Origin of stereocontrol in the construction of the 12-oxatricyclo[6.3.1. 02,7]dodecane ring system by Prins-pinacol reactions

Article abstract of DOI:10.1021/jo9024144

(Chemical Equation Presented) Polycyclic products containing the 12-oxatricyclo[6.3.1.0^2,7]dodecane moiety having either the trans (8a-e) or cis (9a-e) relative configuration of the oxacyclic bridge and the cis angular substituents are formed s

Full text of DOI:10.1021/jo9024144

pubs.acs.org/joc
Origin of Stereocontrol in the Construction of the
2
,7
1
2-Oxatricyclo[6.3.1.0 ]dodecane Ring System by Prins-Pinacol
Reactions
Larry E. Overman* and Paul S. Tanis
Department of Chemistry, 1102 Natural Sciences II, University of California, Irvine,
California 92697-2025
leoverma@uci.edu
Received November 11, 2009
2
,7
Polycyclic products containing the 12-oxatricyclo[6.3.1.0 ]dodecane moiety having either the
trans (8a-e) or cis (9a-e) relative configuration of the oxacyclic bridge and the cis angular
substituents are formed stereospecifically by Prins-pinacol cyclizations of unsaturated R-dithianyl
acetals 14a-e or 15a-e. These results show that the topography (boat or chair) of the Prins
cyclization of the sulfur-stabilized oxocarbenium ions generated from acetals 14a-e or 15a-e is
controlled by the stereoelectronic influence of the allylic substituents, with steric effects playing a
minor role. A complex molecular rearrangement that is terminated by a thio-Prins-pinacol reaction
is also identified.
Introduction
carbocation produced by a Prins cyclization is terminated
2
by a pinacol shift, has proven to be useful in the stereo-
selective synthesis of complex carbocyclic and heterocyclic
3
structures.
The efficient construction of complex molecular struc-
tures from simple materials is one of the primary goals
of synthetic organic chemistry. A particularly expeditious
way to build complexity is to combine two or more dis-
tinct steps into a single transformation. One such pro-
cess, the Prins-pinacol cascade reaction wherein the
We described in 2001 a stereoelectronic model that
allowed the prediction of the stereochemical outcome
of Prins-pinacol syntheses of substituted tetrahydrofur-
1
4
ans. Stereoselectivity in the synthesis of carbocycles has
5
been explained with similar models. However, these
(
1) For selected recent reviews, see: (a) Tietze, L. F.; Brasche, G.;
Gericke, K. M. Domino Reactions in Organic Synthesis; Wiley-VCH:
Weinheim, Germany, 2006. (b) Nicolaou, K. C.; Edmonds, D. J.; Bulger, P. G.
Angew. Chem., Int. Ed. 2006, 45, 7134–7186. (c) Pellissier, H. Tetrahedron
models have only been applied to Prins-pinacol reac-
tions occurring by generally favored chairlike transition
6
structures, with the caveat that steric interactions could
2
006, 62, 1619–1665. (d) Pellissier, H. Tetrahedron 2006, 62, 2143–2173. (e)
M €u ller, T. J. J., Ed. Metal Catalyzed Cascade Reactions. In Topics in
Organometallic Chemistry; Springer-Verlag: Berlin, Germany, 2006;
Vol 19. (f) Chapman, C. J.; Frost, C. G. Synthesis 2007, 1–21. (g) Enders,
D.; Grondal, C.; Huettl, M. R. M. Angew. Chem., Int. Ed. 2007, 46, 1570–
(3) For recent reviews, see: (a) Overman, L. E.; Pennington, L. D. J. Org.
Chem. 2003, 68, 7143–7157. (b) Overman, L. E. Tetrahedron 2009, 65, 6432–
6446.
1
581. (h) Walji, A. M.; MacMillan, D. W. C. Synlett 2007, 1477–1489. (i)
Miyabe, H.; Takemoto, Y. Chem.;Eur. J. 2007, 13, 7280–7286. (j) Padwa,
A.; Bur, S. K. Tetrahedron 2007, 63, 5341–5378. (k) Kirsch, S. F. Synthesis
(4) Cohen, F.; MacMillan, D. W. C.; Overman, L. E.; Romero, A. Org.
Lett. 2001, 3, 1225–1228.
2
008, 3183–3204. (l) Yu, X.; Wang, W. Org. Biomol. Chem. 2008, 6, 2037–
(5) Gahman, T. C.; Overman, L. E. Tetrahedron 2002, 58, 6473–6483.
(6) For reviews, see: (a) Snider, B. B. In The Prins Reaction and Carbonyl
Ene Reactions; Trost, B. M., Fleming, I., Heathcock, C. H., Eds.; Pergamon Press:
New York, 1991; Vol. 2, pp 527-561. (b) Pastor, I. M.; Yus, M. Curr. Org. Chem.
2007, 11, 925–957.
2
046.
(
2) (a) Martinet, P.; Mousset, G.; Michel, M. C. R. Acad. Sci. Paris, Ser.
C 1969, 268, 1303–1306. (b) Martinet, P.; Mousset, G. Bull. Soc. Chim. Fr.
970, 1071–1076.
1
DOI: 10.1021/jo9024144
r 2009 American Chemical Society
Published on Web 12/14/2009
J. Org. Chem. 2010, 75, 455–463 455

Overman and Tanis
JOCArticle
Although the results depicted in Scheme 1 were attrib-
uted initially to steric effects, further experimentation
revealed that additional factors must be operative. If
steric factors were of paramount importance, it was ex-
pected that incorporating additional bulky substituents
on the cyclohexene ring of the cyclization precursor
would further increase the steric clash depicted in
structures C and D, increasing the energy of the chair-
like pathway. However, submission of a mixture of
epimers of bis-dithiane acetal 7 to standard Prins-pinacol
cyclization conditions (SnCl , CH Cl , 0 °C) resulted
2
,7
FIGURE 1. 12-Oxatricyclo[6.3.1.0 ]dodecanes.
potentially override the typically dominant stereoelectro-
nic effects.
4
2
2
in the predominant formation of product 9a, which
In preliminary scouting experiments carried out in the
context of a projected total synthesis of the structurally novel
10
would arise from a chair topography pathway (eq 1).
