N-silyl-tethered radical cyclizations: a new synthesis of gamma-amino alcohols.

Article abstract of DOI:10.1021/ol034288j

[reaction: see text] Various allylic and propargylic amines bearing a protecting group (PG) have been employed in N-silyl-tethered radical cyclizations. The resulting silapyrrolidine adducts could be smoothly oxidized, creating access to gamma-amino alcoh

Full text of DOI:10.1021/ol034288j

ORGANIC
LETTERS
2003
Vol. 5, No. 8
1341-1344
N-Silyl-Tethered Radical Cyclizations: A
New Synthesis of γ-Amino Alcohols
Christophe Blaszykowski, Anne-Lise Dhimane, Louis Fensterbank, and
Max Malacria*
Laboratoire de Chimie Organique, UMR CNRS 7611, UniVersite´ Pierre et Marie
Curie, Case 229, 4 place Jussieu, 75252 Paris Cedex 05, France
malacria@ccr.jussieu.fr
Received February 19, 2003
ABSTRACT
Various allylic and propargylic amines bearing a protecting group (PG) have been employed in N-silyl-tethered radical cyclizations. The resulting
silapyrrolidine adducts could be smoothly oxidized, creating access to γ-amino alcohols. The silylation, radical cyclization, and oxidation
reactions could be consolidated in a one-pot process.
The concept of temporary linking reacting partners in view
of overcoming unfavorable entropy effects, forcing regio-
selectivity, and controlling stereoselectivity has been well-
illustrated over the last two decades.1 Several approaches
have been designed relying, for instance, on pivotal atoms
such as boron (boronate tethers),2 sulfur,3 Lewis acidic salts,4
and more prominently silicon (siloxanes, silaketals, and silyl
ethers).5 All these linkages are based on at least one alcohol
component. In contrast, amines that are key substrates in
organic synthesis have been little involved for that purpose6
and only sporadic reports of N-silyl-tethered Sakurai-type
reaction,7 Diels-Alder cycloaddition,8 Zr-mediated cycliza-
tion,9 hydrosilylation,10 silylformylation,11 and bis-silylation12
have appeared. Interestingly, to the best of our knowledge,
the N-silyl linkage has never been utilized in the context of
radical chemistry. So we turned our attention to the pos-
sibility of developing a nitrogen version of the bromomethyl-
dimethylsilyl (BMDMS) allyl13 and propargyl ether14 radical
cyclizations (Scheme 1). We first examined the preparation
of N-silyl precursors 2, with the concern that aminosilanes
are more fragile substrates than the corresponding ethers,15
(7) Bismara, C.; Di Fabio, R.; Donati, D.; Rossi, T.; Thomas, R. J.
Tetrahedron Lett. 1995, 36, 4283-4286.
(1) For the conceptualization of this, see: Stork, G.; Kim, G. J. Am. Chem.
Soc. 1992, 114, 1087-1088.
(8) Brosius, A. D.; Overman, L. E.; Schwink, L. J. Am. Chem. Soc. 1999,
121, 700-709.
(2) Nicolaou, K. C.; Liu, J.-J.; Yang, Z.; Ueno, H.; Sorensen, E. J.;
Claiborne, C. F.; Guy, R. K.; Hwang, C.-K.; Nakada, M.; Nantermet, P. G.
J. Am. Chem. Soc. 1995, 117, 634-644. (b) Batey, R. A.; Smil, D. V.
Angew. Chem., Int. Ed. 1999, 38, 1798-1800 and references therein.
(3) Bachi, M. D.; Bilokin, Y. V.; Melman, A. Tetrahedron Lett. 1998,
39, 3035-3038. (b) Mascaren˜as, J. L.; Rumbo, A.; Castedo, L. J. Org.
Chem. 1997, 62, 8020-8021 and references therein.
(9) Probert, G. D.; Harding, R.; Whitby, R. J.; Coote, S. J. Synlett 1997,
1371-1374.
(10) Tamao, K.; Nakagawa, Y.; Ito, Y. J. Org. Chem. 1990, 55, 3438-
3439. (b) Tamao, K.; Nakagawa, Y.; Ito, Y. Organometallics 1993, 12,
2297-2308.
(11) Ojima, I.; Vidal, E. S. Organometallics 1999, 18, 5103-5107.
(12) Murakami, M.; Sugimone, M.; Fujimoto, K.; Nakamura, H.;
Andersson, P. G.; Ito, Y. J. Am. Chem. Soc. 1993, 115, 6487-6498.
(13) Nishiyama, H.; Kitajima, T.; Matsumoto, M.; Itoh, K. J. Org. Chem.
1984, 49, 2298-2300. (b) Stork, G.; Kahn, M. J. Am. Chem. Soc. 1985,
107, 500-501.
(14) Magnol, E.; Malacria, M. Tetrahedron Lett. 1986, 27, 2255-2256.
(15) Si-N bond is 77 kcal/mol, about 24 kcal/mol weaker than the Si-O
bond. For a discussion of this, see: Armitage, D. A. In The Chemistry of
the Silicon-Heteroatom Bond; Patai, S., Rappoport, Z., Eds.; John Wiley
& Sons: New York, 1991; pp 365-445.
(4) Bertozzi, F.; Olsson, R.; Fredj, T. Org. Lett. 2000, 2, 1283-1286.
(b) Ward, D. E.; Saeed Abaee, M. Org. Lett. 2000, 2, 3937-3940 and
references therein.
(5) For reviews, see: (a) Bols, M.; Skydstrup, T. Chem. ReV. 1995, 95,
1253-1277. (b) Fensterbank, L.; Malacria, M.; Sieburth, S. M. Synthesis
1997, 813-854. (c) Gauthier, D.; Zandi, K. S.; Shea, K. J. Tetrahedron
1998, 54, 2289-2338.
(6) For an exemple based on phosphonamide species, see: Sprott, K. T.;
McReynolds, M. D.; Hanson, P. R. Org. Lett. 2001, 3, 3939-3942 and
references therein.
10.1021/ol034288j CCC: $25.00 © 2003 American Chemical Society
Published on Web 03/27/2003

