Solvent-dependent diastereoselectivities in reductions of β-hydroxyketones by Sml2

Article abstract of DOI:10.1021/ol049129u

The reductions of a series of β-hydroxyketones by Sml₂ were examined in THF, DME, and CH₃CN using methanol as a proton source. Reductions in THF and DME typically lead to the syn diastereomer with DME providing higher diastereoselect

Full text of DOI:10.1021/ol049129u

ORGANIC
LETTERS
2004
Vol. 6, No. 16
2685-2688
Solvent-Dependent Diastereoselectivities
in Reductions of â-Hydroxyketones by
SmI2
Pramod R. Chopade,† Todd A. Davis,† Edamana Prasad,‡ and
,‡
Robert A. Flowers II*
Department of Chemistry and Biochemistry, Texas Tech UniVersity,
Lubbock, Texas 79409, and Department of Chemistry, Lehigh UniVersity,
Bethlehem, PennsylVania 18015
rof2@lehigh.edu
Received May 12, 2004
ABSTRACT
The reductions of a series of â-hydroxyketones by SmI2 were examined in THF, DME, and CH CN using methanol as a proton source. Reductions
3
in THF and DME typically lead to the syn diastereomer with DME providing higher diastereoselectivities. Reductions in CH CN provided the
3
anti diastereomer predominantly. This study reveals that solvation plays an important role in substrate reduction by SmI2.
The unique place held by SmI2 in the arsenal of synthetic
chemists is a result of its versatility in mediating numerous
fundamentally important reactions in organic synthesis,
including reductions, reductive couplings, and tandem reac-
tions.1,2 Since the introduction of SmI2 by Kagan in 1980,3
the solvent of choice has been tetrahydrofuran (THF). While
THF is a reasonable solvent for many reactions, it is also a
good hydrogen atom donor4 and can terminate certain radical
reactions.5 Other solvents, including tetrahydropyran (THP),6
acetonitrile (CH3CN),7 and even benzene,8 have been re-
ported as media for reactions, but each has certain features
that limit their use.
From a mechanistic vantage point, little is known about
the role of solvation in SmI2-mediated reactions. Examination
of crystal structure data for SmI2 in nitriles,9 dimethoxyethane
(DME),10 and THF11 shows that there is substantial structural
diversity for SmI2-solvates with six-, seven-, and eight-
coordinate complexes reported in the literature. The diversity
of SmI2-solvates and the likely variable affinity of different
solvents for Sm(II) lead to the supposition that the solvent
milieu will have an impact on the reactivity of SmI2 and
† Texas Tech University.
‡ Lehigh University.
(7) Hamann, B.; Namy, J.-L.; Kagan, H. B. Tetrahedron 1996, 52,
14225-14234.
(1) Molander, G. A.; Harris, C. R. Chem. ReV. 1996, 96, 307-338.
(2) Steel, P. G. J. Chem. Soc., Perkin Trans. 1 2001, 2727-2751.
(3) Girard, P.; Namy, J. L.; Kagan, H. B. J. Am. Chem. Soc. 1980, 102,
2693-2698.
(4) Newcomb, M.; Kaplan, J. Tetrahedron Lett. 1988, 29, 3449-3450.
(5) Nagashima, T.; Rivkin, A.; Curran, D. P. Can. J. Chem. 2000, 78,
791-799.
(6) Namy, J.-L.; Colomb, M.; Kagan, H. B. Tetrahedron Lett. 1994, 35,
1723-1726.
(8) Kunishima, M.; Hioki, K.; Ohara, T.; Tani, S. J. Chem. Soc., Chem.
Commun. 1992, 219-220.
(9) Chebolu, V.; Whittle, R. R.; Sen, A. Inorg. Chem. 1985, 24, 3082-
3085.
(10) Evans W. J.; Broomhall-Dillard, R. N. R.; Ziller, J. W. Polyhedron
1998, 17, 3361-3370.
(11) Evans, W. J.; Gummersheimer, T. S.; Ziller, J. W. J. Am. Chem.
Soc. 1995, 117, 8999-9002.
10.1021/ol049129u CCC: $27.50 © 2004 American Chemical Society
Published on Web 07/10/2004