Replacement of SnCl with the monodentate Lewis acid
BF3 OEt2 yielded similar results, ruling out chelation
7
fungal metabolite aspergillin PZ (1), we reported that 12-
4
2,7
3
oxatricyclo[6.3.1.0 ]dodecanes having either a trans or cis
relative configuration of the oxacyclic bridge and the cis
angular substituents could be constructed by Prins-pinacol
reactions (2 or 3, Figure 1). In that report, we concluded
that stereocontrol in the Prins-pinacol reaction hinged upon
steric factors. It was specifically proposed that the steric bulk
of the 1,3-dithiane carbonyl protecting group in cyclization
precursor 4 was responsible for sufficiently destabilizing the
typically favored chair cyclization/rearrangement sequence
between the two dithiane rings as the cause of this un-
expected preference for the formation of pentacyclic pro-
duct 9a.
8
,9
(
C f D f 6) that the boatlike pathway (A f B f 5) was
favored (Scheme 1). The siloxy epimer of 4 was reported to
form an intractable mixture of products under identical
conditions.
SCHEME 1. Steric Rationale for Prins-Pinacol Stereoselec-
tivity
8
To enhance our understanding of the factors that control
stereoselection in Prins-pinacol reactions, additional
experimental studies were undertaken. This article reports
the results of such an investigation that highlights once again
the importance of stereoelectronic factors in dictating the
outcome of Prins-pinacol processes. Additionally, a new
skeletal rearrangement/thio-Prins-pinacol cascade reaction
is described.
Results
Preparation of Prins-Pinacol Substrates. To help identify
which structural features are important in determining
stereoselection in Prins-pinacol reactions of the type
depicted in eq 1, a series of cyclization precursors that
differ at a single position of the cyclohexene ring were
prepared (Scheme 2). In a straightforward sequence,
1
1
cyclohexenyl iodides 10a-e were treated individually
with 1.8 equiv of tert-butyllithium, and the resulting alke-
nyllithium species were coupled at -78 °C in THF with
8
racemic aldehyde 11 to provide a 1.6-2.0:1 mixture
of secondary alcohol epimers 12a-e (S*,R* relative
(7) Zhang, Y.; Wang, T.; Pei, Y.; Hua, H.; Feng, B. J. Antibiot. 2002, 55,
6
93–695.
(
(
8) Overman, L. E.; Velthuisen, E. J. Org. Lett. 2004, 6, 3853–3856.
9) For a quite different approach to the construction of the trans variant
of this oxatricyclic ring system, see: Molander, G. A.; Czak oꢀ , B.; St. Jean,
D. J., Jr. J. Org. Chem. 2006, 71, 1172–1180.
(
10) Velthuisen, E. J. Ph.D. Dissertation, University of California, Irvine,
005.
11) The preparation of cyclohexenyl iodides 10a-e is described in the
Supporting Information.
2
(
4
56 J. Org. Chem. Vol. 75, No. 2, 2010

Overman and Tanis
JOCArticle
SCHEME 2. Preparation of Prins-Pinacol Substrates 14a-e
and 15a-e
Prins-Pinacol Reactions. Prins-pinacol precursors
14a-e and 15a-e were chosen to allow both the size and
electronic properties of the C4 substituents of the cyclohex-
ene ring to be varied. Substrates 14c and 15c contain a spiro-
1
more electron withdrawing than the sulfur substituents of
,3-dioxane group in which the oxygen substituents will be
1
2
substrates 14a and 15a. 1,3-Dithiolane substrates 14b and
5b and 1,3-dioxolane substrates 14d and 15d are sterically
1
smaller than their 1,3-dithiane and 1,3-dioxane congeners,
but maintain similar electronic profiles. Geminal dimethyl
substrates 14e and 15e provide an additional set of reactants
in which only the steric size of the C4 substituents is varied.
All substrates reacted within 30 min upon exposure to 0.5
equiv of SnCl in CH Cl at 0 °C. Precursors having the S*,
4
2
2
R* relative configuration of the vicinal stereocenters (14a-e)
cyclized to provide Prins-pinacol products 8a-e having a
trans relationship between the oxygen bridge and the angular
substituents (Table 1). Depending upon the substituents on
the cyclohexene ring, the yields of product 8 varied greatly.
Prins-pinacol products 8a and 8b were formed in very low
yield from the dithiane and dithiolane substrates, with many
other unidentified products being produced. The low yield
(
7%) of Prins-pinacol product 8a produced from the S*,R*
configuration of the vicinal stereocenters) and 13a-e (S*,
S* relative configuration of the vicinal stereocenters),
respectively. These mixtures of alcohol epimers were sepa-
rated by HPLC, and samples of isomerically pure (>97:3
dr) alcohols were silylated to yield the S*,R* cyclization
substrates 14a-e and their S*,S* siloxy epimers 15a-e in
good overall yields.
precursor 14a is consistent with the results observed earlier in
1
0
reactions of a mixture of siloxy epimers (eq 1). However,
the alternative Prins-pinacol products 9a-e having the
epimeric configuration of the cis angular substituents were
not observed in the H NMR spectra of the crude reaction
mixtures. The relative configuration of Prins-pinacol products
8
constants observed between the vicinal methine hydrogens
H and H , consistent with a dihedral angle of nearly 90°
between these hydrogen atoms (Figure 2).
1
a-e followed from the diagnostic small (<1 Hz) coupling
The relative configurations of alcohols 12a-e and 13a-e
were verified by their individual transformation upon
exposure to p-toluenesulfonic acid or stannic chloride to
cyclic acetals 16a-e (eq 2) and 17a-e (eq 3). The presence or
absence of NOESY crosspeaks between the signals for the
a
b
13
a
b
dioxolane methine hydrogens H and H allowed the relative
configuration of the secondary alcohol stereocenters of the
precursors to be specified.
(12) The revised Pauling electronegativity of sulfur is 2.58 (similar to
carbon, 2.55), whereas for oxygen it is 3.44: Allred, A. L. J. Inorg. Nucl.
Chem. 1961, 17, 215–221.
(13) Karplus, M. J. Am. Chem. Soc. 1963, 85, 2870–2871.
J. Org. Chem. Vol. 75, No. 2, 2010 457

Overman and Tanis
JOCArticle
TABLE 1. Prins-Pinacol Cyclizations of S*,R* Substrates 14a-e
precursor
X
product
a
yield, %
compd
n
compd
b
1
1
1
1
4a
4b
4c
4d
4e
S
S
O
O
Me
3
2
3
2
8a
8b
8c
8d
8e
7
18
54
72
43
FIGURE 2. Conformations of Prins-pinacol products 8 and 9.
