Regio- and enantioselective control in the reactions of α-(N-Carbamoyl)alkylcuprates with allylic phosphates

Article abstract of DOI:10.1021/ol0364576

α-(N-Carbamoyl)alkylcuprates (RCuCNLi or R₂CuLi) react with allylic phosphates to afford homoallylic amines in good chemical yields. Regioselectivity is governed by steric factors in both the cuprate reagent and phosphate substrate and systems can be designed to give either the S _N2′ or S_N2 substitution product cleanly. Excellent enantioselectivities can be achieved with either a scalemic α-di[(N- carbamoyl)-alkyl]cuprate and an achiral phosphate or with a scalemic allylic phosphate and an achiral cuprate reagent.

Full text of DOI:10.1021/ol0364576

ORGANIC
LETTERS
2004
Vol. 6, No. 5
763-766
Regio- and Enantioselective Control in
the Reactions of
r-(N-Carbamoyl)alkylcuprates with
Allylic Phosphates
R. Karl Dieter,* Vinayak K. Gore, and Ningyi Chen
Howard L. Hunter Laboratory, Department of Chemistry, Clemson UniVersity,
Clemson, South Carolina 29634-1905
dieterr@clemson.edu
Received December 18, 2003
ABSTRACT
r-(N-Carbamoyl)alkylcuprates (RCuCNLi or R CuLi) react with allylic phosphates to afford homoallylic amines in good chemical yields.
2
Regioselectivity is governed by steric factors in both the cuprate reagent and phosphate substrate and systems can be designed to give
either the SN2′ or SN2 substitution product cleanly. Excellent enantioselectivities can be achieved with either a scalemic r-di[(N-carbamoyl)-
alkyl]cuprate and an achiral phosphate or with a scalemic allylic phosphate and an achiral cuprate reagent.
Although R-(N-carbamoyl)alkylcuprate chemistry1 is a po-
tentially powerful synthetic methodology for the synthe-
sis of nitrogen heterocycles,2 the utilization of allylic
substrates remains undeveloped.3 Allylation of organo-
metallic reagents offers rich opportunities for C-C bond
construction,4 enantiocontrol using either scalemic sub-
strates4b,c,5 or organometallic reagents,4c,6 and subsequent
functional group interconversions. Efforts to exercise regio-
control in allylic substitution have focused on the organo-
metallic reagent, the leaving group in the allylic substrate,
and solvent influences.4-5,7-12 Generally, reaction conditions
(5) (a) Chong, J. M.; Belelie, J. L. J. Org. Chem. 2001, 66, 5552. (b)
Belelie, J. L.; Chong, J. M. J. Org. Chem. 2002, 67, 3000. (c) Calaza, M.
I.; Hupe, E.; Knochel, P. Org. Lett. 2003, 5, 1059. (d) Harrington-Frost,
N.; Leuser, H.; Calaza, M. I.; Kneisel, F. F.; Knochel, P. Org. Lett. 2003,
5, 2111.
(6) Dieter, R. K.; Topping, C. M.; Chandupatla, K. R.; Lu, K. J. Am.
Chem. Soc. 2001, 123, 5132.
(7) For a review see: (a) Nakamura, E.; Mori, S. Angew. Chem., Int.
Ed. 2000, 39, 3750.
(8) (a) Tseng, C. C.; Yen, S.-J.; Goering, H. L. J. Org. Chem. 1986, 51,
2892. (b) Tseng, C. C.; Paisley, S. D.; Goering, H. L. J. Org. Chem. 1986,
51, 2884. (c) Ba¨ckvall, J.-E.; Selle´n, M.; Grant, B. J. Am. Chem. Soc. 1990,
112, 6615. (d) Persson, E. S. M.; Ba¨ckvall, J.-E. Acta Chim. Scand. 1995,
49, 899.
(1) For a review see: Dieter, R. K. Heteroatomcuprates and R-Hetero-
atomalkylcuprates in Organic Synthesis. In Modern Organocopper Chem-
istry; Krause, N., Ed.; John Wiley & Sons: New York, 2002; Chapter 3, p
79.
