Radical cyclization-fragmentation of ω-haloalkyl cyclobutanones: A modular approach to medium-sized carbocycles

Article abstract of DOI:10.1016/j.tetasy.2003.07.011

A facile radical cyclization-fragmentation sequence of ω-haloalkyl- tethered spirocyclobutanones, which are readily available by the Kulinkovich cyclopropanation of cycloalkene carboxylates and subsequent electrophilic addition to haloalkyl acetals, provides a convenient method for appending seven- and eight-membered rings onto cycloalkene carboxylates. An enantioselective preparation of a medium-sized carbocycle is possible by the use of a non-racemic, C₂-symmetric acetal.

Full text of DOI:10.1016/j.tetasy.2003.07.011

TETRAHEDRON:
ASYMMETRY
Pergamon
Tetrahedron: Asymmetry 14 (2003) 2911–2917
Radical cyclization–fragmentation of v-haloalkyl cyclobutanones:
a modular approach to medium-sized carbocycles
Heong-Sub Oh and Jin Kun Cha*
Department of Chemistry, University of Alabama, Tuscaloosa, AL 35487 USA and
Department of Chemistry, Wayne State University, Detroit, MI 48202 USA
Received 6 June 2003; accepted 10 July 2003
Abstract—A facile radical cyclization–fragmentation sequence of v-haloalkyl-tethered spirocyclobutanones, which are readily
available by the Kulinkovich cyclopropanation of cycloalkene carboxylates and subsequent electrophilic addition to haloalkyl
acetals, provides a convenient method for appending seven- and eight-membered rings onto cycloalkene carboxylates. An
enantioselective preparation of a medium-sized carbocycle is possible by the use of a non-racemic, C2-symmetric acetal.
© 2003 Elsevier Ltd. All rights reserved.
1. Introduction
sized bicyclic compounds, by means of facile radical
fragmentation of spirocyclobutanones containing v-
halo tethers (Scheme 1).8
Free radical-mediated reactions have frequently been
applied in the formation of carbonꢀcarbon bonds.1–3
Among the unique advantages of radical chemistry are
excellent reactivity, yet high levels of predictable regio-
and stereoselectivity, under mild reaction conditions,
and tolerance of several common functional groups.
The utility of free radical cyclization reactions has been
amply demonstrated in the synthesis of carbocyclic and
heterocyclic natural products. Thus, useful methods for
forming medium-sized rings and macrocycles, as well as
five- and six-membered rings, have been developed by
free radical cyclizations. A particularly useful route to
medium-sized carbocycles is based on ring expansion of
strained cyclobutanones via b-scission of alkoxy radi-
cals. For example, Dowd et al. reported an elegant
strategy featuring an inter- or intramolecular [2+2] pho-
tocycloaddition and an alkoxy radical fragmentation
reaction of the resulting, typically fused cyclobutanone;
ring expansion is facilitated by release of ring strain and
does not require a radical stabilizing substituent.4–7 We
herein report an expedient method for the annulation
of seven- and eight-membered rings onto cycloalkene
carboxylates, affording suitably functionalized medium-
* Corresponding author. Tel.: (313) 577-1073; fax: (313) 577-8822;
e-mail: jcha@chem.wayne.edu
Scheme 1.
0957-4166/$ - see front matter © 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetasy.2003.07.011

2912
H.-S. Oh, J. K. Cha / Tetrahedron: Asymmetry 14 (2003) 2911–2917
2. Results and discussion
not be ascertained at this stage, but was subsequently
gleaned from X-ray analysis of the resulting ring expan-
sion product derived from 4b-a (vide infra). The
cognate condensation of 3a and 3b with 2-bromoac-
etaldehyde dimethylacetal yielded 5a and 5b as a nearly
1:1 epimeric mixture at the methoxy stereocenter in 92
and 89% yields, respectively. A cursory survey of other
Lewis acids revealed that TMSOTf, BF3·Et2O, or
Et2AlCl were ineffective in coupling of 3a,b with an
acetal. More recently, however, cyclopropanols 2a,b
were found to undergo clean condensation with certain
bromo- or iodoalkyl acetals by the action of SnCl4 to
afford the respective adducts in comparable or even
better yields. The scope and limitations of the direct
condensation reaction of free vinyl cyclopropanols with
haloalkyl acetals are currently under investigation.