TABLE 2. Prins-Pinacol Cyclizations of S*,S* Substrates 15a-e
1
a
b
Mean yield of pure product from duplicate experiments. Yield was
1
determined by H NMR spectroscopy, using 1-chloro-4-nitrobenzene as
an internal standard.
Substrates 15a-e having the S*,S* relative configuration
of the vicinal stereocenters cyclized to form two major
products: the Prins-pinacol products 9a-e and a byproduct
1
8a-e (Table 2). The relative configuration of products
a-e was signaled by the ∼7.5 Hz coupling constant
9
observed for H and H , as would be expected from the
small dihedral angle between these hydrogens (Figure 2).
In all reactions, the alternate Prins-pinacol products 8a-e
were not detected by H NMR analysis of crude reaction
a
b
1
3
1
mixtures.
Aldehydes 18a-e are the only significant byproducts
produced in the Prins-pinacol cyclizations of substrates
unsaturated acetal
products
a
compd yield, % compd yield, %
a
1
products is the chemical shift of the aldehyde hydrogen in
5a-e. The most striking spectroscopic feature of these
compd
5a
X
n
b
1
S
3
2
3
2
9a
9b
9c
9d
9e
73
46
39
31
68
18a
18b
18c
18d
18e
5
1
the H NMR spectrum, which at δ ∼10.5 ppm is downfield
15b
15c
S
O
11
26
36
12
1
4
of the normal range for aliphatic aldehydes. Single-crystal
X-ray diffraction analysis confirmed the identity of product
15d
15e
O
Me
1
8b, with products 18a,c,d,e being assigned in analogous
5
1
a
b
Mean yield of pure product from duplicate experiments. Yield was
fashion on the basis of their distinctive aldehyde chemical
shifts. Byproducts 18a-e contain a unique pentacyclic ring
system in which the former 1,3-dithiane ring is expanded to a
seven-membered 1,4-dithiepane ring. The oxo bridge span-
ning the carbocyclic seven-membered ring of these bypro-
ducts is also displaced by one carbon from its position in
Prins-pinacol products 8 and 9, with the ring junction
between the carbocyclic seven- and six-membered rings
being trans rather than cis.
1
determined by H NMR spectroscopy from the ratio of the integrals of
the aldehyde signals of 9 and 18.
having a trans relationship between the oxygen bridge and
the angular substituents, whereas substrates 15a-e having
the S*,S* relative configuration lead to 9a-e in which this
relationship is cis. In contrast, no clear relationship between
stereoselection and the size of the cyclohexene C4 substitu-
ents is observed.
3
-5
In our earlier studies,
we found that stereoselection in
Prins-pinacol reactions can be rationalized by the topogra-
phy of the Prins cyclization step, consistent with the very
Discussion
1
6
Among the trends apparent in the results of the
Prins-pinacol reactions summarized in Tables 1 and 2, one
stands out: the configuration of the allylic silyl ether deter-
mines the stereochemical outcome of the Prins-pinacol
reaction. Substrates 14a-e with the S*,R* relative config-
uration of the vicinal stereocenters cyclize to give 8a-e
low activation barriers typically observed for pinacol rear-
17,18
rangements.
The transformation of the oxocarbenium
(16) An alternate explanation for the stereoselection observed in
Prins-pinacol reactions that has not yet been probed thoroughly is the
possibility that the cyclization and rearrangement steps occur with some
degree of concert. Computational studies to evaluate this possibility are
underway.
(
14) Pretsch, E.; B u€ hlmann, P.; Affolter, C. Structure Determination of
Organic Compounds; Springer-Verlag: Berlin, Germany, 2000; pp 12 and
18.
15) Crystallographic data for this compound were deposited at the
Cambridge Crystallographic Data Centre: CCDC 753122. An ORTEP
diagram of compound 18b may be found in the Supporting Information.
As is evident in this diagram, the aldehyde hydrogen atom projects close to
the oxo bridge and directly into the cavity of the molecule.
(17) In the absence of stereoelectronic constraints, pinacol rearrange-
ments have low activation barriers, see: (a) Bartok, M.; Molnar, A. In
Chemistry of Ethers, Crown Ethers, Hydroxyl Compounds and Their Sulfur
Analogues; Patai, S., Ed.; Wiley: New York, 1980; Part 2, pp 722-732.
(b) Rickborn, B. In Comprehensive Organic Synthesis; Trost, B. M., Fleming,
I., Eds.; Pergamon Press: Oxford, UK, 1991; Vol. 3, pp 721-732.
(18) For a particularly revealing example, see: Minor, K. P.; Overman,
L. E. Tetrahedron 1997, 53, 8927–8940.
2
(
4
58 J. Org. Chem. Vol. 75, No. 2, 2010

Overman and Tanis
JOCArticle
SCHEME 3
ion derived from unsaturated acetals 15a-e by either a chair
or boat Prins-pinacol pathway is analyzed in Scheme 3.
The initially generated oxocarbenium ion should be stabilized
the poor hyperconjugative donor C-O σ-bond has poor
overlap with the empty p-orbital in chair carbenium ion
F, whereas this bond has good overlap in boat carbenium ion
H. Thus, stereoelectronic factors favor the formation of
chair hydropyranyl carbenium ion intermediate F. Apparent
also in three-dimensional models for intermediate F is the
lack of destabilizing steric interactions between the dithiane
substituent of the hydropyranyl ring and the cyclohexane
ring.
1
9
by hyperconjugative electron release from the adjacent axial
0,21
2
C-S σ-bond.
state for the Prins cyclization will be late and resemble the
4
As a result, we assume that the transition
4
-hydropyranyl carbenium ion intermediate. The orienta-
tion of the vacant p-orbital and the adjacent siloxy substi-
tuent and geminal hydrogen are quite different in carbenium
ion intermediates F and H that would arise from chair or
boat topography Prins cyclizations, respectively. In the case
of the chair hydropyranyl carbenium ion F, the hyperconju-
gative donor C-H σ-bond has good overlap, whereas the
analogous bond in the boat carbenium ion intermediate
H is nearly orthogonal to the vacant p-orbital. Likewise,
Since our initial stereoelectronic analysis of Prins-pinacol
cyclizations to form tetrahydrofurans was disclosed in 2001,
4
Alder and co-workers published a high-level computational
study of the Prins cyclization of 2-oxonia-1,5-hexadiene to
2
2
form the 4-hydropyranyl carbenium ion. This study high-
lighted the delocalized character and special stability of the
chair 4-hydropyranyl carbenium ion. In this parent series,
the boat analogue of the 4-hydropyranyl cation was calcu-
lated to be more than 12 kcal/mol less stable. In the S*,S*
series, both the stereoelectronic influence of the allylic
substituents and the delocalized nature of a chair 4-hydro-
(
19) The three-dimensional models shown in Schemes 3 and 4 were
generated by the analysis of 15-20 low-energy conformers using MMFF
conformational searches (Spartan ’08 for Macintosh; Wavefunction, Inc.;
Irvine, CA, 2008) on cations derived from Prins cyclization of dimethyl
substrates 14e and 15e, followed by DFT geometry optimization at the
B3LYP/6-31G* level. Minima containing strong deformation of the carbo-
cationic center toward a sulfur atom of the dithiane substituent or ones that
could not be formed directly by the Prins reaction were disregarded.
