Regiochemistry of copper(I)-mediated cyclization reactions of halo-dienamides

Article abstract of DOI:10.1021/jo070680p

(Chemical Equation Presented) Reaction of 2-substituted dienamides with catalytic amounts of copper halide/tripyridylamine (TPA) furnishes either 5-exo or 6-endo products with the outcome dependent upon the radical initiating unit. Reaction of 3-substitut

Full text of DOI:10.1021/jo070680p

mediates the cyclization of 3,3-disubstituted enamides 1a via a
4-exo atom transfer reaction mode to give 2a6 and CuBr/Me6-
tren facilitates the cyclization of 2,3-disubstituted enamides 3a
via a 5-endo radical polar crossover process to give 4a (Scheme
1).4b During the course of these studies, we noted that it was
not possible to mediate the reaction of corresponding 3- and
2-monosubstituted derivatives (1b and 2b, R2 ) H) using these
and forcing (refluxing toluene) conditions.This prompted us to
explore radical cyclizations onto dienamides. We postulated that
cyclization of suitably substituted dienamides might facilitate
these reactions by stabilizing thermodynamically the cyclized
radicals as allyl species 7 and 8. While the regiochemical
outcome of radical cyclizations onto monoenamides has been
well-documented,7 to our knowledge, no information as to the
regiochemistry of addition to dienamides of type 5 or 6 has
appeared in the literature. While we predicted that cyclization
of radical 5 would lead to the 4-exo product 7, we could not
rule out cyclization via alternative 5-endo, 6-exo, or 7-endo
reaction pathways. Similarly for the cyclization of radicals of
type 6, alternative 4-exo, 5-exo, or 6-endo pathways could not
be dismissed. While 7-endo cyclization of 5 and 6-endo
cyclization of 6 also led to allylic radicals 9 and 10, these modes
of cyclization are rare and we did not expect to observe them
(Scheme 2).
Regiochemistry of Copper(I)-Mediated
Cyclization Reactions of Halo-dienamides
Andrew J. Clark,*,† Joanna V. Geden,† Stephen Thom,‡ and
Paul Wilson†
Department of Chemistry, UniVersity of Warwick, CoVentry,
West Midlands, CV4 7AL, U.K., and AstraZeneca Research
and DeVelopment, Charnwood, Bakewell Road,
Loughborough, Leicestershire, LE11 5RH, U.K.
msrir@csV.warwick.ac.uk
ReceiVed April 23, 2007
Reaction of 2-substituted dienamides with catalytic amounts
of copper halide/tripyridylamine (TPA) furnishes either 5-exo
or 6-endo products with the outcome dependent upon the
radical initiating unit. Reaction of 3-substituted dienamides
produces â-lactams via a 4-exo cyclization with termination
of the reaction occurring via either halogen atom transfer,
trapping with oxygen, elimination, or radical-radical cou-
pling depending upon the diene.
SCHEME 1
Copper(I)-mediated halogen atom transfer radical and radi-
cal-polar crossover cyclizations have attracted considerable
interest in recent years.1 Initially, reactions were restricted to
the cyclization of trichloro derivatives using CuCl and bipyri-
dine2 or TMEDA3 as ligands. Recently, more active catalysts
derived from Me6-tren4 or tripyridylamine (TPA)5 have been
developed which facilitate the cyclization of monobromo
derivatives (e.g., 1 and 3). While the majority of work has
focused on 5-exo cyclizations, we have reported that CuBr/TPA
SCHEME 2. Possible Modes of Cyclization of Dienamides
† University of Warwick.
‡ AstraZeneca R&D.
(1) Clark, A. J. Chem. Soc. ReV. 2002, 1.
(2) (a) Nagashima, H.; Ozaki, N.; Ishii, M.; Seki, K.; Washiyama, M.;
Itoh, K. J. Org. Chem. 1993, 58, 464. (b) Nagashima, H.; Isono, Y.;
Iwamatsu, S. J. Org. Chem. 2001, 66, 315. (c) Udding, J. H.; Tuijp, C. J.
M.; van Zanden, M. N. A.; Hiemstra, H.; Speckamp, J. Org. Chem. 1994,
59, 1993. (d) Bryans, J. S.; Chessum, N. E. A.; Huther, N.; Parsons, A. F.;
Ghelfi, F. Tetrahedron 2003, 59 6221.
(3) (a) Benedetti, M.; Forti, L.; Ghelfi, F.; Pagnoni, U. M.; Ronzoni, R.
Tetrahedron 1997, 41, 14031. (b) De Buyck, L.; Cagnoli, R.; Ghelfi, F.;
Merighi, G.; Mucci, A.; Pagnoni, U. M.; Parsons, A. F. Synthesis 2004,
10, 1680.
(4) (a) Clark, A. J.; De Campo, F.; Deeth, R. J.; Filik, R. P.; Gatard, S.;
Hunt, N. A.; Laste´coue`res, D.; Thomas, G. H.; Verlac J.-B.; Wongtap H.
J. Chem. Soc., Perkin Trans. 1 2000, 671. (b) Clark, A. J.; Dell, C. P.;
Ellard, J. M.; Hunt, N. A.; McDonagh, J. P. Tetrahedron Lett. 1999, 40,
8619. (c) Clark, A. J.; Filik, R. P.; Thomas, G. H. Tetrahedron Lett. 1999,
40, 4885.
Reaction of 3-Substituted Dieneamides: Our initial inves-
tigations focused upon the copper-mediated atom transfer
cyclization reactions of dienamides 11a-d (Scheme 3). We
expected that these compounds would cyclize in a 4-exo mode
based upon precedent from our studies with the corresponding
3,3-disubstituted compounds. Initial imine formation (by con-
(5) (a) De Campo, F.; Laste´coue`res, D.; Verlac J.-B. J. Chem. Soc., Chem.
Commun., 1998, 2117. (b) Clark, A. J.; Dell, C. P.; McDonagh, J. P. C.R.
Acad. Sci. Ser IIC: Chim. 2001, 4, 575.
(6) Clark, A. J.; Battle, G. M.; Bridge, A. Tetrahedron Lett. 2001, 42,
4409.
(7) Ishibashi, H. Chem. Rec. 2006, 6, 23.
10.1021/jo070680p CCC: $37.00 © 2007 American Chemical Society
Published on Web 06/28/2007
J. Org. Chem. 2007, 72, 5923-5926
5923

