Development of an effective chiral auxiliary for hydroxyalkyl radicals

Article abstract of DOI:10.1021/jo010880f

The development of an effective chiral auxiliary for hydroxyalkyl radicals is delineated. Both the 2-tetrahydropyranyl (THP) and tri-O-benzyl-2-deoxy-α-D-glucopyranosyl (GLU) auxiliaries resulted in diastereoselective radical additions to methyl acrylate at -78°C (ds = 6/1 and 11/1, respectively). The developing stereochemistry at the radical center was completely under auxiliary control. Correlation experiments showed that the D-GLU auxiliary led to attack on the radical Si-face. The selectivity of these radical additions dropped-off considerably when the more reactive 2-nitropropene trap was employed. Computational studies suggested that the observed facial selectivity was due primarily to entropic factors in the transition state but that a smaller temperature-dependent enthalpic contribution was also involved. It was hypothesized that incorporation of a quaternary center at C-6 (THP numbering) would restore the facial selectivity with more reactive radical traps by restricting the orientations available to the incoming alkene. In the event, the trans-6-tert-butyltetrahydropyranyl (tBu-THP) auxiliary resulted in very good diastereoselection with 2-nitropropene (ds = 35/1 at -78°C, 15/1 at 0°C, and 8/1 at RT) as did the tri-O-benzyl-6,6-dimethyl-2-α-D-deoxyglucopyranosyl (diMe-GLU) auxiliary during additions to ethyl α-trifluoroacetoxyacrylate (ds = 10/1 at 0°C). A protocol for recovery of the sugar-derived chiral auxiliaries was also established. This work sets the stage for the development of a novel approach to 1, 3, 5...(2n + 1) polyols based on iterative radical homologation as well as the application of these pyranosidic auxiliaries to other synthetically important reactions.

Full text of DOI:10.1021/jo010880f

Develop m en t of a n Effective Ch ir a l Au xilia r y for Hyd r oxya lk yl
Ra d ica ls
Philip Garner,* J ames T. Anderson, Philip B. Cox, Stephen J . Klippenstein, Ray Leslie, and
Noemi Scardovi
Department of Chemistry, Case Western Reserve University, Cleveland, Ohio 44106-7078
ppg@po.cwru.edu
Received August 27, 2001
The development of an effective chiral auxiliary for hydroxyalkyl radicals is delineated. Both the
2-tetrahydropyranyl (THP) and tri-O-benzyl-2-deoxy-R-D-glucopyranosyl (GLU) auxiliaries resulted
in diastereoselective radical additions to methyl acrylate at -78 °C (ds ) 6/1 and 11/1, respectively).
The developing stereochemistry at the radical center was completely under auxiliary control.
Correlation experiments showed that the D-GLU auxiliary led to attack on the radical Si-face. The
selectivity of these radical additions dropped-off considerably when the more reactive 2-nitropropene
trap was employed. Computational studies suggested that the observed facial selectivity was due
primarily to entropic factors in the transition state but that a smaller temperature-dependent
enthalpic contribution was also involved. It was hypothesized that incorporation of a quaternary
center at C-6 (THP numbering) would restore the facial selectivity with more reactive radical traps
by restricting the orientations available to the incoming alkene. In the event, the trans-6-tert-
butyltetrahydropyranyl (tBu-THP) auxiliary resulted in very good diastereoselection with 2-nitro-
propene (ds ) 35/1 at -78 °C, 15/1 at 0 °C, and 8/1 at RT) as did the tri-O-benzyl-6,6-dimethyl-
2-R-D-deoxyglucopyranosyl (diMe-GLU) auxiliary during additions to ethyl R-trifluoroacetoxyacrylate
(ds ) 10/1 at 0 °C). A protocol for recovery of the sugar-derived chiral auxiliaries was also established.
This work sets the stage for the development of a novel approach to 1, 3, 5...(2n + 1) polyols based
on iterative radical homologation as well as the application of these pyranosidic auxiliaries to other
synthetically important reactions.
In tr od u ction
radical reactions. Most of the auxiliary work to date has
focused on chiral equivalents of the carboxylic acid
substituted carbon radical, [•CH(R)CO2H], some notable
examples of which (structures 4-7) are depicted in
Figure 1. In each of these systems, the preferred orienta-
tion of the auxiliary relative to the prochiral carbon
radical results from steric and/or electronic factors. The
corresponding auxiliary-substituted acrylamides serve as
good models that enable one to understand the confor-
mational preferences of these chiral radicals.3 Upon
completion of the radical homologation reaction, the
carboxylic acid moiety in 3 can be released by hydrolysis
and the chiral auxiliary recycled. Following the devel-
opmental path of closed-shell processes such as the
Diels-Alder reaction and Lewis acid-mediated carbonyl
additions, recent efforts have resulted in enantioselective
free radical reactions.4 Despite these advances, a truely
general approach to the control of acyclic stereochemistry
during radical reactions still remains an elusive goal. In
this context, the development of a readily available and
easily recoverable chiral auxiliary for hydroxyalkyl radi-
cals would represent a particularly valuable synthetic
advance since it would complement the existing meth-
odology.
Radical-based C-C bond forming reactions offer a
powerful alternative to their polar counterparts and,
accordingly, occupy an important role in the arsenal of
organic synthesis.1 The different reactivity patterns
associated with carbon radicals (versus carbanions and
carbocations), especially in terms of functional group
compatibility, augurs well for the application of radicals
to complex synthetic problems. Furthermore, an im-
proved understanding of radical structure/reactivity now
makes the design of successful radical chain processes
relatively straightforward. Whereas early work on radical
chemistry focused on issues related to chemoselectivity
and regioselectivity, attention has now shifted to the
control of stereochemistry.2
Both substrate derived and auxiliary derived chirality
have been used to control the stereochemical course of
(1) (a) Hart, D. J . Science 1984, 223, 883. (b) Giese, B. Radicals in
Organic Synthesis: Formation of Carbon-Carbon Bonds; Pergamon
Press: New York, 1986. (c) Ramaiah, M. Tetrahedron 1987, 43, 3541.
(d) Curran, D. P. Synthesis 1988, 417, 489. (e) J asperse, C. P.; Curran,
D. P.; Fevig, T. L. Chem. Rev. 1991, 91, 1237. (f) Renaud, P.; Sibi, M.
Radicals in Organic Synthesis; Vol. 1, Wiley-VCH: Weinheim, 2001.
(2) (a) Giese, B. Angew. Chem., Int. Ed. Engl. 1989, 28, 969. (b)
Porter, N. A.; Giese, B.; Curran, D. P. Acc. Chem. Res. 1991, 24, 296.
(c) Smadja, W. Synlett 1994, 1. (d) Stereochemistry of Radical Reac-
tions: Concepts, Guidelines, and Synthetic Applications; Curran, D.
P.; Porter, N. A.; Giese, B., Eds.; VCH: Weinheim, 1996. (e) Renaud,
P.; Gerster, M. Angew. Chem., Int. Ed. 1998, 37, 2563.
(3) Sibi, M. P.; J i, J . Angew. Chem., Int. Ed. Engl. 1996, 35, 190.
(4) Sibi, M. P.; Porter, N. A. Acc. Chem. Res. 1999, 32, 163.
10.1021/jo010880f CCC: $22.00 © 2002 American Chemical Society
Published on Web 07/25/2002
J . Org. Chem. 2002, 67, 6195-6209
6195

Garner et al.
In order for the proposed radical homologation meth-
odology to be useful for asymmetric synthesis, it would
have to fulfill the following criteria:
(1) The chiral auxiliary must be readily available and
easily introduced.
(2) The chiral auxiliary must exert high and predictable
diastereoselection during the radical addition.
(3) Radical generation must be accomplished efficiently
under mild conditions from readily available precursors.
(4) The chiral radical equivalent must undergo clean
addition (C-C bond formation) to a variety of syntheti-
cally useful traps.
F IGURE 1. Carboxamide/imide-based chiral auxiliaries.
(5) The auxiliary should be recoverable and possibly
multifunctional.
We now present a full account of our efforts in this
area, which have resulted in the development of a
practical chiral auxiliary for hydroxyalkyl radicals. While
we were completing our initial studies,6 Curran reported
two examples of stereocontrolled selenium transfer to
7-phenylmenthyloxyalkyl radicals.7 This was, to our
knowledge, the first example of a chiral auxiliary for
hydroxyalkyl radicals. Although high levels of diastereo-
selectivity were observed at the selenoacetal center (C-
Se bond formation), this auxiliary suffered from the fact
that the ether linkage precludes its removal under mild
conditions. Furthermore, no examples of C-C bond
formation were reported.
F IGURE 2. Auxiliary-mediated hydroxyalkyl radical homolo-
gation.
Resu lts a n d Discu ssion
Design Hyp oth esis. The success with carbonyl sub-
stituted R-carbon radicals stems from the fact that
established conformational control elements of enolates
(A-strain, chelation) could be used to introduce diastereo-
facial bias. Furthermore, known transformations of car-
boxylic acids could be used for both the attachment
and removal of the chiral auxiliary. In considering an
analogous approach to chiral hydroxyalkyl radical equiva-
lents, it occurred to us that the tetrahydropyran (THP)
group might serve as an effective scaffold for the design
of a chiral auxiliary. The THP group and its carbohydrate
relatives are readily installed and removed from alcohols
under mildly acidic conditions.8 The anomeric carbon is
necessarily chiral, its configuration can be controlled, and
it provides a structural conduit for transmitting stereo-
chemical information to the prochiral radical center. The
nearest analogies that we could find for a THP-based
chiral auxiliary for hydroxyalkyl radicals were the chiral
γ-alkoxyallylboronates of Wuts9 and Hoffmann10 as well
as the glycosylated dienes of Lubineau11 and Stoodley.12
Roush showed that carbohydrates could be used to control
γ-alkoxyallylstannane additions to aldehydes.13 More
F IGURE 3. Strategy for polyol synthesis via iterative radical
homologation.
Although the work cited above allowed control over
facial selectivity with prochiral carbon radicals R to
carbonyl groups, the development of an analogous chiral
hydroxyalkyl radical equivalent has lagged behind. We
thus set out to formulate a plan that would accomplish
the reaction sequence embodied in Figure 2. Generation
of the chiral hydroxylalkyl radical equivalent 9 would be
effected via the functionality “X” in the precursor 8. The
radical 9 would then be intercepted in a diastereoselective
manner by an appropriate olefinic radical trap (C-C bond
formation) eventually leading to the construct 10. Re-
moval of the auxiliary would, at this stage, reveal a chiral
secondary alcohol 11. While the advantage of synthesiz-
ing isolated secondary alcohols in this manner may not
be immediately obvious, it must be stated that our long-
term goal was to employ a radical trap that embodies
the precursor structure (see 9 + 13 in Figure 3). In this
case, the process becomes iterative and suggests a novel
approach to the synthesis of 1,3,5...(2n + 1) polyol
systems that constitute a number of macrocyclic antibi-
otic structures such as roflamycoin (12).5 This strategy
presumes that the chiral auxiliary attached to the
R-carbon will dominate the stereochemical course of the
reaction over any substrate control that might result from
the inherent chirality of “R” in the growing chain.
(6) Garner, P. P.; Cox, P. B.; Klippenstein, S. J . J . Am. Chem. Soc.
1995, 117, 4183.
(7) Curran, D. P.; Geib, S. J .; Kuo, L. H. Tetrahedron Lett. 1994,
35, 6235.
(8) For a review of the use of carbohydrates as recoverable chiral
auxiliaries, see: Kunz, H.; Ru¨ck, K. Angew. Chem., Int. Ed. Engl. 1993,
32, 336.
(9) Wuts, P. G. M.; Bigelow, S. S. Chem. Commun. 1984, 736.
(10) Metternich, R.; Hoffmann, R. W. Tetrahedron Lett. 1984, 25,
4095.
(11) Lubineau, A.; Queneau, Y. Tetrahedron 1989, 45, 6697, and
prior work cited therein.
(12) Beagley, B.; Larsen, D. S.; Pritchard, R. G.; Stoodley, R. J . J .
Chem. Soc., Perkin Trans. 1 1990, 3113, and prior work cited therein.
(5) A complimentary radical-based approach to 1,3-diols based on
the [1,2]-Wittig rearrangement of â-alkoxyalkyl allyl ethers has been
reported. This reaction is believed to proceed via an achiral radical
anion [-O-CH•-CHdCH2] See: Schreiber, S. L.; Goulet, M. T.
Tetrahedron Lett. 1987, 28, 1043.
6196 J . Org. Chem., Vol. 67, No. 17, 2002