potentially alter product distributions. To test this hypothesis,
SmI2 was characterized in THF, CH3CN, and DME and the
reduction of a series of â-hydroxyketones was examined.
This study reveals that solvation plays an important role in
substrate reduction by SmI2.
Solutions of SmI2 in THF, CH3CN, and DME were
prepared by reaction of I2 with excess Sm metal. The
solutions were stored in a drybox until used. Concentrations
of SmI2 in each solvent were determined by iodometric
titration. The solubilities of SmI2 in THF and CH3CN were
0.1 and 0.05 M, respectively. Solubility in DME was initially
0.1 M, but precipitation was evident after a few days. The
solution of SmI2 in DME was allowed to stand, and the
concentration was monitored as a function of time and
reached a steady state of 0.02 M after a few days. Sonication
in a cleaning bath temporarily dissolved precipitated SmI2-
DME.
Figure 1. UV-vis spectra of SmI2 in THF (dot), DME (dash-
dot), and CH3CN (solid).
Next, vapor pressure osmometry (VPO) was utilized to
determine the solution MW of SmI2 in DME. This study
showed that SmI2 is clearly monomeric in DME with an
aggregation number of 1.09 ( 0.02. This finding was
consistent with previous studies showing that SmI2 is a
solvated monomer in THF12 and CH3CN.13
To characterize SmI2 further in the three solvents, UV-
vis spectroscopy was utilized. The spectra are displayed in
Figure 1. The spectrum of SmI2 in THF shows the typical
absorptions at 350, 418, 557, and 618 nm. The spectrum of
SmI2 in CH3CN shows broad absorptions at 440 and 680
nm. The DME solvate of SmI2 more closely resembles the
spectrum in THF, although there are some differences. The
low-energy bands are slightly blue-shifted, with the highest
wavelength band having a higher intensity.
A recent paper by Dorenbos assigned the absorption bands
in Sm(II) complexes to f to d (4fn to 4fn-15d1) transitions.14
The analysis in the Dorenbos work was based on the spectra
of lanthanides generated in inorganic salts in the solid state.
Though care must be taken in comparing the electronic
transitions in solutions with the solid state, the apparent
changes in the absorption maxima as well as the relative
intensity of these peaks in different solvents suggest that the
transition energy and probability are affected by solvation
of Sm(II). The X-ray crystal structures of SmI2 in these
solvents indicate that the solvent shell geometry is different
in each solvent.9-11 If the solvent can perturb the accessible
d orbitals of Sm(II), it is reasonable to assume that the change
in solvent shell geometry and affinity could lead to unique
d-orbital pertubation, resulting in f to d transitions of different
energy and probability. Nonetheless, examination of the
spectra in Figure 1 clearly shows that each SmI2-solvate is
unique.
V13 (vs saturated Ag/AgNO3), respectively. The potential of
SmI2 in DME was measured employing a glassy carbon
electrode, a saturated Ag/AgNO3 reference, and a platinum
wire auxiliary electrode. The electrolyte was either LiI or
tetra-n-butylammonium hexafluorophosphate. The E1/2 of
SmI2 in DME was found to be -1.62 ( 0.05 V. The CV is
similar to those generated in THF and CH3CN. Comparison
of SmI2 in all three solvents shows very little difference
between them in terms of their thermodynamic reducing
power (within experimental error).
The characterization of SmI2 in THF, CH3CN, and DME
is consistent with the following: (1) SmI2 is monomeric in
all three solvents; (2) dissolution of SmI2 in each solvent
provides unique UV-vis spectra; and (3) the E1/2 values are
nearly the same within experimental error. The affinity of
solvent for Sm(II) and the ability of various functional groups
to displace solvent (or I-) and interact with Sm(II) through
chelation can potentially alter the energies of intermediates
and activated complexes along the reaction coordinate,
resulting in different product distributions.15
To address this issue, the reduction of a series of
â-hydroxyketones was examined in all three solvents. The
seminal work of Keck showed that the reduction of â-hy-
droxyketones by SmI2 is sensitive to substitution, proton
source, and temperature.16 On the basis of these findings, it
was proposed that these substrates may be sensitive to
changes in the solvent milieu as well. Initial experiments
utilized 25 equiv of MeOH (based on [SmI2]) as a proton
source in all three solvents. The diastereoselectivity of the
reductions were determined using GC and 1H NMR, and the
identities of the diastereomers were established using the
protocol reported by Rychnovsky.17 The results are shown
in Table 1. Examination of the data shows some very
interesting trends. In global terms, reductions in THF and
The final characterization of reductants utilized cyclic
voltammetry (CV) to estimate the redox potential of SmI2
in DME. Potentials for SmI2 in THF and acetonitrile have
been reported and are -1.58 ( 0.0412 and -1.44 ( 0.05
(15) Prasad, E.; Flowers, R. A., II. J. Am. Chem. Soc. 2002, 124, 6357-
(12) Shotwell, J. B.; Sealy, J. M.; Flowers, R. A., II. J. Org. Chem. 1999,
64, 5251-5255.
6361.
(16) Keck, G. E.; Wager, C. A.; Sell, T.; Wager, T. T. J. Org. Chem.
1999, 64, 2172-2173.
(17) Rychnovsky, S.; Yang, G.; Powers, J. J. Org. Chem., 1993, 58,
5251-5255.
(13) Kuhlman, M. L.; Flowers, R. A., II. Tetrahedron Lett. 2000, 41,
8049-8052.
(14) Dorenbos, P. J. Phys.: Conden. Matter 2003, 15, 575-594.
2686
Org. Lett., Vol. 6, No. 16, 2004