SCHEME 3. Preparation of 11a-d
SCHEME 6. Cyclization of Trichloro Analogues
SCHEME 4. Cyclization Reactions of 11a,b
19 (presumably from elimination of the undetected 3° bromide)
and the dimer 20 (as a 1:1 mixture of diastereomers).
Interestingly, reaction of the analogues 21a-c (Scheme 6)
was significantly slower than that of their 11a-d counterparts
(18 h for 21a compared to 2 h for 11a). For the reaction of
21a, a significant amount of the oxidized aldehyde 23 was
detected (22a:23 ) 7:1). This ratio could be reversed if the
reaction was carried out under oxygen (22a:23 ) 1:2); however,
the yield dropped to 26%.8 Reaction of derivative 21b (a 5:1
mixture of rotamers at room temperature) gave a complex
mixture of decomposition products on treatment with 30 mol
% of CuCl/TPA in refluxing toluene for 20 h. The known amide
13b was the only identifiable compound isolated from the crude
mixture (albeit impure).9 On the basis of this result, dichloro-
dienamides were not evaluated further. Cyclization of 21c
proceeded in the same manner as the analogous 13d, albeit in
a better yield. In addition to the two products 24 and 25, a trace
amount of the 3° chloride 22b was isolated (5%), indicating
that the diene 24 is probably produced via thermal elimination
of HCl from 22b.
SCHEME 5. Cyclization Reactions of 11c,d
Encouraged by the higher yields obtained in the cyclization
of the trichloro derivative 21b, we briefly explored the scope
of cyclization onto the hindered derivative 26 (obtained from
â-cyclocitral). Reaction yielded the 6-exo product 27 (as one
diastereomer as determined by NOE) and the 4-exo product 28
in a 3:2 ratio in poor combined yield (31%; Scheme 7). The
rest of the mass balance was baseline material.
Surprised by the regiochemical outcome of cyclization of 26
(where the 6-exo product predominated), we next investigated
the regiochemical outcome and relative rate of cyclization onto
the dieneamide functionality 29 in a competition experiment
(Scheme 8). In this case, cyclization onto the diene 29 would
be competitive with a traditional 5-exo cyclization. We also
compared the outcome of this reaction with that of the 3,3-
disubstituted alkene 30. While cyclization of 30 at room
temperature furnished an inseparable mixture of the products
(33:32, 5:1, 97%), the only product arising from reaction of 29
was the corresponding 5-exo product 31 (91%), indicating that
cyclization onto the 4-exo position of the diene fragment is
slower than both the 5-exo reaction and the 4-exo cyclization
densation of aldehydes 12a-d with p-methoxybenzylamine)
followed by Et3N-mediated acylation with 2-bromo-2-methyl-
propionyl bromide produced the desired cyclization precursors
11a-d. The low yield (12%) for the synthesis of the dienamide
11a was due to its relative instability when compared to that of
the other compounds (13a, 20% was isolated during the
purification of 11a).
Treatment of a 0.12 M solution of 11a-d in toluene with 30
mol % of CuBr/TPA furnished the corresponding â-lactams 14-
20 (Schemes 4 and 5). The compounds 11a,b underwent fairly
rapid reactions (2 h) to give the primary allyl bromides 14 and
15 in 60 and 57% yields, respectively. Trapping of the
intermediate cyclized allyl radicals only occurred at the 1°
position in both cases.
Compounds 11c,d (which contained substitution at the
terminal carbon of the diene system) underwent much slower
reactions (19 h) and gave rise to mixtures of products 16-20.
Reaction of 11c (2:1 E:Z) led to three â-lactam products arising
from atom transfer 16 (as a 1:1 mixture of diastereomers),
elimination 17, and oxidation 18. Heating a pure sample of the
secondary bromide 16 at reflux in toluene produced a 3.5:1 ratio
of 16:17 after 6 h, providing tentative evidence that the
eliminated product 17 arises via thermal elimination from the
bromide 16. Reaction of 11d produced the corresponding diene
(8) Presumably caused by oxidation of the CuCl/TPA complex to the
corresponding Cu(II) peroxo complex: Chishiro, T.; Shimazaki, F.; Tani,
F.; Tachi, Y.; Naruka, Y.; Karasawa, S.; Hayani, S.; Maeda, Y, Angew.
Chem., Int. Ed. 2003, 42, 2788.
(9) Gani, V.; Viout, P. Tetrahedron 1978, 34, 1337.
5924 J. Org. Chem., Vol. 72, No. 15, 2007