Chiral Auxiliary for Hydroxyalkyl Radicals
F IGURE 5. Nicolaou’s oxidative deglycosylation of ampho-
tericin B.
F IGURE 4. Ground state conformation rationale for stereo-
control with a pyranosidic auxiliary. Structures 15 and 16 were
generated by performing geometry-optimized ab initio calcula-
tions at the UHF/6-31G* level with Spartan 5.1.3.
influence that the chiral auxiliary has on the competing
Si- and Re-transition state energies. This, in fact, turns
out to be the case with radicals such as 14 (vide infra).
At this point in our studies, however, it was not at all
clear that tetrahydropyranyloxyalkyl radicals such as 14
could even be generated and trapped. Despite the fact
that radical fragmentations are relatively rare, we feared
that such a reaction could provide a path to a tetrahydro-
pyranyl radical, examples of which are well-known (see
ref 2a). In fact, just such a fragmentation was proposed
by Nicolaou to explain the oxidative deglycosidation of
amphotericin B (17) upon exposure to NBS (Figure 5).19
This concern turned out to be unfounded, at least with
the systems we have studied so far.
recently, Tius has reported the use of glucose as well as
a camphor derived lactol (with an embedded THP sub-
structure) as chiral auxiliaries for oxygenated allenes.14
This line of reasoning led us to consider pyranosidic
chiral auxiliaries for hydroxyalkyl radicals having the
general structure 14 (Figure 4). In analogy to the known
solid-state and solution conformational preferences of
R-methyl glycosides,15 we expected that the C(2)-O(1′)
rotamer with a O(1)-C(2)-O(1′)-C(2′) dihedral angle of
∼60° would be preferred since it minimized an unfavor-
able gauche interaction with C(3) and benefits from the
exo-anomeric effect. This expectation is supported by ab
initio calculations (Figure 4). There is also spectroscopic
evidence indicating that oxygen substituted alkyl radicals
are slightly pyramidalized due to a stabilizing n f SOMO
interaction.16 Thus, the true structure of radical 14 may
best be described as an equilibrium between two pyra-
midalized radicals 15 and 16 of nearly equal energy. If
radical addition proceeds via a relatively early transition
state (TS), the reaction diastereoselectivity would be
governed (at least in part) by the relative proportions of
Si-pyramidialized radical 15 and its Re-counterpart 16.
An example of such “kinetic quenching” of diastereomeric
radicals has been reported by Rychnovsky and co-
workers.17 Of course, the Curtin-Hammett principle
would be in effect if the difference in activation energies
is greater than the barrier to interconversion between
these ground state conformers.18 In this case, the reaction
diastereoselectivity would be dictated primarily by the
Ra d ica l Ch em istr y
Convenient access to pyranosidylalkyl radicals was to
be achieved using chemistry developed by Barton and co-
workers (Scheme 1). Thus, R-hydroxyacid derivatives
corresponding to 21 were converted to their correspond-
ing Barton esters 23 by one of two methods. In the initial
stages of our investigations, we prepared the Barton
esters by simply treating the carboxylic acid with 2-mer-
captopyridine-N-oxide and N,N′-dicyclohexylcarbodiimide
(DCC) in dichloromethane (DCM). However, after en-
countering difficulty in getting this reaction to go to
completion with hindered carboxylic acid substrates
(sterically encumbered R or auxiliary groups), we devel-
oped the so-called HOTT reagent, S-(1-oxido-2-pyridinyl)-
1,1,3,3-tetramethylthiouronium hexafluorophosphate (22),
for the synthesis of “difficult” Barton esters.20 This
reagent eventually became the method of choice for all
of our Barton esterifications. Because of their inherent
instability toward visible light, the bright yellow Barton
esters were not isolated but generated in the dark and
used immediately for the next step.
(13) Roush, W. R.; VanNieuwenhze, M. S. J . Am. Chem. Soc. 1994,
116, 8536.
(14) Harrington, P. E.; Tius, M. A. J . Am. Chem. Soc. 2001, 123,
8509, and prior work cited therein.
(15) (a) J effrey, G. A.; Pople, J . A.; Binkley, J . S.; Vishveshwara, S.
J . Am. Chem. Soc. 1978, 100, 373. (b) Wu, T.-C.; Goekjian, P. G.; Kishi,
Y. J . Org. Chem. 1987, 52, 4819. This paper suggests that steric
destabilization is more important than exo-anomeric stabilization in
determining the conformational preferences of glycosides.
(16) McKelvery, R. D.; Sugawara, T.; Iwamura, H. Magn. Reson.
Chem. 1985, 23, 330.
Upon photolysis using a sunlamp in the presence of
an excess of electron deficient olefin, the Barton ester 23
undergoes decarboxylation to give the radical 9. This
(19) Nicolaou, K. C.; Chakraborty, T.. K.; Ogawa, Y.; Daines, R. A.;
Simpkins, N.. S.; Furst, G. T. J . Am. Chem. Soc. 1988, 110, 4660.
(20) Garner, P.; Anderson, J . T.; Dey, S.; Youngs, W. J .; Galat, K.
J . Org. Chem. 1998, 63, 5732.
(17) Buckmelter, A. J .; Powers, J . P.; Rychnovsky, S. D. J . Am.
Chem. Soc. 1998, 120, 5589.
(18) Roth, M.; Damm, W.; Giese, B. Tetrahedron Lett. 1996, 37, 351.
J . Org. Chem, Vol. 67, No. 17, 2002 6197

Garner et al.
SCHEME 1
F IGURE 6. Olefinic radical “traps” used in this study.
independently subjected to Barton’s radical decarbox-
ylation conditions23 in the presence of methyl acrylate.
Barton esterification was accomplished using DCC. Pho-
tolysis of the diastereomeric Barton esters 34 and 35 led
to the formation of enantiomeric radicals 36 and ent-36,
respectively, which could be trapped with methyl acrylate
to give the enantiomeric addends 37 and ent-37 (X ) SPy)
in good overall yield along with small amounts of double
addition telomer. Telomer formation (see structure 27 in
Scheme 1), which results from the further addition of
addend radicals such as 25 to the alkene 24, was
generally observed when methyl acrylate was the radical
trap. Unfortunately, the lower reactivity of methyl acryl-
ate toward radical 36 made it necessary to use 5 equiv
of this trap in these experiments. These represent
optimized reaction conditions for this trap. The diastereo-
selectivity of these radical additions was best determined
after reductive cleavage of the 2-thiopyridyl ethers to
afford 37 and ent-37 (X ) H). The absolute configuration
of these adducts was established by subsequent correla-
tion experiments (see below). The diastereoselectivity
associated with the formation of 37 mirrored that found
for ent-37 under the same reaction conditions. As ex-
pected, higher diastereoselectivities were observed at
lower reaction temperatures (rxn T, ds: rt, 4/1; -20 °C,
5/1; -78 °C, 6/1; see Supporting Information for details).24
That the diastereomeric relationship between acids 32
and 33 was transformed into an enantiomeric relation-
ship in adducts 37 and ent-37 showed that the stereo-
chemical course of these reactions was under auxiliary
control and, thus, independent of the starting R-hydroxy-
ester configuration.
nucleophilic carbon-centered radical then adds to the
electron deficient olefin 24 to give the electron-deficient
addend radical 25, which can either transfer a thiopyridyl
group from unreacted 23 or collapse with [•SPy] to give
the R-thiopyridyl ether 26. Possible side reactions include
further addition of 25 to the trap 24 (telomerization)
leading to species such as 27 as well as “radical decar-
boxylation” to give the hemithioacetal 28. The extent that
these side reactions compete with the desired intermo-
lecular addition reaction was found to be both trap and
auxiliary dependent. For a particular radical-trap pair,
the rate of the desired radical addition was found to
depend on both electronic (SOMO-LUMO gap) and steric
factors. Throughout these studies, we are assuming that
the rate of auxiliary loss would be approximately the
same for each diastereomer.
Three different radical traps were employed in this
auxiliary development study (Figure 6). Methyl acrylate
(29), 2-nitropropene (30), and ethyl R-trifluoroacetoxy-
acrylate (31)21 have all been used successfully as traps
for radicals derived from Barton ester precursors. The
latter alkene is particularly interesting in that it pos-
sesses a structure that make it functionally equivalent
to the hypothetical iterative trap 13. As will be discussed
below, the diastereoselectivity observed with a particular
chiral auxiliary also varies as a function of the radical
trap. To facilitate comparison, the results of our radical
trapping experiments with each of these traps are
collected separately in Tables 1-3.
Com p u ta tion a l Stu d ies
To better understand the interplay of these stereocon-
trol elements, ab initio calculations were performed on
the competing Re- and Si-transition structures (TSs).25
Energies and canonical partition functions (employing
rigid rotor harmonic oscillator assumptions) were deter-
mined for both pyramidalized radicals corresponding to
36 and six transition states (TS) corresponding to the
addition of CH2dCH2 to both faces of 36 in three different
torsional orientations. The three separate transition
structures for addition to the Si-face were found to be
P r elim in a r y Tr a p p in g Stu d ies. To test the general
validity of our proposal, we prepared the diastereomeric
carboxylic acids 32 and 33 (Scheme 2) from the chro-
matographically separable THP ethers of (S)-methyl
lactate. Mash had established that the faster eluting
ester 33 possessed the (2S,1′S) configuration whereas the
slower eluting ester configuration was (2S,1′R).22 These
esters were saponified and the resulting carboxylic acids
(23) Barton, D. H. R.; Crich, D.; Kretzschmar, G. J . Chem. Soc.,
Perkin Trans. 1 1986, 39.
(24) The diastereomer ratios at -20 and -78 °C differ from those
reported in our original Communication (ref 6). We attribute this
difference to unintentional diastereomer enrichment during “rough”
flash chromatography of the intermediate thiopyridyl ethers as opposed
to the desulfurized products.
(25) Quantum mechanical calculations were performed at the Har-
tree-Fock level with a STO-3g basis set using the GAUSSIAN 92
Program Package, Revision A: Frisch, M. J .; Trucks, G. W.; Head-
Gordon, M.; Gill, P. M. W.; Wong, M. W.; Foresman, J . B.; J ohnson, B.
G.; Schlegel, H. B.; Robb, M. A.; Replogle, S.; Gomperts, R.; Andres, J .
L.; Raghavachari, K.; Binkley, J . S.; Gonzalez, C.; Martin, R. L.; Fox,
D. J .; Defrees, D. J .; Baker, J .; Stewart, J . J . P.; Pople, J . A. Gaussian
Inc., Pittsburgh, PA, 1992.
(21) Barton, D. H. R.; Chern, C.-Y.; J aszberenyi, J . Cs. Tetrahedron
1995, 51, 1867.
(22) Mash, E. A.; Arterburn, J . B.; Fryling, J . A.; Mitchell, S. H. J .
Org. Chem. 1991, 56, 1088.
6198 J . Org. Chem., Vol. 67, No. 17, 2002

Chiral Auxiliary for Hydroxyalkyl Radicals
TABLE 1. Ra d ica l Ad d ition s to Meth yl Acr yla te
TABLE 2. Ra d ica l Ad d ition s to 2-Nitr op r op en e
within 0.1 kcal/mol of each other while those for addition
to the Re-face were 0.2, 0.3, and 1.0 kcal/mol higher in
energy than the lowest value for Si-addition. Further-
more, the torsional motions in the Re-TS are, on average,
much more strongly hindered than those associated with
the Si-TS (Figure 7). As a result, entropic effects strongly
J . Org. Chem, Vol. 67, No. 17, 2002 6199

Garner et al.
TABLE 3. Ra d ica l Ad d ition s to Eth yl r-Tr iflu or oa cetoxya cr yla te
SCHEME 2
favor (i.e., by a factor of 6) Si-attack; yielding a net
calculated preference of 12/1 for the (R,R)-diastereomer
at room temperature and 17/1 at -78 °C.26 These entropic
effects are simply indicative of the greater probability of
CH2dCH2 having an appropriate orientation for Si-attack
and suggest caution when attempting to explain kinetic
preferences in terms of enthalpic factors alone.27
To visualize the connection between steric hindrance
and entropy, one may consider the “cone of acceptance”
defined by rotating the incoming alkene about the
axis of the newly forming bond. This cone represents
all possible (reactive) orientations of the alkene as it
approaches the radical center. Auxiliary substituents
at C-5 restrict the number of reactive orientations
available to the alkene during Re-approach. Since no
such restriction occurs during Si-approach, it is pre-
ferred. We stress that this TS model is only meant to be
used as a guide for auxiliary design and may not be
appropriate for all situations. Since the calculations were
performed on a hypothetical reaction (radical + ethylene),
(26) The small values of the energetic differences between the
transition states for the Si-face and Re-face additions suggests that
the diastereoselectivity for this reaction should be quite small (ds )
1.9/1) when
a prequilibrium between the two radicals and only
energetic effects are considered. The barrier for inversion of these
pyramidalized radicals through a planar configuration is calculated
to be on the order of 2.0 kcal/mol, which is consistent with our
observation that the diastereomeric Barton esters led to identical
diastereomer ratios. For experimental support of this notion, see:
Hoffmann, R.; Ru¨ckert, T.; Bru¨ckner, R. Tetrahedron Lett. 1993, 34,
297.
(27) For an alternative approach, see: Damm, W.; Giese, B.;
Hartung, J .; Hasskerl, T.; Houk, K. N.; Hu¨ter, O.; Zipse, H. J . Am.
Chem. Soc. 1992, 114, 4067.
6200 J . Org. Chem., Vol. 67, No. 17, 2002

