ph
Ph
Ph
Me
Et
i-Pr
t-Bu
Ph
Me
Et
i-Pr
t-Bu
Ph
12:1
29:1
99:1
4:1
1.5:1
99:1
>99:1
>99:1
1:41
2.6:1
99:1
1:2
Comparison with other reported protocols using SmI2
reveals some interesting trends and differences among SmI2-
based approaches to the reduction of â-hydroxyketones.
When R1 ) t-Bu and R2 ) Ph (entry 5, Table 1), poor
diastereoselectivities were obtained using SmI2/H2O/Et3N and
SmI2/MeOH in THF7 or DME8 suggesting that this substitu-
tion pattern may be inherently difficult to reduce diastereo-
selectively. When R1 ) Me and R2 ) Ph (entry 12, Table
1), excellent diastereoselectivities providing the anti dia-
stereomer were obtained with SmI2/MeOH in THF and DME,
whereas reduction by SmI2/H2O/Et3N provided only modest
selectivity for the anti diastereomer. Substrates in entries 1,
6, and 11 of Table 1 are reduced with greater diastereo-
selectivity by SmI2/H2O/Et3N than SmI2 in THF/MeOH.
While ease of reduction, workup, and in some cases higher
diastereoselectivities are obtained in reductions of â-hydroxy-
ketones with SmI2/H2O/Et3N in THF compared to SmI2/
MeOH, it is important to assess various mechanistic scenarios
responsible for reaction outcomes so that practitioners can
make judicious choices best suited to their system of interest.
The reduction of ketones by SmI2 in the presence of proton
sources likely proceeds through a House-type mechanism,11
and recent mechanistic work has shown that the rate-limiting
step is the first proton transfer to the initially formed ketyl
radical anion.12,13 While reduction of â-hydroxyketones may
be somewhat more complex since the presence of an internal
proton source (â-OH) may also act as a donor, initial rate
studies are also consistent with the initial proton transfer to
the ketyl radical anion as the rate-limiting step.14 The radical
produced after protonation of the ketyl is reduced to a
carbanion by a second equivalent of Sm(II). The stereo-
defining step in the present reductions is likely to be the
final proton transfer to the carbanion.
10
11
12
Me
Me
t-Bu
Ph
a Substrate (â-hydroxyketone, 1 mmol) was placed in a flame-dried
round-bottom flask and dissolved in 10 mL of anhydrous THF, and the
solution was cooled to 0 °C. A mixture of SmI2 (2.5 equiv) in THF,
triethylamine (5 equiv), and deaerated water (6.25 equiv) was added
dropwise to the reaction mixture. The reaction mixture was allowed to stir
at 0 °C until the color of the reaction changed from blue to yellow.
proposed that rapid precipitation of Sm(OH)3 and a tertiary
ammonium salt, Et3NH+ I-, provides the driving force for
the reduction.6
To expand the applicability of the SmI2/H2O/Et3N reagent
and to determine its general utility in important single-
electron-transfer-promoted reactions, the reduction of â-hy-
droxyketones to 1,3-diols was studied. The seminal work of
Keck has shown that the reduction of â-hydroxyketones is
sensitive to proton donor source and temperature.7 Recent
work in our laboratory has shown that solvation also plays
an important role in determining the stereochemical outcome
of these reductions.8 A limited survey of the effects of
substitution and solvent on reduction of â-hydroxyketones
in three solvents showed that, in most cases, DME provided
superior diastereoselectivities over THF (both solvents lead-
ing to the syn product), whereas reductions in CH3CN led
to anti-1,3-diols. Since the nature of substitution in substrates
is known to influence outcomes in SmI2/H2O/Et3N-mediated
reactions,6 a series of â-hydroxyketones were synthesized
to examine the role of substituents in detail.
Table 1 contains the series of â-hydroxyketones examined
in this study. The â-hydroxyketones were synthesized by the
aldol reaction of ketones and aldehydes or the L-proline-
catalyzed asymmetric aldol reaction.9 Substrates were sub-
jected to reduction using SmI2/H2O/Et3N. These reactions
were performed at 0 °C in an inert atmosphere of N2, and
all reactions were completed within 5 min of addition of the
There is strong synthetic and mechanistic evidence to
support the presence of chelation along the reaction coor-
dinate of Sm(II)-mediated reduction and reductive coupling
(10) Rychnovsky, S. D.; Yang, G.; Powers, J. P. J. Org. Chem. 1993,
58, 5251.
(11) House, H. O. In Modern Synthetic Reactions, 2nd ed.; W. A.
Benjamin: Menlo Park, CA, 1972.
(12) Dahle´n, A.; Hilmersson, G. Tetrahedron Lett. 2001, 42, 5565.
(13) Chopade, P. R.; Prasad, E.; Flowers, R. A., II. J. Am. Chem. Soc.
2004, 126, 44.
(7) Keck, G. E.; Wager, C. A.; Sell, T.; Wager, T. T. J. Org. Chem.
1999, 64, 2172.