2.1. Synthesis of spiro cyclobutanones
In the pioneering work by Dowd et al.,4,5 the requisite
cyclobutanone substrates were prepared by the [2+2]
cycloaddition reaction of ketenes or keteniminium
salts.9 This [2+2] cycloaddition was amenable to both
intermolecular and intramolecular reactions, but lim-
ited primarily to the preparation of fused cyclobu-
tanones. It occurred to us that
a
vinylogous
Mukaiyama10 condensation of vinyl cyclopropanols or
silyl ethers with readily available haloacetals would
offer an efficient and complementary route to spiro
cyclobutanones bearing v-haloalkyl tethers. Vinyl
cyclopropanols can be viewed as vinylogous enols, and
their synthetic utility in electrophilic addition reactions
was first demonstrated by the Wasserman and Trost
groups and, more recently, in our laboratory.11–15 Our
interest in vinyl cyclopropanols as useful synthetic
intermediates stemmed from their ready availability by
the Kulinkovich reaction16,17 of carboxylic esters and
other derivatives; vinyl cyclopropanols 2a,b had been
previously prepared in our laboratory in 95 and 86%
yields, respectively, by treatment of 1a and 1b with
ethylmagnesium bromide in the presence of titanium
isopropoxide.18–20
2.2. Radical ring expansion
In a typical example of the free radical-mediated
cyclization–fragmentation
sequence,
n-tributyltin
hydride was slowly added to a solution (12–14 mM) of
4a in refluxing benzene in the presence of AIBN to
afford 6a as a single isomer in nearly quantitative yield.
The parallel sequence involving 4b was next evaluated,
especially because the resulting bicyclo[6.3.0]undecane
skeleton is found in increasing numbers of bioactive
natural products.21 Again, free radical-mediated
cyclization–fragmentation of 4b-a and 4b-b proceeded
smoothly to give 6b-a and 6b-b (84 and 93% yields),
respectively, as the sole product in each case.
In the initial study, cyclopropanols 2a and 2b were
converted to the corresponding TMS ethers 3a and 3b
by standard methods (75–88% yield), TiCl4-promoted
condensation of which with 3-bromopropionaldehyde
dimethylacetal afforded 4a as a 5:3 mixture of two
isomers (71% yield) and a 1:1.3 mixture of 4b-a and
4b-b (56% yield), respectively. The ring junction stereo-
chemistry of 4a,b as shown in Scheme 1 was tentatively
assigned on the basis of the stereoelectronic require-
ments for electrophilic addition. The relative configura-
tion of the methoxy substituent in these products could
The operative mechanism is thought to involve addition
of the initially formed primary radical A to the cyclobu-
tanone carbonyl group to generate the reactive alkoxy
radical B (Scheme 2). Ring opening, which is driven by
the relief of strain present in the four-membered ring,
affords the fused bicyclic radical C. This radical C is
expected to suffer 1,5-hydrogen transfer leading to the
Scheme 2.

H.-S. Oh, J. K. Cha / Tetrahedron: Asymmetry 14 (2003) 2911–2917
2913
Bu3SnD. From individual radical allylation of 4b-a and
4b-b with allyltributyltin, 12 (as a 7:1 diastereomeric
mixture) and 13 (as a 5:1 diastereomeric mixture) were
obtained in 70 and 52% yields, respectively. The full
stereochemistry of 4b-a, 4b-b, and 12 was unequivocally
established by single-crystal X-ray analysis of carba-
mate 14; the latter was prepared from the major
diastereomer of 12 having the allyl group in the b-
configuration by reduction with NaBH4, followed by
acylation of the major alcohol with p-bromophenyliso-
cyanate. The X-ray structure of 14 is depicted in
ORTEP presentation (Fig. 1).
stabilized a-acyl radical D, which finally undergoes
hydrogen abstraction from tributyltin hydride to fur-
nish the bicyclic ketone products 6a and 6b. By analogy
to a related system examined by Dowd,22 the formation
of the trans isomers would be preferred to the alternate
cis ring junction isomers (i.e. C over C%; 1,5-hydrogen
transfer of Ha versus Hb).
A deuterium labeling experiment using tributyltin deu-
teride and also the radical allylation reaction by allyl-
tributyltin were undertaken next to substantiate the
proposed reaction mechanism involving 1,5-hydrogen
transfer (Scheme 3). Reduction of 4a with tributyltin
deuteride afforded 8 having the deuterium a to the
carbonyl group as a 4:1 diastereomeric mixture (94%).
Treatment of 4a with allyltributyltin 10 resulted in the
placement of an allyl group adjacent to the carbonyl
group to give 9, in 96% yield, as a ꢀ6:1 separable
mixture of two diastereomers.
