Carbocyclization by radical closure onto O-trityl oximes: Dramatic effect of diphenyl diselenide

Article abstract of DOI:10.1021/ja067117t

O-Trityl oximes of 5- and 6-iodoaldehydes undergo radical cyclization to produce oximes when treated in refluxing tetrahydrofuran (THF) with Bu ₃SnH, 1,1′-azobis(cyclohexanecarbonitrile), i-Pr ₂NEt, and diphenyl diselenide (PhSeSePh).

Full text of DOI:10.1021/ja067117t

Published on Web 02/07/2007
Carbocyclization by Radical Closure onto O-Trityl Oximes:
Dramatic Effect of Diphenyl Diselenide
Derrick L. J. Clive,* Mai P. Pham, and Rajendra Subedi
Contribution from the Chemistry Department, UniVersity of Alberta, Edmonton,
Alberta, Canada T6G 2G2
Received October 4, 2006; E-mail: derrick.clive@ualberta.ca
Abstract: O-Trityl oximes of 5- and 6-iodoaldehydes undergo radical cyclization to produce oximes when
treated in refluxing tetrahydrofuran (THF) with Bu3SnH, 1,1′-azobis(cyclohexanecarbonitrile), i-Pr2NEt, and
diphenyl diselenide (PhSeSePh).
Scheme 1a
Introduction
Several years ago, work in this laboratory led to the
development of a method for stannane-mediated radical cy-
clization onto O-trityl oxime ethers,1 as summarized in
Scheme 1.2 This process differs from the classical cyclization
of hexenyl radicals (eq 1) in the essential fact that the acceptor
double bond is preserved in the product (cf. 4 of Scheme 1 and
6 of eq 1). Attempts were made3 in the initial study to extend
the chemistry of Scheme 1 to the more useful case of
carbocyclization, in which the cyclizing chain does not contain
a heteroatom. However, it was found that carbocyclization,
unlike the formation of lactones, is not general; it worked in a
few cases, though not in others. We have now reexamined such
reactions and have identified specific conditions under which
the all-carbon process is indeed general, and we report here
details of this synthetic method. The products of the reactions
are oximes of cyclic ketones and so should serve as precursors
to the corresponding amines and ketones.
a Oxime geometries are arbitrary.
than that of the corresponding all-carbon system. We also
expected tautomerization of the intermediate nitroso compound
to be facilitated by the adjacent aromatic ring. In the event, a
modest yield (48%) of the desired oxime 10 was obtained (eq
3), and the yield was improved (69%) by using tetrahydrofuran
Initial Attempts To Extend the O-Trityl Oxime Cyclization
to Carbocycles. Initial attempts to effect carbocyclization along
the lines of eq 2 were unpromising, as the desired oxime did
(THF) as solvent in the presence of 2-pyridonesan additive
that was introduced in the expectation that it might facilitate
(through hydrogen bonding) tautomerization of the intermediate
nitroso compound to the oxime. Repetition of the experiment
in THF, but with i-Pr2NEt instead of 2-pyridone, and ABC [1,1′-
azobis(cyclohexanecarbonitrile)5] instead of AIBN (2,2′-azobi-
sisobutyronitrile), raised the yield to 91%.
not appear to be formed. Therefore we examined a potentially
less demanding casesthe ether 9 (eq 3)schosen because it was
known4 that 5-exo cyclization with oxygen in the chain is faster
(1) Clive, D. L. J.; Subedi, R. Chem. Commun. 2000, 237-238.
(2) For evidence of the intermediacy of nitroso compounds in such reactions,
see Bella, A. F.; Slawin, A. M. Z.; Walton, J. C. J. Org. Chem. 2004, 69,
5926-5933.
(3) Subedi, R. Ph.D. Thesis, University of Alberta, Canada, 2002.
(4) Beckwith, A. L. J.; Blair, I.; Phillipou, G. J. Am. Chem. Soc. 1974, 96,
1613-1614.
(5) ABC (trade name V-40) has a longer half-life than AIBN: Yoshikai, K.;
Hayama, T.; Nishimura, K.; Yamada, K.; Tomioka, K. J. Org. Chem. 2005,
70, 681-685.