Chiral Auxiliary for Hydroxyalkyl Radicals
SCHEME 3
F IGURE 7. Calculated TSs for the Si- and Re-addition of 36
to ethylene. Only one of three torsional orientations about the
developing C-C bond is shown.
uration of the starting hydroxyester 40 is a moot point
and racemic R-hydroxyesters (conveniently available by
the cyanohydrin synthesis) may also be used.
Subjection of the carboxylic acid 40a to the Barton
chemistry again resulted in the smooth formation of a
chiral radical 42a which could be trapped with methyl
acrylate (29) (Scheme 3). This reaction afforded the
adduct 43a in 61% yield along with 16% of double
addition telomer. The diastereoselectivity of this radical
addition at -78 °C was found to be 11/1 after reductive
removal of the thiopyridyl group. This diastereomer ratio
was determined by careful HPLC analysis of the crude
desulfurization reaction mixture obtained from crude
43a . To unambiguously determine the absolute config-
uration of the newly formed stereocenter, 44a was
subjected to acidic methanolysis to afford the known (R)-
γ-lactone 45 and the methyl 2-deoxyglucoside 46 (which
could, in theory, be recycled). Synthetic 45 prepared from
a 5.8/1 mixture of 44 and its (4S)-diastereomer exhibited
an [R]D of +21.3° (c 0.25, CH2Cl2) [lit.29 +30.1° (c 0.85,
CH2Cl2)]. A similar reaction on the THP-adduct ent-37
(X ) H) produced the antipodal (S)-γ-lactone. These
correlation experiments confirmed that the empirically
observed diastereofacial selectivity of these radical ad-
ditions was indeed consistent with our computationally
derived TS model.
Ra d ica l Ap p r oa ch to Ald ols. The use of another
established radical trap with Barton ester 41 resulted
in a conceptually novel approach to asymmetric aldol
synthesis (Scheme 4).30 Thus, photolysis of the Barton
ester 41 in the presence of 2-nitropropene (30) produced
the primary adducts 47 as a mixture of diastereomers.
Conversion of the geminal nitro thioether function to a
ketone was best accomplished by first filtering the crude
reaction mixture through silica gel (to remove the N,N′-
dicyclohexylurea byproduct) then exposing the crude
F IGURE 8. Synthesis of 2-deoxyglycosidic substrates.
one cannot expect a quantitative correlation between
the theoretical and observed diastereoselectivities (with
electron-deficient alkenes). However, trends that are
observed can be considered valid. We conclude that
the temperature dependence of the reaction diastereo-
selectivity is primarily due to enthalpic factors but that
the (T-independent) preference for Si-attack results from
entropic effects.
Su ga r -Der ived Au xilia r ies. Next, experiments were
conducted to see if acetal mediated acyclic stereocontrol
would extend to the use of carbohydrates as recoverable
chiral auxiliaries. We chose to investigate 2-deoxyglucose
based auxiliaries because of their synthetic accessibility.
Since our results with the THP auxiliary were consistent
with preferential addition to the Si-face of radical 36, we
presumed that additional pyranoside substituents at C-3,
C-4, and C-5 (carbohydrate numbering) would not inter-
fere with the absolute sense of asymmetric induction. The
2-deoxy-R-glucoside 39 was prepared by the addition of
methyl lactate to tri-O-benzyl-D-glucal (38) following the
procedure of Mioskowski and Falck (Figure 8).28 The use
of Ph3P‚HBr as a catalyst in this reaction is essential in
that it leads to the desired R-2-deoxyglycoside without
competing Ferrier rearrangement. Saponification of ester
39 gave the carboxylic acid 40 in good overall yield.
Because the stereochemical course of the radical addi-
tions are under auxiliary control (see above), the config-
(29) Mori, K. Tetrahedron 1975, 31, 3011.
(30) Garner, P.; Leslie, R.; Anderson, J . T. J . Org. Chem. 1996, 61,
6754.
(28) Bolitt, V.; Mioskowski, C.; Lee, S.-G.; Falck, J . R. J . Org. Chem.
1990, 55, 5812.
J . Org. Chem, Vol. 67, No. 17, 2002 6201

Garner et al.
SCHEME 4
F IGURE 9. Chiral auxiliary reformation.
adducts to reductive Nef conditions.31 The protected
aldols 48 could be obtained in excellent overall yields
from the starting carboxylic acids (better than 90% per
step in optimized cases). The kinetic diastereomer ratios
were readily determined by comparing the crude 1H NMR
spectra with those of deliberately prepared mixtures. The
diastereoselectivity observed with trap 30, while still
respectable, was even lower than that observed during
analogous additions to methyl acrylate. Only marginal
improvements in selectivity were observed when the
steric bulk of R was increased, and the reaction temper-
ature was lowered (Table 2, entries 2-5). Note that the
diastereoselectivity was even worse when the unsubsti-
tuted THP auxiliary was used with this trap (entry 6).
Glycoside cleavage with concomitant production of the
free aldol represented a potential difficulty because of the
ease with which these aldol products can suffer dehydra-
tion. Attempted use of our previously elucidated condi-
tions for auxiliary removal (PPTS/MeOH) failed to give
any detectable aldol product. However, exposure of 48
to PhSH + BF3‚OEt2, according to Danishefsky’s prece-
dent,32 resulted in clean conversion to the free aldols 50
and the 2-deoxythioglycosides 51 (as a mixture of ano-
mers). The negative optical rotations observed for the free
aldols indicated that the newly formed stereocenter had
the R-configuration33 as was expected from our previous
radical addition results with methyl acrylate. Free aldols
could be obtained in >90% ee (Mosher ester analysis)34
starting from protected aldols that had been purified
(enriched) by flash chromatography. To complete the
auxiliary recovery cycle (Figure 9), the thioglycosides 51
were oxidized with OXONE and the resulting sulfoxides
52 heated to induce sulfenic acid elimination. This two-
step procedure afforded the starting tri-O-benzyl-D-glucal
38 in very good yield.
F IGURE 10. Postulated effect of C-6 substitution on the Re-
TSs of radical additions to methyl acrylate (left) and 2-nitro-
propene (right).
F IGURE 11. Synthesis of tBu-THP auxiliary substrates.
a consequence of an earlier TS, with commitment to bond
formation occurring beyond the effect of the Glu and THP
C-6 substituents (shown qualitatively in Figure 10). This
hypothesis suggested that the incorporation of an R-sub-
stituent at C-6 other than H would result in higher
diastereoselectivities. Unfortunately, such a substitution
would also introduce a 1,3-diaxial interaction and disrupt
the preference for the R-anomer or chair conformation
as required by our model. The incorporation of a trans-
tert-butyl group at C-6 of the auxiliary, on the other hand,
would hinder Re-approach for even the reactive 2-nitro-
propene trap while maintaining the desired chair con-
formation. Although one might expect the benzyloxy-
methyl substituent of the glucose derived auxiliary to be
large enough to hinder the Re-approach of any trap, this
is true for only one of the three available staggered
rotamers about the exocyclic C-C bond. No such ambi-
guity exists with the tert-butyl substituent, which neces-
sarily places a methyl group in the path of the incoming
trap irrespective of exocyclic C-C bond rotation.
To test this idea, an asymmetric synthesis of the tert-
butyl substituted lactol 53 was devised building on the
work of Crimmins (see Supporting Information for syn-
thesis details).35 This compound served as an effective
auxiliary precursor (Figure 11). Anomeric mixture 53
reacted smoothly with an excess of inexpensive (S)-ethyl
lactate in the presence of PPTS to give the R-anomer 54
in good yield. These conditions ensured that unreacted
lactol would not compete with ethyl lactate for the
Au xilia r y Design Revisited . According to our TS
model (Figure 7), it is the R-hydrogen substituent at C-6
(THP numbering) that interferes with Re-approach of the
trap to radical 14 resulting in the observed Si-selectivity.
We reasoned that the drop-off in diastereoselectivity
observed with the more reactive 2-nitropropene trap was
(31) McMurry, J . E.; Melton, J . J . Org. Chem. 1973, 38, 4367.
(32) Halcomb, R. L.; Boyer, S. H.; Wittman, M. D.; Olson, S. H.;
Denhart, D. J .; Liu, K. K. C.; Danishefsky, S. J . J . Am. Chem. Soc.
1995, 117, 5720.
(33) Cf. Fauve, A.; Veschambre, H. J . Org. Chem. 1988, 53, 5215.
(34) Dale, J . A.; Dull, D. L.; Mosher, H. S. J . Org. Chem. 1969, 34,
2543.
6202 J . Org. Chem., Vol. 67, No. 17, 2002

Chiral Auxiliary for Hydroxyalkyl Radicals
SCHEME 5
oxonium intermediate to give the “lactol dimers” which
appeared to be less effective oxonium ion precursors. The
O-protected lactate ester 54 could then be saponified to
give the carboxylic acid 55 which was subjected to Barton
esterifiation and photolysis in the presence of either
2-nitropropene or methyl acrylate to give 57 or 58 (40-
50% overall yield) after reductive desulfurization or
hydrolysis (Scheme 5).
the radical addition reaction were disappointing (42%
yield, ds ) 2/1) with this auxiliary. We tentatively
ascribed this result to the conformational flexibility of
the furanoside ring as well as the splay that results from
the five- versus six-membered ring. Both of these factors
cause the radical center to be further away from the
quaternary carbon which is expected to block the ap-
proach to one face of the radical.
2n d Gen er a tion Su ga r Au xilia r y. Even though the
tert-butyl-substituted auxiliary represented an effective
solution to the stereocontrol problem, the isolated yields
of the products were only moderate. The lowered yields
were traced to the noticeable lability of adducts possess-
ing this auxiliary when compared to adducts with the
2-deoxyglucose derived auxiliary. When the reductive
hydrolysis was not properly buffered during the radical
aldol sequence, the tert-butyl auxiliary was eliminated
and lactol 53 was isolated in 46% yield. The analogous
sequence with the sugar auxiliary could be performed
under unbuffered conditions (pH ∼ 1) without any
problem. Furthermore, the tert-butyl auxiliary protected
aldol 57 had to be stored at -20 °C to avoid decomposi-
tion whereas the sugar protected aldol 48 could be kept
at 5 °C indefinitely without any problem. Clearly, an
auxiliary was needed that merged the favorable stability
of the 2-deoxyglucose derived auxiliary with the enhanced
stereocontrol associated with the tert-butyl-substituted
THP auxiliary.
As the results in Scheme 5 and Table 2 (entries 7-9)
indicate, the tert-butyl substituted auxiliary leads to a
dramatic enhancement of the diastereofacial selectivity
during addition of the lactate derived radical to 2-nitro-
propene.36 Significantly, the ratio of protected aldol
diastereomers 57 was 35/1 at -78 °C, which corresponds
to a diastereomeric excess (de) of 94%. This represents a
7-fold increase in diastereofacial selectivity over that
observed with the glucose derived auxiliary at the same
reaction temperature. Considerable stereodifferentiation
was observed even when the addition reaction was
performed at room temperature, with the diastereo-
selectivity higher than that observed with the glucose
derived auxiliary at -78 °C! The radical addition to
methyl acrylate also benefits from the use of this new
auxiliary, with adduct 48 being formed with a diastereo-
selectivity of 9/1 at RT (Table 1, entry 5). As is usual with
this trap, 12% of the telomer resulting from radical
addition to 2 equiv of acrylate was also isolated. The
configuration of the major product was correlated with
(S)-4-methylbutyrolactone as described above thus con-
firming that the absolute sense of stereocontrol with this
auxiliary conforms to our TS model.
A carbohydrate literature search revealed a possible
candidate for an auxiliary that could be accessed starting
from an inexpensive sugar derivative. The uronate
derived glucal 63 could be easily prepared on a large scale
starting from commercially available, inexpensive D-glu-
curono-6,3-lactone (61).38 This compound was considered
We also investigated the use of the camphor derived
“NOE auxiliary” (Table 2, entry 10) since it embodies a
â-quaternary carbon at a position analogous to C-6 of our
THP auxiliary. The starting acid 59 was prepared from
the commercially available “NOE-lactol dimer” as de-
scribed above for the tert-butyl-substituted auxiliary.37
Unfortunately, both the yield and diastereoselectivity of
(35) Crimmins, M. T.; Al-awar, R. S.; Vallin, I. M.; Hollis, W. G.;
O’Mahony, R.; Lever, J . G.; Bankaitis-Davis, D. M. J . Am. Chem. Soc.
1996, 118, 7513.
(36) Garner, P.; Anderson, J . T. Tetrahedron Lett. 1997, 38, 6647.
(37) We thank Mr. Subhakar Dey for preparing this material.
F IGURE 12. Synthesis of 2nd generation sugar (diMe-GLU)
auxiliary.
J . Org. Chem, Vol. 67, No. 17, 2002 6203