Our attention was next directed at the construction of
the bicyclo[5.4.0]undecane and bicyclo[5.3.0]decane
rings by the analogous ring expansion sequence of 5a
and 5b (Scheme 4). Individual treatment of 5a-a and
5a-b with n-Bu3SnH and AIBN in refluxing benzene
gave 7a-a and 7a-b, respectively, in excellent (96ꢀ97%)
yields; the annulation products were obtained as >15:1
and 4:1 diastereomeric mixtures. In contrast to the
above-mentioned eight-membered ring annulation reac-
tion, 1,5-hydrogen transfer was anticipated to be
unlikely during the corresponding formation of a seven-
membered ring due to geometrical constraints. Appar-
ently, H-abstraction (from n-Bu3SnH) by the tertiary
radical intermediate at the ring junction proceeded with
unexpectedly high diastereoselectivity, and the cis stereo-
chemistry was tentatively assigned for the ring junction
of the major isomers of 7a-a and 7a-b. This moderate
to exceptional degree of stereoselectivity is especially
noteworthy, as Dowd had previously reported the loss
of stereocontrol at the ring junction in related cyclobu-
tanones 15 and 16 in the non-stereoselective prepara-
tion of 17 and 18.22 Elucidation of the subtle, yet
important factors that are responsible for the
diastereoselectivity in these systems, as well as the full
stereochemical determination, must await further stud-
In parallel, 11 was obtained as a 2:1 diastereomeric
mixture in 69% yield by treatment of 4b-a with n-
Scheme 3.
Figure 1.

2914
H.-S. Oh, J. K. Cha / Tetrahedron: Asymmetry 14 (2003) 2911–2917
ies. The analogous sequence involving 5b-a and 5b-b
was next examined under identical conditions: in each
case, only 19-b and 19-b were obtained in 80–90%
yields. The desired ring expansion pathway (F“G) was
thwarted by direct reduction by n-Bu3SnH or, possibly,
1,5-hydrogen transfer (F“H); further experiments with
n-Bu3SnD are necessary to determine whether 1,5-
hydrogen transfer is operative. The n-Bu3SnH experi-
ments with 5a and 5b illustrate a delicate balance
between the two competing reaction manifolds avail-
able to the initial primary radical.
requires two straightforward transformations involving
condensation of readily available vinyl cyclopropanols
or TMS ethers with known v-haloalkyl acetals and
subsequent radical ring expansion of the resulting v-
haloalkyl-tethered cyclobutanones. In the first step, a
simple, yet versatile coupling of the two modules gener-
ates three contiguous stereocenters, one of which
becomes a tertiary radical following the ring construc-
tion. Enantio- and diastereoselective construction of
these stereocenters would greatly enhance the synthetic
utility of this new annulation procedure. Following the
seminal contributions by W. S. Johnson,23 C2-symmet-
ric chiral acetals have emerged as useful tools in asym-
metric induction.24–26
2.3. Enantioselective annulation
Particularly noteworthy is the concise and convergent
nature of the present annulation procedure: this
method for assembling seven- or eight-membered rings
Toward this end, we chose (R,R)-(+)-hydrobenzoin 20
as a chiral auxiliary because of its commercial availabil-
ity, ease of removal and subsequent preparation of
enantiopure 21 in good yield (Scheme 5). Titanium
chloride-promoted condensation of 21 with 3a afforded
a 6:1 mixture of two isomeric cyclobutanones (82%);
the major product was tentatively assigned to have the
stereochemistry as shown in 22 on the basis of mecha-
nistic considerations in terms of an open-chain transi-
tion state.10 A working model I is shown, but is a
matter of speculation at this juncture. The provisional
stereochemical assignment needs to be confirmed by
X-ray analysis. The use of a bulkier alkyl acetal 21,
vis-a`-vis a dimethyl acetal (e.g. 3-bromopropionalde-
hyde dimethylacetal), could reinforce the preference for
an extended, open-chain transition state, which could
account for enhanced diastereoselectivity. The ring
Scheme 4.
Scheme 5.

H.-S. Oh, J. K. Cha / Tetrahedron: Asymmetry 14 (2003) 2911–2917
2915
expansion reaction of 22 was then achieved by slow
addition of Bu3SnH to give 23 in 96% yield. Finally, the
chiral auxiliary was removed cleanly by hydrogenolysis
(H2, Pd/C) to afford 24, [h]D=−8.3 (c 0.004, CHCl3), in
93% yield. Thus, the feasibility of an enantioselective
formation of an eight-membered ring under the aegis of
a chiral acetal was demonstrated in this preliminary
investigation. Other readily available chiral acetals will
be evaluated in due course for further refinement of the
cyclobutanone-based enantio- and diastereoselective
annulation strategy.
1.98–2.05 (m, 4H), 5.66 (m, 1H); 13C NMR (90 MHz,
CDCl3) l 1.3, 13.0, 22.7, 23.0, 25.3, 25.6, 60.3, 122.0,
138.3.
1
3b. H NMR (360 MHz, CDCl3) l 0.11 (s, 9H), 0.79
(m, 2H), 0.89 (m, 2H), 1.87 (m, 2H), 2.55 (m, 2H), 2.33
(m, 2H), 5.52 (m, 1H); 13C NMR (90 MHz, CDCl3) l
1.30, 14.0, 23.9, 32.5, 32.7, 56.5, 123.8, 146.8.
