Designed Chiral Acyl Radical Equivalents. Preparation and Cyclizations of Disymmetrically Substituted 1,3-Dioxabicyclo[4.4.0]decan-2-yl Radicals

Article abstract of DOI:10.1021/jo961750n

The diastereoselectivity of 5-exo-trigonal cyclizations of 2-(4-penten-1-yl)-1,3-dioxolan-2-yl and 2-(4- penten-1-yl)-1,3-dioxan-2-yl radicals is investigated. When dioxolanes or dioxanes derived from C₂ symmetrically substituted diols are employed the diastereoselectivity is poor. In the dioxanyl series this is a consequence of the cyclization occurring through a twist-boat conformer. Disymmetrically substituted dioxanyl radicals, derived from the alcohols 21 and 41, are, however, constrained to chairlike conformations and accordingly give rise to highly diastereoselective cyclizations. Conditions are described for the hydrolysis of the resulting spiroacetals and for determination of the ee of the resulting 2-methylcyclopentanones.

Full text of DOI:10.1021/jo961750n

J . Org. Chem. 1997, 62, 275-286
275
Design ed Ch ir a l Acyl Ra d ica l Equ iva len ts. P r ep a r a tion a n d
Cycliza tion s of Disym m etr ica lly Su bstitu ted
1,3-Dioxa bicyclo[4.4.0]d eca n -2-yl Ra d ica ls
D. Stien,†,‡ R. Samy,‡ R. Nouguier,‡ D. Crich,*,‡ and M. P. Bertrand*,†
LCMO, Associe´ au CNRS, Faculte´ des Sciences Saint-J e´roˆme, Av. Normandie-Niemen, 13397 Marseille
Cedex 20, France, and Department of Chemistry, University of Illinois at Chicago, 845 West Taylor Street,
Chicago, Illinois 60607-7061
Received September 11, 1996X
The diastereoselectivity of 5-exo-trigonal cyclizations of 2-(4-penten-1-yl)-1,3-dioxolan-2-yl and 2-(4-
penten-1-yl)-1,3-dioxan-2-yl radicals is investigated. When dioxolanes or dioxanes derived from
C2 symmetrically substituted diols are employed the diastereoselectivity is poor. In the dioxanyl
series this is a consequence of the cyclization occurring through a twist-boat conformer. Disym-
metrically substituted dioxanyl radicals, derived from the alcohols 21 and 41, are, however,
constrained to chairlike conformations and accordingly give rise to highly diastereoselective
cyclizations. Conditions are described for the hydrolysis of the resulting spiroacetals and for
determination of the ee of the resulting 2-methylcyclopentanones.
In tr od u ction
Resu lts a n d Discu ssion
The obvious starting point for this investigation was
the examination of the 5-exo-trig cyclizations of C2-
symmetric 1,3-dioxan-2-yl radicals.6 Such R,R-dialkoxy
radicals may be generated for addition to CC multiple
bonds by a variety of pathways.7 Hydrogen atom ab-
straction from acetals by t-BuO• radicals,7a or by excited
states of benzophenone,7b is reported to give preparatively
useful yields in intermolecular additions but was less
successful when applied to the cyclizations of dioxolanyl
and dioxanyl radicals.7c However, the generation of 1,3-
dioxolan-2-yl and dioxan-2-yl radicals bearing an ap-
propriate ethylenic tether can readily be achieved via
intramolecular hydrogen atom abstraction with vinyl
radicals, as established by Curran,7d,e and this was the
strategy adopted in this study. The various substrates
(3) were readily obtained through standard malonate and
acetalization techniques (Scheme 1).
Stereoselective radical reactions have been vigorously
pursued over the last decade. Many successful enantio-
selective radical additions and cyclizations have now been
performed via the temporary incorporation of a chiral
auxiliary.1,2 Recent demonstrations of the compatibility
of Lewis acids with certain types of radical reactions
suggest that further progress in radical stereoselectivity
is to be anticipated; indeed, examples of the successful
use of designed Lewis acids in the promotion of diaste-
reoselective radical reactions have already been docu-
mented.3 For a number of years we have been interested
in the development of acyl radicals and their application
in synthesis.4 A logical extrapolation is the design of
chiral acyl radical equivalents. In this context, we
describe in detail here our studies of the cyclizations of
chirally substituted 1,3-dioxolan-2-yl and 1,3-dioxan-2-
yl radicals.5
5-exo Cycliza tion s of 1,3-Dioxola n -2-yl Ra d ica ls.
Three substrates (3a -c) were prepared from readily
available C2-symmetric 1,2-diols. These (0.05 M) were
then subjected to cyclization under a standard set of
experimental conditions involving the in situ generation
of Bu3SnH by reduction of Bu3SnCl (0.005 M) with
NaBH3CN in 2-methyl-2-propanol,8 at 80 °C in the
presence of AIBN. After completion, the crude reaction
mixtures were filtered on silica gel to remove stannyl
residues. The spirocyclic products 4 and 5 were thus
obtained in admixture with the comigrating reduction
product 6. After ozonolysis,7e or treatment with catalytic
† LCMO.
‡ University of Illinois at Chicago.
X Abstract published in Advance ACS Abstracts, J anuary 1, 1997.
(1) For reviews see: (a) Porter, N. A.; Giese, B.; Curran, D. P. Acc.
Chem. Res. 1991, 24, 296. (b) Smajda, W. Synlett 1994, 1. (c) Curran,
D.; Porter, N.; Giese, B. Stereochemistry of Radical Reactions: Con-
cepts, Guidelines and Synthetic Applications; VCH: New York, 1995.
(2) Recent advances: (a) Hanessian, S.; Yang, H.; Schaum, R. J .
Am. Chem. Soc. 1996, 118, 2507. (b) Garner, P. J .; Cox, P. B.;
Klippenstein, S. J . J . Am. Chem. Soc. 1995, 117, 4183. (c) Garner, P,
J .; Leslie, R.; Anderson, J . T. J . Org. Chem. 1996, 61, 6754.
(3) (a) Toru, T.; Watanabe, Y.; Tsusaka, M.; Ueno, Y. J . Am. Chem.
Soc. 1993, 115, 10464. (b) Nagano, H.; Kuno, Y. J . Chem. Soc., Chem.
Commun. 1994, 987. (c) Curran, D. P.; Kuo, L. H. J . Org. Chem. 1994,
59, 3259. (d) Renaud, P.; Moufid, N.; Kuo, L. H.; Curran, D. P. J . Org.
Chem. 1994, 59, 3547. (e) Nishida, M.; Ueyama, E.; Hayashi, H.;
Ohtake, Y.; Yamaura, Y.; Yanaginuma, E.; Yonemitsu, O.; Nishida,
A.; Kawahara, N. J . Am. Chem. Soc. 1994, 116, 6455. (f) Yamamoto,
Y.; Onuki, S.; Yumoto, M.; Asao, N. J . Am. Chem. Soc. 1994, 116, 421.
(g) Renaud, P.; Bourquard, T.; Gerster, M.; Moufid, N. Angew. Chem.,
Int. Ed. Engl. 1994, 33, 1601. (h) Nishida, M.; Hayashi, H.; Yamura,
Y.; Yanaginuma, E.; Yonemitsu, O.; Nishida, A.; Kawahara, N.
Tetrahedron Lett. 1995, 36, 269. (i) Sibi, M. P.; J i, J . J . Am. Chem.
Soc. 1996, 118, 3063. (j) Nishida, M.; Hayashi, H.; Nishida, A.;
Kawahara, N. J . Chem. Soc., Chem. Commun. 1996, 579. (k) Wu, J .
H.; Radinov, R.; Porter, N. A. J . Am. Chem. Soc. 1995, 117, 11029. (l)
Sibi, M. P.; J asperse, C. P.; J i, J . J . Am. Chem. Soc. 1995, 117, 10779.
(4) Lead references to acyl radicals in synthesis: (a) Crich, D.; Yao,
Q. J . Org. Chem. 1996, 61, 3566. (b) Crich, D.; Chen, C.; Hwang, J .-T.;
Yuan, H.; Papadatos, A.; Walter, R. I. J . Am. Chem. Soc. 1994, 116,
8937. For reviews see: (c) Na´jera, C.; Yus, M. Org. Prep. Proc. Int.
1995, 27, 383. (d) Crich, D.; Yuan, H. in Advances in Free Radical
Chemistry; Rawal, V. K., Ed.; J AI: Greenwich, 1997; Vol. 2, in press.
(5) Preliminary communication: Bertrand, M.; Crich, D.; Nouguier,
R.; Samy, R.; Stien, D. J . Org. Chem. 1996, 61, 3588.
(6) For reviews on C2 symmetry in asymmetric synthesis: (a)
Whitesell, J . K. Chem. Rev. 1989, 89, 1581. (b) Alexakis, A.; Mangeney,
P. Tetrahedron: Asymmetry 1990, 1, 477.
(7) (a) Rao, B. V.; Chan, J . B.; Moskowitz, N.; Fraser-Reid, B. Bull.
Soc. Chim. Fr. 1993, 130, 229. (b) Fraser-Reid, B.; Hicks, D. R.; Walker,
D. L.; Iley, D. E.; Yunker, M. B.; Tam, S. Y.-K.; Anderson, R. C.
Tetrahedron Lett. 1975, 297. (c) Nishida, A.; Nishida, M.; Yonemitsu,
O. Tetrahedron Lett. 1990, 31, 7035. (d) Curran, D. P.; Chen, M.-H.;
Spletzer, E.; Seong, C. M.; Chang, C.-T. J . Am. Chem. Soc. 1989, 111,
8872. (e) Curran, D. P.; Shen, W. J . Am. Chem. Soc. 1993, 115, 6051.
(f) Bogen, S.; J ournet, M.; Malacria, M. Synlett. 1994, 958. For another
methodology see: (g) Renaud, P.; Vionnet J .-P.; Vogel, P. Tetrahedron
Lett. 1991, 32, 3491.
(8) (a) Corey E. J .; Suggs, J . W. J . Org. Chem. 1975, 40 , 2554. (b)
Stork, G.; Sher, P. M. J . Am. Chem. Soc. 1986, 108, 303.
S0022-3263(96)01750-1 CCC: $14.00 © 1997 American Chemical Society

276 J . Org. Chem., Vol. 62, No. 2, 1997
Stien et al.
Sch em e 1a
F igu r e 2.
Sch em e 3
a
Key: (a) O3, Me2S; (b) HC(OEt)3, TsOH; (c) NaH, CH2dC(Br)-
CH2Br; (d) CSA, toluene, ∆; (e) H2SO4, SiO2; (f) TMSOTf, -80 f
40 °C.
Sch em e 2a
cyclization of 1,3-dioxolan-2-yl radicals led us to inves-
tigate C2-symmetric 1,3-dioxan-2-yl radicals. Such sp3-
hybridized, σ-type species adopt a chairlike conformation
with the unpaired spin axial.11 In this they are closely
analogous to the 1-alkoxy-1-glycopyranosyl radicals,
which are trapped selectively from the axial direction.12
On this basis we reasoned that 1,3-dioxan-2-yl radicals
bearing an appropriate unsaturated chain at the radical
center would react preferentially along the axial direction
in 5-hexenyl cyclizations. Use of a C2-symmetric system
would result in only two diastereomeric transition states
(Figure 2), one of which would be significantly disfavored
on steric grounds. Racemic substrates 3d -f were sub-
jected to the standard cyclization, and workup, conditions
leading to the results outlined in Scheme 3.
a
Key: (a) Bu3SnCl, NaBH3CN, AIBN, t-BuOH, 80 °C; (b)
Bu3SnH, AIBN, benzene, hν, 20 °C.
Apart from a few percent of the uncyclized reduction
product 10, these reactions each led to three spirocyclic
diastereomers 7-9. These could not be isolated sepa-
rately; consequently, the 1H NMR spectral and GC
analyses were conducted on the mixture. From the NMR
data it was clear that in each case the major component
(9) of the mixture had undergone a structural modifica-
tion in the acetal moiety. Thus, in the trans-dimethyl-
dioxane series, as typified by 3d , two distinct methyl
doublets, reflecting their axial and equatorial nature, are
present at δ 1.15 and 1.32. In contrast, in the cis-series,
for example, 9d , the equatorial nature of the two corre-
sponding methyl groups is evident from the close simi-
F igu r e 1.
OsO4/NaIO4, and subsequent chromatography, a mixture
of 4 and 5 was obtained free of 6 and subjected to analysis
by either capillary GC or 1H NMR spectroscopy. As
reported in Scheme 2, no stereoinduction was observed
with either 3a or 3b, possibly because 1,3-dioxolanyl
radicals are pyramidal,9 which reduces steric interactions
between the R group and the double bond in the assumed
chairlike cyclization transition state (Figure 1).10 Nev-
ertheless, the use of a bulky, tertiary R group (3c)
resulted in a 52% diastereomeric excess at 80 °C and 64%
at 20 °C.
larities of their chemical shifts (δ 1.10 and 1.14).
A
parallel observation relates to the corresponding protons
(11) (a) Malatesta, V.; McKelvey, R. D.; Babcock, B. W.; Ingold, K.
U. J . Org. Chem. 1979, 44, 1872. (b) Malatesta, V. Ingold, K. U. J .
Am. Chem. Soc. 1981, 103, 609. (c) Gregory, A. R.; Malatesta, V. J .
Org. Chem. 1980, 45, 122. (d) Hayday, K.; McKelvey, R. D. J . Org.
Chem. 1976, 41, 2222. (e) Beckwith, A. L. J .; Easton, C. J . J . Am. Chem.
Soc. 1981, 103, 615. (f) Beckwith, A. L. J .; Brumby, S. J . Chem. Soc.,
Perkin Trans. 2 1987, 1801.
(12) Lead references: (a) Crich, D.; Ritchie, T. J . Tetrahedron 1988,
44, 2319. (b) Crich, D.; Hermann, F. Tetrahedron Lett. 1993, 34, 3385.
(c) Crich, D.; Sun, S.; Brunckova, J . J . Org. Chem. 1996, 61, 605. (d)
Yamazaki, N.; Eichenberger, E. Curran, D. P. Tetrahedron Lett. 1994,
35, 6623. (e) Kahne, D.; Yang, D.; Lim, J . J .; Miller, R.; Paguaga, E. J .
Am. Chem. Soc. 1988, 110, 8716. (f) Sugai, T.; Shen, G.-J .; Ichikawa,
Y.; Wong, C.-H. J . Am. Chem. Soc. 1993, 115, 413.
5-exo Cycliza tion s of 1,3-Dioxa n -2-yl Ra d ica ls.
The failure to induce high diastereoselectivity in the
(9) (a) Brunton, G.; Ingold, K. U.; Roberts, B. P.; Beckwith, A. L. J .;
Krusic, P. J . J . Am. Chem. Soc. 1977, 99, 3177. (b) Beckwith, A. L. J .;
Tindal, P. K. Aust. J . Chem. 1971, 24, 2099.
(10) Note, however, that boatlike transition structures lead to the
same prediction for stereoinduction.