9
10.1021/ja067117t CCC: $37.00 © 2007 American Chemical Society
J. AM. CHEM. SOC. 2007, 129, 2713-2717
2713

A R T I C L E S
Clive et al.
unlikely to be the only solvent effect since a few of our O-trityl
oximes do cyclize in an aromatic solvent (9, 48% in PhMe; 13,
43% in PhMe; 15, 74% in PhH).
Evaluation of First-Generation Conditions: Use of THF,
Hu1nig’s Base, and 1,1′-Azobis(cyclohexanecarbonitrile).
Although the result with 7 was poor (41% in THF, as stated
earlier), we decided to examine a number of other substrates
under the conditions that worked best with 13, and we chose
several carbohydrate-derived oximes; these would lead to
optically pure cyclopentanone derivatives bearing a variety of
functional groups.
Oxime 15 cyclized efficiently (eq 6), but it appears to be an
inherently favorable case, as the simple use of AIBN in PhH
without added base gave a yield of 74%. The related, but more
flexible, oxime 17 also cyclized in good yield (eq 7), as did the
corresponding perbenzylated compound 19 (eq 8), again with
an excess of stannane to ensure consumption of the starting
material.
The all-carbon system 11 was examined next under several
different conditions, some of which are shown in eq 4.
Compound 7 was also cyclized by treatment with Bu3SnH (2
equiv) and AIBN (0.3 equiv) but in THF as solvent; with this
procedure, cyclopentanone oxime was isolated in 41% yield.
At this point, we examined the possibility of carrying out a
6-exo trigonal cyclization and we found it convenient to use
compound 13 (eq 5); while the resulting ring does contain a
heteroatom, its location would not facilitate tautomerization of
the intermediate nitroso compound (cf. 3) to any significant
extent. A large number of different conditions were tried (Table
1); those shown in eq 5 gave the highest yield. In all our
cyclization studies we have used the standard method of
simultaneous slow addition of stannane and initiator (by double
syringe pump) and have kept initial concentrations of the
substrate at ca. 0.02 M.
The experiments summarized in Table 1 cover a range of
solvent dielectric constants and boiling points, the presence or
absence of Lewis acids (which might coordinate to the oxime
nitrogen), and the presence or absence of additives (N-
methylpyrrolidinone, 2-pyridone, pyridine, Hu¨nig’s base), which,
together with solvent polarity, might facilitate tautomerization6
of the intermediate C-nitroso compound. The observations
summarized in Table 1 suggested that the most promising
conditions are the use of THF as a solvent, i-Pr2NEt as base,
and ABC as initiator. When the THF-Hu¨nig’s base-ABC
procedure was applied to 7, the yield was 41%, the same value
as that obtained with AIBN in THF without added base (see
above). We assume that the base does indeed facilitate tau-
tomerization of an intermediate nitroso compound to the oxime,
but the reason for the solvent effect (THF versus PhMe, entries
i and iii) is not clear. The ethereal solvents examined [THF,
1,2-dimethoxyethane (DME), dioxane, tetrahydropyran] all have
the potential for generating radicals by loss of hydrogens R to
oxygen,7 and it is known8 that the use of benzene (and
presumably toluene) as a solvent for radical reactions can result
in homolytic addition to the solvent; the resulting cyclohexa-
dienyl radicals do not propagate the desired chain reaction.
While such interference may occur in the present case, it is
Under the conditions used for the reactions shown in eqs 6-8,
the oximes 219-24 gave little, if any, of the desired cyclization
products. In the case of 22 and 23, the presence of geminal
disubstitution had been expected to facilitate ring closure but,
evidently, other factors were at work.
When equimolar amounts of 21 and 17 were subjected to
our standard conditions [Bu3SnH (4 equiv), ABC (0.2 equiv),
THF, and i-Pr2NEt], the former was largely recovered (65%),
while the latter gave the normal cyclization product (76%), and
so we conclude that the failure of some compounds to cyclize
is not due to contamination by adventitious radical inhibitors.