Garner et al.
SCHEME 6
to be an ideal auxiliary precursor since C-6 could be
converted to a tertiary alcohol by a double Grignard
addition to give the auxiliary glucal 63 after O-benzyla-
tion (Figure 12). This compound could be processed as
described for 38 f 40 (Figure 8) to give the ethyl ester
64 and then, after saponification, the carboxylic acid 65.
When carboxylic acid 65 was subjected to HOTT-
mediated Barton esterification and the resulting solution
photolyzed at 0 °C for 20-25 min in the presence of ethyl
R-trifluoroacetoxyacrylate (31), a clean reaction was
observed (Scheme 6, Table 3).39 The auxiliary-protected
γ-hydroxy-R-ketoester 69 was isolated in 90% overall
yield after silica gel mediated hydrolysis of the hemi-
thioacetal moiety in the initially formed adduct 68 It
should be noted that somewhat lower yields were ob-
served when either saturated NaHCO3 or K2CO3/EtOH
were used for hydrolysis as described by Barton. Control
experiments confirmed that adduct 69 was susceptible
to base-catalyzed elimination of the auxiliary lactol. The
diastereoselectivity of this radical addition reaction was
determined to be 10/1, which is very good considering
that the reaction was performed at 0 °C. For comparison,
same reaction with the unmodified glucal auxiliary gave
a 2/1 mixture of the diastereomeric adduct 70 at this
same temperature. As expected from our earlier studies,
higher diastereoselectivities (>20/1) were observed when
the radical homologation of 65 was performed at -78 °C.
Exp er im en ta l Section
Melting points were determined using a Mel-Temp Capillary
melting point apparatus and are not corrected. All moisture-
sensitive reactions were performed in an inert, dry atmosphere
of Ar or N2. Reagent grade solvents were used for either
chromatography or extraction. The following solvents and
reagents were purified beyond commercial reagent grade: THF
was distilled under Ar or N2 from a purple solution of sodium-
benzophenone ketyl. Benzene and methylene chloride were
distilled over CaH2 (0-1 mm grain size) under an inert
atmosphere (Ar or N2). Acetonitrile, diisopropylethylamine,
pyridine, thiophenol, toluene, and triethylamine were distilled
over CaH2 and stored over activated 4 Å molecular sieves.
Analytical thin-layer chromatography (TLC) was performed
using either Merck silica gel 60 F-254 plates or J T-Baker
Si250F plates (0.25 mm thickness). The plates were visualized
first with UV illumination followed by charring with 5%
p-anisaldehyde in 95:5:1 EtOH-AcOH-H2SO4, 2% vanillin in
98:2 EtOH-H2SO4, or “Verghn’s reagent”41 (12.5 g of am-
monium molybdate and 0.5 g of ceric sulfate dissolved in 250
mL of 10% aq H2SO4). Flash chromatography was performed
using silica gel (230-400 mesh). The solvent compositions
reported for all chromatographic separations are on a volume/
volume (v/v) basis.
Syn th esis of Ca r boxylic Acid Su bstr a tes. (2S,2′S)- a n d
(2S,2′R)-2-(Tetr a h yd r op yr a n -2-yloxy)p r op ion ic Acid (32
a n d 33). To a 0.4 M stirred solution of the ester in THF was
added 4 equiv of 1 N NaOH at 0 °C. After 1 h at this
temperature, TLC analysis (3:2 hexanes/EtOAc) revealed that
the starting material was consumed. The stirred solution was
then acidified at 0 °C by the slow addition of 1 N HCl to pH )
4 (pH paper) and extracted three times with EtOAc. The
aqueous phase was then acidified to pH ) 3 and extracted
three times with EtOAc. The combined organic layers were
dried with MgSO4, filtered, and concentrated at reduced
pressure to give the carboxylic acid (yield ) 53-71%) as a clear
oil which was used without further purification. Rf 0.15 (3:2
hexanes/EtOAc).
Con clu sion s
The development of an effective recoverable chiral
auxiliary for hydroxyalkyl radicals has been achieved.
The key breakthrough in the auxiliary design process
came as a result of an ab initio modeling of the competing
transition states associated with the radical addition
reaction. This work sets the stage for the development
of a novel approach to 1, 3, 5...(2n + 1) polyols based on
iterative radical homologation as well as the application
of pyranosidic chiral auxiliaries to other synthetically
important closed-shelled processes.40
(2S)-[[2-Deoxy-3,4,6-tr is-O-(ph en ylm eth yl)-r-D-a r a bin o-
h exop yr a n osyl]oxy]p r op a n oic Acid , Meth yl Ester (39a ).
To a solution of tri-O-benzyl-D-glucal (38) (2.45 g, 5.88 mmol)
and (S)-methyl lactate (0.88 mL, 8.8 mmol) in freshly distilled
CH2Cl2 (15 mL) was added Ph3P‚HBr (100 mg, 0.300 mmol),
and the solution was stirred at ambient temperature, under
(38) (a) Fehlhaber, H. W.; Snatzke, G.; Vlahov, I. Liebigs Ann. Chem.
1987, 637. (b) Wyss, P. C.; Kiss, J .; Arnold, W. Helv. Chim. Acta 1975,
58, 637.
(40) For a recent application of pyranosidic chiral auxiliaries to the
Diels-Alder reaction, see: Garner, P.; Anderson, J . T.; Turske, R. A.
Chem. Commun. 2000, 1579.
(39) Garner, P.; Anderson, J . T. Org. Lett. 1999, 2, 1057.
(41) Corey, E. J .; Roberts, B. E. J . Am. Chem. Soc. 1997, 119, 12425.
6204 J . Org. Chem., Vol. 67, No. 17, 2002

Chiral Auxiliary for Hydroxyalkyl Radicals
Ar, for 2 h. The reaction was then quenched with sat aq
NaHCO3 and the resulting suspension extracted with CH2Cl2
(3 × 75 mL), the organic layers were combined, dried (MgSO4),
and filtered, and the solvent was removed to provide a thick
oily residue that was purified by flash chromatography (17%
EtOAc-hexanes) to give the pure glucoside 39a (2.58 g, 83%)
[2S-(2r,6â)]-2-[[6-(1,1-Dim et h ylet h yl)t et r a h yd r o-2H -
p yr a n -2-yl]oxy]p r op a n oic Acid , Eth yl Ester (54). The
lactol 53 (425 mg, 2.69 mmol), (S)-ethyl lactate (6.36 g, 6.10
mL, 53.8 mmol), and a catalytic amount of PPTS (30 mg) were
dissolved in freshly distilled dry benzene (18 mL) and heated
to reflux under Ar. The reaction was complete (TLC) after 1
h. The contents were cooled to room temperature and then
poured into sat aq NaHCO3 (20 mL) and extracted with EtOAc
(3 × 50 mL). The organic layers were combined and washed
with brine (20 mL), dried over MgSO4, filtered through cotton,
and then concentrated by rotary evaporation. The crude
product was purified by column chromatography (SiO2) using
4% Et2O-hexanes to give 54 as a clear oil, 69% yield. Rf 0.66,
20% EtOAc-hexanes; 1H NMR (300 MHz, CDCl3) δ 4.92 (d, J
) 2.8 Hz, 1 H, H-2′), 4.38 (q, J ) 7.0 Hz, 1 H, H-2), 4.19 (2H,
OCH2CH3), 3.35 (dd, J ) 11.8, 1.9 Hz, 1 H, H-6′), 1.90-1.72
(m, 2 H), 1.66-1.53 (3H), 1.44 (d, J ) 7.1 Hz, 3 H, H-3), 1.38-
1.22 (m + t, J ) 7.1 Hz, 4 H, OCH2CH3), 0.89 (s, 9 H, t-Bu);
13C NMR (75 MHz, CDCl3) δ 173.5 (CO), 96.2, 76.5, 68.8, 60.6,
33.6, 29.4, 25.5, 24.9, 18.6, 18.1, 14.1; EIMS (m/z): [M - H]+
calcd for C14H25O4, 257.1753; found, 257.1765; [M - t-Bu]+
calcd for C10H17O4, 201.1127; found, 201.1128.
as a thick colorless oil. Rf 0.33, 25% EtOAc-hexanes; [R]25
+
D
58.9° (c 5.05, CH2Cl2); 1H NMR (300 MHz, CDCl3) δ 7.40-
7.18 (m, 15 H, 3 × Ph), 5.06 (br d, J ) 3 Hz, 1 H, H-1′), 4.90
(d, J ) 11.1 Hz, 1 H, 0.5 PhCH2), 4.68 (s, 2 H, PhCH2), 4.63
(d, J ) 12 Hz, 1 H, 0.5 PhCH2), 4.53 (d, J ) 11.1 Hz, 1 H, 0.5
PhCH2), 4.48 (d, J ) 12.0 Hz, 1 H, 0.5 PhCH2), 4.19 (q, J )
6.8 Hz, 1 H, H-2), 4.04 (ddd, J ) 11.7, 9.6, 5.7 Hz, 1 H, H-3′),
3.94 (dt, J ) 10.5, 2.1 Hz, 1 H, H-5′), 3.77 (m, 2 H, H-4′, H-6′),
3.67 (s, 3 H, OMe), 3.54 (dd, J ) 10.5, 2.1 Hz, 1 H, H-6′), 2.27
(ddd, J ) 12.6, 5.7,1.0 Hz, 1 H, H-2′ eq), 1.91 (ddd, J ) 12.6,
9.6, 3 Hz, 1 H, H-2′ ax), 1.39 (d, J ) 6.7 Hz, 3 H, Me); 13C
NMR (75 MHz, CDCl3) δ 173.2 (CO), 138.8, 138.7, 138.2
(aromatic C), 128.4, 128.3, 127.90, 127.88, 127.8, 127.7,
127.63, 127.58 (aromatic PhCH), 97.6 (anomeric CH), 78.2,
77.5 (CH), 74.9, 73.5 (CH2), 72.8 (CH), 72.1 (CH2), 71.4 (CH),
68.7 (CH2), 53.0 (CH3), 35.7 (CH2), 18.0 (CH3); IR (CHCl3)
3000, 2940, 2820, 1740, 1440, 1350, 1200, 1090 cm-1; EIMS
(m/z): [M - PhCH2]+ calcd for C24H29O7, 429.1913; found,
429.1899.
[2S-(2r,6â)]-2-[[6-(1,1-Dim et h ylet h yl)t et r a h yd r o-2H -
p yr a n -2-yl]oxy]p r op a n oic Acid (55). The protected R-hy-
droxy ester 54 (472 mg, 1.83 mmol) was dissolved in THF (15
mL) and stirred at 0 °C. To the stirring solution was added
1.0 N NaOH (13 mL) over 5 min. The resulting solution was
stirred overnight (ca. 13 h) under Ar at room temperature
when TLC showed the reaction to be complete. The contents
were cooled to 0 °C and acidified to pH ) 4 by slowly adding
1.0 N HCl. The acidified reaction contents were poured onto
brine (50 mL) and extracted with EtOAc (3 × 80 mL). The
organic layers were combined and washed with brine (50 mL),
dried over MgSO4, filtered through cotton, and then concen-
trated to give 55 as a thick clear oil, quantitative yield. 1H
NMR (300 MHz, CDCl3) δ 4.95 (d, J ) 2.6 Hz, 1 H, H-2′), 4.43
(q, J ) 7.0 Hz, 1 H, H-2), 3.34 (dd, J ) 11.5, 1.7 Hz, 1 H, H-6′),
1.90-1.70 (m, 2 H), 1.68-1.54 (m, 3 H), 1.49 (d, J ) 7.1 Hz, 3
H, H-3), 1.38-1.24 (m, 1 H), 0.89 (s, 9 H, t-Bu); 13C NMR (75
MHz, CDCl3) δ 179.1 (CO), 96.6, 76.7, 68.7, 33.7, 29.3, 25.8,
24.9, 18.5, 18.0; EIMS (m/z): M+ calcd for C12H22O4, 230.1518;
found, 230.1511.
Gen er a l P r oced u r e for th e Sa p on ifica tion of Ester s
39a -c. The protected R-hydroxy ester 39 was dissolved in
MeOH (0.15 M) and vigorously stirred at 0 °C. To the stirring
solution was added aq NaOH solution (1.0 M, 7 equiv) over 5
min. The resulting solution was warmed to room temperature
and stirred overnight (ca. 13 h) when the reaction was judged
to be complete by TLC. The contents were cooled to 0 °C and
acidified to pH ) 4 by slowly adding aq HCl solution (1.0 M).
The acidified mixture was poured onto brine and extracted
with EtOAc (3×). The organic layers were combined and
washed with brine, dried over MgSO4, filtered through cotton,
and then concentrated by rotary evaporation and then vacuum
pumped to give the acid 40 as a thick clear oil.
2-[[2-Deoxy-3,4,6-t r is-O-(p h en ylm et h yl)-r-D-a r a bin o-
h exop yr a n osyl]oxy]h exa n oic Acid (40b). A 1:1 mixture of
diastereomers: 1H NMR (300 MHz, CDCl3) (excluding CO2H)
δ 7.39-7.12 (m, 15 H, Ar), 5.05 (d, J ) 2.6 Hz, 0.5 H, H-1′),
5.01 (d, J ) 2.7 Hz, 0.5 H, H-1′), 4.92-4.83 (m, 1 H), 4.71-
4.39 (5 H), 4.29 (t, J ) 6.2 Hz, 0.5 H), 4.21-4.14 (m, 0.5 H),
4.08-3.95 (m, 2 H), 3.82-3.72 (m, 1.5 H), 3.72-3.45 (m, 3.5
H), 2.45 (dd, J ) 13.1, 5.1 Hz, 0.5 H), 2.31 (dd, J ) 13.2, 4.8
Hz, 0.5 H), 1.86-1.68 (m, 3 H), 1.49-1.34 (m, 5 H), 0.99-0.83
(m, 3 H); 13C NMR (APT) (75 MHz, CDCl3) δ 178.3 (+), 177.9
(+), 138.5 (+), 138.3 (+), 138.0 (+), 128.4 (-), 128.16 (-),
128.13 (-), 128.11 (-), 127.98 (-), 127.95 (-), 127.92 (-),
127.87 (-), 127.86 (-), 127.8 (-), 127.72 (-), 127.69 (-), 127.66
(-), 127.6 (-), 98.8 (-), 96.5 (-), 78.1 (-), 77.3 (-), 75.2 (+),
74.9 (+), 73.45 (+), 73.36 (+), 72.0 (+), 71.7 (-), 68.7 (+), 68.5
(+), 35.5 (+), 32.4 (+), 32.0 (+), 27.5 (+), 27.1 (+), 22.4 (+),
22.3 (+), 13.9 (-), 13.84 (-), 13.83 (-).
2-[[2-Deoxy-3,4,6-t r is-O-(p h en ylm et h yl)-r-D-a r a bin o-
h exop yr a n osyl]oxy]-3-m eth ylbu ta n oic Acid (40c). Data
for more polar diastereomer on TLC (EtOAc-hexanes): 1H
NMR (300 MHz, CDCl3) δ 7.37-7.16 (m, 15 H, Ar), 5.03 (d, J
) 2.6 Hz, 1 H, H-1′), 4.89 (d, J ) 10.8 Hz, 1 H, 0.5 PhCH2),
4.72-4.61 (m, 3H, 1.5 PhCH2), 4.52 (m, 2 H, PhCH2), 4.14 (d,
J ) 4.4 Hz, 1 H, H-2), 3.99 (ddd, J ) 11.3, 8.8, 4.8 Hz, 1 H,
H-3′), 3.78 (m, 2 H), 3.70-3.57 (m, 2 H), 2.44 (dd, J ) 12.1,
5.0 Hz, 1 H, H-2′eq), 2.16 (m, 1 H, H-3), 1.77 (m, 1 H, H-2′
ax), 1.01 (d, J ) 6.8 Hz, 3 H, MeCHMe), 0.96 (d, J ) 6.9 Hz,
3 H, MeCHMe); 13C NMR (APT) (75 MHz, CDCl3) δ 176.6 (+),
138.54 (+), 138.47 (+), 138.0 (+), 128.33 (-), 128.28 (-), 128.27
(-), 127.98 (-), 127.95 (-), 127.92 (-), 127.89 (-), 127.86 (-),
127.8 (-), 127.70 (-), 127.67 (-), 127.66 (-), 127.6 (-), 127.54
(-), 127.48 (-), 100.0 (-), 83.2 (-), 78.0 (-), 74.8 (+), 73.3
(+), 72.0 (+), 71.9 (-), 68.5 (+), 35.5 (+), 31.3 (-), 18.5 (-),
17.5 (-).
1,5-An h yd r o-2,7-d id eoxy-6-C-m eth yl-3,4,6-tr is-O-(p h en -
ylm eth yl)-D-a r a bin o-h ep t-1-en itol (63). A dry THF (0.44
M) solution of the triester 62 was added to a 0 °C 3.0 M THF
solution of MeMgCl (7 equiv) over a period of 5 min under Ar.
The resulting solution was allowed to warm to room temper-
ature and stirred for 90 min when TLC (70% EtOAc/hexanes)
showed the reaction to be complete. The reaction was quenched
at 0 °C by the careful addition of 5 mL of saturated aq NH4Cl
solution. The reaction mixture was poured onto 25 mL of H2O
and extracted 4 × 40 mL EtOAc. The organic layers were
combined and washed with 50 mL of brine, dried over Na2-
SO4, filtered, and concentrated. No purification, 87% crude
yield. A dry THF solution (30 mL) of this crude triol (4.05 g,
23.2 mmol) was added to a THF suspension (20 mL) of NaH
(3.35 g, 139 mmol, rinsed with 3 × 10 mL pentane to remove
the oil) under Ar over 10 min. The resulting mixture was
stirred for 15 min at RT. To the reaction mixture was added
Bu4NI (0.43 g, 1.2 mmol) followed by benzyl bromide (13.8 mL,
116 mmol) over a 10 min period. The resulting mixture was
stirred overnight (15 h) when TLC (15% EtOAc-hexanes)
showed that the reaction was complete. The reaction mixture
was poured onto ice (75 mL) and extracted with Et2O (3 ×
100 mL). The combined organic layers were washed with brine
(75 mL), dried over Na2SO4, filtered through cotton, and then
concentrated using rotary evaporation and vacuum pumping.
The crude yellow product was purified by column chromatog-
raphy using 4% EtOAc-hexanes to give 63 as a light yellow
oil, 87% yield. Rf 0.29, 10% EtOAc-hexanes; 1H NMR (300
MHz, CDCl3): δ 7.35-7.20 (m, 15 H, 3 × Ph), 6.44 (d, J ) 4.6
Hz, 1 H, H-1), 4.91 (m, 1 H, H-2), 4.89 (d, 1 ) 11.2 Hz,1 H, 0.5
J . Org. Chem, Vol. 67, No. 17, 2002 6205