Designed Chiral Acyl Radical Equivalents
J . Org. Chem., Vol. 62, No. 2, 1997 277
Sch em e 4
identify 7 and 8. Consequently, in a first-degree ap-
proach, for instance in the case of 3d , the diastereose-
lectivity of the cyclization step could be either (6 + 74):
20 ) 80:20 or (20 + 74):6 ) 94:6. Furthermore, in
addition to this difficulty, it became obvious that a
complete mechanistic scheme should take into account
the possibility that both diastereoisomeric cyclopentyl-
methyl radicals (B and D) might rearrange according to
Scheme 6. Indeed, there is evidence in the literature that
radicals of types B and D can undergo 1,5-hydrogen shifts
(Scheme 7).14
Sch em e 5a
Although the rearrangement of D (Scheme 6) appeared
less likely than that of B and, indeed, than that of the
comparable, less strained [5.5]spirocyclic system in Scheme
7, it was necessary to check this hypothesis. If both B
and D underwent 1,5-hydrogen transfer the chances of
good stereoinduction in the optically pure series would
be low since the ensuing radicals C and E would lead to
enantiomeric forms of 9 (Scheme 6). Furthermore, since
no separation of 9 and ent-9 could be achieved on chiral
phases,15 a direct measure of the selectivity was impos-
sible, even when starting from an enantiomerically pure
substrate. To approach this problem a series of cycliza-
tions were conducted at different concentrations of
Bu3SnH, based on the assumption that B but not D was
subject to the second 1,5-hydrogen atom transfer. Thus,
the cyclization selectivity (7 + 9):8, or (8 + 9):7 in view
of the uncertainty in the identity of 7 and 8, should be
independent of [Bu3SnH] but the ratio of 7:9 (or 8:9)
should vary in a linear manner with [Bu3SnH] so
enabling a distinction to be made between 7 and 8. In
the event (Table 1), the data were ambiguous, suggesting
that both B and D underwent intramolecular hydrogen
atom abstractions. Nevertheless, examination of molec-
ular models suggested that radical B should undergo 1,5-
hydrogen transfer much more rapidly than D, even if the
six-membered ring were to adopt a boat conformation.
Finally, the cyclization stereoselectivity was deduced
with the aid of a full kinetic analysis of the system.
Assuming the concentrations of all the radical species in
the medium to be stationary, the following equations
giving the ratios [7] + [8] + [9]/[7] and [7] + [8] + [9]/[8]
can be derived (Figure 3). Given the reasonable assump-
a
Key: (a) Bu3SnCl, NaBD3CN, AIBN, t-BuOH, 80 °C.
at positions 4 and 6 of the dioxanyl ring. In 3d the axial
proton is a doublet of doublet of quartet at δ 3.80 and
the equatorial a multiplet at δ 4.08-4.27, whereas in 9d
the corresponding two protons are grouped in a single
multiplet at δ 3.77-3.97. This result can only be
explained by a radical translocation consecutive to the
5-exo cyclization (Scheme 4). To confirm this mechanism
3d was cyclized with catalytic Bu3SnCl and NaBD3CN
as overall reductant. The 2H NMR spectrum of the
reduced products showed that 80% of the deuterium was
attached to the carbon R to the oxygen and that only 20%
of the product was deuterated at the cyclization terminus
(Scheme 5).13 Similar labeling experiments were con-
ducted on the bromide 3a , leading to the conclusion that
no corresponding 1,5-hydrogen shift had occurred in the
dioxolanyl series, once again bearing witness to the
considerable susceptibility of 1,5-hydrogen atom abstrac-
tions to transition state geometry (Scheme 5).12c,d
These unanticipated results posed several new ques-
tions. As no separation of the diastereomeric products
could be achieved, it was not possible to unambiguously
Sch em e 6

278 J . Org. Chem., Vol. 62, No. 2, 1997
Stien et al.
Sch em e 7
Ta ble 1. Red u ction of 3d a t Differ en t Tin Hyd r id e
Con cen tr a tion s
product ratio
[Bu3SnH] ×
(7d +
9d )/8d
(8d +
9d )/7d
103 (M)
7d
8d
9d
10d
5.4
13.0
15.6
19.4
38.9
45.4
5.8
9.9
10.7
11.3
12.8
11.8
18
61.7
40.1
35.7
30.8
16.2
14.7
14.5
28.9
32.7
37.7
54.8
58.6
3.75
2.37
2.22
2.08
1.79
1.78
13.74
6.18
5.29
4.51
2.53
2.51
21.1
20.9
20.2
16.2
14.9
14,16
tion that k7 , k9 and, symmetrically, that k8 , k10
the two ratios should depend linearly on 1/[Bu3SnH].
In the event both relationships held, so justifying the
above hypothesis and enabling the assignment of 7 and
8. Graphical and mathematical resolution of the data
(Supporting Information) led to the ratio k1/k2 ) 1.3 for
the cyclization of 3d . This selectivity was even lower
than that previously obtained with 1,3-dioxolan-2-yl
radicals in the case of 3c.
F igu r e 3.
Sch em e 8
Attempts to slow down the rearrangements of radicals
B and D by taking advantage of the primary kinetic
isotopic effect were carried out with acetals derived from
(()-2,4-dideuterio-2,4-pentanediol and from (()-3,5-di-
deuterio-2,2,6,6-tetramethyl-3,5-heptanediol.17 Unfortu-
nately, no significant modification in the distribution of
the products was observed.
In a further series of experiments substrates 11,
designed with a view to eliminating the problematic 1,5-
hydrogen transfer by benzylic stabilization of the cyclized
radical, were prepared by standard methods. When these
substrates were subjected to the standard cyclization
conditions isomerization of the auxiliary was avoided, but
the diastereoselectivities were disappointingly low (Scheme
8). Additional complications are introduced with the use
of a nonterminal vinyl bromide as in 11a ,b owing to the
uncertain nature of the alkene geometry in the crucial
cyclization reaction. Simple, unconjugated, vinyl radicals
are sp2-hybridized with the single electron in an sp2-
hybrid (σ-like) orbital. They exist as mixtures of two
geometric isomers that interconvert via a low barrier.18
If this inversion occurs more rapidly than the 1,5-
hydrogen atom abstraction, the geometry of the alkene
must arise from a balancing of steric interactions between
the vinyl radical substituents and the chiral auxiliary.
Examination of molecular models leads to the conclusion
that for the vinyl radicals derived from 11a ,b, 1,5-
hydrogen atom abstraction will lead preferentially to the
cis-alkene (Scheme 9).19,20 However, it is likely that this
selectivity will not be complete and that mixtures of (E)-
and (Z)-alkenes will be generated.
(13) (a) The deuteration is only partial under these experimental
conditions since deuterium and hydrogen can exchange between
t-BuOH and NaBD3CN; see: Kreevoy, M. M.; Hutchins, J . E. C. J .
Am. Chem. Soc. 1969, 91, 4329. (b) Two broad singlets were registered
in the 2D spectrum at 3.69 and 0.98 ppm, respectively (4:1 relative
integration).
(14) (a) Sugimura, T.; Goto, S.-I.; Koguro, K.; Futagawa, T.; Misaki,
S.; Morimoto, Y.; Yasuoka, N.; Tai, A. Tetrahedron Lett. 1993, 34, 505.
(b) Sugimura, T.; Tai, A.; Koguro, K. Tetrahedron 1994, 50, 11647.
(15) Attempts to achieve the separation of 9 and ent-9 by GC on
FS-hydrodex â-PM or FS-Lipodex A (Macherey Nagel) and by HPLC
on Chiracel ODH (Daicel) were unsuccessful.
(16) For examples of selective axial trapping of closely related
radicals see: (a) Rychnovsky, S. D.; Skalitzky, D. J . Synlett 1995, 555.
(b) Rychnovsky, S. D.; Plzak, K.; Pickering, D. Tetrahedron Lett. 1994,
35, 6799. (c) Giese, B. Angew. Chem., Int. Ed. Engl. 1989, 28, 969.
(17) For the use of primary isotope effects to retard 1,5- and even
intermolecular hydrogen atom abstractions see: (a) Clive, D. L. J .;
Cantin, M.; Khodabocus, A.; Kong, X.; Tao, Y. Tetrahedron 1993, 49,
7917. (b) Warth, J . L.; Frank, B. L.; Christner, D. F.; Absalon, M. J .;
Stubbe, J .; Kozarich, J . W. Biochemistry 1993, 32, 2601. (c) Sugiyama,
H.; Fujimoto, K.; Saito, I.; Kawashima, E.; Sekine, T.; Ishido, Y.
Tetrahedron Lett. 1996, 37, 1805.
(18) For a compilation of ESR spectroscopic data for vinyl radicals
see: Neugebauer, F. A. in Lanholt-Bornstein New Series; Fischer, H.,
Ed.; Springer-Verlag: Berlin, 1989; Vol. 17b, p 492. Also see: Sima-
mura, O. Top. Stereochem. 1969, 4, 1.
(19) A kinetically equivalent scheme applies to more conjugated
π-type vinyl radicals such as the R-phenylvinyl system, which adopts
a linear geometry with the single electron in an sp-hybrid (π-like)
orbital with maximum stabilization from overlap with the conjugated
moiety. Stereoselectivity is again determined by steric factors at the
transition state for quenching.