Table 1. Reduction of â-Hydroxyketones by SmI2 in THF, CH3CN, and DME
a GC yields. b MeOH (25 equiv, based on [SmI2]) was used in the THF and DME studies. c Variable concentrations of MeOH were used in the acetonitrile
studies. d From ref 16.
DME show the same trends, while reductions in CH3CN
always provide the anti 1,3-diol as the major diastereomer.
Quantitative reductions were obtained in THF and DME,
while good yields (70-80%) were obtained in CH3CN. Use
of greater amounts of SmI2 in CH3CN did not provide
significantly better yields in the reductions.
In comparison to SmI2 in THF, reduction of â-hydroxy-
ketones (entries 1, 2, and 4, Table 1) by SmI2 in DME
provided significantly higher diastereoselectivities. The eight-
coordinate solvate provided by the bidentate DME could
potentially increase the steric demands of the reductant
enough to provide the observed increase in diastereoselec-
tivity.
Although reduction of all substrates by SmI2 in CH3CN
provides anti diastereoselectivity, the selectivity was modest
using 25 equiv of MeOH. The different stereochemical
outcome for reductions in CH3CN indicated the possibility
of an alternate mechanistic pathway for the SmI2-mediated
conversion in this solvent. Another factor that could poten-
tially have mechanistic implications is the presence of the
proton source, MeOH. Proton sources are known to have an
impact on the stereochemical outcome of numerous SmI2-
based reactions,16,18 and recent mechanistic studies have
shown that proton donor concentration can have an impact
on the mechanistic pathway of the reduction.19,20
To examine the interplay between MeOH and solvent on
SmI2, UV-vis spectra of the reductant in each solvent was
monitored in the presence of increasing amounts of MeOH.
In THF and DME, no changes were visible in the spectra
upon addition of up to 200 equiv of MeOH. Conversely,
addition of MeOH to SmI2 in CH3CN (Figure 2) showed
changes with as little as 2 equiv of the proton donor,
suggesting coordination of the alcohol to Sm.21 These
findings indicate that MeOH has a much higher affinity for
the oxophilic Sm in CH3CN. In THF and DME, higher
concentrations of MeOH are required to displace bulk solvent
from the inner sphere of Sm.20
Since coordination of MeOH to SmI2 could disrupt
chelation of â-hydroxyketones, a number of reductions were
run at lower concentration of the proton donor in CH3CN.
In all cases, it was found that 10 or fewer equivalents of
MeOH provided higher diastereoselectivities in CH3CN than
(18) (a) Edmonds, D. J.; Muir, K. W.; Procter, D. J. J. Org. Chem. 2003,
68, 3190-3198. (b) Hutton, T. K.; Muir, K. W.; Procter, D. J. Org. Lett.
2003, 5, 4811-4814.
(19) Hoz, S.; Yacovan, A.; Bilkis, I. J. Am. Chem. Soc. 1996, 118, 261-
262.
(20) Chopade, P. R.; Prasad, E.; Flowers, R. A., II. J. Am. Chem. Soc.
2004, 126, 44-45.
(21) Dahlen, A.; Hilmersson, G.; Knettle, B. W.; Flowers, R. A., II. J.
Org. Chem. 2003, 68, 4870-4875.
Org. Lett., Vol. 6, No. 16, 2004
2687

and DME, with the latter solvent generally providing better
selectivity. Reductions in CH3CN provided anti diols in good
yields, and smaller amounts of MeOH were found to give
better diastereoselectivities. Further characterization of SmI2
by UV-vis spectroscopy showed evidence for coordination
of MeOH to Sm in CH3CN, but not in THF or DME.
Although the discreet mechanistic details are not clear at this
point, the SmI2-MeOH complex is proposed to play a role
in different reactivity patterns displayed in CH3CN.
The affinity of solvent for SmI2 likely plays a direct role
in the ability of various substrates or additives (alcohols) to
displace solvent from the metal and interact with it through
direct coordination to the inner sphere. If this supposition is
correct, the data communicated herein shows that different
combinations of solvent, additive, and substrate could have
an impact on the outcome of SmI2-mediated reactions.
Mechanistic (rate) studies are currently being carried out to
examine the details of these reactions to provide synthetic
chemists with the knowledge necessary to make judicious
choices in solvent and additive to effect desired outcomes
in these reactions. The results of these studies will be reported
in due course.
Figure 2. Uv-vis spectra of samarium diiodide (5 mM) (solid)
and samarium diiodide in the presence of 2 equiv of methanol (dash
dot) in acetonitrile.
the original experiments conducted with 25 equiv (Table 1).
Although the diastereoselectivities in CH3CN are still modest,
reduction of 3-hydroxy-1-phenylbutane-1-one (entry 4, Table
1) to the corresponding diol provides a syn:anti ratio of 1:10.
In summary, physical characterization of SmI2 by a variety
of techniques has shown that its dissolution in THF, DME,
and CH3CN provides a unique reductant in each, presumably
through differences in solvation. The impact of changes in
the solvent milieu was examined through the reduction of a
series of â-hydroxyketones using MeOH as a proton source.
The same trends in diastereoselectivity were apparent in THF
Acknowledgment. R.A.F. is grateful to the National
Science Foundation (CHE-0413845) and Lehigh University
for support of this work. We thank Dr. Rebecca S. Miller
for her useful comments on the manuscript.
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.
OL049129U
2688
Org. Lett., Vol. 6, No. 16, 2004