The O-benzyl oxime corresponding to 21 cyclized normally (no
base, Bu3SnH, ABC, THF, 61%) to the expected10 O-benzyl-
hydroxylamine.
(6) (a) Donaruma, L. G. J. Org. Chem. 1958, 23, 1338-1340. (b) Di Giacomo,
A. J. Org. Chem. 1965, 30, 2614-2617.
(7) Cf. Yoshimitsu, T.; Tsunoda, M.; Nagaoka, H. J. Chem. Soc. Chem.
Commun. 1999, 1745-1746.
(9) While the desired cyclization product from 21 would have a heteroatom in
the newly formed ring, the location of the heteroatom would not (greatly)
facilitate tautomerization of the intermediate nitroso compound.
(8) Beckwith, A. L. J.; Bowry, V. W.; Bowman, W. R.; Mann, E.; Parr, J.;
Storey, J. M. D. Angew. Chem., Int. Ed. 2004, 43, 95-98.
9
2714 J. AM. CHEM. SOC. VOL. 129, NO. 9, 2007

Radical Closure onto O-Trityl Oximes
A R T I C L E S
Table 1. Radical Cyclization of 13
Bu3SnH equiv
initiator (equiv)
additive (equiv)
solvent
temp
yield of 14, %
i
ii
iii
iv
v
vi
vii
viii
ix
4
4
4
4
4
4
4
2
4
1.5
2
4
4
4
4
4
4
4
4
4
AIBN (0.4)
AIBN (0.4)
AIBN (0.4)
AIBN (0.8)
AIBN (0.6)
AIBN (0.63)
AIBN (0.6)
AIBN (0.2)
AIBN (0.6)
Et3B (1.3), air
Et3B (2.5), air
AIBN (0.6)
AIBN (0.6)
AIBN (0.6)
AIBN (0.1)
AIBN (0.1)
ABC (0.1)
PhMe
THF
THF
THF
MeCN
DME
THF
THF
THF
PhH
reflux
43a
71a
71a
69a
52a
67a
54a,b
c
2-pyridone (1.3)
reflux
reflux
reflux
reflux
reflux
reflux
reflux
reflux
room temp
room temp
reflux
reflux
reflux
reflux
reflux
reflux
reflux
reflux
reflux
N-methylpyrrolidinone (1.5)
pyridine (1.4)
BF3‚Et2O (4)
Ti(OPr-i)4 (4)
39a
d
x
xi
BF3‚Et2O (3)
THF
e
xii
xiii
xiv
xv
xvi
xvii
xviii
xix
xx
dioxane
tetrahydropyran
1:1 THF-DMSO
2-butanone
EtOAc
THF
THF
THF
THF
45a
68a
43a,f
49a (61g)a
36a
86h
86a
92h
68a
pyridine (4)
pyridine (4)
i-Pr2NEt (4)
Bu3SnCl (4)
ABC (0.1)
ABC (0.1)
AIBN (0.6)
a Stannane and initiator were added in two approximately equal portions, as considerable starting material remained after the initial 10-h reflux period.
b After recrystallization to remove traces of AIBN. c No product detected (tlc). d Only a trace of product detected (tlc), even after additional Bu3SnH (1.5
equiv) and Et3B (1.3 equiv) were added. e Mainly starting material recovered. f Not corrected for recovered starting material (10%). g Based on conversion;
16% of starting material recovered. h The excess of reagents was used initially, rather than being added in portions.
Iodide 25 cyclized in THF in 68% yield upon treatment with
Bu3SnH (3 equiv) in the presence of ABC (2 equiv) and i-Pr2-
NEt (4 equiv); the yield was somewhat lowered (57%) by using
lesser amounts of stannane (1.2 equiv) and ABC (1 equiv).
tions but obtained the hydroxylamine 27 (eq 9), indicating that
Ph2CH• is not a sufficiently good radical leaving group.