Designed Chiral Acyl Radical Equivalents
J . Org. Chem., Vol. 62, No. 2, 1997 279
Sch em e 9
F igu r e 4.
A further cyclization was also conducted with a view
to probing the effect of a trisubstituted alkenes on the
selectivity of cyclization (eq 1), but no measureable
selectivity was observed.
Consideration of Schemes 8 and 9 and eq 1 reveals that
cyclization onto cis-styrenes is somewhat less selective
than that to terminal alkenes and that any trans-
substituent on the alkene completely eliminates selectiv-
ity.21 We were thus led to the conclusion that the lack
of stereoselectivity in the entire dioxanyl series originates
from the ability of the dioxane ring to adopt a twist-boat
conformation of type H (Figure 4) in which 1,3-diaxial-
type interactions between the developing bond and the
axial R group in either of the chairs F and G are
avoided.22 We therefore focused our attention on impos-
ing a chairlike conformation on the dioxanyl radical
throughout the course of the cyclization. On the grounds
that the rigid trans-fused bicyclic system would reduce
the incidence of boat and/or twist-boat conformers, trans-
1,3-dioxabicyclo[4.4.0]decan-2-yl transition states I were
targeted (Figure 4).
In I the substituent R is the primary stereodirecting
group and R′′ a group intended to further disfavor boat
and twist-boat conformers. The group R′ () R) was
originally intended simply to be an artifact of the
synthesis enabling ready preparation of the tertiary
alcohol by Grignard reaction with an ester. However,
subsequent computational studies reveal it to play a
major role in destabilizing twist boats. Thus, with the
F igu r e 5.
aid of the MM2 force field within the Chem3D Pro
program, we have modeled chair and twist-boat conform-
ers for the series of dioxanes J -L (Figure 5). In the fully
methylated series (J ) the chair is found to be some 5
kcal‚mol-1 lower in energy than the twist boat. Removal
of the “equatorial” methyl group (K, ≡ I, R′ ) H) puts
the chair and twist-boat conformers within a 1 kcal‚mol-1
range. In the absence of the angular methyl group (L,
≡ I, R′′ ) H) the chair conformation is only 1.7 kcal‚mol-1
lower in energy than the twist boat. Thus, the combined
importance of the “equatorial” and angular methyl groups
in imposing the chair conformation on J is readily
appreciated. Although we have not attempted to model
transition states for the cyclizations, it seems reasonable
to expect a similar influence of R′ and R′′ on the
conformation of the dioxanyl radical.
The racemic diol 21 was prepared as illustrated in
Scheme 10. Thus, â-keto ester 16 was transformed
through alkylation and subsequent reduction with LAH
into diol 18, the relative configuration of which had been
determined by X-ray crystallography.23 Two consecutive
oxidation steps24 led to ester 20, which was converted into
21 by treatment with MeMgI (Scheme 10). Commercial25
(-)-trans-p-menthane-3,8-diol (22)26 was selected to probe
the importance of the angular methyl group.
(20) In principle, this question may be addressed through the use
of higher stannane concentrations such that, following 1,5-hydrogen
atom abstraction, the dioxanyl radical is trapped by the stannane
rather than by cyclization. In practice, such an approach is likely to
result simply in the quenching of the vinyl radical by the stannane.
Such experiments were therefore not attempted.
(21) For substrates 11a ,b, 14, 21, 39, and 45 there is also the
possibility that, due to benzylic stabilization of the cyclized radicals
rendering the cyclization more exothermic, the earlier transition state
with its longer developing bond results in
a reduction in energy
differential between chairs F and G. We thank Professor P. J . Garner,
Case Western Reserve University, for this insight; see ref 2b,c.
(22) trans-4,6-Disubstituted-2,2-dimethyl-1,3-dioxanes adopt twist-
boat conformations: (a) Rychnovsky, S. D.; Rogers, B.; Yang, G. J . Org.
Chem. 1993, 58, 3511. (b) Rychnovsky, S. D.; Yang, G.; Powers, J . P.
J . Org. Chem. 1993, 58, 5251. (c) For conformational analysis of 1,3-
dioxanes: Antenuis, M. J . O.; Tavernier, D.; Borremans, F. Heterocycles
1976, 4, 293.
(23) Smith, W. B. J . Org. Chem. 1979, 44, 1631.
(24) (a) Anelli, P. L.; Biffi, C. Montanari, F.; Quici, S. J . Org. Chem.
1987, 52, 2559. (b) Inch, T. D.; Ley, R. V.; Rich, P. J . Chem. Soc. C
1968, 1693.
(25) Aldrich Chemical Co., Milwaukee, WI.

280 J . Org. Chem., Vol. 62, No. 2, 1997
Stien et al.
Sch em e 10
bridgehead proton, a double doublet with J ) 11 and 5
Hz at δ 3.68 for the precursor 23, was a broad singlet at
δ 3.85 in 25. Further confirmation was obtained from
experiments with Bu3SnD. The minor product, which did
not incorporate deuterium at the bridgehead position in
the presence of Bu3SnD, retained the trans-dioxadecaline
configuration.27 It was tentatively assigned structure 26
and results either from the “equatorial” trapping of the
dioxolanyl radical by the double bond or, more likely,
from a twist-boat conformation of the auxiliary in the
transition structure R (Scheme 12). We note that the
angular methyl group apparently induces a high stereo-
selectivity in this last process. A single diastereomer of
26 resulting from R was detected, the alternative twist-
boat S being sterically disfavored.28 This assignment was
supported by reactions run at different [Bu3SnCl] con-
centrations wherein the 25/26 ratio was not significantly
modified. Therefore, the two products could not derive
from a common intermediate radical of the type O. Since
intermediate Q was precluded on grounds of prohibitive
steric strain, 26 could only be formed via radical T
(Scheme 12). Under the same conditions, 29 gave a
mixture of four spirocyclic products. Two of them (31 and
32) resulted from the axial trapping of the dioxanyl
radical by the double bond. The radical O rearranged,
as in the previous case, to the bridgehead radical via
hydrogen migration. However, in the absence of the
angular methyl group the final hydrogen abstraction from
tin hydride was no longer selective and led to a mixture
of cis- and trans-dioxadecalins.27 As previously shown,
Sch em e 11a
1
the H NMR resonances of the bridgehead protons (td, J
) 10.6, 4.2 Hz at δ 3.39 in 29; broad s at δ 4.21 in 31; td,
J ) 10.6, 4.4 Hz at δ 3.46 ppm in 32) were diagnostic.
Confirmation of the proposed mechanism was again
obtained from experiments conducted with Bu3SnD. Two
other epimeric products 34 (1:1), resulting from equato-
rial trapping by the double bond (or from twist-boat
conformers), accounted for 30% of the mixture (both
incorporated label at the methyl group). Thus, as
expected, appropriately substituted 1,3-dioxabicyclo[4.4.0]-
decan-2-yl radicals improved the stereoinduction with
respect to C2-symmetric 1,3-dioxan-2-yl radicals. In the
best case, however, a 70% ee could be expected after
hydrolysis.
Diol 41 was designed as a readily available, optically
pure equivalent of diol 21 that, furthermore, might
improve stereoinduction due to more severe steric strain
and to the replacement of the angular methyl group by
a tertiary alkyl group.29 It was prepared according to
Scheme 13 from ketopinic acid 3630 and readily converted
to the dioxane 44 by reaction of its bis-TMS ether with
aldehyde 2 under the Noyori-type conditions.31
a
Key: (a) Bu3SnCl, NaBH3CN, t-BuOH, AIBN, 80 °C; (b)
Bu3SnD, AIBN, PhH, syringe pump addition, 80 °C.
When 44 was cyclized under the standard conditions
at 80 °C, a mixture of three products (46, 48, 49) was
isolated in 74% yield (Scheme 14). The two major
products resulted from the preferred cyclization followed
by 1,5-hydrogen atom transfer and nonselective quench-
Transacetalization of bromide 1 with 21 and 22 led to
substrates 23 and 29, respectively, the cyclizations of
which were performed under standard conditions at
reflux in 2-methyl-2-propanol with the results reported
in Scheme 11.
The major product 25 isolated from 23 arises from a
fast 1,5-shift of the bridgehead hydrogen followed by
trapping of the resulting radical P leading to a cis-
dioxadecalin, according to the general mechanism in
Scheme 12. The 1H NMR spectrum provided clear
evidence for the cis-fused nature of 25. Thus, the
(27) Greene, F. D.; Lowry, N. N. J . Org. Chem. 1967, 32, 882.
(28) Unfortunately, no supplementary information could be gained
from NOESY spectra.
(29) More exact, optically pure analogs of 21 might be readily
prepared from natural products such as hederagenin and aphidicolin,
which incorporate the essential diol functionality in 18; however, this
was not pursued due to the relative inaccessibility of such terpenoids.
(30) Bartlett, P. D.; Knox, L. Organic Syntheses; Wiley: New York,
1973; Collect. Vol. V, p 689.
(26) For the configuration of 22 and its stereoisomers see: (a)
Asakawa, Y.; Mutsuda, R.; Tori, M.; Hashimoto, T. Phytochemistry
1988, 27, 3861. (b) Shishibori, T. Bull. Chem. Soc. J pn. 1968, 41, 1170.
(31) Tsunoda, T.; Suzuki, M.; Noyori, R. Tetrahedron Lett. 1980, 21,
1357.

Designed Chiral Acyl Radical Equivalents
J . Org. Chem., Vol. 62, No. 2, 1997 281
Sch em e 12
Sch em e 13
Sch em e 14a
a
Key: (a) Bu3SnCl, NaBH3CN, 80 °C, t-BuOH; (b) Bu3SnH, hν,
20 °C, PhH.
A series of phenyl-substituted substrates 24, 30, and
45 were also prepared from bromide 43 and from diols
21, 22, and 41, respectively, and subjected to the stan-
dard cyclization conditions in 2-methyl-2-propanol at
reflux. In each case (Schemes 11 and 14), the second 1,5-
hydrogen transfer was effectively suppressed but, as
noted for 11 above, the cyclizations were somewhat less
selective than in the corresponding simple, unsubstituted
series presumably for the same reasons. Nevertheless,
when 45 was subjected to cyclization with photochemical
initiation at 20 °C a highly selective cyclization was
observed in which the ratio of products was 94:6, repre-
senting a de of 88% (Scheme 14).
ing. This was confirmed by an experiment with Bu3SnD,
when both incorporated the label at the ring junction.
Compound 46 retained the exo-borneol structure as
indicated by the doublet of doublet nature of the ring
junction proton (δ 3.79). This assignment is further
corroborated by a strong crosspeak in the NOESY
spectrum between the Me doublet at δ 0.95 and the ring
junction proton (δ 3.79). Compound 48 is clearly an endo-
borneol derivative as evidenced by the downfield nature
of the ring junction proton (δ 4.52) and its multiplicity
(ddd) with the additional 4J (W) coupling. The third,
minor product (49) retains the exo-borneol structure (δ
3.80, dd) and does not incorporate deuterium at the ring
junction in the presence of Bu3SnD. This minor product
must arise from “equatorial” quenching of a twist-boat
conformer comparable to R in Scheme 12. When the
cyclization was conducted photochemically at 20 °C, only
46 and 48 were formed (Scheme 14). As both 46 and 48
arise from the same cyclized radical the selectivity in the
cyclization step for 44 at 20 °C is essentially total.
Moreover, hydrolysis of the mixture of 46 and 48 should
result in the optically pure ketone 51.
Finally, we turned to hydrolysis of the cyclized product
mixtures and determination of the ee of the resulting
ketones. Racemic samples of ketones 51 and 52 were
prepared and used to devise appropriate techniques for
the determination of ee’s. Screening of numerous chiral
lanthanide shift reagents failed to reveal conditions for
1
base-line resolution of (()-51 in the H NMR spectrum,
and we eventually turned our attention to the Pirkle
chiral solvating agent 53 (Chart 1). Although this
reagent is most commonly applied to more polar species,