Garner et al.
PhCH2), 4.74 (d, 1 ) 11.0 Hz,1 H, 0.5 PhCH2), 4.64 (d, 1 )
11.6 Hz,1 H, 0.5 PhCH2), 4.60-4.52 (m, 3 H, 1.5 PhCH2), 4.31
(dt, J ) 8.6, 1.8 Hz, 1 H, H-3), 3.99 (dd, J ) 9.2, 6.7 Hz, 1 H,
H-4), 3.84 (d, J ) 9.2 Hz, 1 H, H-5), 1.42 (s, 3 H, Me), 1.40 (s,
3 H, Me); 13C NMR (75 MHz, CDCl3): δ 145.2, 139.7, 138.4,
138.3, 128.4, 128.24, 128.17, 128.0, 127.8, 127.6, 127.5, 127.3,
127.0, 100.3, 82.3, 78.3, 76.8, 76.5, 73.9, 70.8, 64.3, 24.6, 23.4;
EIMS (m/z): [M - PhCH2]+ calcd for C22H25O4, 353.1753;
found, 353.1730.
(2S)-2-[[2,7-Did eoxy-6-C-m et h yl-3,4,6-t r is-O-(p h en yl-
m eth yl)-r-D-a r a bin o-h eptopyr an osyl]oxy]pr opan oic Acid,
Eth yl Ester (64). (An adaptation of the procedure described
above for 39 was employed.) Rf 0.47, 25% EtOAc-hexanes; 1H
NMR (300 MHz, CDCl3): δ 7.34-7.19 (m, 15 H, 3 × Ph), 5.07
(br t, J ) 4 Hz, 1 H, H-1′), 4.81-4.58 (m, 6 H, 3 × PhCH2),
4.27 (q, J ) 6.8 Hz, 1 H, CH3CH), 4.18 (m, 2 H, OCH2CH3),
4.00 (ddd, J ) 10.3, 5.9, 4.6 Hz, 1 H, H-3′), 3.87-3.79 (m, 2 H,
H-4′, H-5′), 2.25 (dt, J ) 13.4, 4.3 Hz, 1 H, H-2′ eq), 1.88 (ddd,
J ) 13.9, 9.6, 4.5 Hz, 1 H, H-2′ ax), 1.40-1.38 (m, 9 H, 2 ×
Me, CH3CH), 1.27 (t, J ) 7.1 Hz, 3 H, OCH2CH3); 13C NMR
(75 MHz, CDCl3): δ 172.4, 140.0, 138.7, 138.6, 128.3, 128.1,
127.8, 127.6, 127.5, 127.2, 127.1, 126.8, 95.7, 78.6, 77.4, 76.7,
76.1, 73.3, 71.8, 71.2, 64.3, 60.8, 34.2, 25.1, 23.1, 17.5, 14.2;
EIMS (m/z): calcd [M - PhCH2]+ for C27H35O7, 471.23828;
found, 471.23788; calcd [M - C(CH3)2OCH2Ph]+ for C24H29O6,
413.1964; found, 413.1962.
temperature until the reaction was judged to be complete by
IR analysis (see above).
Ra d ica l Ad d ition s to Meth yl Acr yla te. Methyl acrylate
(29) (5 equiv) was added to the fluorescent yellow suspension
of Barton ester (in most cases, prepared by method A above)
and the mixture irradiated using a 275 W sunlamp at -78 °C
until the Barton ester had been consumed as evidenced by TLC
analysis. The reaction mixture was then filtered through
Celite, and the volatiles removed to provide a residue that was
either processed as described below to determine the kinetic
diastereoselectivity or subjected to flash chromatography to
afford the pure adducts.
Ra n ey Nick el Activa tion . A 20 g amount of Nickel-
aluminum alloy was added to 100 mL of 6 N NaOH over 45
min with vigorous stirring at 0 °C under Ar, when the ice bath
was removed and the mixture was stirred at room temperature
for 4 h. The basic supernatant was then decanted, and the
dark solid was washed with water (20 × 300 mL) and MeOH
(3 × 300 mL), stirring the mixture for about 1 min for each
washing. At the end of this procedure, the catalyst was
suspended in 120 mL of MeOH under Ar.
Ra n ey Nick el Red u ction . The ester was dissolved in
MeOH, the freshly activated suspension of Raney-Ni in MeOH
(ca. 20 mL for 100 mg of the starting carboxylic acid) was
added at room temperature under argon, and the mixture was
vigorously stirred overnight. After TLC analysis indicated that
the starting material was consumed, the mixture was filtered
through a pad of Celite and concentrated under reduced
pressure to give a crude product that was purified by flash
chromatography.
(2S)-2-[[2,7-Did eoxy-6-C-m et h yl-3,4,6-t r is-O-(p h en yl-
m eth yl)-r-D-a r a bin o-h eptopyr an osyl]oxy]pr opan oic Acid
(65). (An adaptation of the procedure described above for 55
1
was employed.) H NMR (300 MHz, CDCl3): δ 7.33-7.18 (m,
15 H, 3 × Ph), 5.11 (t, J ) 4.3 Hz, 1 H, H-1′), 4.73-4.56 (m, 6
H, 3 × PhCH2), 4.32 (q, J ) 6.9 Hz, 1 H, CH3CH), 3.94 (ddd,
J ) 8.4, 5.1, 4.2 Hz, 1 H, H-3′), 3.87-3.77 (m, 2 H, H-4′, H-5′),
2.19 (dt, J ) 13.7, 4.0 Hz, 1 H, H-2′ eq), 1.99 (ddd, J ) 13.3,
8.2, 5.0 Hz, 1 H, H-2′ ax), 1.43 (d, J ) 6.8 Hz, 3 H, CH3CH),
1.36 (s, 3 H, Me), 1.33 (s, 3 H, Me); 13C NMR (APT) (75 MHz,
CDCl3): δ 176.9 (+), 139.8 (+), 138.3 (+), 128.4 (-), 128.2 (-),
128.1 (-), 127.7 (-), 127.4 (-), 127.1 (-), 126.9 (-), 95.9 (-),
77.9 (-), 76.6 (-), 73.1 (+), 71.7 (+), 71.1 (-), 64.4 (+), 33.4
(+), 24.7 (-), 22.9 (-), 17.4 (-); IR (neat): 1745 cm-1 (CdO);
EIMS (m/z): [M - PhCH2]+ calcd for C25H31O7, 444.2070; found
444.2099.
Ba r ton Ester ifica tion Rea ction s. A. DCC P r oced u r e.
To a 0.1 M solution of carboxylic acid and 2-mercaptopyridine-
N-oxide (1.5 equiv) in freshly distilled CH2Cl2 was added DCC
(1.5 equiv) under Ar. The mixture was stirred at RT in the
dark for ca. 1.5 h or until the reaction was judged to be
complete by TLC (50% EtOAc-hexanes).
B. Or igin a l HOTT P r oced u r e. A solution of carboxylic
acid (0.1 M), Et3N, or Hu¨nig’s base (4.0 equiv), and DMAP (0.1
equiv) in dry THF was added to a two-neck flask containing
HOTT (1.5 equiv) in the dark under an Ar atmosphere. This
mixture was stirred in the dark until the reaction was judged
to be complete by IR spectroscopy. When IR monitoring was
not possible (because of overlapping carbonyl bands), the
reaction was allowed to proceed for an arbitrary time and
presumed to be complete if no starting acid was detected after
the trapping step. In general, IR monitoring was performed
by injecting 75 µL of the reaction mixture into an IR solution
cell (NaCl, 0.1 mm gap) in a very dimly lit room. A second
solution cell containing THF was used as a reference. The
bright yellow Barton ester solution was then submitted to one
of the trapping reactions.
C. Im p r oved HOTT P r oced u r e. A dry THF solution of
the acid was added to an aluminum foil-covered flask contain-
ing a dry CH3CN solution of HOTT (1.5 equiv) and DMAP (0.1
equiv) at room temperature under Ar. It was later determined
that DMAP was not necessary for fast and complete Barton
esterification when this mixed solvent system was used. The
final concentration of acid was 0.1 M in (3:1) THF-CH3CN.
Et3N (3 equiv) was added to the mixture and the resulting
solution stirred in the dark for 30 min under Ar at room
Det er m in a t ion of Dia st er eom er R a t ios of Met h yl
Acr yla te Ad d u cts. The diastereoselectivity (ds) of the radical
reactions leading to adducts 37/ent-37, 44, and 60 were
determined using either NMR or HPLC based analytical
techniques. In the case of 37 and ent-37, the crude 2-thiopy-
ridyl ether mixture was filtered through silica gel collecting
all fractions containing material within Rf 0.6 and 0.4 (3/2
hexanes-EtOAc). The resulting mixture was then subjected
to Raney-Ni reduction. The desulfurized products were purified
by flash chromatography taking care to avoid inadvertent
diastereomer enrichment and the diastereoselectivites deter-
1
mined by H NMR.
With 44 and 60, the crude radical reaction mixture was
subjected to “rough” flash chromatography, taking care to
collect all monoadducts (which had identical Rf values in 25%
EtOAc-hexanes). This mixture of primary adducts was sub-
jected to Raney-Ni reduction as described above and the crude
product mixtures analyzed by 1H NMR spectroscopy or reverse
phase HPLC (C-18 Dynamax column, gradient elution using
90% f 20% H2O-MeCN at 1% /min, then 20% f 17% over
20 min, flow rate ) 1 mL/min).
In the case of radical reactions leading to adduct 58, the
following GC conditions were used to determine the diastereo-
selectivity: Column: HP-5 (cross-linked 5% PhMe Silicone);
30 m × 0.32 mm × 2.5 µm film thickness; initial oven temper-
ature: 60 °C for 10 min; injector temperature: 250 °C; ramp:
10 °C/min; final oven temperature: 250 °C for 5 min; detector
temperature: 250 °C; retention times: major diastereomer:
22.3 min, minor diastereomer 21.8 min.
37 a n d en t-37 (X ) SP y): Rf 0.48, 3/1 hexanes-EtOAc; 37
(X ) SPy), [R]22D +21.3° (c 1.2, CHCl3); ent-37 (X ) SPy), [R]22
D
-22.6° (c 0.9, CHCl3); 1H NMR (300 MHz, CDCl3) δ 8.38 (m, 1
H, Py), 7.45 (m, 1 H, Py), 6.95 (m, 1 H, Py), 4.85-4.50 (m, 2
H, H-1′, H-3), 3.9 (m, 2 H, H-1, H-5′), 3.69 (2s, 3 H, CO2Me),
3.4 (m, 1 H, H-5′), 2.31-1.41 (m, 8 H, H-2′, H-3′, H-4′, H-2),
1.28 (d, 3 H, J ) 6.3 Hz, CH3); 13C NMR (75 MHz, CDCl3) δ
173.2 (CO), 157.1 (PyC), 149.5, 149.4, 136.23, 136.20, 122.4,
122.3, 120.1, 120.0 (Py CH), 100.6, 99.8 (anomeric CH), 73.2,
72.7 (CH), 62.8 (CH2), 52.49, 52.45 (CH3), 43.4, 43.2 (CH), 39.4,
39.3, 31.0, 30.9, 25.49, 25.46 (CH2), 22.4, 21.9 (CH3), 20.0, 19.9
(CH2); IR (CHCl3) 3000, 2940, 1730, 1570, 1410, 1120, 1020
6206 J . Org. Chem., Vol. 67, No. 17, 2002