Representative procedure for a vinylogous Mukaiyama-
type condensation: To a solution of 3a (0.50 g, 2.38
mmol) in CH2Cl2 (4 mL) at −40°C were added sequen-
tially 3-bromopropionadehyde dimethylacetal (0.63 g,
3.1 mmol) and TiCl4 (0.59 g, 3.1 mmol). The reaction
mixture was stirred at the same temperature for 2 h.
The reaction was then quenched by addition of wet
ether (20 mL) and allowed to warm to rt. The mixture
was washed with brine (10 mL), dried over MgSO4,
filtered, and concentrated in vacuo to give the crude
product. Purification by column chromatography on
silica gel (using 10:1 hex–EtOAc as eluent) afforded 4a
as a mixture of two isomers, 0.30 g (44%) and 0.19 g
(27%).
3. Conclusion
A concise method for appending medium-sized rings to
cycloalkene carboxylates has been developed by
employing a tandem radical cyclization–fragmentation
sequence of bromoalkyl-tethered spirocyclobutanones.
The requisite substrates for ring expansion were readily
prepared by
a modular approach involving the
Kulinkovich cyclopropanation of cycloalkene carboxyl-
ates and subsequent electrophilic addition of the result-
ing alkenyl cyclopropanols or TMS ethers to bro-
moalkyl acetals. Also included is an enantioselective
preparation of a medium-sized bicyclic compound by
employing a nonracemic, C2-symmetric acetal.
Spectra data for the more polar isomer (on TLC) of 4a:
1
IR (film) 1774 cm−1; H NMR (360 MHz, CDCl3) l
0.97 (m, 1H), 1.26 (m, 2H), 1.47 (m, 1H), 1.59–1.73 (m,
5H), 1.85–1.99 (m, 2H), 2.07 (m, 1H), 2.26 (m, 1H),
2.87–3.01 (m, 2H), 3.18 (dt, J=9.0, 4.7 Hz, 1H), 3.28
(s, 3H), 3.44 (dd, J=8.1, 7.3, 2H); 13C NMR (90 MHz,
CDCl3) l 18.6, 22.0, 25.4, 26.3, 28.8, 34.2, 34.3, 41.5,
44.3, 56.5, 67.1, 81.3, 215.0.
4. Experimental
Representative procedure for the Kulinkovich cyclopropa-
nation: To a solution of methyl 1-cyclohexene-1-car-
boxylate 1a (0.71 mL, 5.1 mmol) in THF (23 mL) was
added titanium tetraisopropoxide (1.5 mL, 4.9 mmol).
A 1 M THF solution (an excess; ꢀ4 equiv.) of freshly
prepared ethylmagnesium bromide was then added at
5–10°C over a period of 1 h (via syringe pump). TLC
(using 4:1 hex–EtOAc as eluent) indicated no starting
ester, and water (1.7 mL) was then added slowly to
quench the reaction. After the resulting mixture had
been stirred for 30 min, Celite was added and the
resulting mixture was stirred for an additional 3–4 h at
rt. The mixture was dried over MgSO4 for 30 min and
filtered through a pad of Celite. The filtrate was con-
centrated under reduced pressure. Purification by
column chromatography (using 7:1 to 4:1 hex–EtOAc)
afforded 655 mg (93%) of 2a as a colorless oil.
Spectral data for the less polar isomer: IR (film) 1766
cm−1; 1H NMR (360 MHz, CDCl3) l 1.22 (m, 1H),
1.48–1.68 (m, 9H), 1.74–1.82 (m, 2 H), 2.25 (dt, J=
11.0, 8.4 Hz, 1H), 2.91 (t, J=8.4 Hz, 2H), 3.33 (s, 3H),
3.44–3.57 (m, 3H); 13C NMR (90 MHz, CDCl3) l 23.1,
23.3, 25.7, 25.9, 31.1, 35.8, 36.1, 41.6, 46.3, 58.2, 67.1,
80.2, 215.2.
1
4b-a. IR (film) 1774 cm−1; H NMR (360 MHz, CDCl3)
l 1.58 (m, 2H), 1.68–1.95 (m, 5H), 2.07 (m, 2H), 2.29
(m, 2H), 2.82 (ddd, J=17.6, 9.7, 5.0 Hz, 1H), 3.13
(ddd, J=17.6, 10.2, 7.8 Hz, 1H), 3.37 (s, 3H), 3.42–3.52
(m, 3H); 13C NMR (90 MHz, CDCl3) l 21.2, 24.1, 27.3,
30.5, 35.9, 38.7, 43.0, 44.9, 57.6, 72.5, 79.9, 215.5.