282 J . Org. Chem., Vol. 62, No. 2, 1997
Stien et al.
Ch a r t 1
dioxan-2-yl radicals for disymmetrically substituted 1,3-
dioxabicyclo[4.4.0]decan-2-yl radicals. We have demon-
strated that it is necessary to maintain the 1,3-dioxane
ring in a chair conformation in order to achieve a high
stereoinduction in the cyclization reaction. Conforma-
tional rigidity of the masked acyl radical is of prime
importance, as is further underlined by the recent related
work of Nishida.34 The 2-alkylcyclopentanone 51 was
prepared with a very high enantiomeric excess. Further
developments will be reported in due course.
it has been successfully used for the NMR resolution of
simple lactones.32 In the event, successful base-line
resolution of one wing of an AB quartet assigned to
Exp er im en ta l Section
1
HaHa′ was readily achieved in the H NMR spectrum of
Gen er a l P r oced u r es. 1H NMR spectra were recorded in
CDCl3 at 200, 300, or 400 MHz and 13C NMR spectra in CDCl3
at 50, 75, or 100 MHz as indicated. Chemical shifts (δ) are in
ppm downfield from tetramethylsilane, and coupling constants
(J ) are in Hz. All solvents were dried and distilled by standard
techniques. LC is column chromatography on silica gel.
Aceta liza tion P r oced u r es. Meth od A. A solution of 1
(0.14 mmol/mL of solvent), the appropriate diol (1.1 equiv),
and camphorsulfonic acid (0.1 equiv) in toluene was heated
at reflux for 2 h. After distillation of the toluene-ethanol
azeotrope, the residue was diluted with Et2O and washed
successively with saturated aqueous NaHCO3 and with water.
The organic layer was dried (MgSO4) and concentrated in
vacuo and the residue purified by LC. Meth od B.31 In a
typical experiment, TMSOTf (21 µL, 0.11 mmol, 0.1 equiv) was
added at -80 °C to a solution of 2 (340 mg, 1.06 mmol) and
the bis(trimethylsilyl)-O-protected diol (1.06 mmol, 1 equiv)
in dichloromethane (9 mL). The reaction mixture was allowed
to warm slowly to the specified temperature, and then a few
drops of pyridine were added immediately followed by satu-
rated aqueous NaHCO3. The aqueous phase was extracted
with dichloromethane, and the combined organic layers were
dried (Na2SO4). After removal of the solvent in vacuo, the
residue was purified by LC.
(S *,S *)-2-Br om o-4,4-b is(e t h oxyca r b on yl)-5-(4,5-d i-
m eth yl-1,3-d ioxola n -2-yl)-1-p en ten e (3a ). Method A. The
reaction of 1 (280 mg, 0.71 mmol) with (()-2,3-butanediol led
to 3a (230 mg) in 83% yield: 1H NMR (200 MHz) δ 1.16-1.20
(m, 3H), 1.23-1.27 (superimposed m, 3H), 1.27 (t, J ) 7.1,
6H), 2.39 (d, J ) 4.9, 2H), 3.29 (d, J ) 0.5, 2H), 3.50-3.58 (m,
2H), 4.12-4.25 (m, 4H), 5.17 (t, J ) 4.9, 1H), 5.61 (d, J ) 1.5,
1H), 5.77 (m, 1H); 13C NMR (50.3 MHz) δ 13.6, 16.3, 16.6, 35.6,
43.1, 54.0, 61.2, 61.3, 77.5, 79.5, 100.0, 122.2, 127.2, 169.8.
Anal. Calcd for C16H25BrO6: C, 48.97; H, 6.43. Found: C,
48.90; H, 6.38.
(S ,S )-2-B r o m o -4,4-b is (e t h o x y c a r b o n y l)-5-[4,5-b is -
[(b en zyloxy)m et h yl]-1,3-d ioxola n -2-yl]-1-p en t en e (3b ).
Method A. The reaction of 1 (395 mg, 1.0 mmol) with (S,S)-
dibenzylthreitol led after purification by LC (10-20% ethyl
acetate/pentane) to 3b (471 mg, 78%): 1H NMR (200 MHz) δ
1.22 (t, J ) 7.3, 6H), 2.43 (d, J ) 4.6, 2H), 3.28 (m, 2H), 3.54-
3.62 (m, 4H), 3.88-4.10 (m, 2H), 4.10-4.22 (m, 4H), 4.54 (s,
2H), 4.55 (s, 2H), 5.18 (t, J ) 4.6, 1H), 5.57 (d, J ) 1.5, 1H),
5.74 (m, 1H), 7.31 (m, 10H); 13C NMR (50.3 MHz) δ 13.7, 35.2,
43.4, 54.1, 61.5, 70.1, 70.3, 73.2, 73.3, 77.1, 77.7, 101.9, 122.5,
127.1, 127.3, 127.5, 128.3, 137.7, 170.0. Anal. Calcd for
C30H37O8Br: C, 59.59; H, 6.17. Found: C, 59.57; H, 6.17.
(R ,R)-2-Br om o-4,4-b is(e t h oxyca r b on yl)-5-[4,5-b is(1-
m eth oxy-1-m eth yleth yl)-1,3-dioxolan -2-yl]-1-pen ten e (3c).
Method B: 40 °C, 30 min. TMS triflate (14 µL, 0.07 mmol)
was added to a solution of the bis(trimethylsilyl) ether of (R,R)-
2,5-dimethoxy-2,5-dimethylhexane-3,4-diol (226 mg, 0.64 mmol)
and 2 (207 mg, 0.64 mmol) in CH2Cl2 (5 mL). After workup
and LC (5-10% ethyl acetate/pentane), 3c was isolated (260
mg, 80%): 1H NMR (200 MHz): δ 1.07 (s, 3H), 1.08 (s, 3H),
1.18 (s, 3H), 1.20 (s, 3H), 1.25 (t, J ) 7.2, 6H), 2.38 (d, J ) 5.5,
2H), 3.17 (s, 3H), 3.20 (s, 3H), 3.27 (AB quartet, J AB ) 15.4,
2H), 3.76 (d, J ) 3.6, 1H), 4.2 (d, J ) 3.6, 1H), 4.08-4.27 (m,
(()-51 in CDCl3 solution in the presence of a slight excess
of (S)-53. Numerous variations on standard hydrolysis
conditions were then applied to the cyclization mixtures
from 44 and the ee of the resulting ketone determined.
Unfortunately, in most instances, 51 was found to be
almost or completely racemic, suggesting considerable
susceptibility to enolization under even very mild acidic
conditions. However, the use of PdCl2(CH3CN)2 in ac-
etone33 was found to minimize this problem. Application
of these conditions to the mixture of 46, 48, and 49
derived from the cyclization of 44 at 80 °C gave an 89:11
enantiomeric ratio of ketones, while the mixture obtained
by conducting the cyclization at 20 °C resulted in a 95:5
mixture of enantiomers. These enantiomeric ratios are
marginally lower than those calculated according to the
ratio of 46:48:49 in the mixtures employed indicating
that, although racemization has been minimized, it has
not been completely suppressed. Ketone 51 obtained in
this manner from 46:48:49 was found to be dextrotatory
and to have a negative Cotton effect in the CD spectrum
consistent with it being enriched in the (R)-enantiomer.
This in turn is fully consonant with the above rationale
for diastereoselectivity in the cyclization of 44. Predict-
ably, therefore, hydrolysis of the cyclization mixture from
29 gave 51 moderately enriched in the (-)-enantiomer.
The same hydrolysis protocol was applied to the mixtures
of 47 and 50, giving the benzylcyclopentanone 52 in
excellent yield. Disappointingly, the benzylic methylene
resonances of 52 obscured the HaHa′ AB quartet and so
prevented application of the same convenient direct NMR
determination of enantiomeric purity. Eventually, we
had recourse to the reduction of 52 with NaBH4 at 0 °C
to the mixture of alcohols 54, which was converted to the
corresponding Mosher esters and analyzed by 1H and 19F
NMR. In this manner, the ee of 52 obtained by cycliza-
tion of 45 at 80 and 20 °C followed by Pd(II)-catalyzed
hydrolysis was found to be 38 and 62%, respectively. It
will be noted that the liberation of ketone 52 from the
various cyclization mixtures appears to have occurred
with significantly greater racemization than was the case
with 51. We suspect that this is an artifact of the method
used to measure ee and that significant racemization is
taking place in the course of the borohydride reduction,
meaning that the values reported for 52 are only lower
limits.
In summary, this paper reports the progression of our
concept of asymmetric acyl radical equivalents. The
design of these reactive species led us to abandon
successively C2-symmetric 1,3-dioxolan-2-yl and 1,3-
(32) (a) Pirkle, W. H.; Sikkenga, D. L.; Pavlin, M. S. J . Org. Chem.
1977, 42, 384. (b) Pirkle, W. H.; Adams, P. E. J . Org. Chem. 1978, 43,
378.
(33) Lipshutz, B. H.; Pollart, D.; Monforte, J .; Kotsuki, H. Tetrahe-
dron Lett. 1985, 26, 705.
(34) Nishida, A.; Kawahara, N.; Nishida, M.; Yonemitsu, O. Tetra-
hedron 1996, 52, 9713.