Chiral Auxiliary for Hydroxyalkyl Radicals
cm-1; HRMS m/z calcd for C16H23NO4S [M]+ 325.1348, obsd
325.1343.
Au xilia r y Rem ova l a n d γ-La cton e F or m a tion . The
glucoside 44 (210 mg, 0.38 mmol) was dissolved in MeOH, a
catalytic amount of PPTS added, and the resulting solution
was refluxed overnight. The reaction mixture was then
quenched with satd NaHCO3 solution, extracted with Et2O,
dried (MgSO4), and filtered and the solvent removed to provide
the crude product mixture that was purified by flash chroma-
tography (2/1 petroleum ether-Et2O) to supply the γ-lactone
45 (25 mg, 66%) and methyl glucoside 46 (160 mg, 93%).
[R-(R*,S*)]-4-[(Tetr a h yd r o-2H-p yr a n -2-yl)oxy]pen ta n o-
ic Acid , Meth yl Ester (37, X ) H): Rf 0.59, 4/1 hexanes-
EtOAc; 1H NMR (300 MHz, CDCl3) δ 4.60 (t, J ) 2.7 Hz, 1 H,
H-2′), 3.75 (sextet, J ) 6.0 Hz, 1 H, H-4), 3.67 (s, 3 H, OMe),
3.49 (m, 1 H, H-6′), 2.37 (t, J ) 7.8 Hz, 2 H, H-2), 1.80-1.50
(m, 8 H), 1.24 (d, J ) 6.3 Hz, 3 H, H-5); 13C NMR (75 MHz,
CDCl3) δ 173.2 (CO), 99.8 (anomeric CH), 73.3 (CH), 62.9
(CH2), 51.6 (CH3) 31.5, 31.0, 30.0, 25.5 (CH2), 21.5 (CH3), 20.0
(CH2); IR (CH2Cl2) 2940, 1730, 1430, 1360, 1330, 1190, 1160,
1130, 1060 cm-1; EIMS (m/z): [M - H]+ calcd for C11H19O4,
215.1283; found, 215.1279.
(R)-4-Meth ylbu tyr ola cton e (45). Starting from 44 with
ds ) 5.8/1, [R]25D ) +21.3° (c 0.25, CH2Cl2). 1H NMR (300 MHz,
CDCl3) δ 4.56 (sextet, 1 H, J ) 6.3 Hz, H-4), 2.55 (m, 2 H,
H-2), 2.35 (sextet, 1 H, 6.3 Hz, H-3), 1.84 (m, 1 H, H-3), 1.41
(d, 3 H, J ) 6.3 Hz, Me); IR (neat) 2960, 2920, 1780, 1450
THP -Au xilia r y Telom er Ch a r a cter iza tion : Rf 0.14, 3/1
hexanes-EtOAc; MS (EI) m/z, 411 (10%, [M]+), 326 (30%, [M
- C5H9O]+), 310 (100%, [M - C5H9O2]+), 301 (<10%, [M -
SPy]+), 269 (<10%, [M - C8H14O2]+); HRMS (EI) m/z calcd for
cm-1
.
Meth yl 2-deoxy-3,4,6-tr is-O-(ph en ylm eth yl)-r-D-a r a bin o-
h exop yr a n osid e (46): Rf 0.50, 3/1 hexanes-EtOAc; 1H NMR
(300 MHz, CDCl3) δ 7.35 (15 H, 3 × Ph), 4.92 (d, J ) 10.8 Hz,
1H, 0.5 PhCH2), 4.88 (br d, J ) 2.4 Hz, 1H, H-1′), 4.70-4.50
(5H, 2.5 PhCH2), 4.00 (ddd, J ) 11.5, 9.6, 5.7 Hz, 1H, H-3′),
3.80-3.50 (4H, H-4′, H-5′, 2 × H-6′), 3.34 (s, 3H, OCH3), 2.30
(ddd, J ) 12.6, 5.7, 1.0 Hz, 1H, H-2′ eq), 1.85 (ddd, J ) 12.6,
9.6, 3.0 Hz, 1H, H-2′ ax); 13C NMR (75 MHz, CDCl3) δ 138.6,
138.2, 128.5, 128.42, 128.38, 128.1, 128.0, 127.9, 127.81,
127.78, 127.7, 127.6 (PhCH), 98.6 (anomeric CH), 78.3, 77.7,
(CH), 75.0, 73.5, 71.8 (CH2), 70.7 (CH), 69.0 (CH2), 54.7 (CH3),
35.5 (CH2); IR (CHCl3) 3000, 2920, 2840, 1490, 1445, 1355
cm-1; EIMS (m/z): M+ calcd for C28H32O5, 448.2250; found,
448.2240.
C
20H29NO6S [M]+ 411.1716, obsd 411.1717.
(S)-4-Meth ylbu tyr ola cton e (en t-45). To a 0.2 M solution
of ent-37 (X ) H) in MeOH was added a catalytic amount of
PPTS, and the stirred mixture was refluxed overnight at which
point TLC (4/1 hexanes-EtOAc) analysis revealed that the
starting material was consumed. The mixture was poured into
a saturated aqueous solution of NaHCO3 and extracted five
times with EtOAc. The organic layer was washed with brine,
dried with MgSO4, and concentrated by rotary evaporation,
keeping the bath at 3 °C because the lactone is volatile. The
crude mixture was purified by flash chromatography. The
separation was monitored by 1H NMR since the product could
not be detected by TLC. Yield ) 55%. Starting from ent-37 (X
[2S-[2r(S*),6â]]-4-[[6-(1,1-Dim et h ylet h yl)t et r a h yd r o-
2H-p yr a n -2-yl]oxy]p en ta n oic Acid Meth yl Ester (58): Rf
) H) with ds ) 6/1, [R]25 ) -21.4° (c 0.70, CH2Cl2). (Spectral
D
1
data reported below for compound 45.)
0.67, 3/1 hexanes-EtOAc; H NMR (300 MHz, CDCl3) δ 4.89
(s, 1 H, H-2′), 3.79 (sextet, J ) 6.1 Hz, 1 H, H-4), 3.67 (s, 3 H,
OMe), 3.42 (dd, J ) 12.1, 1.7 Hz, 1 H, H-6′), 2.35 (t, J ) 7.7
Hz, 2 H, H-2), 1.84-1.70 (m, 3 H), 1.62-1.52 (m, 4 H), 1.35-
1.20 (m + d, J ) 6.3 Hz, 4 H, H-5), 0.88 (s, 9 H, t-Bu); 13C
NMR (APT) (75 MHz, CDCl3) δ 174.1 (+) (CO), 96.5 (-), 75.9
(-), 71.0 (-), 51.5 (-), 33.8 (+), 30.8 (+), 30.1 (+), 29.7 (+),
25.9 (-), 25.1 (+), 20.9 (-), 18.3 (+); EIMS (m/z): [M + H]+
calcd for C15H29O4, 273.2064; found, 273.2068; M+ calcd for
3,5-Did eoxy-4-O-[2-d eoxy-3,4,6-tr is-O-(p h en ylm eth yl)-
r-D-a r a bin o-h exop yr a n osyl]-2-S-(2-p yr id in yl)-2-th iop en -
ta n oic Acid , Meth yl Ester (43a ): Rf 0.65, 2/1 hexanes-
1
EtOAc; H NMR (300 MHz, CDCl3) δ 8.48 (m, 1 H, Py), 7.48
(m, 1 H, Py), 7.39-7.14 (m, 16 H, Py, 3 × Ph), 7.10 (m, 1H,
Py), 5.10 (br d, 1 H, J ) 2.7 Hz, H-1′), 4.95-4.48 (m, 7 H, 3 ×
PhCH2, H-2), 4.07-3.60 (m, 9 H, H-3′, H-4′, H-5′, H-6′, H-1,
CO2Me), 2.40-1.40 (m, 4 H, H-2′ eq, H-2′ ax, 2 × H-2), 1.28
(d, 3 H, J ) 6.0 Hz, CH3); 13C NMR (75 MHz, CDCl3) δ 173.2,
172.9 (CO), 157.0 (PyC), 149.53, 149.47 (CH), 138.86, 138.60,
138.20 (C), 136.33, 136.25 (PyCH), 128.51, 128.46, 128.39,
128.35, 128.13, 128.0, 127.93, 127.89, 127.8, 127.7, 127.6
(ArCH), 122.5, 122.4, 120.2, 120.1 (PyCH), 99.3, 98.6 (anomeric
CH), 78.4, 77.7 (CH), 75.1, 75.0 (CH2), 73.8 (CH), 73.53 (CH2),
73.50 (CH), 71.8, 71.6 (CH2), 71.1, 71.0 (CH), 68.99, 68.96
(CH2), 52.6, 52.5 (CH3), 43.3, 43.1 (CH), 39.4, 39.3, 35.8, 35.7
(CH2), 22.0, 21.6 (CH3); IR (CHCl3) 3000, 2910,1750, 1570,
1450, 1410, 1020 cm-1; EIMS (m/z): M+ calcd for C38H43NO7S,
657.2760; found, 657.2742.
C
15H28O4, 272.1988; found, 272.1835.
2-[(2S)-2-[[(2R,6R)-6-(1,1-Dim et h ylet h yl)t et r a h yd r o-
2H-p yr a n -2-yl]oxy]p r op yl]p en ta n ed ioic Acid Dim eth yl
Ester (ter t-Bu tyl-Au xilia r y Telom er ). Mixture of diaster-
eomers (1:1); Rf 0.40, 4/1 hexanes-EtOAc; 1H NMR (300 MHz,
CDCl3) δ 4.88 (br d, J ) 2.2 Hz, 1 H, H-2′), 4.83 (br d, J ) 2.2
Hz, H-2′, 1 H), 3.71 (sextet, J ) 5.6 Hz, 1 H, H-6), 3.67 (s, 2 ×
OMe), 3.41 (dd, J ) 11.9, 2.5 Hz, 1 H, H-6′), 2.61-2.42 (m, 1
H), 2.32 (t, J ) 7.2 Hz, 2 H), 2.02-1.66 (m, 8 H), 1.66-1.41
(m, 10 H), 1.31-1.19 (m + d, J ) 6.4 Hz, 10 H), 0.87 (s), 0.86
(s); 13C NMR (75 MHz, CDCl3) δ 176.0, 175.7, 173.2, 97.3, 96.6,
77.1, 75.9, 71.1, 70.9, 51.53, 51.48, 41.6, 40.7, 38.8, 38.7, 33.8,
33.7, 31.6, 31.5, 30.2, 30.0, 29.6, 29.3, 27.8, 27.7, 25.9, 25.8,
25.1, 21.44, 21.37, 18.2; EIMS (m/z): [M - OCH3]+ calcd for
Glu ca l-Au xilia r y Telom er Ch a r a cter iza tion : Rf 0.16,
3/1 hexanes-EtOAc; MS (EI) m/z, 411 (40%, [M]+), 328 (100%,
[M - C27H29O4 + 2H]+), 326 (<10%, [M - C27H29O4]+), 310
(<10%, [M - C27H29O5]+); HRMS (EI) m/z calcd for C42H49NO9S
[M]+ 743.3128, obsd 743.3070.
C
18H31O5, 327.2171; found, 327.2176.
Ra d ica l Ad d ition s to 2-Nitr op r op en e. The reaction flask
containing the preformed Barton ester (generally prepared
using procedure “A”) was cooled to the desired temperature,
and 2-nitropropene (3 equiv) was added. The resulting mixture
was irradiated with a 275 W sunlamp until the intermediate
Barton ester was consumed as judged by TLC. The reaction
mixture was then filtered through ca. 1 cm of Celite and the
resulting filter cake washed with CH2Cl2. The volatiles were
removed to give the crude adduct as an orange oil that was
filtered through a column of silica gel eluting with 20%
EtOAc-hexanes to remove residual dicyclohexylurea and other
baseline material.
Bu ffer ed Red u ctive Hyd r olysis (Keton e F or m a tion ).
To a THF solution (ca. 0.4 M) of the primary radical adducts
was added freshly prepared 15% aq TiCl3 solution (25-30
equiv) buffered to pH 5-6 using NH4OAc (6 equiv). The
reaction mixture was heated to ca. 48 °C and stirred under
(4R)-4-[[2-Deoxy-3,4,6-t r is-O-(p h en ylm et h yl)-r-D-a r a -
bin o-h exop yr a n osyl]oxy]p en ta n oic Acid , Meth yl Ester
(44): Rf 0.48, 3/1 hexanes-EtOAc; [R]25D + 52.1° (c 0.83, CH2-
1
Cl2); H NMR (300 MHz, CDCl3) δ 7.40-7.10 (m, 15 H, 3 ×
Ph), 5.04 (br d, J ) 3.0 Hz, 1 H, H-1′), 4.91 (d, J ) 10.8 Hz, 1
H, 0.5 PhCH2), 4.65 (m, 3 H, 1.5 PhCH2), 4.50 (m, 2 H, PhCH2),
3.98 (ddd, J ) 11.5, 9.4, 5.1 Hz, 1 H, H-3′), 3.90-3.60 (m, 8 H,
H-4′, H-5′, 2 × H-6′, H-4, CO2Me), 2.36 (t, J ) 7.6 Hz, 2 H,
H-2), 2.25 (dd, J ) 13.5, 5.1 Hz, 1 H, H-2′ eq), 1.87-1.67 (m,
3 H, H-3, H-2′ ax), 1.20 (d, J ) 6.3 Hz, 3 H, Me); 13C NMR (75
MHz, CDCl3) δ 174.1 (CO), 138.8, 138.5, 138.2 (C), 128.4, 128.1,
127.9, 127.7, 127.6 (PhCH), 97.7 (anomeric CH), 78.4, 77.8
(CH), 75.1 (CH2), 70.9 (CH), 68.9 (CH2), 51.7 (CH3), 35.9, 31.5,
30.1 (CH2), 21.2 (CH3); IR (CHCl3) 3000, 1725, 1530, 1460, 1430
cm-1; HRMS-FAB (m-nitrobenzyl alcohol, m/z): M+ calcd for
C
33H40O7, 548.2774; found, 548.2771.
J . Org. Chem, Vol. 67, No. 17, 2002 6207