(2) (a) Li, S.-L.; Dieter, R. K. J. Org. Chem. 2003, 68, 969. (b) Dieter,
R. K.; Watson, R. Tetrahedron Lett. 2002, 43, 7725. (c) Dieter, R. K.; Lu,
K. J. Org. Chem. 2002, 67, 847. (d) Dieter, R. K.; Yu, H. Org. Lett. 2000,
2, 2283.
(3) Dieter, R. K.; Velu, S. E.; Nice, L. E. Synlett 1997, 1114.
(4) For reviews of organocopper-mediated allylic substitution, see: (a)
Lipshutz, B. H.; Sengupta, S. Org. React. 1992, 41, 135. (b) Breit, B.;
Demel, P. Copper-Mediated Diastereoselective Conjugate Addition and
Allylic Substitution Reactions. In Modern Organocopper Chemistry; Krause,
N., Ed.; John Wiley & Sons: New York, 2002; Chapter 6, p 188. (c) Sofia,
A.; Karlstro¨m, E.; Ba¨ckvall, J.-E. Copper-Mediated Enantioselective
Substitution Reactions. In Modern Organocopper Chemistry; Krause, N.,
Ed.; John Wiley & Sons: New York, 2002; Chapter 8, p 259.
(9) Ba¨ckvall, J.-E.; Persson, E. S. M.; Bombrun, A. J. Org. Chem. 1994,
59, 4126.
(10) (a) Valverde, S.; Bernabe´, M.; Go´mez, A. M.; Puebla, P. J. Org.
Chem. 1992, 57, 4546. (b) Calo´, V.; Lopez, L.; Carlucci, W. F. J. Chem.
Soc., Perkin Trans. 1 1983, 2953.
(11) (a) Goering, H. L.; Kantner, S. S.; Seitz, E. P., Jr. J. Org. Chem.
1985, 50, 5495. (b) Goering, H. L.; Kantner, S. S. J. Org. Chem. 1983, 48,
721. (c) Goering, H. L.; Seitz, E. P., Jr.; Tseng, C. C. J. Org. Chem. 1981,
46, 5304.
10.1021/ol0364576 CCC: $27.50 © 2004 American Chemical Society
Published on Web 02/06/2004

favoring formation of R2CuM (M ) Li, MgX) reagents
facilitate formal SN2 substitution, while conditions favoring
formation of RCuXM (X ) Cl, CN; M ) Li, MgX) promote
SN2′ substitution involving allylic rearrangement. Use of
Lewis acid additives enhances SN2′ selectivity.7,8 This general
pattern holds for reactions involving copper(I)-catalyzed
Grignard reagents with allylic esters,8 halides,8c,9 and sulfides
of benzothiozole-2-thiol.10 Reactions of organocopper(I) or
lithium cuprate reagents with allylic halides display the same
general patterns.11 Et2O as a solvent generally favors SN2′
substitution, while THF favors SN2 substitution, although in
Et2O the choice of Cu(X) (X ) Cl, Br, I, CN)8c-d can
determine selectivity. Allylic phosphates12 strongly favor anti
SN2′ substitution with cuprates prepared from CuCN, while
CuOTf and CuSCN favor SN2 substitution. These results
illustrate the sensitivity of organocopper-mediated reactions
to solvent effects,8c-d,13a-c Cu(I) salt,8b,12,13d and the nature
of the reagent.4b,c,5,8b-d
corresponding alkoxides [LDA, -50 °C, THF] with (PhO)2P-
(O)Cl. These allylic phosphates, along with 1e, were unstable
to chromatographic purification (e.g., silica gel, alumina, or
florisil), and the crude preparations were used. The requisite
scalemic allylic alcohols needed for phosphates 2a,b were
obtained by asymmetric addition of 1-(5-phenyl)pentynylzinc
to isobutrylaldehyde (82%, 92% ee)14 followed by reduction
to either the trans15 or cis16 allylic alcohols. trans-Carveol17
was converted to phosphate 3, while cis-carveol18 was
converted to phosphate 4.