SCHEME 7. Cyclization of 26
SCHEME 9. Cyclization of 2-Substituted Dienamides
SCHEME 8. Competition Experiments
SCHEME 10. Cyclization of 2-Substituted Dienamides
of an analogous 3,3-disubstituted alkene. Cyclization of 30 at
reflux furnished the thermodynamic product 33 only (95%).
Reaction of 2-Substituted Dieneamides: Trichloro deriva-
tives with substituents at the 2-position in enamides (e.g., 34a)
have been reported to undergo cyclization exclusively in the
5-endo mode, while analogous compounds with less activated
initiating functionality (e.g., 34b) fail to undergo any cyclization.4b
Thus we next investigated the cyclization reactions of 35a-c
(Scheme 9), and in order to potentially increase the bias toward
the 5-endo mode of cyclization, we introduced substituents at
the 5-exo and 6-endo positions to retard competitive reaction
at these points. Cyclization of the bromide 35b furnished a 72%
yield of the 5-exo cyclization product 36 (5:2 mixture of
diastereomers) with no 5-endo products being detected.
In conclusion, we have investigated the regiochemical
outcome of Cu(I)-mediated cyclization of 2- and 3-substituted
dienamides. For 3-dienamides derived from aldehydes (11a-
d, 21a-b) cyclization proceeds via a 4-exo cyclization mode
mirroring the observed outcome of the related 3,3-disubstituted
enamides (e.g., 1a). Termination of the reaction via a 1° radical
generally leads to halogen atom transfer, while termination by
a 2° radical leads to mixtures of atom transfer, elimination, and
oxygen trapping, making the reactions less synthetically useful.
Cyclization leading to 3° radicals leads to elimination and
dimerization. On the other hand, reaction of 2-substituted
dienamides (35b,c and 41) derived from ketones did not proceed
via the expected 5-endo mode of cyclization (based upon
precedence with the reaction of 1b), instead proceeding via the
5-exo mode predominantly. The nature of the initiating radical
functionality in 35a-c had a profound outcome on the regio-
chemistry with 35a proceeding via the 6-endo cyclization mode.
Cyclization of the 2° bromide 35c also delivered the 5-exo
product 37 (2:1 mixture of diastereomers, 67%) but with a minor
amount (7%) of the formal 6-endo product 38. Two of the four
possible diastereomers of 37 were produced and tentatively
assigned as 37a,b based upon NOE data. In both cases, the
expected 5-endo cyclization mode was not observed. This was
surprising in light of the exclusive 5-endo cyclization of 34a
which has a similar steric demand. While reports of 6-endo
cyclizations have appeared in the literature,7 they are relatively
rare.10 So it was interesting to observe that these were the major
products arising from the cyclization of 35a (50%). Finally we
investigated the cyclization of the dienamide 41 which was
found to convert to a single product 42 in 71% yield when
treated with CuCl and TPA in toluene at reflux (Scheme 10).
Experimental Section
General Procedure for the Synthesis of Dienamides 11a-d,
21a,b, 26, 29, 31, and 32 Derived from Aldehydes. The amine
(4-methoxybenzylamine or allylamine) (40 mmol) was added to
the appropriate aldehyde (40 mmol) at 0 °C. After 15 min, the
mixture was dissolved in diethyl ether (100 mL), dried over MgSO4,
(10) Cyclization onto imines: (a) Tajino, M.; Otsuka, N.; Fukuyama,
T.; Matsubara, H.; Ryu, I. J. Am. Chem. Soc. 2006, 128, 7712. (b)
Cyclization of vinyl radicals: (c) Quirante, J.; Vila, X.; Palona, L.; Guiu,
J. M.; Bonjoch, J. Tetrahedron 2007, 63, 1372.
J. Org. Chem, Vol. 72, No. 15, 2007 5925

