Cobalt-catalyzed tandem radical cyclization and cross-coupling reaction: Its application to benzyl-substituted heterocycles [14]

Full text of DOI:10.1021/ja0100423

5374
J. Am. Chem. Soc. 2001, 123, 5374-5375
Scheme 1
Cobalt-Catalyzed Tandem Radical Cyclization and
Cross-Coupling Reaction: Its Application to
Benzyl-Substituted Heterocycles
Katsuyu Wakabayashi, Hideki Yorimitsu, and
Koichiro Oshima*
Table 1. Cobalt-Catalyzed Phenylative Radical Cyclizationa
Department of Material Chemistry
Graduate School of Engineering, Kyoto UniVersity
Yoshida, Sakyo-ku, Kyoto 606-8501, Japan
ReceiVed January 4, 2001
substrate
X
R1
Br n-C4H9
n-C4H9
R2 R3
R4
R5 product
yieldb
Radical reaction mediated by group 14 metal hydrides, mainly
tin hydrides, has been extensively investigated and widely used
in organic synthesis.1 Accordingly, radical reaction with other
metallic reagents is a fascinating and developing area in synthetic
radical chemistry.2-9 The most attractive feature of this strategy
is the generation of a new carbon-metal bond by the capture of
a carbon-centered radical, derived from a certain radical trans-
formation, with a metallic reagent. Sequential ionic reaction offers
multibond-forming events, that is, a heterogenerative process10
(Scheme 1). However, further carbon-carbon bond formation is
not always easy because of the low stability or reactivity of the
resulting organometallics. Electrophiles used for elongating a
carbon chain are allyl halides and reactive carbonyl compounds
such as acid chloride, aldehyde, and ketone. In a few cases,
activated carbon-carbon multiple bonds4 and alkyl halides6 could
be employed as electrophiles. On the other hand, cross-coupling
reaction with alkenyl or aryl halides seems difficult. Although
we have also developed radical cyclization of 6-halo-1-hexene
derivatives mediated by organomagnesium,11 -manganese,12 and
-iron13 reagents, the following carbon-carbon bond formation was
quite limited. Attempts to combine a cyclopentylmethylmetallic
species with iodobenzene resulted in very poor yields of the cross-
coupling product. Here we wish to report a new route to
phenylation of a cyclopentylmethy group. Cobalt-catalyzed
coupling reaction of halo acetal with a phenyl Grignard reagent
proceeded smoothly, involving radical cyclization prior to cou-
pling. Coupling reaction under cobalt catalysis is rare14,15 com-
pared with nickel, palladium, and copper catalysts,16 and remains
to be studied.
1a
1b
1c
1d
1e
1f
H
H
H
H
H
H
H
n-C5H11
n-C5H11
n-C5H11
n-C5H11
H
2
2
2
3
4
4
5
6
80% (55/45)
78% (55/45)
N.R.
71% (51/49)
84% (62/38)
84% (60/40)
51% (single)
22% (91/9)
I
H
H
Cl n-C4H9
Br (CH2)3
Br (CH2)3
H
H
Me Me
Me Me
H
H
I
(CH2)3
Br (CH2)3
(CH2)3
H
1g
1h
H
H
Me
H
I
a Substrate (0.5 mmol), CoCl2(dppe) (0.05 mmol), PhMgBr (1.1
mmol), and THF (1 mL) were employed. b Isolated yield. Diastereomer
ratios are in parentheses.
Scheme 2
Bromo acetal 1a was treated with phenylmagnesium bromide
in THF at 0 °C in the presence of CoCl2(dppe)17,18 under argon.
After being stirred for 30 min, the reaction mixture was quenched
with saturated ammonium chloride solution. Usual workup
followed by silica gel column purification provided phenylated
cyclic acetal 2 in 80% yield (Table 1, entry 1).