(8) Chopade, P. R.; Davis, T. A.; Prasad, E.; Flowers, R. A., II. Org.
Lett. 2004, 6, 2685.
(9) List, B.; Lerner, R. A.; Barbas, C. F., III. J. Am. Chem. Soc. 2000,
122, 239.
(14) Chopade, P. R. Ph.D. Dissertation, Texas Tech University, Lubbock,
TX, 2004.
120
Org. Lett., Vol. 7, No. 1, 2005

Figure 1. Electrostatic interaction between lone pairs on carbanion intermediate and neighboring oxygen.
reactions of â-substituted ketones, and this is likely to be
important in the reductions described above.7,15-17 The
majority of the reactions examined in the present case provide
the syn diastereomer as the major product. After analysis of
numerous chelation models, the results of the reductions
described herein are consistent with the importance of
electrostatic interactions in defining the final stereochemistry
of the 1,3-diol product (Figure 1). Reaction of a â-hydrox-
yketone with SmI2/H2O/Et3N through two electron transfers
(from 2 equiv of SmI2) and one proton transfer (from H2O)
leads to two potential intermediates shown as A and B in
Figure 1. Protonation of A leads to the formation of the syn-
1,3-diol, whereas protonation of B leads to the formation of
the anti-1,3-diol. In A, the electron lone pair of the oxygen
and the lone pair of the adjacent carbanion are nearly
antiperiplanar and as a result the electrostatic repulsion
between them is not significant. Conversely, in B, the lone
pairs are only separated by approximately 60°, leading to
increased electrostatic repulsion. On the basis of this analysis,
the transition state (TS) from A is of lower energy than the
TS from B.
system before configurational change takes place at the
carbon center. Alternatively, distortion of the intermediate
due to the interaction between the lone pair of the carbanion
and the samarium (as shown for A in Figure 1) could move
the nonbonding orbitals on carbon and oxygen out of plane,
providing stabilization as well.
The mechanistic scenarios described above do not explain
the change in diastereoselectivity when R2 ) Ph. After
careful analysis, it is our hypothesis that the electron-deficient
Sm3+ cation could possibly interact with the π-system of
the benzene ring (Figure 2). Since this interaction is only
A different explanation for the syn stereoselectivity is
possible as well. The axially oriented radical precursor of A
is expected to be stabilized due to favorable overlap of the
singly occupied orbital and the nonbonding orbital on the
neighboring oxygen. Upon reduction of the radical, copla-
narity of the nonbonding orbitals on the carbon and
neighboring oxygen, while Coulombically favorable, leads
to an unfavorable orbital interaction.18 This intermediate
(resembling A) could be protonated quickly by water in the
Figure 2. Depiction of possible intermediate leading to poor
stereoselectivity or reversal of stereoselectivity when R2 ) Ph.
possible in B, it could act to stabilize B depending on the
substituent at R1 and as a result provide an alternative
(15) Molander, G. A.; Harris, C. R. J. Org. Chem. 1998, 63, 812.
(16) Kawatsura, M.; Hosaka, K.; Matsuda, F.; Shirahama, H. Synlett
1995, 729.
(18) Skrydstrup, T.; Jarreton, O.; Maze´as, D.; Urban, D.; Beau, J.-M.
(17) Prasad, E.; Flowers, R. A., II. J. Am. Chem. Soc. 2002, 124, 6357.
Chem. Eur. J. 1998, 4, 655.
Org. Lett., Vol. 7, No. 1, 2005
121

mechanistic (stereodifferentiating) pathway for the reduction.
Additional detailed mechanistic studies on a variety of
substrates will be necessary to further test this hypothesis.
In conclusion, reduction of a series of â-hydroxyketones
by SmI2/H2O/Et3N in THF provided good to excellent
selectivity for the syn-diol with a wide range of substrates.
All reductions provided quantitative yields and were com-
pleted within a few minutes. Furthermore, precipitation of
byproducts eliminates the need for further purification, thus
simplifying the workup compared to standard Sm(II)-based
reductions. Analysis of various transition state models
proceeding through chelation control are consistent with the
importance of electrostatic interactions controlling the dia-
stereoselectivity of the 1,3-diol in most instances. Regardless
of the exact mechanistic details of the present reductions,
the data presented herein show the utility and ease of the
SmI2/H2O/Et3N reducing system in the reduction of â-hy-
droxyketones. Examination of the use of this novel reducing
system in reductive coupling reactions is currently underway,
and the results of these studies will be reported in due course.
Acknowledgment. R.A.F. is grateful to the National
Science Foundation (CHE-0413845 and CHE-0423972) for
support of this work. G.H. is grateful to the Swedish Research
Council (2001-1713 and 2003-2442) for financial support.
We thank Professor David Birney for many fruitful discus-
sions related to this work. We also thank reviewers A and B
for their insightful comments.
Supporting Information Available: General methods,
experimental protocols, and spectroscopic data. This material
is available free of charge via the Internet at http://pubs.acs.org.
OL047835P
122
Org. Lett., Vol. 7, No. 1, 2005