Designed Chiral Acyl Radical Equivalents
J . Org. Chem., Vol. 62, No. 2, 1997 283
4H), 5.27 (t, J ) 5.5, 1H), 5.59 (d, J ) 1.5, 1H), 5.72-5.75 (m,
1H); 13C NMR (50.3 MHz) δ 13.7, 18.7, 21.1, 22.1, 35.7, 43.1,
49.0, 49.3, 54.4, 61.5, 75.6, 77.1, 81.3, 83.9, 102.2, 122.2, 127.4,
170.0. Anal. Calcd for C22H37BrO8: C, 51.87; H, 7.32.
Found: C, 51.84; H, 7.35.
(S *,S *)-2-Br om o-4,4-b is(e t h oxyca r b on yl)-5-(4,6-d i-
m eth yl-1,3-d ioxa n -2-yl)-1-p en ten e (3d ). Method A. The
reaction of 1 (580 mg, 1.47 mmol) with (()-2,4-pentanediol led
to 3d (543 mg, 91%): 1H NMR (200 MHz) δ 1.15 (d, J ) 6.1,
3H), 1.26 (t, J ) 7.3, 6H), 1.32 (d, J ) 7.1, 3H), 1.14-1.34
(superimposed m, 1H), 1.80 (ddd, J ) 13.2, 11.6, 6.1, 1H), 2.32
(d, J ) 5.1, 2H), 3.25 (broad s, 2H), 3.80-4.03 (m, 2H), 4.03-
4.30 (m, 4H), 4.99 (t, J ) 5.1, 1H), 5.59 (d, J ) 1.5, 1H), 5.71
(m, 1H); 13C NMR (50.3 MHz) δ 13.7, 16.8, 21.5, 36.3, 36.4,
43.5, 54.3, 61.3, 67.5, 67.9, 91.3, 122.0, 127.3, 170.0. Anal.
Calcd for C17H27BrO6: C, 50.13; H, 6.68. Found: C, 50.59; H,
6.83.
(S*,S*)-2-Br om o-4,4-b is(et h oxyca r b on yl)-5-(4,6-d iiso-
p r op yl-1,3-d ioxa n -2-yl)-1-p en ten e (3e). Method A. The
reaction of 1 (1.186 g, 3.0 mmol) with (()-2,6-dimethyl-3,5-
heptanediol led, after purification by LC (5% ethyl acetate/
pentane) to 3e (770 mg, 55%): 1H NMR (200 MHz) δ 0.84 (d,
J ) 6.3, 3H), 0.86 (d, J ) 6.8, 3H), 0.91 (d, J ) 6.8, 3H), 0.96
(d, J ) 6.6, 3H), 1.24 (t, J ) 7.3, 3H), 1.25 (t, J ) 7.3, 3H),
1.49-1.79 (m, 3H), 2.14-2.30 (m, 1H), 2.38 (d, J ) 5.1, 2H),
3.19 (dd, J ) 15.1, 0.5, 1H), 3.35 (dd, J ) 15.1, 0.5, 1H), 3.28-
3.39 (m, 1H), 3.42-3.52 (m, 1H), 4.10-4.25 (m, 4H), 4.86 (t, J
) 5.1, 1H), 5.60 (d, J ) 1.7, 1H), 5.77 (m, 1H); 13C NMR (50.3
MHz) δ 13.7, 17.7, 18.1, 19.3, 19.3, 26.9, 27.9, 32.6, 36.1, 43.1,
54.3, 61.3, 61.4, 77.0, 78.2, 92.7, 122.1, 127.4, 170.0. Anal.
Calcd for C21H35BrO6: C, 54.53; H, 7.63. Found: C, 54.41; H,
7.69.
(S*,S*)-2-Br om o-4,4-bis(eth oxyca r bon yl)-5-(4,6-ter t-bu -
tyl-1,3-d ioxa n -2-yl)-1-p en ten e (3f). Method A. 1 (197.5 mg,
0.5 mmol) and (()-2,2,6,6-tetramethyl-3,5-heptanediol (113
mg, 0.6 mmol) in benzene (25 mL) gave, after 2 h and LC (6/1
hexane/ethyl acetate), 3f as an oil (203 mg, 83%): 1H NMR
(300 MHz) δ 0.82 (s, 18H), 1.23 (t, J ) 5.1, 6H), 1.58 (m, 2H).
2.30 (dd, J ) 3.9, 14.7, 1H), 2.44 (dd, J ) 7.3, 14.8, 1H), 3.27
(m, 3H), 4.17 (m, 4H), 4.97 (dd, J ) 3.8, 7.3, 1H), 5.59 (d, J )
1.51, 1H), 5.73 (s, 1H); 13C NMR (75 MHz) δ 13.8, 13.9, 25.3,
25.8, 26.9, 33.8, 34.1, 35.6, 43.2, 54.4, 61.4, 61.5, 74.7, 80.5,
96.9, 121.9, 127.8, 170.0, 170.1; ν (CH2Cl2) 1734 cm-1. Anal.
Calcd for C23H39BrO6: C, 56.21; H, 8.00. Found: C, 56.21; H,
7.76.
Gen er a l P r oced u r e for Ra d ica l Cycliza tion s. AIBN
(0.2 equiv) was added in portions over 5 h to a refluxing
solution of 3 (0.05 M) in degassed 2-methyl-2-propanol,
containing Bu3SnCl (0.1 equiv) and NaBH3CN (2 equiv). After
evaporation of the solvent different workups were performed
depending on the substrate: direct purification of the residue
by LC; ozonolysis and then LC; oxidation with OsO4/NaIO4 in
THF-H2O (4:1) followed by extraction with dichloromethane
and purification by LC. Deuteration experiments were con-
ducted by slowly adding (3 h) with a syringe pump a solution
of Bu3SnD (1.5 equiv) and AIBN (0.15 equiv) in benzene (∼1
mL) to a refluxing solution of substrate (0.05 M) in benzene.
The reaction mixture was heated 2 h more after the end of
the addition and then concentrated.
Cycliza tion of 3a . Cyclization of 3a (212 mg, 0.54 mmol)
led, after purification of the crude product by LC (0 to 5% ethyl
acetate/pentane), to a mixture of 4a , 5a , and 6a (122 mg, 72%).
After reductive ozonolysis and a second purification by LC (5%
ethyl acetate/pentane) a 1:1 mixture of 4a and 5a (74 mg, 45%)
was isolated.
Dieth yl 1,7,8-tr im eth yl-6,9-d ioxa sp ir o[4.4]n on a n e-3,3-
d ica r boxyla te (4a , 5a ): 1H NMR (200 MHz) δ 0.94 (d, J )
6.6, 3H), 1.19-1.28 (m, 12H), 1.88 (dd, J ) 12.4, 5.3, 0.5H),
1.94 (dd, J ) 12.6, 4.0, 0.5H), 2.05-2.25 (m, 1H), 2.33 (d, J )
14.2, 0.5H), 2.47 (d, J ) 14.2, 0.5H), 2.38-2.48 (m superim-
posed, 1H), 2.59 (d, J ) 14.2, 0.5H), 2.61 (d, J ) 14.2, 0.5H),
3.43-3.71 (m, 2H), 4.10-4.25 (m, 4H); 13C NMR (50.3 MHz) δ
11.8, 13.9, 16.0, 16.2, 17.2, 38.1, 38.3, 40.4, 40.6, 43.5, 44.5,
55.5, 55.8, 61.2, 61.4, 77.6, 78.5, 79.1, 79.2, 115.0, 115.3, 171.5,
172.0, 172.2. Anal. Calcd for C16H26O6: C, 61.11; H, 8.34.
Found: C, 61.09; H, 8.30.
Cycliza tion of 3b. Cyclization of 3b (200 mg, 0.33 mmol)
led, after purification of the crude product by LC (0-10% ethyl
acetate/pentane), to a mixture of 4b, 5b, and 6b (130 mg, 75%).
After reductive ozonolysis and a second purification by LC
(10% ethyl acetate/pentane) a 58:42 mixture of 4b and 5b (74
mg, 45%) was isolated.
Dieth yl (2S,3S,9S*)-2,3-bis[(ben zyloxy)m eth yl]-6-m eth -
yl-1,4-d ioxa sp ir o[4.4]n on a n e-8,8-d ica r boxyla te (4b, 5b):
1H NMR (200 MHz) δ 0.93 (d, J ) 6.6, 3H), 1.18-1.27 (m, 6H),
1.90 (dd, J ) 12.7, 3.2, 0.42H), 1.96 (dd, J ) 12.7, 2.9, 0.58H),
2.13-2.32 (m, 1H), 2.32-2.53 (m, 2H), 2.64 (d, J ) 14.4,
0.42H), 2.67 (d, J ) 14.4, 0.58H), 3.55-3.64 (m, 4H), 3.81-
4.15 (m, 2H), 4.10-4.22 (m, 4H), 4.55-4.57 (m, 4H), 7.28-
7.32 (m, 10H); 13C NMR (50.3 MHz) δ 11.8, 11.82, 13.9, 38.0,
38.3, 40.0, 40.4, 43.2, 44.0, 55.6, 55.7, 61.3, 61.4, 69.7, 69.9,
70.1, 70.5, 70.7, 73.2, 73.3, 76.6, 78.0, 78.1, 78.3, 116.7, 117.1,
127.4, 127.5, 128.3, 137.8, 137.9, 171.5, 172.0, 172.1. Anal.
Calcd for C30H38O8: C, 68.41; H, 7.28. Found: C, 68.45; H,
7.31.
Cycliza tion of 3c. Cyclization of 3c (102 mg, 0.2 mmol)
led after purification by LC (0-10% ethyl acetate/pentane) to
a mixture of 4c and 5c (70 mg, 81%) in a 73:27 ratio (GC),
only traces of 6c were detected. When a mixture of 3c (20.4
mg, 0.04 mmol), Bu3SnH (16 µL, 0.06 mmol), and AIBN (1.3
mg, 0.008 mmol) in solution in benzene (4 mL) was irradiated
for 3 h at rt, 4c and 5c (82:18) were isolated (15.4 mg, 64%)
after ozonolysis.
Dieth yl (2R,3R,9R*)-2,3-bis(1-m eth oxy-1-m eth yleth yl)-
6-m eth yl-1,4-dioxaspir o[4.4]n on an e-8,8-dicar boxylate (4c,
5c): 1H NMR (200 MHz) δ 0.93 (d, J ) 6.8, 2.1H), 0.96 (d, J )
6.8, 0.9H), 1.10-1.29 (m, 18H), 1.77-1.98 (m, 1.3H), 2.10-
2.30 (m, 0.7H), 2.35-2.70 (m, 2.3H), 2.77 (d, J ) 14.2, 0.7H),
3.17, 3.18, 3.20 (3 x s, 6H), 3.84 (d, J ) 5.9, 0.7H), 3.89 (d, J
) 5.1, 0.3H), 3.98 (d, J ) 5.9, 0.7H), 4.01 (d, J ) 5.1, 0.3H),
4.02-4.27 (m, 4H). Anal. Calcd for C22H38O8: C, 61.37; H,
8.90. Found: C, 61.72; H, 9.06.
Cycliza tion of 3d . Cyclization of 3d (244 mg, 0.6 mmol)
led, after purification of the crude product by LC (10% ethyl
acetate/pentane) to a mixture of 7d , 8d , and 9d (140 mg, 71%)
in a 6/20/74 relative ratio and to 10 (20 mg, 10%).
Dieth yl 1,7,9-tr im eth yl-6,10-d ioxa sp ir o[4.5]d eca n e-3,3-
d ica r boxyla te (m a jor isom er 9d ): 1H NMR (200 MHz) δ
0.95 (d, J ) 6.6, 3H), 1.10 (d, J ) 5.9, 3H), 1.14 (d, J ) 6.1,
3H), 1.24 (t, J ) 7.1, 3H), 1.25 (t, J ) 7.1, 3H), 1.30-1.40
(superimposed m, 1H), 1.49 (dt, J ) 12.9, 2.4, 1H), 1.91 (dd, J
) 12.5, 11.5, 1H), 1.98-2.13 (m, 1H), 2.22 (d, J ) 13.7, 1H),
2.48 (dd, J ) 12.7, 7.3, 1H), 3.05 (d, J ) 13.7, 1H), 3.77-3.97
(m, 2H), 4.17 (q, J ) 7.1, 4H); 13C NMR (50.3 MHz) δ 11.9,
13.9 (2), 21.6 (2), 37.4, 37.6, 40.3, 42.8, 56.4, 61.3, 61.4, 65.6,
67.6, 106.2, 171.7, 172.3. Anal. Calcd for C17H28O6: C, 62.16;
H, 8.60. Found: C, 62.01; H, 8.62.
(S*,S*)-4,4-Bis(eth oxyca r bon yl)-5-(4,6-d im eth yl-1,3-d i-
oxa n -2-yl)-1-p en ten e (10d ): 1H NMR (200 MHz) δ 1.15 (d, J
) 6.1, 3H), 1.24 (t, J ) 7.1, 6H), 1.31 (d, J ) 7.1, 3H), 1.10-
1.30 (superimposed m, 1H), 1.80 (ddd, J ) 13.2, 11.5, 6.1, 1H),
2.20 (AB part of an ABX pattern, 2H), 2.71 (dt, J ) 7.3, 0.9,
2H), 3.81-3.97 (m, 1H), 4.08-4.27 (m, 5H), 4.96 (t, J ) 5.2,
1H), 5.06 (m, 1H), 5.13 (m, 1H), 5.58-5.79 (m, 1H). Anal.
Calcd for C17H28O6: C, 62.18; H, 8.59. Found: C, 62.19; H,
8.57.
Cycliza tion of 3e. Cyclization of 3e (185 mg, 0.4 mmol)
led, after purification of the crude product by LC (0-3% ethyl
acetate/pentane), to a mixture of 7e, 8e, and 9e (134 mg (34
mg of a pure sample of 9e were separated), 87%) in a 3/14/83
relative ratio and to 10e (17 mg, 11%).
Dieth yl 7,9-d iisop r op yl-1-m eth yl-6,10-d ioxa sp ir o[4.5]-
d eca n e-3,3-d ica r boxyla te (m a jor isom er 9e): 1H NMR (200
MHz) δ 0.88 (d, J ) 6.5, 3H), 0.89 (d, J ) 6.5, 6H), 0.91 (d, J
) 6.5, 3H), 0.98 (d, J ) 6.6, 3H), 1.27 (t, J ) 7.1, 6H), 1.27
(superimposed m, 1H), 1.46 (dt, J ) 12.4, 2.4, 1H), 1.52-1.72
(m, 2H), 1.92 (dd, J ) 12.5, 10.0, 1H), 2.02-2.20 (m, 1H), 2.20
(d, J ) 14.6, 1H), 2.47 (dd, J ) 12.5, 7.3, 1H), 3.05 (d, J )
14.6, 1H), 3.33-3.53 (m, 2H), 4.19 (q, J ) 7.1, 2H), 4.14-4.26
(superimposed m, 2H); 13C NMR (50.3 MHz) δ 11.7, 13.9, 17.5,
17.8, 17.9, 18.1, 30.3, 32.9, 33.1, 37.7, 43.0, 56.4, 61.2, 61.4,