Various substrates were examined, and the results are sum-
marized in Table 1, Scheme 2 and Scheme 3. Halo acetal bearing
a terminal alkene moiety underwent phenylative cyclization to
give the corresponding benzyl-substituted tetrahydrofuran deriva-
tive in good to excellent yield (Table 1). It is worth noting that
the stereochemistry of the products was quite similar to that in
the previous reports of radical reaction.4,11-13,19 This observation
is highly suggestive of the same transition state of the cyclization
(1) (a) Curran, D. P.; Porter, N. A.; Giese, B. Stereochemistry of Radical
Reactions; VCH Verlagsgesellschaft mbH.; Weinheim, 1996. (b) Giese, B.;
Kopping, B.; Go¨bel, T.; Dickhaut, J.; Thoma, G.; Kulicke, K. J.; Trach, F.
Org. React. 1996, 48, 301-856.
(2) SmI2: Curran, D. P.; Totleben, M. J. J. Am. Chem. Soc. 1992, 114,
6050-6058 and references therein.
(3) PhLi: Bailey, W. F.; Carson, M. W. J. Org. Chem. 1998, 63, 9960-
9967.
(4) Et2Zn-mediated reaction with Pd or Ni catalyst. Further carbon-carbon
bond formation was carried out after converting a zinc species into organo-
copper: (a) Knochel, P. Synlett 1995, 393-403. (b) Stadtmu¨ller, H.; Vaupel,
A.; Tucker, C. E.; Stu¨demann, T.; Knochel, P. Chem. Eur. J. 1996, 2, 1204-
1220.
(5) Cobaloxime and related cobalt complexes: (a) Pattenden, G. Chem.
Soc. ReV. 1988, 17, 361-382. (b) Ali, A.; Harrowven, D. C.; Pattenden, G.
Tetrahedron Lett. 1992, 33, 2851-2854. (c) Torii, S.; Inokuchi, T.; Yukawa,
T. J. Org. Chem. 1985, 50, 5875-5877. (d) Okabe, M.; Tada, M. J. Org.
Chem. 1982, 47, 5382-5384.
(14) (a) Cahiez, G.; Avedissian, H. Tetrahedron Lett. 1998, 39, 6159-
6162. (b) Avedissian, H.; Be´rillon, L.; Cahiez, G.; Knochel, P. Tetrahedron
Lett. 1998, 39, 6163-6166. (c) Nishii, Y.; Wakasugi, K.; Tanabe, Y. Synlett
1998, 67-69. (d) Kharasch, M. S.; Fields, E. K. J. Am. Chem. Soc. 1941, 63,
2316-2320. (e) Elsom, L. F.; Hunt, J. D.; McKillop, A. Organomet. Chem.
ReV. A 1972, 8, 135-152.
(15) Coupling reaction of organic halides with stoichiometric cobalt-ate
complexes, such as Me4CoLi2, was reviewed: Kauffmann, T. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 386-403.
(16) ComprehensiVe Organic Synthesis; Trost, B. M., Fleming, I., Heath-
cock, C. H., Eds.; Pergamon Press: New York, 1991; Vol. 3, Chapter 2.1-
2.5.
(6) Catalytic Cp2TiCl2 with Grignard reagent: (a) Terao, J.; Saito, K.; Nii,
S.; Kambe, N.; Sonoda, N. J. Am. Chem. Soc. 1998, 120, 11822-11823. (b)
Nii, S.; Terao, J.; Kambe, N. J. Org. Chem. 2000, 65, 5291-5297.
(7) CrCl2: Takai, K.; Matsukawa, N.; Takahashi, A.; Fujii, T. Angew. Chem.,
Int. Ed. 1998, 37, 152-155.
(17) The catalyst CoCl2[1,2-bis(diphenylphosphino)ethane] was prepared
by mixing THF solutions of CoCl2 and of dppe. It is crucial for success that
even a trace of water should be removed. When hygroscopic CoCl2 was
insufficiently dried, the nonphenylated cyclic product was the major product,
probably because an active cobalt catalyst for phenylation was not formed in
the presence of water. In each experiment, CoCl2 and dppe were dried carefully
under reduced pressure (0.5 Torr) by heating with a hair-dryer for 30 min
immediately before use. Also see Supporting Information.