1
4b-b. IR (film) 1774 cm−1; H NMR (360 MHz, CDCl3)
Representative procedure for silylation of vinylcyclo-
propanols: To a mixture of 2a (0.39 g, 2.82 mmol) and
2,6-lutidine (3.5 g, 280 mmol) in CH2Cl2 (15 mL) at 0°C
was added TMSOTf (1.7 mL, 8.5 mmol). After the
reaction mixture had been stirred for 1 h at the same
temperature, saturated aqueous NaHCO3 solution (10
mL) was added. The aqueous layer was extracted with
Et2O. The combined organic extracts were washed with
water and brine, then dried (MgSO4), and concentrated
under reduced pressure. Purification by silica gel
column chromatography (10:1 hex–EtOAc containing
1% Et3N) afforded 3a (523 mg, 88%) as a colorless
liquid: 1H NMR (360 MHz, CDCl3) l 0.12 (s, 9H), 0.69
(m, 2H), 0.77 (m, 2H), 1.55 (m, 2H), 1.62 (m, 2H),
l 1.29 (m, 1H), 1.60–1.87 (m, 5H), 1.94–2.09 (m, 2H),
2.14–2.24 (m, 2H), 2.39 (m, 1H), 2.88–3.05 (m, 2H),
3.29 (s, 3H), 3.37 (m, 1H), 3.45 (m, 2H); 13C NMR (90
MHz, CDCl3) l 21.2, 23.4, 27.9, 28.9, 34.9, 37.9, 43.1,
48.3, 55.8, 72.3, 80.3, 215.8.
Representative procedure for free radical-mediated
cyclization-fragmentation: A solution of n-Bu3SnH (62
mg, 0.2 mmol) and AIBN (2 mg, 0.01 mmol) in benzene
(5 mL) was added slowly over a period of 8 h (syringe
pump) to a solution of 4a (the more polar isomer; 20
mg, 0.07 mmol) in benzene (5 mL) at reflux. The
reaction mixture was cooled to rt and concentrated
under reduced pressure. The resulting crude product

2916
H.-S. Oh, J. K. Cha / Tetrahedron: Asymmetry 14 (2003) 2911–2917
was dissolved in CH2Cl2 (20 mL) and washed with 10%
potassium fluoride solution (10 mL), dried (MgSO4),
filtered, and concentrated. Purification by column chro-
matography on silica gel (15:1 hex–EtOAc) gave 6a
Representative procedure for radical allylation: A solu-
tion of allyltributyltin (102 mg, 0.3 mmol) and AIBN (6
mg, 0.04 mmol) in benzene (1 mL) was added, over a
period of 6 h (syringe pump), to a solution of 4a (the
more polar isomer; 30 mg, 0.1 mmol) in benzene (1 mL)
at reflux. The reaction mixture was cooled to rt and
concentrated under reduced pressure. The residue was
then dissolved in CH2Cl2 (30 mL) and washed with 10%
potassium fluoride solution (10 mL), dried over
MgSO4, filtered, and concentrated to give the crude
product. Purification by column chromatography on
silica gel (15:1 hex–EtOAc) afforded 9 (24 mg, 96%) as
1
(14.3 mg, 99%): IR (film) 1699 cm−1; H NMR (360
MHz, CDCl3) l 1.02–1.47 (m, 6H), 1.51–1.90 (m, 6H),
2.09 (ddd, J=12.3, 6.6, 3.7 Hz, 1H), 2.18 (m, 1H), 2.29
(ddd, J=15.6, 6.5, 3.6 Hz, 1H), 2.44 (dt, J=2.7, 12.3
Hz, 1H), 2.59 (m, 1H), 2.92 (dt, J=3.1, 11.7 Hz, 1H),
3.20 (m, 1H), 3.34 (s, 3H); 13C NMR (90 MHz, CDCl3)
l 23.7, 26.6, 26.7, 30.1, 31.7, 32.9, 35.2, 38.9, 41.7, 44.8,
58.1, 85.4, 218.3; MS m/z 210 (M+).
1
a 5:1 diastereomeric mixture: IR (film) 1704 cm−1; H
1
6b-a. IR (film) 1698 cm−1; H NMR (360 MHz, CDCl3)
NMR (360 MHz, CDCl3) l 1.09–1.44 (m, 6H), 1.52–
1.75 (m, 5H), 1.78 (dd, J=5.3, 3.2 Hz, 1H), 1.81–1.91
(m, 2H), 2.10 (m, 1H), 2.41 (dt, J=14.1, 6.8 Hz, 1H),
2.45–2.57 (m, 2H), 2.78 (m, 1H), 3.25 (s, 3H), 3.41 (m,
1H), 5.01 (dd, J=9.9, 1.5 Hz, 1H), 5.03 (dd, J=17.2,
1.5 Hz, 1H), 5.72 (m, 1H); 13C NMR (90 MHz, CDCl3)
l 26.3, 26.6 (2 C’s), 32.5, 34.6, 35.8, 36.7, 38.9, 42.0 (2
C’s), 46.1, 56.9, 83.5, 116.9, 136.1, 216.4; MS m/z 250
(M+).