salts.17 So we examined the possibility of running this
sequence at lower temperatures by using photochemical
initiation conditions. This proved to be beneficial since 4a
was obtained in a better yield (62%), accompanied by only
18% of 1a. Optimization of the conditions was achieved by
running an irreversible deprotonation of the amine with
n-BuLi and using photochemical initiation (69% of 4a,
Scheme 2).
Scheme 1. N-Silyl-Tethered Radical Cyclization
We then pursued the use of an allyl partner as a radical
acceptor, varying the nature of the PG on amines 1, and
focused on the preparation of γ-amino alcohols 5 after final
Tamao oxidation (Table 1). Our initial result (entry 1) with
Table 1.
and that the bromomethyl group could also be a supplemen-
tary source of instability.16 The latter was in part confirmed
by a preliminary experiment: while N-Boc allylamine could
be cleanly silylated with trimethylsilyl chloride (70% after
chromatography), only untractable material was obtained
with BMDMS chloride. Moreover, no precursor 2 could be
isolated with a benzyl protecting group (PG), whereas the
tosyl one showed moderate stability. In contrast, N-TMS and
N-phenyl precursors have been obtained efficiently (> 70%
yield).
These findings suggested to us that it could be necessary
to conduct the silylation, radical cyclization, and further
heterocycle 3 treatment in a one-pot operation. This principle
was initially tested on tosylallylamine 1a as shown in Scheme
2. Thus, silylation of 1a with BMDMSCl in the presence of
entry
amine 1, PG
5, yield (%)
1
2
3
4
5
6
1a , Ts
1a , Ts
1b, Ph
1b, Ph
1c, Bn
1d , Boc
5a , 23a
5a , 63b
5b, 18c
5b, 65d
5c,
5d , 75
a Siloxane dimer was isolated in 44% yield. b HBF4.OEt2 treatment was
run prior to Tamao oxidation. c Bromoanilines were also isolated in 44%
yield. d KF/MeOH treatment was run prior to Tamao oxidation.
a Ts protecting group showed that the preparation of a
γ-amino alcohol derivative was possible; however, the yield
was modest (23% 5a), due to the reluctant oxidation of the
siloxane (-Si-O-Si- dimer)8,18 that is formed during the
Tamao oxidation. So we had to have recourse to a HBF4
treatment prior to the oxidative treatment in order to cleave
the N-Si bond and generate a fluorosilane that could be
smoothly oxidized as proved by the increased yield of 5a in
entry 2 (63%). The aniline precursor 1b (entry 3) also gave
a satisfactory yield of cyclized adducts. However, the
expected product 5b (18%) was accompanied by 44%
monobrominated adducts in para and ortho positions (3:1
Scheme 2
1
ratio) on the aniline ring. Examination of the H NMR of
the crude product of the radical cyclization showed that this
bromination pathway was not taking place at that stage.
Presumably, the Tamao conditions would promote the
Et3N and DMAP, followed by a slow addition of tributyltin
hydride in thermal initiation conditions and methyllithium
treatment provided the expected branched amine 4a in 28%
overall yield, with 47% recovery of 1a. Although this result
demonstrated the feasability of the one-pot reaction, we
suspected that the aminosilane decomposed to some extent
in these thermal conditions in the presence of ammonium
(17) For a discussion of the cleavage of aminosilanes with ammonium
salts, see: Anderson, H. H. J. Am. Chem. Soc. 1951, 73, 5802-5803. See
also refs 15 and 16a.
(18) Tamao, K.; Ishida, N. J. Organomet. Chem. 1984, 269, C37-C39.
(b) Tamao, K. J. Synth. Org. Chem. Jpn. 1988, 46, 861-878.
(19) For related oxidative brominations, see: Narender, N.; Srinavasu,
P.; Ramakrishna Prasad, M.; Kulkarni, S. J.; Raghavan, K. V. Synth. Comm.
2002, 32, 2313-2318 and references therein.
(16) Survey of literature data shows that only very few of these substrates
with no functionality have been described; see: (a) Niederpru¨m, H.; Simmler,
W. Chem. Ber. 1963, 96, 965-975. (b) Yoder, C. H.; Cader, B. M. J.
Organomet. Chem. 1982, 233, 275-279 and references therein.
(20) Compound 2e has been prepared by two consecutive N-silylations
of allylamine (72% yield) and been found to be stable to aqueous workup
and bulb to bulb distillation. For a rationalization of this stability, see:
Schorr, M.; Schmitt, W. Phosphorus, Sulfur, Silicon 1992, 68, 25-35.
1342
Org. Lett., Vol. 5, No. 8, 2003