(18) Several ligands such as tmeda, dppm, dppe, dppp, dppb, PPh3, and
P(OPh)3 were examined. Among them, dppe was extremely efficient. The
use of other ligands resulted in lower yields of the phenylated product and
gave a nonphenylated byproduct (30-50%). CoCl2 itself and CoCl(PPh3)3
did not give satisfactory results. Employing Co(dmgH)2PyCl (see ref 5) resulted
in recovery of the starting material.
(8) Cp2TiCl: (a) Gansa¨uer, A. Synlett 1998, 801-809. (b) RajanBabu, T.
V.; Nugent, W. A. J. Am. Chem. Soc. 1994, 116, 986-987.
(9) Mn: Takai, K.; Ueda, T.; Ikeda, N.; Moriwake, T. J. Org. Chem. 1996,
61, 7990-7991.
(10) Wender, P. A. Chem. ReV. 1996, 96, 1-2.
(11) Inoue, A.; Shinokubo, H.; Oshima, K. Org. Lett. 2000, 2, 651-653.
(12) (a) Inoue, R.; Nakao, J.; Shinokubo, H.; Oshima, K. Bull. Chem. Soc.
Jpn. 1997, 70, 2039-2049. (b) Nakao, J.; Inoue, R.; Shinokubo, H.; Oshima,
K. J. Org. Chem. 1997, 62, 1910-1911.
(13) Hayashi, Y.; Shinokubo, H.; Oshima, K. Tetrahedron Lett. 1998, 39,
63-66.
10.1021/ja0100423 CCC: $20.00 © 2001 American Chemical Society
Published on Web 05/10/2001

Communications to the Editor
J. Am. Chem. Soc., Vol. 123, No. 22, 2001 5375
Scheme 3
Scheme 4
Table 2. Radical Cyclization with Various Arylmagnesium
Reagents
Ar
yield of acetal /%
yield of lactone /%
varying amounts of PhMgBr. Treatment of 1a (1 mmol) with the
cobalt complex, prepared from 1 mmol of CoCl2(dppe) and 2
mmol of PhMgBr, provided a trace of 2. Most of 1a was
recovered, and biphenyl was quantitatively obtained. Accordingly,
Co(II)Ph2(dppe) was unstable under the reaction conditions and
underwent reductive elimination.23 Three equivalents of PhMgBr
also gave biphenyl (1 mmol), and 1a remained unchanged.
Employing a complex generated by CoCl2(dppe) (1 mmol) and
4 mmol of PhMgBr dramatically changed the outcome. The
phenylated product 1a was obtained in reasonable yield (31%),
along with biphenyl (1.2 mmol). Therefore, the cobalt species
that is active for this reaction would be the 17-electron-ate
complex [Co(0)Ph2(dppe)](MgBr)2, which would potentially exist
as an equilibrium mixture of a mono-ate complex and the di-ate
complex. Judging from these results, we are tempted to propose
a draft mechanism for the catalytic reaction as shown in Scheme
4. The reaction of CoCl2(dppe) with 4 equiv of PhMgBr gives
[Co(0)Ph2(dppe)](MgBr)2 (17) with concomitant production of 1
equiv of biphenyl. The zero-valent-ate complex undergoes a single
electron transfer24 to a substrate to yield an anion radical of the
substrate and cobalt(I) complex 18. The immediate loss of
bromide from the anion radical affords 5-hexenyl radical inter-
mediate, which is transformed into a cyclopentylmethyl radical.
Then, the cobalt species 18 would recombine with the carbon-
centered radical to form divalent cobalt species 19. The following
reductive elimination provides the product and the Co(0) complex
20, which is reconverted into 17 by the action of the remaining
PhMgBr.
C6H5
2
80
65
14
97
83
95
-
3-CF3-C6H4
4-MeO-C6H4
2-MeO-C6H4
2-Me-C6H4
2-thienyl
15a
15b
15c
15d
15e
16a
16b
16c
16d
16e
81
<5
<15
63
-
75
step in the present reaction as in the free radical reaction. Allylic
alcohols, constituting the substrates, were not detected in the
reaction mixture. Thus, â-alkoxy elimination, which could be
facilitated by halogen-metal exchange, did not take place.