Second-Generation General Conditions: Effect of Diphe-
nyl Diselenide on the Cyclization. As our attempt to incor-
porate the release of less persistent species than triphenylmethyl
radicals was unsuccessful, we decided to examine the effect of
generating PhSeH in the reaction mixture,14 our expectation
being that the selenol would rapidly capture triphenylmethyl
radicals. To this end, iodide 25 (Table 2, entry i) was cyclized
in the presence of PhSeSePh.14 When this experiment was
conducted in PhH [with AIBN (0.1 equiv) and no base] the
yield was 38%, but in THF [with ABC (1 equiv) and i-Pr2NEt
(4 equiv)] the yield rose to 91%. These conditions were found
to be generalseven bromide 21 cyclizedsand were applied to
all the examples listed in Table 2. In a test experiment with 25,
it was established that there is no advantage to using more than
1 equiv of ABC (Table 2, entry 1, footnote c).
The effect of replacing Bu3SnH by two other reducing agents
was examined. Treatment of 21 with (Me3Si)3SiH11 and ABC,
with or without i-Pr2NEt, in refluxing THF led to recovery (93-
98%) of the starting bromide; and the same bromide was again
recovered (82%) when it was exposed to the action of N-
ethylpiperidine hypophosphite12 [C8H15N‚H3PO2] and AIBN in
refluxing THF. A mixture of 21 and its corresponding O-benzyl
oxime was treated with N-ethylpiperidine hypophosphite and
ABC (0.6 equiv) in refluxing THF; only the benzyl oxime
appeared to cyclize (27% yield).
If our assumption is correct that buildup of a persistent radical
prevents chain propagation, and hence cyclization, then the
beneficial effect produced by the presence of PhSeSePh (which
leads to PhSeH on reaction with Bu3SnH) is easily understand-
able. Benzeneselenol has a much greater reactivity (several
orders of magnitude) toward carbon radicals compared to the
reactivity of Bu3SnH;15 consequently, even persistent radicals
are reduced,14 and the resulting phenylselenyl radicals rapidly
generate stannyl radicals from Bu3SnH to re-form PhSeH. In
this way, the buildup of persistent radicals (which react slowly
with stannanes) is suppressed by the presence of small quantities
of PhSeSePh, and so chain propagation is not disrupted.
A control experiment in which the O-trityl oxime of benzal-
dehyde13 was treated with Bu3SnH, AIBN, and i-Pr2NEt in
refluxing THF showed that the carbon-nitrogen double bond
is not reduced, as the starting oxime was recovered (95%).
Suspecting that formation of a persistent radical (Ph3C•) might
be involved in the failure of some compounds to cyclize, we
subjected the O-benzhydryl oxime 26 to our cyclization condi-
The iodo ester 23 was used for five additional experiments,
whose outcome is listed in Table 3, together with the result
(entry i) for our optimum conditions. The tabulated observations
confirm the superiority of the conditions (see Table 2, entry i)
(10) Review: Fallis, A. G.; Brinza, I. M. Tetrahedron 1997, 52, 17543-17594.
(11) Chatgilialoglu, C. Organosilanes in Radical Chemistry; Wiley: Chichester,
U.K., 2004.
(12) Cf Graham, S. R.; Murphy, J. A.; Kennedy, A. R. J. Chem. Soc., Perkin
Trans. 1 1999, 3071-3073.
(13) Lutz, W. B. J. Org. Chem. 1971, 36, 3835-3837.
(14) Crich, D.; Hwang, J.-T. J. Org. Chem. 1998, 63, 2765-2770 and references
cited therein.
(15) (a) Chatgilialoglu, C.; Ingold, K. U.; Scaiano, J. C. J. Am. Chem. Soc.
1981, 103, 7739-7742. (b) Newcomb, M.; Choi, S.-Y.; Horner, J. H.
J. Org. Chem. 1999, 64, 1225-1231.
9
J. AM. CHEM. SOC. VOL. 129, NO. 9, 2007 2715

A R T I C L E S
Clive et al.