Garner et al.
Ar overnight. The reaction was monitored by TLC (40%
EtOAc-hexanes) and IR spectroscopy to determine CdO
formation (1710 cm-1) and -NO2 disappearance (1550 cm-1).
When the reaction was judged complete, the mixture was
poured into water and extracted with Et2O then EtOAc. The
combined organic layers were washed with sat aq NaHCO3
and brine, dried with Na2SO4, filtered, and concentrated to
give the O-protected aldol as an oily residue which was purified
by flash column chromatography (SiO2) eluting with EtOAc-
hexanes.
(4R)-4-[[2-Deoxy-3,4,6-t r is-O-(p h en ylm et h yl)-r-D-a r a -
bin o-h exop yr a n osyl]oxy]-2-p en ta n on e (48a ): Rf 0.44, 3/2
hexanes-EtOAc; 1H NMR (300 MHz, CDCl3) δ 7.35-7.15 (m,
15 H, 3 × Ph), 5.1 (br d, J ) 2.6 Hz, 1 H, H-1′), 4.9 (d, J )
10.8 Hz, 1 H, 0.5 PhCH2), 4.68-4.48 (m, 5 H, 2.5 PhCH2), 4.15
(sextet, J ) 6.1 Hz, 1 H, H-4), 3.95 (ddd, J ) 13.8, 8.9, 5.1 Hz,
1 H, H-3′), 3.85-3.57 (m, 4 H), 2.70 (dd, J ) 16.1, 7.7 Hz, 1H,
H-3), 2.40 (dd, J ) 16.1, 4.9 Hz, 1H, H-3), 2.18-2.11 (m + s,
4 H, H-2′eq, COCH3), 1.69 (ddd (apparent td), J ) 12.3, 12.3,
3.7 Hz, 1 H, H-2′ ax), 1.20 (d, J ) 6.4 Hz, 3 H, CHCH3); 13C
NMR (APT) (75 MHz, CDCl3) δ 207.0 (+, CO), 138.7 (+), 138.5
(+), 138.2 (+), 128.3 (-), 128.0 (-), 127.9 (-), 127.8 (-), 127.70
(-), 127.67 (-), 127.64 (-), 127.61 (-), 127.5 (-), 98.1 (-,
anomeric C), 78.3 (-), 77.5 (-), 75.0 (+), 73.5 (+), 71.7 (+),
71.4 (-), 71.1 (-), 68.9 (+), 50.6 (+), 35.9 (+), 31.1 (-), 31.0
(-), 21.6 (-); EIMS (m/z): M+ calcd for C32H38O6, 518.2668;
found, 518.2767.
(-), 127.9 (-), 127.83 (-), 127.80 (-), 127.76 (-), 127.7 (-),
127.63 (-), 127.60 (-), 127.57 (-), 127.5 (-), 98.4 (-, anomeric
C), 79.8 (-), 78.4 (-), 77.5 (-), 75.0 (+), 73.4 (+), 71.6 (+),
71.4 (-), 68.9 (+), 43.6 (+), 36.0 (+), 31.7 (-), 18.5 (-), 17.0
(-); EIMS (m/z): [M - H]+ calcd for C34H42O6, 545.2903; found,
545.2597.
Ald ol Dep r otection . To a 0.1 M solution of O-protected
aldol 48a in dry CH2Cl2 at -78 °C were added thiophenol (5
equiv) and BF3‚OEt2 (1.5 equiv) under Ar. The reaction
mixture was stirred for 2 h at -78 °C after which the cooling
bath was removed allowing the reaction mixture to warm to
room temperature and stirred for 1 h. When the reaction was
complete, as judged by TLC, the contents were poured into
sat aq NaHCO3 and extracted with CH2Cl2. The organic layers
were combined and dried with Na2SO4, filtered, and then
concentrated to an oil which was purified by gradient flash
chromatography (10% f 50% f 100% EtOAc-hexanes) to give
the thioglycoside 51 and then the deprotected aldol 50a .
Glu ca l Refor m a tion . To a 0.02 M solution of 51 in 3:1:1
MeOH-THF-H2O at 0 °C was added 2 equiv of OXONE with
stirring. When judged complete by TLC, the reaction was
quenched with 10% NaHSO3 solution and extracted with
EtOAc. The organic layers were combined and washed with
sat aq NaHCO3 and brine, dried over Na2SO4, and then
concentrated. The residue was dissolved in dry benzene to
make a 0.06 M solution which was heated to reflux until TLC
showed depletion of the intermediate sulfoxide 52 after ca. 90
min. The reaction contents were cooled to room temperature
and then poured into sat aq NaHCO3 and extracted with
EtOAc. The organic layers were combined and washed with
brine, dried over Na2SO4, and then concentrated. Purification
of tri-O-benzyl-D-glucal (38) was effected by flash chromatog-
raphy (11% EtOAc-hexanes).
(()-4-[[2-Tetr a h yd r o-2H-p yr a n -2-yl]oxy]-2-p en ta n on e
(49). Inseparable (3:1) mixture of diastereomers, data for major
diastereomer: Rf 0.45, 3/2 hexanes-EtOAc; 1H NMR (300
MHz, CDCl3): δ 4.67 (m, 1 H, H-2′), 4.20 (sextet, J ) 6.2 Hz,
1 H, H-4), 3.90 (m, 1 H, H-6′), 3.50 (m, 1 H, H-6′), 2.75 (dd, J
) 15.9, 7.6 Hz, H-3), 2.44 (dd, J ) 15.9, 5.0 Hz, H-3), 2.16 (s,
3 H, H-1), 1.85-1.62 (m, 2 H), 1.58-1.45 (m, 4 H), 1.27 (d, J
(4R)-4-Hyd r oxy-2-p en ta n on e (50a ) (from a 4.8/1 mixture
of 48a /d ia -48a ): Rf 0.28, 3/2 hexanes-EtOAc; [R]20 -32° (c
1
J
) 6.3 Hz, 3 H, H-5); Diagnostic H NMR (300 MHz, CDCl3)
0.48, CHCl3) corresponds to 51% ee (3/1 de); lit.42 [R]25 -60°
J
data for minor diastereomer: δ 2.19 (s, 1 H, H-1), 1.17 (d, J )
6.2 Hz, 1 H, H-5); 13C NMR (APT) (75 MHz, CDCl3): δ 207.3
(+), 99.5 (-), 96.5 (-, minor diastereomer), 70.8 (-), 68.4 (-,
minor), 62.8 (+), 62.6 (+, minor), 51.2 (+, minor), 50.6 (+),
31.9 (+), 25.3 (+), 21.9 (-), 19.9 (+), 19.8 (+, minor), 19.7 (-);
EIMS (m/z): M+ calcd for C10H18O3, 186.1256; found, 186.1245.
(95% ee); 1H NMR (300 MHz, CDCl3) δ 4.20 (m, 1 H, H-4),
3.10 (br, 1 H, OH), 2.64 (dd, J ) 17.6, 3.3 Hz, 1 H, H-3), 2.55
(dd, J ) 17.6, 8.8 Hz, 1 H, H-3), 2.18 (s, 3 H, H-1), 1.20 (d, J
) 7.0 Hz, 3 H, CHCH3); EIMS (m/z): M+ calcd for C5H10O2,
102.0681; found, 102.0667.
(4R)-4-H yd r oxy-2-oct a n on e (50b ) (from >95% pure
48b): Rf 0.16, 3/1 hexanes-EtOAc; [R]25D -36° (c 0.81, CHCl3)
(>90% ee according to Mosher ester study); 1H NMR (300 MHz,
CDCl3) δ 4.03 (m, 1 H, H-4), 3.00 (br s, 1 H, OH), 2.63 (dd, J
) 17.6, 3.1 Hz, 1 H, H-3), 2.54 (dd, J ) 17.6, 8.8 Hz, 1 H, H-3),
2.18 (s, 3 H, COCH3), 1.53-1.26 (m, 6 H, 3 × CH2), 0.90 (t, J
) 7.1 Hz, 3 H); EIMS (m/z): M+ calcd for C8H16O2, 144.1150;
found, 144.1136.
(4R)-4-[[2-Deoxy-3,4,6-t r is-O-(p h en ylm et h yl)-r-D-a r a -
bin o-h exop yr a n osyl]oxy]-2-octa n on e (48b): Rf 0.47, 3/2
hexanes-EtOAc; 1H NMR (300 MHz, CDCl3) δ 7.36-7.16 (m,
15 H, 3 × Ph), 5.10 (br d, J ) 2.9 Hz, 1 H, H-1′), 4.88 (d, J )
10.8 Hz, 1 H, 0.5 PhCH2), 4.68-4.48 (m, 5 H, 2.5 PhCH2), 4.05
(quintet, J ) 6.2 Hz, 1 H, H-4), 3.91 (ddd, J ) 13.5, 8.7, 4.8
Hz, 1 H, H-3′), 3.87-3.57 (m, 4 H), 2.65 (dd, J ) 15.9, 7.5 Hz,
1 H, H-3), 2.46 (dd, J ) 15.9, 4.8 Hz, 1 H, H-3), 2.18-2.13 (s
+ m, 4 H, COCH3, H-2′eq), 1.70 (ddd (apparent td), J ) 12.3,
12.3, 3.8 Hz, 1 H, H-2′ ax), 1.64-1.46 (m, 2 H, 2/9 n-Bu), 1.35-
1.20 (m, 4 H, 4/9 n-Bu), 0.90-0.80 (m, 3 H, 3/9 n-Bu); 13C NMR
(APT) (75 MHz, CDCl3) δ 207.3 (+, CO), 138.7 (+), 138.4 (+),
138.1 (+), 128.3 (-), 128.1 (-), 128.02 (-), 127.99 (-), 127.96
(-), 127.86 (-), 127.83 (-), 127.81 (-), 127.7 (-), 127.64 (-),
127.61 (-), 127.58 (-), 127.55 (-), 127.5 (-), 97.5 (-, anomeric
C), 78.3 (-), 77.5 (-), 75.0 (+), 74.8 (-), 73.4 (+), 71.6 (+),
71.3 (-), 68.8 (+), 47.9 (+), 35.9 (+), 34.9 (+), 31.1 (-), 27.4
(+), 22.6 (+), 13.9 (-); EIMS (m/z): [M - H]+ calcd for
P h en yl 2-Deoxy-3,4,6-tr is-O-(p h en ylm eth yl)-1-th io-D-
a r a bin o-h exop yr a n osid e (51). A 2/1 mixture of anomers:
Rf 0.67, 3/2 hexanes-EtOAc; 1H NMR (300 MHz, CDCl3) δ
7.54-7.19 (m), 5.68 (d, J ) 4.9 Hz), 4.91 (d, J ) 10.9 Hz, 0.5
H), 4.90 (d, J ) 10.9 Hz, 0.5 H), 4.76-4.43 (m, 0.5 H), 4.30
(dd, J ) 7.9, 1.8 Hz, 0.5 H), 3.95 (ddd, J ) 13.3, 8.7, 4.8 Hz,),
3.85-3.50 (m, 3 H), 2.49-2.42 (m, 1 H), 2.18-2.08 (m, 1 H),
1.83-1.52 (m, 2 H), 1.30-1.26 (m, 1 H); 13C NMR (APT) (75
MHz, CDCl3) δ 138.43 (+), 138.36 (+), 138.2 (+), 138.1 (+),
135.0 (+), 134.1 (+), 131.4 (-), 131.24 (-), 131.21 (-), 131.18
(-), 131.16 (-), 129.1 (-), 129.01 (-), 128.98 (-), 128.95 (-),
128.92 (-), 128.89 (-), 128.87 (-), 128.7 (-), 128.61 (-), 128.58
(-), 128.55 (-), 128.52 (-), 128.49 (-), 128.46 (-), 128.45 (-),
128.37 (-), 128.35 (-), 128.3 (-), 128.1 (-), 128.04 (-), 128.01
(-), 127.98 (-), 127.92 (-), 127.90 (-), 127.86 (-), 127.8 (-),
127.73 (-), 127.70 (-), 127.49 (-), 127.48 (-), 127.32 (-),
127.30 (-), 127.2 (-), 127.0 (-), 84.0 (-), 82.1 (-), 80.7 (-),
79.3 (-), 78.4 (-), 77.94 (-), 77.90 (-), 77.88 (-), 75.0 (+),
73.4 (+), 71.9 (+), 71.7 (-), 69.5 (+), 68.9 (+), 36.9 (+), 36.3 (+).
[2S-[2r(S*),6â]]-4-[[6-(1,1-Dim et h ylet h yl)t et r a h yd r o-
2H-p yr a n -2-yl]oxy]-2-p en ta n on e (57): Rf 0.47, 4/1 hex-
C
35H44O6, 559.3060; found, 559.2977.
(4R)-4-[[2-Deoxy-3,4,6-t r is-O-(p h en ylm et h yl)-r-D-a r a -
bin o-h exop yr a n osyl]oxy]-5-m eth yl-2-h exa n on e (48c): Rf
0.47, 3/2 hexanes-EtOAc; 1H NMR (300 MHz, CDCl3) δ 7.35-
7.17 (m, 15 H, 3 × Ph), 5.10 (d, J ) 3.2 Hz, 1 H, H-1′), 4.88 (d,
J ) 10.8 Hz, 1 H, 0.5 PhCH2), 4.66-4.47 (m, 5 H, 2.5 PhCH2),
3.93 (m, 1 H, H-4), 3.90-3.74 (m, 3 H), 3.68-3.55 (m, 2 H),
2.60 (dd, J ) 16.1, 8.7 Hz, 1 H, H-3), 2.38 (dd, J ) 16.1, 3.5
Hz, 1 H, H-3), 2.14 (s + m, 4 H, COCH3, H-2′eq), 1.99 (m, 1 H,
H-5), 1.69 (dt, J ) 12.2, 12.2, 3.7 Hz, 1 H, H-2′ ax), 0.88 (d, J
) 6.9 Hz, 3 H, MeCHMe), 0.82 (d, J ) 6.9 Hz, 3 H, MeCHMe);
13C NMR (APT) (75 MHz, CDCl3) δ 207.7 (+, CO), 138.7 (+),
138.4 (+), 138.2 (+), 128.3 (-), 128.1 (-), 128.02 (-), 127.99
(42) See ref 29.
6208 J . Org. Chem., Vol. 67, No. 17, 2002