Therefore, a mechanism involving halogen-cobalt exchange
followed by intramolecular carbocobaltation might be improb-
able.20 Radical reaction is a preferable methodology to construct
a quaternary carbon, and this is indeed the case of 1g. Bicyclic
acetal 5 was obtained as a single isomer. However, the reaction
of 1h having a nonsubstituted allyloxy group was sluggish. The
expected product 6 was obtained in 22% yield in addition to
allylbenzene. Chloro acetal 1c did not undergo the reaction.
Nitrogen-containing substrate 7a was also subjected to cy-
clization in the presence of the cobalt catalyst to yield 3-ben-
zylpyrrolidine derivative 8a (Scheme 2). Synthesis of carbocycle
8c, where the Thorpe-Ingold effect does not work, was achieved
in good yield.21 However, in the case of cyclization onto an
internal alkene moiety, incorporation of a phenyl group was not
successful (Scheme 3). Treatment of 9 with PhMgBr in the
presence of CoCl2(dppe) gave a mixture of regio- and stereoiso-
mers in regard to the double bond formed, accompanied with the
nonphenylated cyclic acetal 11. When 12 was employed, the only
product obtained was 13 via regioselective â-hydride elimination.
Not only PhMgBr but also a wide range of aryl Grignard
reagents, including a 2-thienylmagnesium reagent, could be
employed (Table 2). Jones oxidation of the products yielded
â-arylmethyl-γ-lactones, which can be useful building blocks of
some lignans.22 Unfortunately, introduction of 2-substituted aryl
groups was problematic. Steric effect of the substituents on a metal
center is likely to upset the catalytic cycle.
In summary, cobalt-catalyzed sequential radical cyclization and
cross-coupling reaction provides a new methodology of multibond
formation in a single operation. This method promises high
efficiency in the construction of arylmethyl-substituted hetero-
cycles.
Acknowledgment. This work was supported by Grants-in-Aid for
Scientific Research (Nos. 12305058 and 10208208) from the Ministry
of Education, Culture, Sports, Science and Technology, Government of
Japan. H.Y. acknowledges the JSPS Research Fellowship for Young
Scientists.
To get deeper insight into the reaction mechanism, the reaction
of 1a with a stoichiometric cobalt complex was examined with
Supporting Information Available: Experimental details and char-
acterization data (PDF). This material is available free of charge via the
Internet at http://pubs.acs.org.
(19) Procedure for the stereochemical assignments is described in Sup-
porting Information.
(20) Stable â-hydroxy- or â-alkoxyalkylcobalt species are known: (a)
Grubb, L. M.; Branchaud, B. P. J. Org. Chem. 1997, 62, 242-243. (b)
Kettschau, G.; Pattenden, G. Tetrahedron Lett. 1998, 39, 2027-2028 and
references therein. These examples seem peculiar to cobaloximes. In our case,
electron-rich cobalt complexes would be active species. Therefore, the â-di-
(alkoxy)alkylcobalt, if formed, probably undergoes fast â-alkoxy elimination.
However, there remains a discussion, that is, which is faster, â-alkoxy
elimination or intramolecular carbocobaltation.
(21) Cyclic product 8c, in addition to a small amount of biphenyl, was the
only product. No uncyclized product, derived from simple cross coupling,
was detected. Methylcyclopentane and 1,5-hexadiene might be major byprod-
ucts.
JA0100423
(22) (a) Doyle, M. P.; Protopopova, M. N.; Zhou, Q.-L.; Bode, J. W.;
Simonsen, S. H.; Lynch, V. J. Org. Chem. 1995, 60, 6654-6655. (b) Ward,
R. S. Nat. Prod. Rep. 1995, 12, 183-205.
(23) In the absence of halo acetal, treating CoCl2(dppe) with 2 equiv of
PhMgBr in THF at 0 °C for 1 min gave biphenyl in 57% yield. Quenching
after 5 min yielded biphenyl quantitatively.
(24) Single-electron transfer from electron-rich-ate complexes to organic
halides is reported: Ashby, E. C.; DePriest, R. N.; Tuncay, A.; Srivastava, S.
Tetrahedron Lett. 1982, 23, 5251-5254. Also see refs 12 and 13.