23
pyranyl carbenium favor the chair Prins-pinacol pathway.
A similar analysis of the Prins-pinacol reaction of
precursors 14a-e is presented in Scheme 4. With these
(
20) For an incisive analysis of the structure of oxocarbenium ions
bearing proximal heteroatoms, and evidence against bridging from an
adjacent sulfur atom to form a stable episulfonium ion intermediate, see:
Beaver, M. G.; Billings, S. B.; Woerpel, K. A. Eur. J. Org. Chem. 2008, 771–
(22) Alder, R. W.; Harvey, J. N.; Oakley, M. T. J. Am. Chem. Soc. 2002,
124, 4960–4961.
7
81.
(
21) (a) Dudley, T. J.; Smoliakova, I. P.; Hoffmann, M. R. J. Org. Chem.
999, 64, 1247–1253. (b) Alabugin, I. V.; Manoharan, M. J. Org. Chem. 2004,
9, 9011–9024.
(23) A similar chair preference was seen in the Prins-pinacol reaction of
the configurationally analogous, structurally less elaborate, precursor
(compound 16 in our earlier study).
1
6
8
J. Org. Chem. Vol. 75, No. 2, 2010 459

Overman and Tanis
JOCArticle
SCHEME 4
substrates, the observed product arises from the boat topo-
graphy pathway (14 f K f L f 8). In this series, the
electron-releasing C-H σ-bond geminal to the siloxy sub-
stituent has poor overlap with the adjacent p-orbital in chair
hydropyranyl cation intermediate J and better overlap in the
corresponding boat intermediate L, whereas the opposite is
the case for the C-O σ-bond. As a result, stereoelectronic
factors favor formation of the boat 4-hydropyranyl carbe-
nium ion intermediate. If this effect is responsible for the
preferential formation of Prins-pinacol products 8a-e, the
stability provided by the greater hyperconjugative stabiliza-
tion in boat hydropyranyl carbenium ion intermediate
and L (Scheme 4) have good overlap between the vacant p-
orbital and the C-C σ-bond that migrates in the pinacol
rearrangement, whereas this orbital overlap is poor in the
alternative carbenium ions H and J. This stereoelectronic
factor would be important if the Prins-pinacol reactions
took place by a concerted pathway, a possibility that merits
1
6
further consideration.
Byproducts 18a-e produced from SnCl -promoted reac-
4
tion of the S*,S* precursors 15a-e are the result of extensive
structural rearrangement. The 1,4-dithiepin and thioacetal
fragments present in these products suggest that both
1
,2-sulfur and 1,2-oxygen migrations are involved in their
2
4,25
L must be quite significant.
It merits note that molecular models suggest that the
stereoelectronically favored carbenium ions F (Scheme 3)
genesis. One plausible sequence is suggested in Scheme 5.
Lewis acid-promoted ionization of acetals 15 would generate
oxocarbenium ion M. From recent studies of Woerpel and
co-workers, the 1,3-dithiane would be expected to stabilize
this cation and likely facilitate its formation by hyperconju-
gation of the axial C-S σ-bond, rather than by bridging to
(
24) DFT (Spartan 08, B3LYP/6-31G*) calculations on chair and boat
conformers of the 9-oxobicyclo[3.3.1]nonanyl carbenium ions i and ii show
that the preference for chair structure i (10.7 kcal/mol) is somewhat less than
2
0,26
form episulfonium ion intermediate N.
1,2-Migration of
2
2
that for the parent 4-hydropyranyl carbenium ion (12.6 kcal/mol) (see the
Supporting Information for details).
this axial sulfur atom via N (presumably a high-energy
intermediate or transition state) would generate the tetra-
2
7
substituted thiocarbenium ion O. This ring expansion
sequence has close precedent in the synthesis of 5,6-dihy-
dro-1,4-dithiepins and 2-alkylidene-1,4-dithiepanes from
(
(
26) Billings, S. B.; Woerpel, K. A. J. Org. Chem. 2006, 71, 5171–5178.
27) For reviews of 1,2-sulfur migrations, see: (a) Fox, D. J.; House, D.;
(
25) For a recent computational study and detailed discussion of the
Warren, S. Angew. Chem., Int. Ed. 2002, 41, 2462–2482. (b) Sromek, A. W.;
Gevorgyan, V. Top. Curr. Chem. 2007, 274, 77–124.
donor ability of C-H bonds to adjacent vacant p-orbitals, see ref 21b.
4
60 J. Org. Chem. Vol. 75, No. 2, 2010

Overman and Tanis
JOCArticle
SCHEME 5. Possible Mechanism of Formation of Byproducts 18a-e
2
8
dithiane precursors. In the present case, rather than losing
an adjacent hydrogen, intermediate O is poised for a 1,2-
oxygen shift to form thiocarbenium ion P. This electrophilic
intermediate is then positioned for capture by the adjacent
cyclohexene π-bond, initiating a thio-Prins-pinacol reac-
tion sequence. If the thio-Prins cyclization occurred by
a topography in which the bulky tert-butyldimethylsilyl
substituent is oriented in a quasiequatorial fashion (P f
epimers 14 and 15 are transformed stereoselectively to two
different products 8 and 9, respectively, provides a dramatic
demonstration of the importance of allylic stereoelectronic
effects in determining the outcome of Prins-pinacol reac-
tions. Depending upon the configuration of this single
stereocenter, the Prins-pinacol reaction occurs by either a
boat- or chairlike sequence.