Table 2. Cyclization of O-Trityloximesa
Table 3. Optimization Studies on Cyclization of 23
entry
solvent
i-Pr2NEt equiv
PhSeSePh equiv
initiator (equiv)
yield of 27, %
i
ii
iii
iv
v
THF
THF
PhH
THF
THF
THF
4
none
4
4
4
4
0.2
0.2
0.2
0.2
0.2
none
ABC (1)
ABC (1)
ABC (1)
AIBN (1)
ABC (0.1)
ABC (1)
96
71
63
83
59
56
vi
and on the quantity of initiator, with 1 equiv being needed (cf.
entries i and v); the presence of Hu¨nig’s base (cf. entries i and
ii) is also helpful.
Preparation of Starting O-Trityl Oximes. The O-trityl
oxime ethers used in this work are easily prepared (see
Supporting Information for schemes and experimental details)
from the corresponding aldehydes, which are themselves readily
available. The aldehydes react in high yield (usually >90%)
with TrONH213 in THF (at 65 °C) or CH2Cl2 (room temperature
or reflux); those aldehydes required for the preparation of 7,16
9,17 11,18 13,19 29,20-22 32,20-22 and 3923 were made by the
literature methods cited, while 15, 17, and 19 were prepared
by straightforward routes from known carbohydrates.
Conclusions
Our experiments aimed at extending the process of Scheme
1 to carbocycles allowed us to identify a set of conditions that
is general, as well as high-yielding. The results summarized in
Table 2 clearly establish that radical cyclization of iodo O-trityl
oximes in the presence of PhSeSePh and Hu¨nig’s base, with
THF as a solvent, and 1,1′-azobis(cyclohexanecarbonitrile) as
initiator constitutes an efficient route to five- and six-membered
carbocycles, independent of the substitution pattern. The method
is applicable to flexible chains as well as to those that are
rigidified by the presence of rings; carbohydrates represent a
very convenient source of the oximes that undergo cyclization.
A useful feature of the reaction is that the products contain the
original carbonyl group in the form of its oxime and so should
be amenable to a number of synthetic transformations.
Experimental Section
General Procedure for Preparing O-(Triphenylmethyl)oximes.
13
TrONH2 (1.0 equiv) was added to a stirred solution of the aldehyde
(1.0 equiv) in dry THF, and the mixture was heated at 65 °C for 12 h
under N2, cooled, and evaporated to give a residue that was processed
as described for the individual experiments. In some experiments CH2-
Cl2was used, either at room temperature or at reflux.
General Procedure for Radical Cyclization in the Absence of
Diphenyl Diselenide (Procedure A). The substrate and base were
placed in a round-bottomed flask equipped with a Teflon-coated stirring
bar and a reflux condenser sealed with a rubber septum. The system
was flushed with argon, and dry solvent was injected into the flask.
(16) Kelkar, S. V.; Joshi, G. S.; Reddy, G. B.; Kulkarni, G. H. Synth. Commun.
1989, 19, 1369-1379.
a Except for the differences indicated, the following conditions were
used: Bu3SnH (1.2 equiv), ABC (1 equiv), PhSeSePh (0.2 equiv), i-Pr2NEt
(4 equiv), THF, reflux. b Mixture of geometric isomers. c 89% with 2 equiv
of ABC. d Relative stereochemistry not established. e (E) isomer.
(17) Clive, D. L. J.; Yang, W.; MacDonald, A. C.; Wang, Z.; Cantin, M.
J. Org. Chem. 2001, 66, 1966-1983.
(18) Page, P. C. B.; Rassias, G. A.; Barros, D.; Bethell, D.; Schilling, M. B.
J. Chem. Soc., Perkin Trans. 1 2000, 19, 3325-3334.
(19) Kirmse, W.; Ho¨mberger, G. J. Am. Chem. Soc. 1991, 113, 3925-3934.
(20) Barluenga, J.; Gonza´lez-Bobes, F.; Murgu´ıa, M. C.; Ananthoju, S. R.;
Gonza´lez, J. M. Chem.sEur. J. 2004, 10, 4206-4213.
(21) Buijs, W.; van Elburg, P.; van der Gen, A. Synth. Commun. 1983, 13, 387-
identified by the experiments with 25. While the use of diphenyl
diselenide (cf. entries i and vi) is a key factor (as described
above), there is clearly a significant dependence on the solvent
(cf. entries i and iii), with THF being an effective choice,
392.