Deprotonation of carbamates 5a-c (Scheme 1) according
to Beak’s procedure19 followed by treatment with THF-
Scheme 1
Recent studies suggest that anti SN2′ regioselectivity may
be enhanced by use of allylic phosphates,5a-c,12 allylic
perfluorobenzoates,5d CuCN-derived cuprates5,8c-d,12 or zinc
cuprate5c-d reagents alone or in combination. With these
caveats in mind, the development of regio- and stereo-
selective reactions between allylic phosphates and R-(N-
carbamoyl)alkylcuprates was undertaken. We now report that
R-(N-carbamoyl)alkylcuprates undergo allylic substitution
reactions in high yield and that excellent SN2 or SN2′
regioselectivity can be obtained by judicious choice of
cuprate reagent, R-(N-carbamoyl)alkyl ligand, and reaction
conditions. Utilization of scalemic R-(N-carbamoyl)alkyl-
cuprates or scalemic allylic phosphates affords good to
excellent enantiocontrol.
Allylic phosphates 1a-e (Figure 1) were prepared by
soluble CuCN‚2LiCl afforded either the lithium alkylcyano-
cuprates 6a-c (i.e., RCuCNLi) or the lithium dialkylcuprates
7a-c (i.e., R2CuLi). Formation of scalemic cuprates from
5b required asymmetric deprotonation in Et2O followed by
addition of THF soluble CuCN‚2LiCl to afford a 1:1 Et2O/
THF solvent mixture of the stereogenic cuprate reagent.6 The
reaction of cuprate reagents 6a-c or 7a-c with allylic
phosphates gave homoallylic amines (Scheme 1, Table 1)
in modest to excellent yields.
reaction of the allylic alcohol [THF, Et3N, 25 °C]5a with
Reaction of cuprate reagents with allylic phosphates 1a,b
affords the same substitution product via either reaction
pathway. The alkylcyanocuprates 6a,b gave higher yields
of carbamoyl alkenes 8a,b than the corresponding lithium
dialkylcuprates 7a,b upon reaction with 1a (Table 1, entries
1-4). In the reactions of lithium dialkylcuprates 7a,b with
1a, homocoupling dimers [i.e., Me(Boc)NCH2CH2N(Boc)Me
(37%) and bis-N-Boc-pyrrolidine (22%), respectively] were
Figure 1. Allylic phosphates.
(13) (a) House, H. O.; Lee, T. V. J. Org. Chem. 1978, 43, 4369. (b)
Bertz, S. H.; Dabbagh, G. Tetrahedron 1989, 45, 425. (c) Christenson, B.;
Hallnemo, G.; Ullenius, C. Tetrahedron 1991, 47, 4739. (d) Bertz, S. H.;
Gibson, C. P.; Dabbagh, G. Tetrahedron Lett. 1987, 28, 4251.
(14) Jiang, B.; Chen, Z.; Tang, X. Org. Lett. 2002, 4, 3451.
(15) Andre’s, J. M.; Pedrosa, R. Tetrahedron: Asymmetry 1998, 9, 2493.
(16) Burgstahler, A. W.; Widiger, G. J. Org. Chem. 1973, 38, 3652.
(17) Dupuy, C.; Luche, J. L. Tetrahedron 1989, 45, 3437.
(PhO)2P(O)Cl in 65-90% yields after purification. Allylic
phosphates 2a,b, 3, and 4 were prepared by reaction of the
(12) (a) Yanagisawa, A.; Nomura, N.; Yamamoto, H. Tetrahedron 1994,
50, 6017 and references therein. (b) Yanagisawa, A.; Nomura, N.; Noritake,
Y.; Yamamoto, H. Synthesis 1991, 1130.
(18) Ireland, R. E.; Maienfisch, P. J. Org. Chem. 1988, 53, 640.
(19) Beak, P.; Basu, A.; Gallaghar, D. J.; Park, Y. S.; Thayumanavan,
S. Acc. Chem. Res. 1996, 29, 552 and references therein.
764
Org. Lett., Vol. 6, No. 5, 2004

cuprate reagent with the less reactive phosphate, gives higher
chemical yields. When the R2CuLi reagents 7a-c give higher
chemical yields than the RCuCNLi reagents, it may also
reflect formation of RCuCNLi upon reaction of R2CuLi with
allylic phosphates.