A new molecular reorganization was identified that yields
products 18 that contain a previously unknown oxygen-
bridged 2,6-dithiabicyclo[5.5.0]dodecane ring system. A
complex cascade sequence that involves oxocarbenium ion
formation, followed by sequential 1,2-sulfur and 1,2-oxygen
migrations and eventually a thio-Prins-pinacol reaction, is
suggested to explain the formation of these products.
2
9
Q), the observed products 18a-e would result.
The yield of the thio-Prins-pinacol byproduct 18 was
higher in reactions of the 1,3-dioxane and 1,3-dioxolane
precursors 15c,d (Table 2). Because of inductive electron
withdrawal by the oxygen substituents, the alkene π-bond in
these acetal precursors would be expected to be less nucleo-
philic. Thus, the higher yields of the thio-Prins-pinacol
byproducts 18c,d suggest that oxocarbenium ion M parti-
tions between Prins cyclization to eventually generate
Prins-pinacol products 9 and 1,2-sulfur migration to gener-
ate thiocarbenium ion O and ultimately thio-Prins-pinacol
products 18.
3
0
Experimental Section
rel-(R)-(1,4-Dithiaspiro[4.5]dec-7-en-8-yl)(rel-(7S,9S)-7-meth-
oxy-9-methyl-8-oxa-1,5-dithiaspiro[5.5]undecan-9-yl)methanol
(
7
(
12b) and rel-(S)-(1,4-Dithiaspiro[4.5]dec-7-en-8-yl)(rel-(7S,9S)-
-methoxy-9-methyl-8-oxa-1,5-dithiaspiro[5.5]undecan-9-yl)methanol
13b). To a stirring solution of t-BuLi (10.9 mL, 1.05 M in
pentane) in THF (64 mL) at -78 °C was added a solution of
vinyl iodide 10b (1.90 g, 6.37 mmol) in THF (8 mL) dropwise by
syringe over 5 min. After 15 min, a solution of aldehyde 11 (1.52 g,
5.74 mmol) in THF (8 mL) was added in similar fashion. After 1 h,
the cooling bath was removed and the reaction was allowed to
warm to rt and was maintained at rt for 1 h. The reaction mixture
Conclusions
Stereoelectronic effects, not steric factors, are responsible
for the stereoselectivity observed in Prins-pinacol reactions
employed by us previously and in the present study to
8
assemble polycyclic products containing the 12-oxatricyclo-
[
2
6.3.1.0 ]dodecane moiety. The observation that siloxy
,7
was treated with saturated aqueous NH Cl (60 mL), and the layers
4
were separated. The aqueous layer was extracted with Et O, and
2
the combined organic layers were washed with brine, dried
(28) (a) Saigo, K.; Hashimoto, Y.; Fang, L.; Hasegawa, M. Heterocycles
989, 29, 2079–2082. (b) Afonso, C. A. M.; Barros, M. T.; Godinho, L. S.;
1
Maycock, C. D. Synthesis 1991, 575–580 and references cited therein.
29) The formation of products 18 is the first example of a thio-
(30) Representative experimental procedures for preparing a cyclization
4
substrate and its rearrangement upon reaction with SnCl . General experi-
(
Prins-pinacol sequence in which the initial thio-Prins cyclization generates
a seven-membered ring.
mental details, experimental procedures, and characterization for other new
compounds can be found in the Supporting Information.
J. Org. Chem. Vol. 75, No. 2, 2010 461

Overman and Tanis
1424 cm ; HRMS (ES) m/z calcd for C H NaO S (M þ
JOCArticle
over Na SO , filtered, and concentrated. The residue was purified
-
1
2
4
18 26
2 4
þ
by silica gel flash chromatography (25:73:2-30:68:2 ethyl
N) to yield a 1.6:1 mixture of 12b and 13b,
respectively, as an off-white solid (1.89 g, 75%): R 0.1 (20:80 ethyl
Na) 425.0714, found 425.0725.
acetate-hexanes-Et
3
tert-Butyl(rel-(R)-(1,4-dithiaspiro[4.5]dec-7-en-8-yl)(rel-(7S,-
9S)-7-methoxy-9-methyl-8-oxa-1,5-dithiaspiro[5.5]undecan-9-
yl)methoxy)dimethylsilane (14b). To a stirring solution of allylic
alcohol 12b (280 mg, 0.63 mmol), N,N-dimethyl-4-aminopyr-
idine (5.0 mg, 0.041 mmol), and pyridine (260 μL, 3.2 mmol)
in CH Cl (6.3 mL) at -78 °C was added TBDMSOTf (290 μL,
f
acetate-hexanes).