(22) We used HI (cf. ref 16) to make the parent iodo alcohol, instead of P2I4
(ref 21). For oxidation to the iodo aldehyde (ref 20) we used PDC.
(23) Gibson, S. E. (ne´e Thomas); Guillo, N.; Middleton, R. J.; Thuilliez, A.;
Tozer, M. J. J. Chem. Soc., Perkin Trans. 1997, 1, 447-455.
9
2716 J. AM. CHEM. SOC. VOL. 129, NO. 9, 2007

Radical Closure onto O-Trityl Oximes
A R T I C L E S
The solution was heated to reflux, and solutions of Bu3SnH and initiator
in the corresponding solvent were injected simultaneously by syringe
pump over 10 h. Refluxing was continued for an arbitrary period of 2
h after the addition. The mixture was cooled, and the solvent was
evaporated to give a residue that was processed as described for the
individual experiments.
followed with 23 (280 mg, 0.457 mmol), PhSeSePh (28.0 mg, 0.0897
mmol), and Hu¨nig’s base (0.32 mL, 1.837 mmol) in THF (30 mL);
Bu3SnH (0.15 mL, 0.5576 mmol) in THF (10 mL); and ABC (120
mg, 0.4911 mmol) in THF (10 mL). After evaporation of the solvent,
flash chromatography of the residue over silica gel (2 × 20 cm), with
elution by 30% EtOAc-hexane, gave 34 (106.7 mg, 96%) as a mixture
(ca. 1:1) of Z and E isomers (1H NMR): FTIR (CH2Cl2cast) 3254,
General Procedure for Radical Cyclization in the Presence of
Diphenyl Diselenide (Procedure B). The substrate and PhSeSePh (ca
0.2 mol/mol of substrate) were placed in a round-bottomed flask with
a condenser fused on. The flask contained a Teflon-coated stirring bar
and was sealed with a rubber septum. Dry THF and Hu¨nig’s base (ca
4 mol/mol of substrate) were injected and the solution was purged with
Ar via a long needle (dipping into the solvent) for 0.5 h. (We found
that yields were lower if oxygen was present.) The solution was heated
at 95 °C (oil bath), and solutions of Bu3SnH (ca 1.2 mol/mol of
substrate) and ABC (ca 1 mol/mol of substrate), each in dry THF, were
injected simultaneously by syringe pump over 10 h (Ar atmosphere).
Refluxing was continued for an arbitrary period of 2 h after the addition.
The mixture was cooled and the solvent was evaporated to give a
residue that was processed as described for the individual experiments.
2,3-Dihydro-1H-inden-1-one Oxime (12) (from 39). General
procedure B for radical cyclization was followed, with 39 (44.0 mg,
0.085 mmol), PhSeSePh (5.3 mg, 0.017 mmol), and Hu¨nig’s base (0.059
mL, 0.34 mmol) in dry THF (10 mL); Bu3SnH (0.027 mL, 0.10 mmol)
in THF (3 mL); and ABC (21.0 mg, 0.086 mmol) in THF (3 mL).
After evaporation of the solvent, flash chromatography of the residue
over silica gel (2 × 30 cm), with elution by 30% EtOAc-hexane, gave
1224,25 (12.0 mg, 94%) as a single isomer, identical to material made
from 11.
1
2983, 1731, 1368 cm-1; H NMR (CDCl3, 400 MHz) δ 1.18-1.255
(overlapping t, 6H), 2.31 (q, J ) 7.8 Hz, 2H), 2.51 (t, J ) 7.9 Hz,
1H), 2.60 (t, J ) 8.1 Hz, 1H), 2.97 (s, 1H), 3.08 (s, 1H), 4.10-4.22
(overlapping q, 4H), 8.78-8.90 (two overlapping br s, 1H); 13C NMR
(CDCl3, 100.6 MHz) δ 13.8 (q), 25.7 (t), 28.9 (t), 31.6 (t), 32.0 (t),
34.7 (t), 37.8 (t), 58.5 (s), 58.6 (s), 61.7 (t), 162.6 (s), 162.8 (s), 170.6
(s), 170.8 (s); exact mass (electrospray) m/z calcd for C
+ H) 244.11795, found 244.11828.