284 J . Org. Chem., Vol. 62, No. 2, 1997
Stien et al.
74.1, 76.3, 105.9, 171.7, 172.5. Anal. Calcd for C21H36O6: C,
65.58; H, 9.44. Found: C, 65.54; H, 9.34.
Cycliza tion of 11b. Cyclization of 11b (84 mg, 0.15 mmol)
followed by treatment with catalytic OsO4 and NaIO4 (160 mg,
0.75 mmol) and purification by LC (95/5 hexane/ethyl acetate)
gave 12b and 13b as an inseparable mixture of diastereomers
(44 mg, 60%).
(S*,S*)-4,4-Bis(eth oxyca r bon yl)-5-(4,6-d iisop r op yl-1,3-
d ioxa n -2-yl)-1-p en ten e (10e): 1H NMR (200 MHz) δ 0.83 (d,
J ) 6.6, 3H), 0.86 (d, J ) 6.8, 3H), 0.91 (d, J ) 6.8, 3H), 0.96
(d, J ) 6.6, 3H), 1.23 (t, J ) 7.2, 3H), 1.24 (t, J ) 7.2, 3H),
1.50-1.80 (m, 3H), 2.12-2.36 (m, 3H), 2.62-2.85 (m, 2H), 3.33
(ddd, J ) 11.2, 6.1, 2.9, 1H), 3.47 (ddd, J ) 12.0, 8.8, 1.5, 1H),
4.02-4.25 (m, 4H), 4.81 (dd, J ) 5.9, 3.9, 1H), 5.02-5.16 (m,
2H), 5.61-5.82 (m, 1H). Anal. Calcd for C21H36O6: C, 65.60;
H, 9.44. Found: C, 65.63; H, 9.46.
Cycliza tion of 3f. Cyclization of 3f (166 mg, 0.34 mmol)
gave after treatment with catalytic OsO4 and NaIO4 (364 mg,
1.7 mmol) and purification by LC (95/5 hexane/ethyl acetate)
7f, 8f, and 9f as an inseparable mixture of diastereomers (90
mg, 65%).
Dieth yl 1-Ben zyl-7,9-d i-ter t-bu tyl-6,10-d ioxa sp ir o[4.5]-
d eca n e-3,3-d ica r boxyla te (12b a n d 13b): 1H NMR (300
MHz) δ 0.84 (s, 9H), 0.86 (s, 9H), 0.90 (s, 9H), 0.92 (s, 9H),
1.21 (m, 12H), 1.57 (m, 4H), 2.10 (m, 3H), 2.44 (m, 4H), 2.51
(m, 3H), 2.65 (d, J ) 13.8, 2H), 3.06 (t, J ) 10.0, 2H), 3.44 (m,
4H), 4.14 (m, 8H), 7.14 (m, 6H), 7.28 (m, 4H); 13C NMR (75
MHz) δ 13.8, 13.9, 25.0, 25.2, 25.3, 25.5, 27.3, 27.5, 33.4, 33.9,
34.1, 34.5, 35.2, 35.6, 41.2, 41.9, 48.5, 55.4, 55.8, 61.2, 61.4,
75.1, 75.3, 75.5, 75.8, 108.2, 108.4, 125.6, 128.1, 128.9, 140.9,
141.1, 171.7, 172.2; ν (CH2Cl2) 1731 cm-1
. Anal. Calcd for
C29H44O6: C, 71.28; H, 9.08. Found: C, 71.12; H, 9.11.
(S *,S *)-2-Br om o-4,4-b is(e t h oxyca r b on yl)-5-(4,6-d i-
m et h yl-1,3-d ioxa n -2-yl)-1,1-d ip h en yl-1-p en t en e (14).
Method A. 2-Bromo-4,4-bis(ethoxycarbonyl)-6,6-diethoxy-1,1-
diphenyl-1-hexene (99 mg, 0.18 mmol) and (()-2,4-pentanediol
(23 mg, 0.22 mmol) gave 14 (90 mg, 90%): 1H NMR (300 MHz)
δ 0.87 (d, J ) 7.0, 3H), 1.08 (d, J ) 6.1, 3H), 1.20 (m, 6H),
1.35 (m, 1H), 1.67 (dd, J ) 12.2, 6.1, 1H), 2.28 (dd, J ) 14.5,
3.8, 1H), 2.41 (dd, J ) 14.5, 6.9, 1H), 3.50 (d, J ) 15.3, 1H),
3.52 (m, 1H), 3.82 (d, J ) 15.3, 1H), 4.09 (m, 5H), 4.37 (dd, J
) 6.9, 3.9, 1H), 7.26 (m, 10H); 13C NMR (75 MHz) δ 13.7, 16.4,
21.6, 36.4, 36.6, 39.6, 55.1, 61.3, 61.4, 67.1, 67.8, 90.7, 121.5,
127.3, 127.7, 128.0, 128.3, 128.5, 129.2, 129.7, 140.0, 143.8,
146.4, 170.2, 170.7. Anal. Calcd for C29H35BrO6: C, 62.26;
H, 6.31. Found: C, 62.15; H, 6.22.
Dieth yl 7,9-d i-ter t-bu tyl-1-m eth yl-6,10-d ioxa sp ir o[4.5]-
d eca n e-3,3-d ica r boxyla te (7f, 8f, a n d 9f): 1H NMR (300
MHz) δ 0.82 (s, 9H), 0.85 (s, 9H), 0.93 (d, J ) 6.6, 3H), 1.23
(m, 6H), 1.29 (t, J ) 2.6, 2H), 1.91 (d, J ) 12.5, 1H), 2.07 (m,
2H), 2.38 (dd, J ) 12.8, 7.1, 1H), 3.06 (d, J ) 13.9, 1H), 3.23
(dd, J ) 11.3, 2.3, 1H), 3.34 (dd, J ) 11.3, 2.6, 1H), 4.16 (m,
4H); 13C NMR (75 MHz) δ 11.4, 13.9, 25.1, 25.4, 25.6, 33.8,
34.1, 37.9, 38.0, 43.3, 56.3, 61.2, 61.4, 75.5, 79.0, 105.8, 171.6,
172.8; ν (CH2Cl2) 1731 cm-1
. Anal. Calcd for C23H40O6: C,
66.96; H, 9.77. Found: C, 66.73; H, 9.69.
(S *,S *)-2-Br om o-4,4-b is(e t h oxyca r b on yl)-5-(4,6-d i-
m eth yl-1,3-dioxan -2-yl)-1-ph en yl-1-pen ten e (11a). Method
A. 2-Bromo-4,4-bis(ethoxycarbonyl)-6,6-diethoxy-1-phenyl-1-
hexene (236 mg, 0.5 mmol) and (()-2,4-pentanediol (63 mg,
0.6 mmol) gave 11a (205 mg, 85%) as an inseparable mixture
of E- and Z-isomers: 1H NMR (300 MHz) δ 0.76 (d, J ) 7.6,
3H), 1.03 (d, J ) 6.1, 3H), 1.22 (m, 15H), 1.33 (d, J ) 6.8, 3H),
1.62 (m, 2H), 1.81 (dd, J ) 12.7, 6.2, 2H), 2.18 (dd, J ) 14.4,
3.6, 1H), 2.31 (m, 1H), 2.37 (d, J ) 5.1, 2H), 3.25 (m, 2H), 3.44
(s, 2H), 3.49 (d, J ) 15.5, 1H), 3.76 (d, J ) 15.4, 1H), 3.92 (m,
2H), 4.17 (m, 9H), 5.06 (t, J ) 5.5, 1H), 6.89 (s, 1H), 7.15 (s,
1H), 7.35 (m, 8H), 7.50 (d, J ) 7.3, 2H); 13C NMR (75 MHz) δ
13.7, 13.9, 16.2, 16.9, 21.5, 21.6, 36.2, 36.3, 36.4, 36.6, 37.0,
45.4, 54.5, 54.9, 61.4, 67.1, 67.6, 67.0, 90.5, 91.5, 120.2, 123.9,
127.6, 127.8, 128.0, 128.5, 128.7, 128.8, 133.0,135.7, 136.6,
170.1, 170.3, 170.4. Anal. Calcd for C23H31BrO6: C, 57.15;
H, 6.46. Found: C, 56.99; H, 6.40.
Cycliza tion of 11a . Cyclization of 11a (106 mg, 0.22 mmol)
in 2-methyl-2-propanol (4.4 mL, 0.05 M), followed by treatment
with catalytic OsO4 and NaIO4 (235 mg, 1.1 mmol) and
purification by LC (95/5 hexane/ethyl acetate), gave 12a and
13a as an inseparable mixture of diastereomers (55 mg, 62%).
Diet h yl 1-b en zyl-7,9-d im et h yl-6,10-d ioxa sp ir o[4.5]-
d eca n e-3,3-d ica r boxyla te (12a a n d 13a ): 1H NMR (300
MHz) δ 1.21 (m, 24H), 1.48 (m, 1H), 1.62 (t, J ) 6.9, 4H), 1.99
(m, 1H), 2.13 (d, J ) 13.9, 2H), 2.39 (m, 7H), 2.78 (d, J ) 13.8,
1H), 2.97 (m, 1H), 3.10 (d, J ) 13.8, 1H), 3.99 (m, 3H), 4.14
(m, 9H), 7.22 (m, 10H); 13C NMR (75 MHz) δ 14.0, 21.2, 21.4,
21.7, 21.9, 33.8, 35.1, 37.8, 40.1, 41.0, 41.8, 47.9, 50.0, 56.1,
61.4, 61.5, 61.8, 63.4, 65.3, 66.2, 105.7, 108.5, 125.5, 125.7,
128.1, 128.2, 128.9, 140.7, 141.2, 171.8, 172.3; ν (CH2Cl2) 1726
cm-1. Anal. Calcd for C23H32O6: C, 68.29; H, 7.97. Found:
C, 68.16; H, 7.81.
(S*,S*)-2-Br om o-4,4-bis(eth oxyca r bon yl)-5-(4,6-d i-ter t-
bu tyl-1,3-d ioxa n -2-yl)-1-p h en yl-1-p en ten e (11b). Method
A. 72 (236 mg, 0.5 mmol) and (()-2,2,6,6-tetramethyl-3,5-
heptanediol (113 mg, 0.6 mmol) gave 11b (255 mg, 90%) as
an inseparable mixture of E- and Z-isomers: 1H NMR (300
MHz) δ 0.67 (s, 9H), 0.78 (s, 9H), 0.86 (s, 9H), 0.88 (s, 9H),
1.19 (t, J ) 7.2, 6H), 1.28 (t, J ) 7.3, 6H), 1.62 (t, J ) 5.3,
2H), 2.21 (dd, J ) 4.3, 14.7, 1H), 2.50 (m, 3H), 3.42 (m, 6H),
4.07 (m, 4H), 4.20 (m, 8H), 4.43 (dd, J ) 4.3, 7.6, 1H), 5.05
(dd, J ) 4.0, 7.2, 1H), 6.96 (s, 1H), 7.14 (s, 1H), 7.31 (m, 8H),
7.50 (d, J ) 8.4, 2H); 13C NMR (75 MHz) δ 13.8, 13.9, 22.6,
25.1, 25.4, 25.7, 25.9, 26.7, 26.9, 31.6, 33.5, 33.8, 33.9, 34.1,
35.7, 35.9, 36.8, 41.5, 45.1, 54.7, 54.9, 61.3, 61.6, 62.9, 74.3,
74.8, 79.9, 80.7, 96.0, 97.2, 120.6, 124.2, 127.6, 127.8, 18.0,
128.5, 128.7, 130.0, 135.8, 136.0, 136.6, 169.9, 170.2, 170.3,
170.6; ν (CH2Cl2) 1739 cm-1. Anal. Calcd for C29H43BrO6: C,
61.37; H, 7.64. Found: C, 61.06; H, 7.79.
Cycliza tion of 14. Cyclization of 14 (67 mg, 0.12 mmol)
followed by treatment with catalytic OsO4 and NaIO4 (130 mg,
0.60 mmol) and purification by LC (95/5 hexane/ethyl acetate)
gave 15a and 15b as an inseparable mixture of diastereomers
(32 mg, 62%).
Diet h yl 1-b en zh yd r yl-7,9-d im et h yl-6,10-d ioxa sp ir o-
[4.5]d eca n e-3,3-d ica r boxyla te (15a a n d 15b): 1H NMR (300
MHz) δ 0.81 (d, J ) 6.4, 6H), 0.90 (d, J ) 6.2, 6H), 1.21 (m,
12H), 1.36 (m, 4H), 1.86 (m, 1H), 1.95 (d, J ) 13.7, 1H), 2.17
(d, J ) 14.1, 1H), 2.29 (dd, J ) 13.8, 8.4, 1H), 2.42 (dd, J )
13.7, 7.9, 1H), 2.62 (d, J ) 13.9, 1H), 2.74 (d, J ) 13.9, 1H),
2.90 (m, 1H), 3.05 (d, J ) 14.0, 1H), 3.16 (m, 2H), 3.33 (m,
1H), 3.86 (m, 1H), 4.11 (m, 11H), 7.11 (m, 4H), 7.29 (m, 16H);
13C NMR (75 MHz) δ 14.0, 20.8, 21.1, 21.8, 22.0, 36.3, 37.0,
37.4, 38.2, 42.7, 43.0, 50.6, 51.2, 53.0, 55.4, 56.0, 61.4, 62.4,
63.2, 64.0, 106.2, 107.6, 125.5, 126.0, 127.8, 128.0, 128.1, 128.2,
128.5, 129.0, 144.5, 144.7, 145.1, 171.6, 171.7, 172.2, 172.4.
Anal. Calcd for C29H36O6: C, 72.48; H, 7.54. Found: C, 72.44;
H, 7.32.
(2R*,4a R*,8a R*)-2-[4-Br om o-2,2-bis(et h oxyca r b on yl)-
4-p en t en -1-yl]-4,4,4a -t r im et h ylh exa h yd r ob en zo-1,3-d i-
oxin (23). Method A. The reaction of 1 (255 mg, 0.644 mmol)
with 21 led to 23 as a single epimer (270 mg, 88%), after the
crude product was purified by LC (3% ethyl acetate-pen-
tane): 1H NMR (300 MHz) δ 1.00-1.10 (m, 1H), 1.05 (s, 3H),
1.06 (s, 3H), 1.25 (t, J ) 7.2, 6H), 1.30 (s, 3H), 1.34-1.64 (m,
6H), 1.70-1.78 (m, 1H), 2.34 (d, J ) 5.4, 2H), 3.21 (dd, J )
15.2, 0.7, 1H), 3.30 (dd, J ) 15.2, 0.7, 1H), 3.68 (dd, J ) 11.0,
5.0, 1H), 4.16 (q, J ) 7.2, 2H), 4.16 (q, J ) 7.2, 2H), 5.00 (t, J
) 5.4, 1H), 5.59 (d, J ) 1.5, 1H), 5.74 (m, 1H); 13C NMR (75
MHz) δ 13.5, 13.8, 19.7, 20.8, 24.0, 24.5, 26.7, 30.8, 35.8, 39.2,
43.3, 54.3, 61.2, 76.74, 78.1, 93.2, 122.0, 127.6, 170.1. Anal.
Calcd for C22H35BrO6: C, 55.58; H, 7.42. Found: C, 55.59; H,
7.48.
(2R*,4a R*,8a R*)-2-[4-Br om o-2,2-bis(et h oxyca r b on yl)-
5-ph en yl-4-pen ten -1-yl]-4,4,4a-tr im eth ylh exah ydr oben zo-
1,3-d ioxin (24). Method A. 2-Bromo-4,4-bis(ethoxycarbonyl)-
6,6-diethoxy-1-phenyl-1-hexene (236 mg, 0.5 mmol) and (()-
21 (104 mg, 0.6 mmol) gave 24 (215 mg, 78%) as an inseparable
mixture of E- and Z-isomers: 1H NMR (300 MHz) δ 0.66 (3H,
s), 0.84 (s, 3H), 0.91 (s, 3H), 1.07 (s, 6H), 1.22 (m, 12H), 1.32
(s, 3H), 1.36-1.76 (m, 16H), 2.16 (dd, J ) 14.4, 3.5, 1H), 2.37
(m, 4H),3.45 (m, 4H), 3.70 (dd, J ) 10.9, 4.9, 1H), 3.78 (d, J )
15.5, 1H), 4.00 (m, 1H), 4.14 (m, 8H), 5.06 (t, J ) 5.2, 1H),
6.95 (s, 1H), 7.14 (s, 1H), 7.33 (m, 8H), 7.50 (d, J ) 7.2, 2H);
13C NMR (75 MHz) δ 13.3, 13.6, 13.8, 13.8, 19.2, 19.8, 20.8,