The mixture was separated by HPLC (35:65 ethyl acetate-hexanes
isocratic, 40 mL/min; Peeke KromaSpher 80, 5 μm silica, 300 ꢀ
5
0 mm i.d., λ = 254 nm) to give pure samples of the alcohol
1
2
2
epimers: 12b: t
R
= 25 min; colorless solid; mp 121-122 °C; H
) δ 5.73 (br s, 1H), 4.73 (s, 1H), 3.95
s, 1H), 3.48 (s, 3H), 3.39-3.28 (m, 4H), 2.88 (ddd, J = 14.3, 9.9,
1.3 mmol) dropwise by syringe over 2 min. After a further
10 min, the cooling bath was removed and the reaction was
allowed to warm to rt and was maintained at rt for 18 h. Water
(8 mL) was then added and the layers were separated. The
NMR (500 MHz, CDCl
(
3
3
(
.3 Hz, 1H), 2.82-2.72 (m, 4H), 2.71-2.64 (m, 1H), 2.57-2.50
m, 1H), 2.50 (td, J = 13.1, 4.9 Hz, 1H), 2.34 (td, J = 13.6,
.0 Hz, 1H), 2.31-2.18 (m, 2H), 2.15-1.98 (m, 3H), 1.95-1.85
aqueous layer was extracted with CH
organic layers were washed with brine, dried over Na SO ,
2 2
Cl , and the combined
4
2
4
1
3
(
m, 1H), 1.37-1.31 (m, 1H), 1.34 (s, 3H); C NMR (125 MHz,
CDCl ) δ 135.5 (C), 126.4 (CH), 102.7 (CH), 82.3 (CH), 78.8 (C),
5.7 (C), 56.6 (CH ), 52.3 (C), 42.6 (CH ), 39.0 (CH ), 38.8
filtered, and concentrated in vacuo. The residue was dry-loaded
onto silica gel and purified by silica gel flash chromatography
3
6
(15:84:1 ethyl acetate-hexanes-Et N) to yield 14b as a colorless
3
2
2
3
1
(
(
1
CH
CH
2
), 26.6 (CH
), 25.3 (CH
2
), 26.4 (CH
), 23.4 (CH
2
), 25.9 (CH
2
), 25.8 (CH
2
), 25.5
); IR (film) 3531, 2927, 1424,
foam (320 mg, 86%): R 0.4 (20:80 ethyl acetate-hexanes); H
NMR (500 MHz, CDCl ) δ 5.63 (br s, 1H), 4.68 (s, 1H), 4.01
f
2
2
3
3
-
1
142 cm ; HRMS (ES) m/z calcd for C H NaO S (M þ
(s, 1H), 3.46 (s, 3H), 3.38-3.26 (m, 4H), 3.07-2.97 (m, 2H),
2.79-2.64 (m, 4H), 2.56-2.48 (m, 1H), 2.36-2.27 (m, 1H), 2.15
(ddd, J = 13.9, 8.4, 4.2 Hz, 1H), 2.12-1.90 (m, 6H), 1.60 (ddd,
J = 13.9, 8.4, 4.2 Hz, 1H), 1.28 (s, 3H), 0.89 (s, 9H), 0.04 (s, 3H),
1
9
30
3 4
þ
Na) 457.0976, found 457.0977. 13b: t
R
= 28 min; colorless
) δ 5.56 (br s,
H), 4.72 (s, 1H), 3.73 (s, 1H), 3.50 (s, 3H), 3.38-3.27 (m, 4H),
.88-2.72 (m, 6H), 2.67-2.61 (m, 1H), 2.50-2.30 (m, 3H), 2.32
1
solid; mp 129-132 °C; H NMR (500 MHz, CDCl
1
2
3
1
3
0.00 (s, 3H); C NMR (125 MHz, CDCl ) δ 137.9 (C), 124.5
3
(
td, J = 11.7, 3.9 Hz, 1H), 2.22 (dt, J = 13.7, 4.5 Hz, 1H),
.16-1.87 (m, 4H), 1.38 (dt, J = 13.6, 3.9 Hz, 1H), 1.27 (s, 3H);
(CH), 104.3 (CH), 79.3 (CH), 78.3 (C), 65.6 (C), 56.7 (CH ), 50.9
(C), 42.6 (CH ), 39.3 (CH ), 39.2 (CH ), 39.0 (CH ), 30.3 (CH ),
3
2
2
2
2
2
2
1
3
C NMR (125 MHz, CDCl
CH), 81.1 (CH), 77.9 (C), 65.8 (C), 56.8 (CH
3
) δ 137.0 (C), 125.4 (CH), 103.0
), 51.6 (C), 42.7
29.4 (CH ), 27.0 (CH ), 26.9 (CH ), 26.7 (CH ), 26.1 (CH ), 25.6
2 2 2 2 3
(
(
(
(
3
(CH ), 22.5 (CH ), 18.3 (C), -4.1 (CH ), -4.9 (CH ); IR (film)
2
3
3
3
-1
CH ), 39.0 (CH ), 38.9 (CH ), 38.7 (CH ), 28.6 (CH ), 27.7
2
CH
CH ); IR (film) 3531, 2927, 1425 cm ; HRMS (ES) m/z calcd
2952, 2856, 1472 cm ; HRMS (ES) m/z calcd for
2
2
2
2
þ
2
), 26.1 (CH
2
), 26.0 (CH
2
), 25.33 (CH
2
), 25.29 (CH
2
), 22.9
C H NaO S Si (M þ Na) 571.1840, found 571.1838.
2
5
44
3 4
-
1
3
tert-Butyl(rel-(S)-(1,4-dithiaspiro[4.5]dec-7-en-8-yl)(rel-(7S,-
9S)-7-methoxy-9-methyl-8-oxa-1,5-dithiaspiro[5.5]undecan-9-
yl)methoxy)dimethylsilane (15b). In a similar fashion, allylic
alcohol 13b (140 mg, 0.63 mmol) was silylated, and the crude
product was purified by silica gel flash chromatography (10:90
þ
for C19
H30NaO
3
S
4
(M þ Na) 457.0976, found 457.0977.
rel-(1S,5R,7R)-1-Methyl-7-(1,4-dithiaspiro[4.5]dec-7-en-8-yl)-
,8-dioxaspiro[bicyclo[3.2.1]octane-4,2 -[1,3]dithiane] (16b). To a
0
6
stirring solution of allylic alcohol 12b (30 mg, 0.070 mmol)
in CH Cl (1.0 mL) at rt was added p-TsOH H O (1.0 mg,
.0053 mmol). After 45 min, Et O (1.0 mL)
2
2
3
2
ethyl acetate-hexanes) to yield 15b as a colorless solid (150 mg,
1
0
3
N (300 μL) and H
2
87%): mp 51-54 °C; R
f
0.4 (20:80 ethyl acetate-hexanes); H
were added and the layers were separated. The aqueous layer
was extracted with CH Cl . The combined organic layers were
NMR (500 MHz, CDCl ) δ 5.55 (br s, 1H), 4.91 (s, 1H), 3.91 (s,
3
1H), 3.52 (s, 3H), 3.40-3.28 (m, 4H), 3.06 (ddd, J = 13.9, 8.4,
3.7 Hz, 1H), 2.97 (ddd, J = 14.2, 8.7, 3.7 Hz, 1H), 2.78-2.63 (m,
4H), 2.45 (br s, 2H), 2.22-2.08 (m, 3H), 2.05-1.90 (m, 4H), 1.47
(ddd, J = 13.9, 7.3, 4.2 Hz, 1H), 1.28 (s, 3H), 0.91 (s, 9H), 0.07
2
2
2 2 4
washed with H O, dried over Na SO , filtered, and concen-
trated in vacuo. The residue was purified by silica gel flash
chromatography (20:80 ethyl acetate-hexanes) to yield 16b as a
1
3
colorless oil (24 mg, 85%): R
H NMR (500 MHz, CDCl
f
0.4 (30:70 ethyl acetate-hexanes);
(s, 3H), 0.02 (s, 3H); C NMR (125 MHz, CDCl
125.2 (CH), 105.5 (CH), 83.7 (CH), 78.3 (C), 65.7 (C), 56.9
(CH ), 51.0 (C), 42.7 (CH ), 39.10 (CH ), 39.05 (CH ), 38.96
3
) δ 138.0 (C),
1
3
) δ 6.12 (br s, 1H), 5.54 (s, 1H), 4.04
(
2
2
(
(
(
(
s, 1H), 3.42-3.25 (m, 4H), 2.97 (dd, J = 8.0, 5.3 Hz, 1H),
3
2
2
2
.90-2.63 (m, 5H), 2.37-2.23 (m, 2H), 2.20-2.03 (m, 3H),
(CH ), 30.9 (CH ), 28.8 (CH ), 27.0 (CH ), 26.9 (CH ), 26.2
2
2
2
2
2
1
3
.00-1.85 (m, 4H), 1.65-1.55 (m, 1H), 1.43 (s, 3H); C NMR
) δ 130.9 (C), 122.8 (CH), 101.4 (CH), 86.3
CH), 80.8 (C), 65.6 (C), 51.9 (C), 42.0 (CH ), 39.0 (CH ), 38.9
(CH
-4.9 (CH
m/z calcd for C25
571.1843.