l 1.16 (m, 1H), 1.26–1.73 (m, 6H), 1.82 (m, 2H), 1.96
(m, 1H), 2.11–2.31 (m, 4H), 2.60 (m, 1H), 2.79 (dt,
J=11.5, 3.5 Hz, 1H), 2.99 (dt, J=2.7, 10.3 Hz, 1H),
3.30 (s, 3H); 13C NMR (90 MHz, CDCl3) l 23.4, 25.7,
29.9, 33.2, 36.4, 39.0, 40.0, 41.9, 47.5, 57.0, 84.7, 218.4.
1
6b-b. IR (film) 1701 cm−1; H NMR (360 MHz, CDCl3)
l 1.15 (m, 1H), 1.25–1.68 (m, 6H), 1.78–1.88 (m, 1H),
1.91–1.98 (m, 1H), 2.02–2.30 (m, 5H), 2.44 (m, 1H),
2.78 (dt, J=3.6, 11.7 Hz, 1H), 3.17 (m, 1H), 3.39 (s,
3H); 13C NMR (90 MHz, CDCl3) l 26.5, 26.6, 28.5,
32.2, 36.4, 38.1, 39.9, 41.1, 47.0, 58.0, 82.7, 217.7.
1
12. IR (film) 1711 cm−1; H NMR (360 MHz, CDCl3) l
0.90 (m, 1H), 1.16 (m, 1H), 1.35–1.55 (m, 3H), 1.58–
1.68 (m, 2H), 1.78 (m, 2H), 1.92–2.25 (m, 6H), 2.62 (m,
1H), 2.77 (m, 1H), 2.96 (dt, J=10.0, 2.8 Hz, 1H), 3.28
(s, 3H), 5.04 (dd, J=17.9, 1.2 Hz, 1H), 5.05 (dd,
J=10.1, 1.2 Hz, 1H), 5.71 (m, 1H); 13C NMR (90
MHz, CDCl3) l 25.7, 29.4, 33.4, 36.3, 36.4, 37.9, 38.0,
41.9, 47.6, 48.1, 56.9, 84.3, 117.4, 135.3, 218.9.
1
7a-a. IR (film) 1702 cm−1; H NMR (400 MHz, CDCl3)
l 1.30–1.40 (m, 2H), 1.42–1.54 (m, 4H), 1.58–1.66 (m,
2H), 1.76 (m, 1H), 1.96–2.15 (m, 3H), 2.44–2.47 (m,
2H), 2.74–2.75 (m, 2H), 3.32 (m, 1H), 3.34 (s, 3H); 13C
NMR (100 MHz, CDCl3) l 22.2, 24.4, 25.5, 26.5, 34.1,
35.6, 43.0, 43.3, 43.8, 56.8, 80.3, 212.9; HRMS calcd for
C12H20O2Na (M+) 219.1355, found 219.1354.
1
13. IR (film) 1698 cm−1; H NMR (360 MHz, CDCl3) l
0.90 (m, 1H), 1.15 (m, 1H), 1.42–1.68 (m, 5H), 1.77–
1.98 (m, 4H), 2.00–2.30 (m, 5H), 2.76 (m, 1H), 3.14
(ddd, J=10.7, 5.1, 3.0 Hz, 1H), 3.36 (s, 3H), 5.04 (dd,
J=17.2, 1.4 Hz, 1H), 5.05 (dd, J=10.1, 1.4 Hz, 1H),
5.71 (m, 1H); 13C NMR (90 MHz, CDCl3) l 26.5, 31.8,
32.9, 36.2, 36.4, 38.0, 38.2, 38.4, 46.9, 51.2, 58.1, 82.1,
117.6, 135.0, 219.2.
1
7a-b. IR (film) 1702 cm−1; H NMR (500 MHz, CDCl3)
l 1.24–1.42 (m, 3H), 1.46–1.50 (m, 1H), 1.52 (m, 1H),
1.59–1.62 (m, 3H), 1.80 (m, 1H), 1.90 (m, 1H), 1.99 (m,
1H), 2.13 (m, 1H), 2.41 (dd, J=14.7, 3.0 Hz, 1H), 2.44
(m, 1H), 2.63 (dd, J=16.5, 5.1 Hz, 1H), 2.67 (dd,
J=16.5, 10.1 Hz, 1H), 3.33 (s, 3H), 3.55 (ddd, J=10.1,
5.1, 2.5 Hz, 1H); 13C NMR (100 MHz, CDCl3) l 20.4,
21.7, 25.2, 26.6, 34.5, 38.7, 43.5, 44.2, 45.2, 56.9, 80.4,
211.7; HRMS calcd for C12H20O2Na (M+) 219.1355,
found 219.1358.