filtered, and concentrated in vacuo to give a crude imine that was
used without further purification. To this imine (4 mmol) at 0 °C
in anhydrous CH2Cl2 (20 mL) under nitrogen was added either
trichloroacetyl chloride (4 mmol) or 2-bromoisobutyryl bromide
(4 mmol) followed by triethylamine (4 mmol). After stirring at
0 °C for 3 h, the reaction mixture was washed with water (50 mL)
and brine (50 mL). The organic layer was dried over MgSO4,
filtered, concentrated in vacuo, and purified by column chroma-
tography.
General Procedure for the Synthesis of Dienamides 37a-c
and 47 Derived from Ketones. A mixture of 4-methoxybenzy-
lamine (15 mmol), the appropriate ketone (15 mmol), and anhydrous
ZnCl2 (100 mg) in dry toluene (60 mL) was heated at reflux for 16
h with azeotropic removal of water. Filtration followed by removal
of the solvent in vacuo furnished a crude imine that was used
without further purification. To this imine (4 mmol) at 0 °C in
anhydrous CH2Cl2 (20 mL) under nitrogen was added over 10 min
either trichloroacetyl chloride (4 mmol), 2-bromopropionyl bromide
(4 mmol), or 2-bromoisobutyryl bromide (4 mmol) followed by
N,N-diethylaniline (4 mmol). After stirring for 16 h at room
temperature, the reaction mixture was washed with 1 N HCl (100
mL) and brine (100 mL). The organic layer was dried over MgSO4,
filtered, concentrated in vacuo, and purified by column chroma-
tography.
General Procedure for Cyclization Reactions. To a solution
of the cyclization substrate in dry toluene (0.012 M) under nitrogen
were added CuCl or CuBr (0.3 equiv) and tripyridylamine (0.3
equiv). The mixture was heated at reflux (between 2 and 16 h),
cooled, and filtered through silica with elution with EtOAc. The
combined filtrate was concentrated in vacuo, and the residue was
purified by column chromatography.
Acknowledgment. We acknowledge AstraZeneca for a
studentship (J.V.G.).
Supporting Information Available: Selected experimental
procedures and characterization data for all compounds and
representative 1H NMR for 11, 13-33, and 35-42 (the inseparable
mixture of 32 and 33), and 13C NMR data for 14-20, 22-25, 27,
28, 31, 33, and 36-40. This material is available free of charge
via the Internet at http://pubs.acs.org.
JO070680P
5926 J. Org. Chem., Vol. 72, No. 15, 2007