3
), 25.9 (CH
2
), 25.6 (CH
2
), 24.0 (CH
); IR (film) 2952, 2856, 1472 cm ; HRMS (ES)
3
3
), 18.4 (C), -4.2 (CH ),
-
1
125 MHz, CDCl
3
3
þ
H44NaO
S Si (M þ Na) 571.1840, found
4
2
2
3
CH ), 38.4 (CH ), 31.1 (CH ), 29.0 (CH ), 26.4 (CH ), 26.2
2
2
2
2
2
CH
2
), 25.8 (CH
2
), 24.9 (CH
2
), 23.8 (CH
3
); IR (film) 2917, 2842,
(M þ
Prins-Pinacol Reaction of Unsaturated Acetal 14b with Stan-
nic Chloride. Preparation of rel-(1S,2S,7S,8R)-2-Formyl-1-
-
1
424 cm ; HRMS (ES) m/z calcd for C18
1
Na) 425.0714, found 425.0713.
H
26NaO
S
2 4
þ
2,7
methyl-12-oxabispiro[tricyclo[6.3.1.0 ]dodecane-9,2 -[1,3]dithi-
0
0
rel-(1S,5R,7S)-1-Methyl-7-(1,4-dithiaspiro[4.5]dec-7-en-8-yl)-
0
ane-5,2 -[1,3]dithiolane] (8b). To a stirring solution of TBDMS
(3.8 mL) at 0 °C was
Cl ) drop-
6
,8-dioxaspiro[bicyclo[3.2.1]octane-4,2 -[1,3]dithiane] (17b). In
ether 14b (104 mg, 0.189 mmol) in CH
added SnCl (190 μL, 0.095 mmol, 0.5 M in CH
2
Cl
2
a similar fashion, a solution of allylic alcohol 13b (32 mg,
.074 mmol) in CH Cl (1.0 mL) was treated with p-TsOH H O
4
2
2
0
wise by syringe over 20 s. After a further 30 min, Et N (130 μL,
2
2
3
2
3
(
2.0 mg, 0.010 mmol), and the crude product was purified by
silica gel flash chromatography (20:80 ethyl acetate-hexanes)
to yield 17b as a colorless oil (28 mg, 94%): R 0.35 (30:70 ethyl
acetate-hexanes); H NMR (500 MHz, CDCl ) δ 5.59 (s, 1H),
0.95 mmol) was added. After an additional 5 min, the mixture
was poured into saturated aqueous NaHCO (8 mL) that was
3
f
precooled in ice/water and the layers were separated. The
aqueous layer was extracted with CH Cl , and the combined
1
3
2
2
5
1
2
.55 (br s, 1H), 4.27 (s, 1H), 3.37-3.27 (m, 4H), 2.99 (ddd, J =
2 4
organic layers were washed with brine, dried over Na SO ,
4.6, 8.9, 4.8 Hz, 1H), 2.86-2.61 (m, 5H), 2.41-2.33 (m, 1H),
filtered, and concentrated in vacuo. The residue was purified
by silica gel flash chromatography (15:85 ethyl acetate-hexanes)
1
3
.24-1.90 (m, 8H), 1.53-1.46 (m, 1H), 1.25 (s, 3H); C NMR
(
(
(
(
125 MHz, CDCl ) δ 136.7 (C), 124.7 (CH), 102.5 (CH), 85.1
to yield 8b as a colorless solid (13 mg, 17%): mp 156-158 °C;
3
1
CH), 81.9 (C), 65.4 (C), 52.0 (C), 42.3 (CH
CH ), 38.5 (CH ), 33.5 (CH ), 31.2 (CH
CH ), 25.2 (CH ), 24.8 (CH ), 20.9 (CH ); IR (film) 2919, 1445,
2
), 39.04 (CH
), 26.0 (CH
2
), 39.02
), 25.7
R
CDCl
f
0.3 (30:70 ethyl acetate-hexanes); H NMR (500 MHz,
3
2
2
2
2
2
) δ 9.75 (s, 1H), 4.21 (br s, 1H), 3.40-3.28 (m, 5H),
2.93 (ddd, J = 14.2, 9.4, 3.9 Hz, 1H), 2.88 (ddd, J = 14.6, 8.2, 3.9
2
2
2
3
4
62 J. Org. Chem. Vol. 75, No. 2, 2010

Overman and Tanis
JOCArticle
spectrum of the crude reaction mixture, via a ratio with the
Hz, 1H), 2.77 (ddd, J = 14.6, 7.3, 3.9 Hz, 1H), 2.72 (ddd, J =
3
integration values from isolated product 9b; colorless
1
1
6
1
4.2, 6.2, 3.9 Hz, 1H), 2.40-2.30 (m, 1H), 2.28 (dd, J = 13.7,
.2 Hz, 1H), 2.20 (dd, J = 13.7, 9.8 Hz, 1H), 2.12-1.94 (m, 6H),
.90-1.80 (m, 2H), 1.44 (dd, J = 13.3, 5.9 Hz, 1H), 1.27 (s, 3H);
crystals, mp 99-101 °C; R 0.5 (50% ethyl acetate-hexanes)
f
1
H NMR (500 MHz, CDCl ) δ 10.55 (d, J = 2.1 Hz, 1H),
3
1
3
C NMR (125 MHz, CDCl
3.6 (C), 60.0 (C), 51.6 (C), 42.6 (CH
9.6 (CH ), 37.8 (CH ), 32.0 (CH ), 31.5 (CH
3
) δ 202.6 (CH), 85.1 (C), 84.4 (CH),
), 41.7 (CH), 39.8 (CH ),
), 26.2 (CH ), 26.1
3.84 (ddd, J = 15.4, 11.3, 6.4 Hz, 1H), 3.35-3.25 (m, 4H),
3.16 (d, J = 5.4 Hz, 1H), 3.08 (ddd, J = 15.4, 5.4, 1.1 Hz,
1H), 2.99 (ddd, J = 14.9, 11.4, 6.4 Hz, 1H), 2.79-2.72
(m, 1H), 2.68 (ddd, J = 12.4, 5.4, 2.7 Hz, 1H), 2.41 (td,
J = 13.5, 4.5 Hz, 1H), 2.29-2.15 (m, 3H), 2.04 (dd, J = 13.5,
12.7 Hz, 1H), 2.04-1.90 (m, 4H), 1.86 (dt, J = 13.5, 2.5 Hz,
6
3
2
2
2
2
2
2
2
(
CH ), 25.3 (CH ), 24.7 (CH ), 21.9 (CH ); IR (film) 2935, 2829,
2 2 2 3
-
1
711, 1715, 1424 cm ; HRMS (ES) m/z calcd for C18
2
NaO
H
26
-
þ
2
S
4
(M þ Na) 425.0714, found 425.0709.