(4R, 5R)-2-Bromoethyl-4,5-diphenyl-1,3-dioxolane 21: A
mixture of (R,R)-(+) hydrobenzoin (950 mg, 4.43
mmol) and 3-bromopropionaldehyde diethylacetal (900
mg, 4.27 mmol) was heated at 135–145°C for 2 h and
then cooled. The product was recrystallized from iso-
1
1
8. IR (film) 1702 cm−1; H NMR (360 MHz, CDCl3) l
propanol to give 21 (1.3 g, 91%) as a white solid: H
1.03–1.35 (m, 5H), 1.45 (m, 1H), 1.53–1.70 (m, 5H),
1.72–1.81 (m, 1H), 2.08 (m, 1H), 2.19 (dt, J=5.7, 14.7
Hz, 1H), 2.31 (ddd, J=15.7, 6.6, 3.6 Hz, 1H), 2.45 (m,
1H), 2.61 (m, 1H), 3.21 (m, 1H), 3.36 (s, 3H); 13C NMR
(90 MHz, CDCl3) l 26.6, 26.7, 30.1, 31.6, 32.9, 34.1,
34.8 (t), 38.9, 41.7, 44.8, 58.0, 85.5, 218.3; HRMS calcd
for C13H21DO2 (M+) 211.1683, found 211.1657.
NMR (400 MHz, C6D6) l 2.21 (dt, J=7.3, 4.9 Hz, 2H),
3.30 (t, J=7.3 Hz, 2H), 4.56 (d, J=8.1 Hz, 1H), 4.60
(d, J=8.1 Hz, 1H), 5.44 (t, J=4.9 Hz, 1H), 7.00–7.20
(m, 10H); 13C NMR (100 MHz, C6D6) l 27.2, 38.1,
85.2, 87.1, 103.7, 126.6, 127.0, 128.5, 128.6; GC-MS
m/z 333 (M+).
22 (the major isomer; the less polar fraction): IR (film)
1
1
11. IR (film) 1698 cm−1; H NMR (360 MHz, CDCl3) l
1765 cm−1; H NMR (500 MHz, CDCl3) l 0.86 (m,
1.16 (m, 1H), 1.35–1.73 (m, 6H), 1.80 (m, 2H), 1.95 (m,
1H), 2.13–2.19 (m, 1H), 2.21–2.30 (m, 2H), 2.56 (m,
1H), 2.80 (m, 1H), 2.99 (dt, J=10.1, 2.9 Hz, 1H), 3.30
(s, 3H); 13C NMR (90 MHz, CDCl3) l 23.3, 25.7, 33.3,
36.4, 38.7 (t; another small triplet at 38.6), 38.9, 40.0,
41.9, 47.5, 57.0, 84.7, 217.2; HRMS calcd for
C12H19DO2 (M+) 197.1526; found 197.1557.
1H), 1.14–1.35 (m, 2H), 1.44–1.52 (m, 2H), 1.66–1.74
(m, 3H), 1.78 (m, 1H), 1.96–2.08 (m, 3H), 2.28 (m, 1H),
3.15–3.29 (m, 3H), 3.44 (dt, J=6.1, 10.6 Hz, 1H), 3.49
(dt, J=5.0, 10.6 Hz, 1H), 4.31 (d, J=8.1 Hz, 1H), 4.75
(d, J=8.1 Hz, 1H), 7.01 (m, 2H), 7.07 (m, 2H), 7.14 (m,
3H), 7.25 (m, 3H); 13C NMR (125 MHz, CDCl3) l
18.3, 22.0, 25.3, 26.2, 26.6, 32.4, 34.1, 42.2, 43.1, 67.0,

H.-S. Oh, J. K. Cha / Tetrahedron: Asymmetry 14 (2003) 2911–2917
2917
75.6, 78.2, 84.0, 127.2, 127.8, 128.1, 128.3, 128.6, 128.7,
136.4, 139.2, 218.4.
3. Radicals in Organic Synthesis; Renaud, P., Sibi, M. P.,
Eds.; Wiley-VCH: Weinheim, Germany, 2001.
4. Dowd, P.; Zhang, W. J. Am. Chem. Soc. 1991, 113,
9875–9876.
1
23. IR (film) 1693 cm−1; H NMR (400 MHz, CDCl3) l
0.90–0.13 (m, 6H), 1.44–1.75 (m, 6H), 2.08 (ddd, J=
15.4, 6.5, 2.4 Hz, 1H), 2.20 (m, 1H), 2.32 (ddd, J=15.4,
6.5, 3.3 Hz, 1H), 2.44 (m, 1H), 2.60 (m, 1H), 2.84 (m,
1H), 3.28 (m, 1H), 3.31 (d, J=2.4 Hz, 1H, OH) 4.34 (d,
J=7.3 Hz, 1H), 4.78 (dd, J=7.3, 2.4 Hz, 1H), 7.01–
7.30 (m, 10H); 13C NMR (100 MHz, CDCl3) l 26.4,
26.5, 30.5, 31.0, 32.7, 33.9, 35.0, 39.0, 41.6, 44.6, 78.6,
80.5, 85.0, 127.1, 127.3, 127.9, 128.1, 128.3, 128.5,
138.0, 139.9, 218.0.