11H18NO5 (M
cis-Hexahydropentalen-1(2H)-one Oxime (36). General procedure
B for radical cyclization was followed, with 35 (376.7 mg, 0.797 mmol),
PhSeSePh (49.7 mg, 0.159 mmol), and Hu¨nig’s base (0.56 mL, 3.215
mmol) in dry THF (30 mL); Bu3SnH (0.257 mL, 0.955 mmol) in THF
(9 mL); and ABC (190 mg, 0.778 mmol) in THF (9 mL). After
evaporation of the solvent, flash chromatography of the residue over
silica gel (2 × 30 cm), with elution by 30% EtOAc-hexane, gave 36
(0.0936 g, 85%) as an oil, which was a mixture of Z and E isomers
(1H NMR): FTIR (CH2Cl2,cast) 3244 2949, 2867, 1661 cm-1; 1H NMR
(CDCl3, 400 MHz) δ 1.25-2.70 (m, 11H), 2.92 (ddd, J ) 8.7, 8.7, 4.6
Hz, 0.85H), 3.20 (ddd, J ) 9.1, 9.1, 5.1 Hz, 0.13H), 9.0 (br s, 1H); 13
C
NMR (CDCl3, 100.6 MHz) δ 26.2 (t), 26.7 (t), 27.1 (t), 28.9 (t), 29.3
(t), 30.0 (t), 30.9 (t), 32.4 (t), 32.5 (t), 32.8 (t), 43.10 (d), 43.14 (d),
43.7 (d), 46.8 (d), 170.7 (s); exact mass m/z calcd for C8H13NO
139.09972, found 139.09970.
Acetic Acid [R-(1R,2R,3R)]-2,3-Diacetoxy-4-(hydroxyimino)cy-
clopentyl Ester (18). General procedure A for radical cyclization was
followed, with 17 (131 mg, 0.219 mmol) in THF (15 mL), Bu3SnH
(256 mg, 0.879 mmol) in THF (5 mL), ABC (6 mg, 0.02 mmol) in
THF (5 mL), and i-Pr2NEt (114 mg, 0.879 mmol). Flash chromatog-
raphy of the residue over silica gel (1.7 × 20 cm), with elution by
50% EtOAc-hexane, gave 18 (50.1 mg, 85%) as a foam, which was
a mixture of Z and E isomers: Fourier transform infrared (FTIR) (CH2-
Spiro[4.5]decan-2-one Oxime (38).26 General procedure B for
radical cyclization was followed, with 37 (330 mg, 0.61 mmol),
PhSeSePh (38 mg, 0.12 mmol), and Hu¨nig’s base (0.43 mL, 2.47 mmol)
in dry THF (30 mL); Bu3SnH (0.20 mL, 0.74 mmol) in THF (10 mL);
and ABC (150 mg, 0.61 mmol) in THF (10 mL). After evaporation of
the solvent, flash chromatography of the residue over silica gel (2 ×
30 cm), with elution by 30% EtOAc-hexane, gave 3826 (88 mg, 85%)
as an oil that was a mixture of Z and E isomers (1H NMR): FTIR
1
Cl2 cast) 3396, 2936, 1750, 1429 cm-1; H NMR (CDCl3, 400 MHz)
δ 2.01-2.05 (3 s, 9H), 2.70-2.79 (m, 1H), 2.97 (apparent dd, J )
15.0, 7.5 Hz, 1H), 5.22-5.34 (m, 1.2H), 5.49 (t, J ) 4.5 Hz, 0.85H),
5.68 (dd, J ) 5.0, 2.0 Hz, 0.82H), 6.02-6.08 (m, 0.21H), 8.61-9.18
(br s, 1H); 13C NMR (CDCl3, 100.6 MHz) δ 20.2 (q), 20.3 (q), 20.4
(q), 20.5 (q), 20.6 (q), 20.7 (q), 30.1 (t), 33.0 (t), 65.2 (d), 68.7 (d),
69.9 (d), 70.0 (d), 71.0 (d), 71.4 (d), 156.0 (s), 169.7 (s), 169.8 (s),
170.0 (s); exact mass m/z calcd for C11H15NO7 273.0848, found
273.0849.