Designed Chiral Acyl Radical Equivalents
J . Org. Chem., Vol. 62, No. 2, 1997 285
23.9, 24.2, 24.6, 26.6, 26.8, 30.6, 35.9, 36.1, 36.8, 39.2, 39.3,
45.2, 54.5, 54.9, 61.1, 61.3, 76.2, 76.8, 77.4, 78.2, 92.3, 93.4,
120.5, 124.2, 127.6, 127.8, 128.0, 128.6, 128.7, 128.9, 133.1,
1,3-d ioxin (30). Method A. Reaction of 2-bromo-4,4-bis-
(ethoxycarbonyl)-6,6-diethoxy-1-phenyl-1-hexene (471 mg, 1.0
mmol) and 22 (207 mg, 1.2 mmol) gave 30 (507 mg, 92%) as
an inseparable mixture of E- and Z-isomers: 1H NMR δ 0.98
(m, 15H), 1.22 (m, 23H), 1.68 (m, 4H), 1.88 (m, 2H), 2.24 (m,
2H), 2.38 (dd, J ) 5.1, 2.0, 3H), 2.73 (dd, J ) 10.7, 4.0, 1H),
3.44 (m, 5H), 3.79 (d, J ) 15.5, 1H), 3.98 (m, 1H), 4.16 (m,
8H), 5.00 (t, J ) 5.1, 1H), 6.93 (s, 1H), 7.15 (s, 1H), 7.39 (m,
8H), 7.50 (d, J ) 6.9, 2H); 13C NMR δ 13.7, 13.8, 13.9, 17.7,
18.5, 22.2, 24.9, 24.1, 28.5, 28.7, 31.8, 31.2, 34.4, 34.6, 36.1,
36.2, 36.9, 40.4, 40.6, 49.3, 49.5, 55.1, 61.2, 61.4, 61.5, 74.5,
74.7, 75.0, 75.2, 91.6, 92.5, 120.4, 124.1, 127.7, 127.8, 128.1,
128.6, 128.7, 128.9, 133.1, 135.9, 136.0, 136.6, 170.1, 170.4,
170.6; ν (CH2Cl2) 1731 cm-1. Anal. Calcd for C28H39BrO6: C,
60.98; H, 7.13. Found: C, 60.73; H, 7.15.
Cycliza tion of 29. Cyclization of 29 (E ) CO2Me) (1.03 g,
2.3 mmol) led, after purification of the crude product by LC
(0-2% ethyl acetate/pentane) to a complex mixture of products.
After treatment with NaIO4 (5 equiv) and OsO4 (cat.) in THF/
H2O (4/1) and workup, the crude product was purified by LC
(1-4% ethyl acetate/pentane). A mixture of 31, 32, and 34
(two epimers) in 30/40/15/15 relative proportions was isolated
(606 mg, 72%). Anal. Calcd for C20H32O6: C, 65.19; H, 8.75.
Found: C, 64.70; H, 8.68. The 1H NMR spectrum was too
complex to be fully assigned: 1H NMR (300 MHz) of the
mixture δ 0.82-2.08 (m, 22H), 2.33-2.41 (m, 1.5H), 2.65-2.79
(m, 1H), 3.18-3.23 (m, 0.3H), 3.46 (td, J ) 10.6, 4.4, 0.4H,
31), 3.49-3.57 (m, 0.3H, 32 + 34), 3.69-3.74 (m, 6H), 4.21
(m, 0.3H, 32).
135.9, 136.0, 136.7, 170.0, 170.4, 170.6; ν (CH2Cl2) 1732 cm-1
;
HRMS calcd for C28H39BrO6 551.1832, found 551.1838 (M•+).
Cycliza tion of 23. Cyclization of 23 (100 mg, 0.21 mmol)
led, after workup, subsequent treatment with NaIO4 (5 equiv)
and OsO4 (cat.), and purification by LC (0-3% ethyl acetate/
pentane) to 25 (47 mg, 56%) and to a mixture of 25 and 26 (15
mg, 18%). The relative ratio of 25/26 was determined on the
crude mixture from 1H NMR. When the reaction was per-
formed in the presence of 0.2 and 0.4 equiv of Bu3SnCl, this
ratio became, respectively, 82/18 and 81/19.
1
Aceta l 25: H NMR (300 MHz) δ 0.79 (s, 3H), 1.03 (d, J )
6.7, 3H), 1.04 (s, 3H), 1.23 (t, J ) 7.1, 3H), 1.25 (t, J ) 7.1,
3H), 1.30 (s, 3H), 1.15-1.63 (m, 8H), 1.87 (t, J ) 12.5, 1H),
1.95 (d, J ) 13.6, 1H), 1.92-2.12 (m, 1H), 2.34 (dd, J ) 12.8,
7.8, 1H), 3.18 (d, J ) 13.6, 1H), 3.84 (broad s, 1H), 4.12-4.20
(m, 2H), 4.16 (q, J ) 7.1, 2H); 13C NMR (75 MHz) δ 11.4, 13.9,
16.5, 19.3, 21.2, 24.2, 24.6, 25.9, 27.2, 36.0, 36.1, 41.3, 44.2,
56.5, 61.1, 61.3, 67.7, 77.2, 103.9, 171.8, 172.8. Anal. Calcd
for C22H36O6: C, 66.64; H, 9.15. Found: C, 66.50; H, 9.09. The
essential characteristics of 26 taken from a mixture are as
follows: 1H NMR (300 MHz) δ 0.98 (d, J ) 7.1, 3H), 1.03 (s,
3H), 1.10 (s, 3H), 2.41 (dd, J ) 13.4, 2.1, 1H), 2.46 (m, 1H),
2.66 (d, J ) 13.4, 1H), 2.88 (dd, J ) 13.4, 7.9, 1H), 3.74 (dd, J
) 10.9, 4.9, 1H).
Cycliza tion of 24. Cyclization of 24 (55 mg, 0.10 mmol)
followed by treatment with catalytic OsO4 and NaIO4 (107 mg,
0.50 mmol) and purification by LC (95/5 hexane/ethyl acetate)
gave 27a and 28 as an inseparable mixture of diastereomers
(30 mg, 64%).
Cycliza tion of 30. Cyclization of 30 (500 mg, 0.91 mmol)
and treatment with catalytic OsO4 and NaIO4 (973 mg, 4.55
mmol) followed by purification by LC (95/5 hexane/ethyl
acetate) gave 33 and 35 as an inseparable mixture of diastereo-
mers (279 mg, 65%).
Aceta ls 27a a n d 28: 1H NMR δ 0.90 (m, 4H), 0.98 (s, 6H),
1.05 (s, 6H), 1.15 (s, 3H), 1.22 (m, 12H), 1.37 (s, 3H), 1.47 (m,
12H), 1.87 (m, 6H), 2.70 (m, 8H), 3.70 (dd, J ) 10.4, 5.3, 0.35H),
3.83 (dd, J ) 10.3, 5.2, 0.65H), 4.15 (m, 8H), 7.18 (m, 6H),
7.28 (m, 4H); 13C NMR δ 13.5, 13.7, 14.0, 20.9, 21.1, 23.5, 24.5,
24.7, 25.2, 25.5, 26.9, 27.0, 30.9, 31.2, 35.0, 35.5, 35.9, 37.2,
38.5, 38.8, 44.9, 45.1, 45.3, 55.4, 61.2, 61.5, 61.8, 69.5, 71.7,
78.3, 79.2, 107.5, 108.5, 125.8, 128.4, 129.0, 140.8, 141.2, 171.5,
Aceta ls 33a a n d 35: 1H NMR δ 0.98 (m, 13H), 1.23 (m,
24H), 1.53 (m, 10H), 1.96 (m, 2H), 2.50 (m, 7H), 2.99 (m, 4H),
3.55 (m, 2H), 4.14 (m, 4H), 7.21 (m, 10H); 13C NMR δ 13.9,
14.0, 14.1, 22.0, 22.2, 22.6, 23.2, 23.8, 25.5, 25.6, 29.5, 29.8,
31.0, 31.2, 31.5, 33.6, 34.4, 34.5, 34.9, 35.0, 35.2, 35.4, 40.7,
40.9, 41.2, 41.3, 45.8, 46.6, 46.7, 47.7, 48.2, 48.9, 49.2, 50.3,
51.2, 54.7, 55.9, 61.2, 61.4, 66.8, 68.0, 69.1, 70.3, 74.1, 74.6,
75.2, 104.6, 105.6, 125.6, 125.7, 128.0, 128.2, 129.9, 141.1,
141.2, 171.8, 172.0, 172.4; ν (CH2Cl2) 1735 cm-1. Anal. Calcd
for C28H40O6: C, 71.16; H, 8.53. Found: C, 70.82; H, 8.55.
(-)-2-Br om o-4,4-b is(e t h oxyca r b on yl)-6,6-(10,10-d i-
m eth yl-2-exo,10-bor n a n ed ioxy)-1-h exen e (44). Method
B: 0 °C. TMS triflate (21 µL, 0.11 mmol) was added to a
solution of the bis(trimethylsilyl) diether 42 (3.63 mg, 1.06
mmol) and 2 (340 mg, 1.06 mmol). After workup and purifica-
tion by LC (3% ethyl acetate/pentane), 44 (390 mg, 73%) was
171.6, 172.9, 173.2; ν (CH2Cl2) 1734 cm-1
. HRMS calcd for
C28H40O6 478.2824, found 478.2824 (M•+).
(2R,4a R,7R,8a R)-2-[4-Br om o-2,2-bis(eth oxyca r bon yl)-
4-p e n t e n -1-yl]-4,4,7-t r im e t h ylh e xa h yd r ob e n zo-1,3-d i-
oxin (29, E ) CO2Et). Method A. The reaction of 1 (588 mg,
2.5 mmol) with (-)-trans-p-menthane-3,8-diol (22) led after
purification by LC (2% ethyl acetate-pentane) to 29 as a
mixture of epimers (>95:5; 1 g, 84%): 1H NMR (300 MHz)
major isomer δ 0.91 (d, J ) 6.5, 3H), 0.80-1.09 (m, 3H), 1.63-
1.71 (m, 1H), 1.15 (s, 3H); 1.17 (s, 3H), 1.25 (t, J ) 7.1, 6H),
1.25-1.59 (m, 3H), 1.63-1.71 (m, 1H), 1.83-1.92 (m, 1H), 2.33
(m, 2H), 3.15-3.37 (m, 2H), 3.39 (td, J ) 10.6, 4.2, 1H), 4.09-
4.24 (m, 4H), 4.93 (t, J ) 5.2, 1H), 5.59 (d, J ) 1.6, 1H), 5.73
(m, 1H); 13C NMR (75 MHz) δ 13.67, 18.2, 22.0, 24.9, 28.45,
30.9, 34.3, 35.8, 40.3, 43.2, 49.3, 54.2, 61.1, 61.2, 74.7, 74.9,
92.1, 122.0, 127.3, 169.9, 170.0. Anal. Calcd for C22H35BrO6:
C, 55.58; H, 7.42. Found: C, 55.37; H, 7.36. Due to the
complexity of the spectra of the cyclization products, the
homologous dimethyl ester was prepared.
(2R,4aR,7R,8aR)-2-[4-Br om o-2,2-bis(m eth oxycar bon yl)-
4-p e n t e n -1-yl]-4,4,7-t r im e t h ylh e xa h yd r ob e n zo-1,3-d i-
oxin (29, E ) CO2Me). Method A. The reaction of 1a (E )
CO2Me) dimethyl acetal (0.984 g, 2.9 mmol) with (-)-trans-
p-menthane-3,8-diol (22) led after purification by LC (4% ethyl
acetate-pentane) to 29 as a mixture of epimers (>95:5; 1.032
g, 80%): 1H NMR (300 MHz) δ 0.92 (d, J ) 6.5, 3H), 0.81-
1.09 (m, 3H), 1.15 (s, 3H), 1.18 (s, 3H), 1.26-1.59 (m, 3H),
1.64-1.73 (m, 1H), 1.84-1.92 (m, 1H), 2.33 (m, 2H), 3.22 (dd,
J ) 15.2, 0.7, 1H), 3.30 (dd,J ) 15.2, 0.7, 1H), 3.39 (td, J )
10.6, 4.2, 1H), 3.71 (s, 6H), 4.95 (t, J ) 5.3, 1H), 5.60 (d, J )
1.6, 1H), 5.72 (m, 1H); 13C NMR (75 MHz) δ 18.2, 22.0, 24.9,
28.6, 31.0, 34.8, 36.1, 40.4, 43.5, 49.3, 52.3, 52.3, 54.2, 74.8,
isolated: mp ) 97 °C;
1H NMR (CDCl3, 300 MHz) δ 0.82-
1.08 (m, 2H), 1.04 (s, 3H), 1.21 (s, 6H), 1.25 (t, J ) 7.1, 6H),
1.27 (s, 3H), 1.50-1.78 (m, 5H), 2.33 (d, J ) 5.7, 2H), 3.16
(broad d, J ) 15.3, 1H), 3.29 (broad d, J ) 15.3, 1H), 3.78 (dd,
J ) 7.8, 2.9, 1H), 4.11-4.21 (m, 4H), 4.96 (t, J ) 5.7, 1H),
5.59 (d, J ) 1.6, 1H), 5.71-5.72 (m, 1H);
13C NMR (75 MHz)
δ 13.8 21.6, 23.6, 24.0, 25.6, 27.0 27.9, 35.7, 36.6, 43.1, 47.0,
47.2, 50.3, 54.35, 61.2, 61.3, 76.2, 79.9, 90.6 , 121.9, 127.6,
170.0, 170.1. Anal. Calcd for C24H37BrO6: C, 57.49; H, 7.41.
Found: C, 57.43; H, 7.50.
(-)-2-Br om o-4,4-b is(e t h oxyca r b on yl)-6,6-(10,10-d i-
m eth yl-2-exo,10-bor n a n ed ioxy)-1-p h en yl-1-h exen e (45).
Method B. To a solution containing 43 (87 mg, 0.22 mmol)
and 42 (226 mg, 0.66 mmol) in dichloromethane (5 mL) at -78
°C was added TMSOTf (4.2 µL, 0.022 mmol) and allowed to
warm to room temperature over a period of 10 h. After workup
and purification by LC (95/5 hexane/ethyl acetate), 45 (60 mg,
49%) was isolated as an inseparable mixture of E- and
Z-isomers: 1H NMR δ 0.49 (s, 3H), 0.95 (m, 11H), 1.23 (m,
22H), 1.56 (m, 10H), 2.23 (m, 1H), 2.39 (d, J ) 5.5, 2H), 3.36
(m, 4H), 3.76 (m, 2H), 3.99 (m, 1H), 4.17 (m, 9H), 5.03 (t, J )
5.5, 1H), 6.91 (s, 1H), 7.13 (s, 1H), 7.32 (m, 8H), 7.49 (d, J )
7.1, 2H); 13C NMR δ 13.8, 13.9, 21.8, 22.0, 22.8, 23.8, 24.1,
24.2, 25.5, 25.6, 27.0, 27.1, 27.8, 28.1, 35.9, 36.0, 36.7, 45.0,
46.9, 47.1, 47.2, 47.3, 50.2, 50.4, 54.6, 55.1, 61.1, 61.3, 61.4,
61.5, 75.7, 76.3, 79.5, 80.1, 89.8, 90.9, 120.6, 124.1, 127.6, 127.8,
74.9, 92.0, 122.1,127.2, 170.4, 170.5. Anal. Calcd for C20H31
BrO6: C, 53.70; H, 6.98. Found: C, 54.03; H, 6.98.
-
(2R,4a R,7R,8a R)-2-[4-Br om o-2,2-bis(eth oxyca r bon yl)-
5-p h en yl-4-p en ten -1-yl]-4,4,7-tr im eth ylh exa h yd r oben zo-