5. Dowd, P.; Zhang, W. Chem. Rev. 1993, 93, 2091–2115.
6. Yet, L. Tetrahedron 1999, 55, 9349–9403.
7. Zhang, W. Curr. Org. Chem. 2002, 6, 1015–1029.
8. Oh, H.-S.; Lee, H. I.; Cha, J. K. Org. Lett. 2002, 4,
3707–3709.
9. Tidwell, T. T. Ketenes; Wiley: New York, 1995.
10. Murata, S.; Suzuki, M.; Noyori, R. J. Am. Chem. Soc.
1980, 102, 3248–3249.
11. Wasserman, H. H.; Cochoy, R. E.; Baird, M. S. J. Am.
Chem. Soc. 1969, 91, 2375–2376.
24. To a solution of 23 (8 mg, 0.02 mmol) in EtOH (3
mL) at rt was added 10% palladium on activated
carbon (5 mg). The reaction mixture was hydrogenated
in a Parr hydrogenator under 60 psi for 4 h, filtered
through a pad of Celite and MgSO4, and then concen-
trated in vacuo to give the crude product. Purification
by column chromatography on silica gel (using 7:1
hex–EtOAc as eluent) afforded 24 (3.5 mg, 93%) as a
colorless oil: [h]D=−8.3 (c 0.004, CHCl3); IR (film)
12. Trost, B. M.; Brandi, A. J. Am. Chem. Soc. 1984, 106,
5041–5043.
13. Trost, B. M.; Lee, D. C. J. Am. Chem. Soc. 1988, 110,
6556–6558.
14. Trost, B. M. Top. Curr. Chem. 1986, 133, 3–82.
15. Youn, J.-H.; Lee, J.; Cha, J. K. Org. Lett. 2001, 3,
2935–2938.
16. Kulinkovich, O. G.; de Meijere, A. Chem. Rev. 2000, 100,
2789–2834.
17. Sato, F.; Urabe, H.; Okamoto, S. Chem. Rev. 2000, 100,
2835–2886.
18. Lee, J.; Kim, H.; Cha, J. K. J. Am. Chem. Soc. 1995, 117,
9919–9920.
19. Lee, J.; Kim, H.; Cha, J. K. J. Am. Chem. Soc. 1996, 118,
4198–4199.
20. For a previous synthesis of 2a and 2b, see: Salaun, J.;
Garnier, B.; Conia, J. M. Tetrahedron 1974, 30, 1413–
1421.
21. Mehta, G.; Vishwakarma, S. Chem. Rev. 1999, 99, 881–
930.
22. Zhang, W.; Collins, M. R.; Mahmood, K.; Dowd, P.
Tetrahedron Lett. 1995, 36, 2729–2732.
23. Johnson, W. S.; Harbert, C. A.; Stipanovic, R. D. J. Am.
Chem. Soc. 1968, 90, 5279–5280.
24. Alexakis, A.; Mangeney, P. Tetrahedron: Asymmetry
1990, 1, 477–511.
25. Denmark, S. E.; Almstead, N. G. J. Am. Chem. Soc.
1991, 113, 8089–8110.
1
3398, 1729 (w) cm−1; H NMR (500 MHz, CDCl3) l
0.90 (m, 1H), 1.10–1.23 (m, 3H), 1.24–1.48 (m, 5H),
1.61–1.84 (m, 6H), 1.96 (dt, J=12.2, 3.0 Hz, 1H),
2.09–2.15 (m, 1H), 2.39 (m, 1H), 2.55 (s, br, 1H), 4.02
(br d, J=8.6 Hz, 1H); 13C NMR (125 MHz, CDCl3) l
26.6, 26.9, 32.3, 34.0, 34.5, 34.9, 36.1, 40.3, 41.0, 54.4,
83.3, 108.7; GC-MS m/z 196 (M+).
Acknowledgements
We thank the National Science Foundation (CHE-
0209321) for generous financial support.
References
1. Giese, B. Radicals in Organic Synthesis: Formation of
CarbonꢀCarbon Bonds; Pergamon Press: Oxford, 1986.
2. Motherwell, W. B.; Crich, D. Free Radical Chain Reac-
tions in Organic Synthesis; Academic Press: San Diego,
1992.
26. Sammakia, T.; Smith, R. S. J. Am. Chem. Soc. 1992, 114,
10998–10999.