1
3
Prins-Pinacol Reaction of Unsaturated Acetal 15b with Stannic
Chloride. Preparation of rel-(1S,2R,7R,8R)-2-Formyl-1-methyl-
1H), 1.80-1.72 (m, 1H), 1.29-1.21 (m, 1H), 1.26 (s, 3H);
C
NMR (125 MHz, CDCl ) δ 209.0 (CH), 91.4 (C), 88.6 (C),
3
2
,7
0
1
5
2-oxabispiro[tricyclo[6.3.1.0 ]dodecane-9,2 -[1,3]dithiane-
,2-[1,3]dithiolane] (9b) and rel-(1S,7S,8S,13R,14S)-13-For-
69.1 (C), 62.9 (CH), 55.2 (C), 44.3 (CH
(CH ), 39.1 (CH ), 38.6 (CH ), 38.1 (CH
(CH ), 30.3 (CH ), 28.2 (CH ), 27.3 (CH ), 22.5 (CH ); IR
2
), 42.0 (CH), 40.2
2
2
2
2
), 34.2 (CH ), 33.3
2
1
,7
myl-14-methyl-17-oxa-2,6-dithispiro[tetracyclo[12.2.1.0 .-
0
2
2
2
2
3
8
,13
-1
0
]heptadecane-10,2 -[1,3]dithiolane] (18b). With use of a
(film) 2923, 2856, 2769, 1708 cm ; HRMS (ES) m/z calcd
for C18
þ
procedure identical with that employed for the Prins-
pinacol reaction of 14b, unsaturated acetal 15b (104 mg,
H
26NaO
2
S
4
(M þ Na) 425.0714, found 425.0706.
0
.189 mmol) was exposed to SnCl (190 μL, 0.095 mmol, 0.5 M in
4
Acknowledgment. This research was supported by the
NIH Neurological Disorders and Stroke Institute (NS-12389),
and by a graduate fellowship for PST from Allergan.
We thank Dr. Emile J. Velthuisen (GlaxoSmithKline) and
Professor Keith Woerpel (University of California, Irvine)
for insightful discussion and Dr. John Greaves (UC Irvine)
and Dr. Joseph W. Ziller (UC Irvine) for X-ray and mass
spectral analyses. NMR and mass spectra were obtained at
UC Irvine with instrumentation acquired with the assistance
of NSF and NIH Shared Instrumentation programs.
2 2
CH Cl ) at 0 °C, and the crude product was purified by silica
gel flash chromatography (15:85 ethyl acetate-hexanes) to
yield the following fractions: 9b: 37 mg (48%) as a colorless
solid after recrystallization from CH Cl , mp 175-176 °C;
2
2
1
R 0.25 (30:70 ethyl acetate-hexanes); H NMR (500 MHz,
f
CDCl
3
) δ 9.46 (d, J = 1.9 Hz, 1H), 4.62 (dd, J = 7.4, 2.0 Hz,
1
2
7
1
1
H), 3.42-3.21 (m, 5H), 2.90-2.84 (m, 2H), 2.84-2.74 (m,
H), 2.53 (dd, J = 13.7, 13.5 Hz, 1H), 2.45 (ddd, J = 15.7,
.4, 1.1 Hz, 1H), 2.35-2.23 (m, 3H), 2.17 (ddd, J = 14.7,
2.7, 5.8 Hz, 1H), 2.07-1.90 (m, 4H), 1.66-1.55 (m, 2H),
1
3
.18 (s, 3H); C NMR (125 MHz, CDCl ) δ 204.0 (CH), 85.2
3
(
(
(
(
(
C), 81.6 (CH), 65.3 (C), 57.1 (C), 53.2 (C), 44.5 (CH), 39.7
Supporting Information Available: Experimental proce-
dures, tabulated characterization data, optimized structures,
and relative energies and Cartesian coordinates from DFT
calculations on 9-oxobicyclo[3.3.1]nonanyl carbenium ions i,
CH
CH
2
2
), 39.1 (CH
), 26.6 (CH
2
2
), 38.7 (CH
), 25.7 (CH
2
2
), 36.5 (CH
), 24.8 (CH
2
2
), 32.7 (CH
), 23.0 (CH
2
3
), 31.4
), 20.0
-1
CH ); IR (film) 2935, 2821, 2709, 1717, 1463 cm ; HRMS
2
ES) m/z calcd for C18
þ
H
26NaO
2
S
4
(M þ Na) 425.0714,
1
13
1
ii, and other conformational isomers, and copies of H and C
found 425.0726. 18b: 11% by integration in the H NMR
NMR spectra for new compounds not previously reported, CIF
file for X-ray structure of 18b, and PDB files of the models of
intermediates F, H, J, and L. This material is available free of
charge via the Internet at http://pubs.acs.org.
(
31) Byproduct 18b is highly crystalline, and after crystallization is
difficult to redissolve in standard chromatography solvents. As a result,
some product was lost during purification.
J. Org. Chem. Vol. 75, No. 2, 2010 463