286 J . Org. Chem., Vol. 62, No. 2, 1997
Stien et al.
128.0, 128.6, 128.7, 128.9, 132.9, 135.8, 136.0, 136.6, 170.0,
NMR data demonstrated that the major enantiomer was
reversed with respect to that from the hydrolysis of 46 + 48.
170.3, 170.4; ν (CH2Cl2) 1735 cm-1; HRMS calcd for C30H41
-
BrO6 (M - H)+ 575.2008, found 575.2007.
4,4-Bis(et h oxyca r b on yl)-2-m et h ylcyclop en t a n -1-on e
1
Cycliza t ion of 44. (a) Cyclization of 44 (138 mg, 0.275
mmol) was conducted at 80 °C according to the standard
procedure. After treatment with NaIO4 (5 equiv) and OsO4
(cat.) and then workup, the crude product was purified by LC
(1-3% ethyl acetate/pentane). A 84/11/5 mixture of 46, 48,
and 49 was isolated (86 mg, 74%).
(51): H NMR δ 1.12 (d, J ) 7.0, 3H), 1.25 (t, J ) 7.2, 6H),
1.96 (t, J ) 12.9, 1H), 2.40 (m, 1H), 2.74 (d, J ) 19.1, 1H),
2.86 (m, 1H), 2.98 (d, J ) 19.1, 1H), 4.22 (m, 4H). Anal. Calcd
for C12H18O5: C, 59.49; H, 7.49. Found: C, 59.12; H, 7.44.
Hyd r olysis of 47 a n d 50. To a 94/6 mixture of 47 and 50
(25 mg, 0.05 mmol) in acetone (2 mL) was added Pd(Cl)2-
(CH3CN)2 (0.65 mg, 0.0025 mmol) and the reaction mixture
allowed to stir at room temperature for 48 h. The reaction
mixture was concentrated and purified by LC (95/5 hexane/
ethyl acetate) to yield 52 (13 mg, 82%). This ketone (13 mg,
0.04 mmol) was dissolved in methanol (2 mL) and stirred at
room temperature with sodium borohydride (1.6 mg, 0.04
mmol) for 2 h. The reaction mixture was concentrated and
purified by LC (1/1 hexane/ethyl acetate) to give an 85/15
trans/cis mixture35 of two diastereomeric alcohols (13 mg,
99%): 1H NMR (major isomer) δ 1.23 (m, 6H), 1.90 (dd, J )
8.7, 13.7, 1H), 2.25 (m, 2H), 2.52 (m, 3H), 2.82 (dd, J ) 6.4,
13.7, 1H), 3.99 (q, J ) 6.0, 1H), 4.18 (m, 4H), 7.21 (m, 3H),
7.30 (m, 2H); 13C NMR δ 13.9, 34.8, 37.4, 39.2, 41.6, 43.2, 47.8,
49.1, 57.4, 61.2, 61.7, 73.8, 125.9, 126.2, 128.4, 128.5, 128.8,
140.1, 172.1, 172.9. To these alcohols (13 mg, 0.04 mmol) in
CCl4 (200 µL) and pyridine (200 µL) was added (+)-Mosher’s
acid chloride (20.6 mg, 0.08 mmol) and the reaction mixture
allowed to stir overnight, after which time TLC indicated
complete conversion. The reaction mixture was diluted with
ether (15 mL) and washed with dilute HCl (5 mL), saturated
sodium carbonate (5 mL), and brine (5 mL). The organic
extracts were dried (MgSO4) and concentrated under vacuum.
The 19F NMR spectrum of the extracts consisted of two major
resonances (19/81) at δ 195.9 and 249.1 as well as a number
of minor signals. Purification by LC (4/1 hexane/ethyl acetate)
gave a clean mixture of two Mosher esters (14 mg, 64%)
assigned as the diastereomers arising from esterification of
the scalemic trans-alcohol: 1H NMR δ 1.18 (m), 1.82 (m), 2.29
(dd, J ) 4.9 and 14.8, 1H), 2.51 (m), 2.86 (m), 3.46 (s, 3H),
3.49 (s, 3H), 3.55 (s, 3H), 3.57 (s, 3H), 4.09 (m), 5.16 (1H, m),
7.19 (m), 7.43 (m); 19F-NMR δ 195.9 and 249.1 in a ratio of
19/81 (ee 62%). Application of the same protocol to 52 derived
from hydrolysis of a 80/20 mixture of 47 and 50 gave the same
two 19F NMR signals in a ratio of 31/69 (ee 39%).
(b) A solution containing 44 (75 mg, 0.15 mmol), Bu3SnH
(48 µL, 0.18 mmol), and AIBN (5 mg, 0.03 mmol) in benzene
(15 mL) was irradiated for 3 h at 18 °C. After concentration,
a first purification by LC (0 to 3% ethyl acetate/pentane)
yielded a mixture of 46 and 48 (in a 87/13 ratio) together with
the uncyclized reduction product and traces of starting mate-
rial to be isolated. Reductive ozonolysis of the mixture at -80
°C in CH2Cl2/MeOH (15 mL/4 mL) led after workup and a
second purification by LC (0-3% ethyl acetate/pentane) to a
mixture of 46 and 48 (32 mg, 51%). Anal. Calcd for
C24H38O6: C, 68.22; H, 9.06. Found: C, 67.81; H, 9.20. No
traces of 49 were detected.
Aceta l 46: 1H NMR (400 MHz) δ 0.87-1.05 (m, 2H), 0.95
(superimposed d, J ) 7.2, 3H), 1.03 (s, 3H), 1.17 (s, 3H), 1.22
(s, 3H), 1.25 (t, J ) 7.2, 3H), 1.26 (t, J ) 7.2, 3H), 1.28 (s, 3H),
1.49-1.69 (m, 4H), 1.68 (superimposed dd, J ) 13.7, 3.3, 1H),
1.70-1.80 (m, 1H), 2.40 (d, J ) 13.6, 2H), 2.58 (d, J ) 13.6,
1H), 2.60-2.69 (m, 1H), 2.91 (dd, J ) 13.7, 3.3, 1H), 3.79 (dd,
J ) 7.8, 2.6, 1H), 4.08-4.22 (m, 4H); 13C NMR (75 MHz) δ
13.9, 13.9, 17.4, 21.8, 23.9, 27.0, 27.1, 28.2, 28.3, 37.2, 37.8,
39.3 45.2, 47.0, 47.2, 55.5, 61.0, 61.3, 75.4, 75.5, 107.3, 171.5,
172.6.
Aceta l 48. Characteristic data deduced from spectra of the
mixture: 1H NMR (400 MHz) δ 2.85 (dd, J ) 13.7, 7.4, 1H),
3.19 (d, J ) 13.7, 1H), 4.10-4.21 (m, 4H), 4.50-4.55 (ddd, J
) 10.5, 7.5, 2.2, 1H).
Aceta l 49. Characteristic data deduced from spectra of the
mixture: 1H NMR (400 MHz) δ 2.99 (d, 1H, J ) 13.6), 3.80
(partially masked dd, with the smallest of the two couplings
J ) 2.5, 1H).
Cycliza tion of 45. (a) Cyclization of 45 (115 mg, 0.20
mmol) according to the standard procedure followed by treat-
ment and purification by LC (95/5 hexane/ethyl acetate) gave
47 and 50 as an inseparable 4/1 mixture of diastereomers (60
mg, 61%). (b) When the experiment was conducted at room
temperature the yield of 47+50 (94/6) was 29%.
2-Ben zyl-4,4-b is(et h oxyca r b on yl)cyclop en t a n -1-on e
(52): 1H NMR δ 1.23 (m, 6H), 2.05 (br.s, 1H), 2.59 (m, 2H),
2.68 (s, 1H), 2.75 (s, 1H), 2.99 (d, J ) 18.9, 1H), 3.19 (dd, J )
13.9, 3.4, 1H), 4.19 (m, 4H), 7.15 (d, J ) 8.2, 2H), 7.22 (d, J )
7.1, 1H), 7.29 (m, 2H); 13C NMR δ 13.9, 35.5, 35.7, 44.9, 54.7,
62.0, 126.4, 128.5, 128.8, 138.8, 170.6, 170.9, 214.4; ν (CH2Cl2)
1744, 1729 cm-1. Anal. Calcd for C18H22O5: C, 67.91; H, 6.96.
Found: C, 67.60; H, 6.91.
Aceta ls 47 a n d 50: 1H NMR δ 0.95 (m, 4H), 1.04 (s, 3H),
1.06 (s, 3H), 1.18 (s, 6H), 1.20 (s, 6H), 1.24 (m, 12H), 1.32 (s,
3H), 1.34 (s, 3H), 1.54 (m, 1H), 1.66 (m, 5H), 1.82 (m, 4H),
2.52 (m, 9H), 2.96 (m, 3H), 3.75 (dd, J ) 7.6, 2.6, 1H), 3.86
(dd, J ) 7.6, 2.4, 1H), 4.17 (m, 8H), 7.17 (m, 6H), 7.27 (m, 4H);
13C NMR δ 14.1, 21.9, 24.0, 27.1, 28.4, 28.6, 35.8, 37.3, 37.4,
45.2, 45.6, 47.1, 47.3, 50.1, 55.4, 61.1, 61.6, 75.7, 76.3, 107.2,
125.7, 128.2, 129.0, 141.2, 170.1, 170.4; HRMS calcd for
C30H42O6 498.2981, found 498.2974.
Hyd r olysis of 46 + 48. A mixture of 46/48/49 (84/11/5)
(60 mg, 142 µmol), in solution in acetone (1.4 mL), was stirred
with PdCl2(CH3CN)2 (1.8 mg, 7 µmol). After 24 h at room
temperature, NaHCO3 was added and the solvent was evapo-
rated. Purification by LC (5-10% ethyl acetate/pentane) led
Ack n ow led gm en t. We thank the CNRS and the
NSF (CHE 9625256, 9222697, INT 9216807) for support
of this work. D.C. is a Fellow of the A. P. Sloan
Foundation. P. J . Garner, Case Western Reserve Uni-
versity, is thanked for a helpful discussion.
1
to 51 (30 mg, 87%). The 72% ee was determined by H NMR
in solution in CDCl3 in the presence of the Pirkle alcohol (53).
The enantiomeric ratio 86/14 indicated that some racemization
had occurred concomitantly with the hydrolysis. The CD
spectrum showed a negative CE with two minima (λ ) 302.5,
Su p p or tin g In for m a tion Ava ila ble: Details of the prepa-
ration and characterization of all the substrates, their precur-
sors, and diols 21 and 41; derivation of the rate laws in the
kinetic analysis (15 pages). This material is contained in
libraries on microfiche, immediately follows this article in the
microfilm version of the journal, and can be ordered from the
ACS; see any current masthead page for ordering information.
ꢀ ) -2.03; λ ) 295, ꢀ ) -2.04; [R]mixture ) +56.6° f [R]D
+78.6°).
)
When the 46/48 mixture (12 mg, 0.028 mmol; isolated from
the radical reaction at 18 °C according to the above procedure)
was treated as above, LC purification led to 51 (6 mg, 87%;
1
J O961750N
90% ee based on 400 MHz H NMR in the presence of 53).
Hydrolysis was also performed on the mixture of 31/32/34
(30/40/30) (8.2 mg, 22 µmol), isolated from 29 in acetone-d6
(0.5 mL). The reaction was monitored by NMR. After workup,
the solvent was evaporated and the residue dissolved in CDCl3.
The two enantiomers were obtained in a 76/24 ratio (52% ee).
(35) This tentative assignment of a poorly resolved (1H and 13C
NMR) mixture, not essential to the argument, is based on typical
reductions of model 2-alkylcyclopentanones: Boone, J . R.; Ashby, E.
C. Top. Stereochem. 1979, 11, 53.