oxidation of Bu3SnBr to a Br+ species19 that would be readily
intercepted by the activated aniline moiety. To preclude this
side reaction, a treatment of the crude cyclization product,
prior to Tamao oxidation, with KF in methanol followed by
a filtration over Celite eliminated most of the tin residues
and inorganic salts, which allowed the improvement of the
yield of 5b to 65% (entry 4). Finally, the one-pot process
proved to be completely inapplicable to benzyl precursor 1c
(entry 5) but very versatile in the case of Boc amine 1d,
giving birth to 75% yield of product 5d (entry 6).
NMR spectra of the crude product (Scheme 3). Then, an
unprecedented one-pot Tamao oxidation and Boc protection
sequence furnished the previously synthesized Boc adduct
5d in 16% overall yield (this yield being presumably lowered
by the volatility of intermediate 3e).
More astonishingly, only a trace (<2%) of 6-endo cy-
clization was observed with methallyl precursor 6, which
followed cleanly the 5-exo mode (Scheme 4). This result
strongly contrasts with Nishiyama’s results involving the
corresponding methallyl BMDMS ether precursor, which
reacts in a 2:1 6-endo versus 5-exo reversed selectivity.13a
Presumably, a slower 5-exo-trig process as well as a still
unfavorable 6-endo process would account for the slightly
depressed, but consistent with the literature, yield of this
transformation.
Scheme 3
Finally, propargyl precursors of type 8 could be engaged
in this process to give versatile γ-amino alcohols 11 as a
7:3 mixture of Z/E diastereomers (Scheme 5). This stereo-
chemical assignment was based on NOE measurements and
13C NMR γ-gauche effects.21 In the case of the tosyl
precursor 8a, we faced the same problem of siloxane
formation, which is reluctant to undergo oxidation and results
in a lesser yield of 11a. Additional treatment by HBF4‚OEt2
is not the solution here since it resulted in the formation of
allylamine 12. The overall transformation 8a f 12 is also
worthy of note. Boc precursor 8b reacted cleanly and gave
the same ratio of stereoisomers. This moderate (Z)-stereo-
selectivity originates from a reduction of the vinyl radical,
which displays the minimized allylic strain between the alkyl
chain and the more remote CH2-heteroatom moiety22 (the
C-Si bond is approximately 30% longer than the C-N
bond).
A notable feature of all these reactions is their complete
regioselectivity in favor of the 5-exo-trig process, and this
has to be compared with the results of Nishiyama in the silyl
ether series, who always observed the 6-endo mode on allylic
systems without terminal functionality.13a This was further
Scheme 4
In conclusion, we have reported the first examples of
radical cyclizations of BMDMS amino derivatives. We have
worked out a new and efficient access to γ-amino alcohols
by combining the N-silylation, radical cyclization, and
oxidation steps in a one-pot process. Tuning of the reaction
conditions (notably at the oxidation stage) according to the
nature of the amine protecting group has been devised.
Numerous γ-amino alcohol derivatives exhibit important
confirmed by the radical cyclization of the precursor 2e,20
which provided only the silapyrrolidine 3e as judged by the
Scheme 5
Org. Lett., Vol. 5, No. 8, 2003
1343

biological activities,23 and our efforts are now devoted toward
these targets. Moreover, thanks to the trivalent nature of the
pivotal nitrogen atom, further applications focusing on the
attachment of chiral auxiliaries for an asymmetric version
or grafting to polymers in view of supported synthesis are
under investigation.
Acknowledgment. C.B. thanks the Ministe`re de la
Recherche for an IGERT grant. The authors thank Prof. Eric
Enholm (University of Florida) for stimulating discussions.
(21) Knorr, R.; Hintermeier-Hilpert, H.; Bo¨hrer, P. Chem. Ber. 1990,
123, 1137-1141.
Supporting Information Available: Experimental pro-
cedures and spectral data for all new products. This material
is available free of charge via the Internet at http://pubs.acs.org.
(22) For a model of the reduction of vinyl radicals, see: Journet, M.;
Malacria, M. J. Org. Chem. 1992, 57, 3085-3093.
(23) See for instance: Lee, J.; Park, S.-U.; Kim, J.-Y.; Kim, J.-K.; Lee,
J.; Oh, U.; Marquez, V. E.; Beheshti, M.; Wang, Q.-J.; Modarres, S.;
Blumberg, P. M. Bioorg. Med. Chem. Lett. 1999, 9, 2909-2914.
OL034288J
1344
Org. Lett., Vol. 5, No. 8, 2003