Table 1. Reaction of R-(N-Carbamoyl)alkylcuprates with
Acyclic Allylic Phosphates
The SN2′/SN2 product distributions reflect substitution
patterns in both the cuprate transferable ligand and in the
allylic phosphates, as well as the actual cuprate reagent
employed. With allylic phosphate 1c, there is little difference
in regioselectivity between cuprates 6a,b (RCuCNLi) and
7a,b (R2CuLi), although the difference is larger for the less
sterically hindered cuprates 6a and 7a (entries 9-12). The
primary allylic phosphate with a methyl substituent on the
olefin (i.e., 1d) regioselectively affords the SN2′ substitution
product but with reduced selectivity, and now cuprates 6a,b
give a higher SN2′/SN2 ratio than cuprates 7a,b (13-16).
Almost no regioselectivity is observed with the lithium
dipyrrolidinylcuprate 7b and 1d (entry16), reflecting steric
congestion at both reacting centers. Similarly, trans-cinnamyl
phosphate 1e containing a primary alkyl phosphate and a
phenyl substitutent affords higher SN2′/SN2 ratios with 6a,b
than with 7a,b, and with pyrrolidinylcuprate 7b the SN2
pathway predominates with modest selectivity (entries 17-
20). The benzylic R-(N-carbamoyl)alkylcuprates adhere to
the same pattern but with generally excellent regioselectivity.
Reaction of 6c or 7c with phosphate 1c gives exclusively
the SN2′ substitution product (entries 21-22), while reaction
with 1d gives the SN2 substitution product predominately
with 6c and exclusively with 7c (entries 23-24). Reaction
of the pyrrolidinylcuprates 6b and 7b with allylic phosphates
1d-e generate two new stereogenic centers, yielding mix-
tures of diastereomers ranging between 56:44 and 62:38
(entries 15-16 and 19-20).
Asymmetric deprotonation of N-Boc pyrrolidine generates
a scalemic organolithium reagent that can be converted into
a scalemic organocopper reagent if the transmetalation is
performed at -78 °C in a mixed solvent system employing
THF-soluble CuCN‚2LiCl. Utilizing this protocol, lithium
dipyrrolidinylcuprate 7b reacted with allylic phosphates 1a-c
and 1e to give the R-enantiomers6 with good enantioselec-
tivities (entries 4, 8, 12, 20; er ) 90:10 to 85:15), while no
enantioselectivity was observed with the alkylcyanocuprate
reagent 6b (entries 3, 7, 11, and 19). A complex mixture of
diastereomers and regioisomers precluded enantiomer analy-
sis for the reaction of 7b with 1d (entry 16). No effort was
made to optimize these preliminary results, which suggest
that high enantioselectivities can be achieved with scalemic
R-(N-carbamoyl)alkylcuprates and allylic phosphates.
a N-Boc deprotonation [s-BuLi, THF, or Et2O, (-)-sparteine, -78 °C,
45 min], cuprate formation [-78 °C, 45 min] followed by reaction with
the allyl phosphate [-78 to 25 °C, 12 h]. b Determined by NMR and GC-
MS measurements. c Based upon isolated material purified by colunm
chromatography. d Enantiomeric ratios determined by chiral stationary phase
HPLC on a CHIRALCEL OD column [cellulose tris-(3,5-dimethylphenyl-
carbamate) on silica gel]. e Diastereomeric ratio (dr) ) 61.5:38.5. f No
separation on HPLC. g Dr ) 57:43. h Dr ) 60:40. i For SN2 product. j Dr
) 56:44.
Chirality may be introduced into the allylic phosphate and
potentially exploited for enantiocontrol given the generally
anti SN2′ preference4c for cuprate-mediated allylic substitu-
tions. Reaction of 6a with acyclic allylic phosphates 2a,b
containing either a trans- or cis-substituted olefin gave
modest chemical yields of the substitution product (Table
2, entries 1-4). Higher chemical yields were achieved with
the cyclic allylic phosphates 3-4. Trans-disubstituted cyclo-
hexenyl allylic phosphate 3 gave good enantioselectivity with
6a and poor ees with 6b, while both reagents gave excellent
also isolated. In contrast, the dialkylcuprates 7a,b gave higher
yields than the alkylcyanocuprates 6a,b upon reaction with
1b-e (entries 5-20). The same pattern was observed for
the combination of allylic phosphates 1c,d and benzylic R-(N-
carbamoyl)alkylcuprates 6c and 7c (entries 21-24), with the
latter reagent affording significantly higher chemical yields.