Chiral Auxiliary for Hydroxyalkyl Radicals
anes-EtOAc, [R]22 -108.5° (c 2.67, CHCl3), [R]22 -113.4° (c
diastereomer) 0.38, 25% EtOAc-hexanes; 1H NMR (300 MHz,
CDCl3): δ 7.35-7.19 (m, 15 H, 3 × Ph), 5.12 (t, J ) 3.4 Hz,
1 H, H-1′), 4.79 (d, J ) 10.7 Hz, 1 H, 0.5 PhCH2), 4.65-4.55
(m, 5 H, 2.5 PhCH2), 4.30 (q, J ) 7.1 Hz, 2 H, OCH2CH3), 4.24
(m, 1 H, H-4), 3.90 (ddd, J ) 10.4, 6.4, 4.4 Hz, 1 H, H-3′), 3.81-
3.70 (m, 2 H, H-4′, H-5′), 3.07 (dd, J ) 16.7, 6.7 Hz, 1 H, H-3),
2.92 (dd, J ) 16.8, 5.5 Hz, 1 H, H-3), 2.08 (dt, J ) 13.2, 4.2
Hz, 1 H, H-2′ eq), 1.79 (ddd, J ) 13.9, 9.6, 4.4 Hz, 1 H, H-2′
ax), 1.41-1.33 (m, 9 H, 2 × Me, OCH2CH3), 1.26 (d, J ) 6.3
Hz, 3 H, H-5); 13C NMR (75 MHz, CDCl3): δ 192.7, 168.9,
140.0, 138.5, 128.3, 128.12, 128.05, 127.7, 127.6, 127.4, 127.2,
127.1, 126.8, 96.7, 78.5, 77.3, 76.8, 76.2, 73.4, 71.5, 69.7, 64.4,
62.5, 45.8, 34.2, 25.1, 23.3, 21.7, 13.9; EIMS (m/z): M+ calcd
for C36H44O8, 604.3036; found, 604.3039.
(4R)-4-[[2-Deoxy-3,4,6-t r is-O-(p h en ylm et h yl)-r-D-a r a -
bin o-h exop yr a n osyl]oxy]-2-oxop en t a n oic Acid , E t h yl
Ester (70): A 2/1 mixture of diastereomers; Rf 0.59, 3/2
hexanes-EtOAc; 1H NMR (300 MHz, CDCl3): δ 7.39-7.12 (m,
15 H), 5.13 (d, J ) 3.0 Hz, H-1′), 5.09 (d, J ) 3.2 Hz, H-1′),
4.88 (d, J ) 10.9 Hz), 4.86 (d, J ) 10.7 Hz), 4.70-4.43 (m, 6
H), 4.36-4.19 (m, 3 H), 4.00-3.55 (m, 5 H), 3.19 (dd, J ) 17.1,
7.3 Hz), 3.09 (dd, J ) 17.1, 7.7 Hz), 2.88 (d, 0.5 H), 2.82 (d, 0.5
H), 2.19 (dd, 1 H), 1.82-1.62 (m, 1 H)), 1.35 (t, J ) 7.1 Hz),
1.26 (d, J ) 6.3 Hz), 1.20 (d, 6.1 Hz); EIMS (m/z): M+ calcd
for C34H40O8, 576.2723; found, 576.2706.
D
J
2.67, CHCl3); 1H NMR (300 MHz, CDCl3) δ 4.94 (s, 1 H, H-2′),
4.20 (sextet, J ) 6.3 Hz, 1 H, H-4), 3.35 (dd, J ) 11.5, 1.7 Hz,
1 H, H-6′), 2.71 (dd, J ) 15.5, 6.0 Hz, 1 H, H-3), 2.45 (dd, J )
15.5, 6.3 Hz, 1 H, H-3), 2.16 (s, 3 H, H-1), 1.84-1.65 (m, 1 H),
1.63-1.53 (m, 4 H), 1.34-1.20 (m + d, J ) 6.3 Hz, 4 H, H-5),
0.88 (s, 9 H, t-Bu); 13C NMR (APT) (75 MHz, CDCl3) δ 207.5
(+) (CO), 96.5 (-), 76.0 (-), 68.8 (-), 50.3 (+), 33.8 (+), 31.0
(-), 30.2 (+), 25.9 (-), 25.1 (+), 21.7 (-), 18.2 (+); EIMS (m/
z): [M - H]+ calcd for C14H25O3, 241.1804; found, 241.1809;
[M - t-Bu]+ calcd for C10H18O3, 185.1178; found, 185.1180.
(4S)-4-[(1,10,10-Tr im eth yl-3-oxa tr icyclo[5.2.1.02,6]d ec-
4-yl)oxy]-2-p en ta n on e (60): Rf 0.46 (minor diastereomer),
0.52 (major), 7/3 hexanes-EtOAc; 1H NMR (300 MHz, CDCl3),
major diastereomer: δ 5.39 (d, J ) 4.9 Hz, 1 H, anomeric-H),
4.25-4.18 (m, 2 H), 2.89 (m, 1 H), 2.72 (dd, J ) 15.3, 7.6 Hz,
1 H, H-3), 2.43 (dd, J ) 15.3, 5.3 Hz, 1 H, H-3), 2.17 (s, 3 H,
H-1), 1.88 (m, 1 H), 1.72-1.49 (m, 4 H), 1.40-1.22 (m, 2 H),
1.17 (d, J ) 6.2 Hz, 3H, H-5), 0.97 (s, 3 H, Me), 0.91 (s, 3 H,
Me), 0.86 (s, 3 H, Me); 13C NMR (75 MHz, CDCl3): δ 207.5,
106.3, 89.5, 68.4, 52.3, 51.6, 48.7, 47.5, 40.2, 32.9, 26.6, 20.9,
20.6, 20.0, 18.8, 14.8; EIMS (m/z): M+ calcd for C17H28O3,
280.2038; found, 280.1980; [M - CH3CHOCH2COCH3]+ calcd
for C12H19O, 195.1385; found, 195.1387.
Ra d ica l Ad d ition to Eth yl r-Tr iflu or oa cetoxya cr yla te.
To the preformed Barton ester solution (generally prepared
from substrates 40 and 66 using procedure “C”) was added
trap 31 (3-10 equiv) and the resulting solution cooled to 0
°C. The aluminum foil covering was removed and placed
directly under the flask to reflect the light from the sunlamp.
The stirred reaction mixture was photolyzed for 25-30 min
using a 275 W sunlamp. The reaction was monitored by TLC
(3/2 hexanes-EtOAc). Upon completion of the photolysis
(usually 25 min), irradiation was stopped and the solvent
removed via rotary evaporation. To facilitate the determination
of diastereomer ratios, the crude product was filtered through
a small plug of SiO2 (3-5 cm high × 1.5-3 cm wide), eluting
with 50-100 mL 3/2 hexanes-EtOAc. The crude product was
purified by column chromatography using EtOAc-hexanes.
(4R)-4-[[2,7-Did eoxy-6-C-m et h yl-3,4,6-t r is-O-(p h en yl-
m eth yl)-r-D-a r a bin o-h ep top yr a n osyl]oxy]-2-oxop en ta n -
oic Acid , E t h yl E st er (69): 90% yield (ds ) 10/1). Rf
(major diastereomer) 0.32, 3/1 hexanes-EtOAc; Rf (minor
Ack n ow led gm en t. This work was supported by
grants from the National Institutes of Health and
the National Science Foundation. Thanks to the
Ohio Supercomputing Center for an allocation of CRAY
time.
Su p p or tin g In for m a tion Ava ila ble: Computational de-
tails for the radicals in Figure 4 and transition structures in
Figure 7; experimental details for synthesis of lactol 53 (the
synthesis of compound 62 may be found in the Supporting
1
Information of ref 39), and H NMR spectra of compounds 37,
40b, 40c, 48a , 48b, 48c, 49, 53, 54, 55, 57, 60, 63, 64, 65, and
69. This information is available free of charge via the Internet
at http://pubs.acs.org.
J O010880F
J . Org. Chem, Vol. 67, No. 17, 2002 6209