1
(CH2Cl2 cast) 3239, 3126, 2924, 2855 cm-1; H NMR (CDCl3, 400
MHz) δ 1.25-1.70 (m, 12H), 2.24 (s, 0.7H), 2.35 (s, 1.2H), 2.44 (t,
J ) 7.7 Hz, 1H), 2.51 (t, J ) 7.6 Hz, 1H), 8.30 and 8.35 (two
overlapping br s, 1H); 13C NMR (CDCl3, 100.6 MHz) δ 23.23 (t), 23.25
(t), 25.0 (t), 26.19 (t), 26.23 (t), 28.4 (t), 35.7 (t), 36.4 (t), 37.1 (t),
38.9 (t), 41.5 (t/s), 41.8 (t/s), 42.3 (t/s), 166.7 (s), 167.0 (s); exact mass
m/z calcd for C10H17NO 167.13101, found 167.13098.
Acetic Acid 3-(Hydroxyimino)cyclopentyl Ester (31). General
procedure B for radical cyclization was followed, with 30 (440 mg,
0.835 mmol), PhSeSePh (52.1 mg, 0.1669 mmol), and Hu¨nig’s base
(0.58 mL, 3.33 mmol) in THF (25 mL); Bu3SnH (0.56 mL, 2.08 mmol)
in THF (10 mL); and ABC (0.304 g, 1.244 mmol) in THF (10 mL).
After evaporation of the solvent, flash chromatography of the residue
over silica gel (1.5 × 30 cm), with elution by 30% EtOAc-hexane,
gave 31 (112 mg, 86%) as a mixture of Z and E isomers: FTIR (CH2-
Acknowledgment. We thank the Natural Sciences and
Engineering Research Council of Canada for financial support.
1
Cl2, cast) 3254, 2978, 1737, 1234 cm-1; H NMR (CDCl3, 400 MHz)
Supporting Information Available: Experimental procedures
for 7-11, 12 (from 11), 13-17, 19-27, 30, 32-33, 34 (from
22), 35, 37, 39, 41-45, 47, 49, 51, 53, 54, 56-58, 60-62, trans-
2-[(2-bromocyclohexyl)oxy]acetaldehyde O-(phenylmethyl)-
oxime, and O-benzyl-N-[(3aR,7aR)octahydrobenzofuran-3-yl]-
hydroxylamine; NMR spectra of all new compounds except one
isomer of 32, 39, and 43-45, which were used crude; and X-ray
data for 12. This material is available free of charge via the
Internet at http://pubs.acs.org.
δ 1.88-2.07 (m, 5H), 2.38-2.83 (m, 4H), 5.22-5.33 (m, 1H), 8.58-
8.62 (two overlapping br s, 1H); 13C NMR (CDCl3, 100.6 MHz) δ
21.1 (q), 24.5 (t), 27.7 (t), 30.2 (t), 30.7 (t), 33.9 (t), 37.0 (t), 73.8 (d),
163.4 (s), 163.8 (s), 170.49 (s), 170.53 (s); exact mass (electrospray)
m/z calcd for C7H11NNaO3 180.06311, found 180.06286.
3-(Hydroxyimino)cyclopentane-1,1-dicarboxylic Acid Diethyl Es-
ter (34) (from 23). General procedure B for radical cyclization was
(24) Ribeill, Y.; Genevois-Borella, A.; Vuilhorgne, M.; Mignani, S. Heterocycles
1999, 51, 2977-2982.
(25) Buchanan, G. W.; Dawson, B. A. Can. J. Chem. 1978, 56, 2200-2204.
(26) Rice, L. M.; Dobbs, E. C.; Grogan, C. H. J. Med. Chem. 1965, 8, 825-
829.
JA067117T
9
J. AM. CHEM. SOC. VOL. 129, NO. 9, 2007 2717