Chemical yields thus appear to reflect matching reactivities
between the allylic phosphate and the cuprate reagents. The
combination of the less reactive cuprate reagent with the
more reactive phosphate, as well as that of the more reactive
Org. Lett., Vol. 6, No. 5, 2004
765

electron-withdrawing group (i.e., CN vs R-(N-carbamoyl)).
This parallels the effect of cuprate reagent on regiocontrol
observed for primary allylic phosphates 1d,e (Table 1, entries
13-20, 23-24) where steric effects do not dominate at the
allylic phosphate center.
Table 2. Reaction of R-(N-Carbamoyl)alkylcuprates Derived
from N-Boc-N,N-Dimethylamine with Scalemic Allylic
Phosphates
In summary, we have shown that excellent regioselectivity
can be achieved in the reactions of R-(N-carbamoyl)-
alkylcuprates with allylic phosphates and that the selectivity
is largely governed by steric factors in both the cuprate
reagent and phosphate substrate. Excellent enantioselectivities
can be achieved by either the combination of scalemic
stereogenic R-(N-carbamoyl)alkylcuprate reagents and achiral
allylic phosphates or with achiral carbamoylalkylcuprates and
scalemic allylic phosphates. Reaction of allylic phosphates
with stereogenic R2CuLi R-pyrrolidinylcuprates gives good
ees, while use of the RCuCNLi reagent gives excellent ees
with scalemic allylic phosphates. Thus, good to excellent
regiocontrol and good to excellent enantiocontrol at 2-
pyrrolidinyl stereocenters and at the allylic stereogenic
centers can be achieved in the reactions of R-(N-carbamoyl)-
alkylcuprates with allylic phophates by judicious choice of
reagent and substrate combinations. Generation of adjacent
chiral centers affords poor diastereomeric ratios, and efforts
are underway to control the diastereoselectivity of the
reaction. These preliminary studies suggest that modification
of cuprate reagents, cuprate compositions, and allylic phos-
phates will provide ample opportunities for synthetic ap-
plications yielding high regio- and enantioselectivities.
a N-Boc deprotonation [s-BuLi, THF, or Et2O, (-)-sparteine, -78 °C,
45 min], cuprate formation [-78 °C, 45 min] followed by reaction with
the allyl phosphate [-78 to 25 °C, 12 h]. b Determined by NMR and GC-
MS measurements. c Based upon isolated material purified by colunm
chromatography. d Enantiomeric ratios determined by chiral stationary phase
HPLC on a CHIRALCEL OD column [cellulose tris-(3,5-dimethylphenyl-
carbamate) on silica gel]. e SN2′:SN2 ratios are the same as the ees
determined for the anti SN2′ products. Anti SN2′:Syn SN2′ ) 92:8f to 86:
14g.
Acknowledgment. This work was generously supported
by the National Institutes of Health (GM-60300-01). Support
of the NSF Chemical Instrumentation Program for purchase
of a JEOL 500 MHz NMR instrument is gratefully acknowl-
edged (CHE-9700278).
ees with the cis-disubstituted cyclohexenyl allylic phosphate
4 (entries 5-8). The trans diastereomer 3 has one axial and
one equatorial substituent, while the cis diastereomer 4 has
two axial substituents in the putative conformation required
for anti SN2′ substitution and minor amounts of syn SN2′
products are observed from both 3 and 4. The role of
cuprate-substrate interactions in the disparity of ees between
3 and 4 is unclear.12b Cuprate 6a uniformly gives higher ees
than 7a (Table 2) consistent with a more rapid reductive
elimination when the Cu(III) intermediate contains an
Supporting Information Available: General experimen-
tal procedures for preparation and reactions of R-(N-
carbamoyl)alkylcuprates with allylic phosphates and data
reduction for compounds 8a, 9a,b, 10a,b, 11a,b, 13-18; 13
C
NMR for 11b, 13, 14, 16-18; and HPLC traces for 15-18.
This material is available free of charge via the Internet at
http://pubs.acs.org.
OL0364576
766
Org. Lett., Vol. 6, No. 5, 2004